Evidence-Based Reviews

Psychiatry is experiencing a major paradigm shift.¹ No longer is depression a disease of norepinephrine and serotonin deficiency. Today, we are exploring inflam­mation, methylation, epigenetics, and neuroplasticity as major players; we are using innovative treatment interven­tions such as ketamine, magnets, psilocin, anti-inflammato­ries, and even botulinum toxin.

In 2006, dermatologist Eric Finzi, MD, PhD, reported a case series of 10 depressed patients who were given a single course of botulinum toxin A (BTA, onabotulinum-toxinA) injections in the forehead.² After 2 months, 9 out of the 10 patients were no longer depressed. The 10th patient, who reported improvement in symptoms but not remission, was the only patient with bipolar depression.

As a psychiatrist (M.M.) and a dermatologist (J.R.), we conducted a randomized controlled trial³ to challenge the difficult-to-swallow notion that a cosmetic intervention could help severely depressed patients. After reporting our positive findings and hearing numerous encourag­ing patient testimonials, we present a favorable review on the treatment of depression using BTA. We also present the top 10 questions we are asked at lectures about this novel treatment.

A deadly toxin used to treat medical conditions
Botulinum toxin is one of the deadliest sub­stance known to man.⁴ It was named after the gram-positive bacterium Clostridium botulinum, which causes so-called floppy baby syndrome in infants who eat contam­inated honey. Botulinum toxin prevents nerves from releasing acetylcholine, which causes muscle paralysis.

In the wrong hands, botulinum toxin can be exploited for chemical warfare.⁴ However, doctors are using it to treat >50 medical conditions, including migraine, cervical dystonia, strabismus, overac­tive bladder, urinary incontinence, exces­sive sweating, muscle spasm, and now depression.^5,6 In 2014, BTA was the top cos­metic treatment in the United States, with >3 million procedures performed, generat­ing more than 1 billion dollars in revenue.⁷

The most common site injected with BTA for cosmetic treatments is the glabel­lar region, which is the area directly above and in between the eyebrows (ie, the lower forehead). The glabella comprises 2 main muscles: the central procerus flanked by a pair of corrugators (Figure). When express­ing fear, anger, sadness, or anguish, these muscles contract, causing the appearance of 2 vertical wrinkles, referred to as the “11s.” The wrinkles also can form the shape of an upside-down “U,” known as the omega sign.⁸ BTA prevents contraction of these muscles and therefore prevents the appear­ance of a furrowed brow. During cosmetic procedures, approximately 20 to 50 units of BTA are spread out over 5 glabellar injection sites.⁹ A similar technique is being used in studies of BTA for depression^2,3,10,11 (Figure).

BTA for depression is new to the men­tal health world but, before psychiatrists caught on, dermatologists were aware that BTA could improve quality of life,¹² reduce negative emotions,¹³ and increase feelings of well-being.¹⁴

The evidence
To date, there have been 2 case series,^2,15 3 ran­domized control trials (RCTs),^3,10,11 1 pooled analysis,^16,17 and 1 meta-analysis¹⁸ looking at botulinum for depression (Table 1).^{2,10,11,15-17} In each trial, a single treatment of BTA (ie, 1 doctor’s visit; 29 to 40 units of BTA distributed into 5 glabellar injections sites), was the intervention studied.²

The first case series, by Finzi and Wasserman² is described above. A second case series, published in 2013, describes 50 female patients, one-half depressed and one-half non-depressed, all of whom received 20 units of BTA into the gla­bella.¹⁵ At 12 weeks, depression scores in the depressed group had decreased by 54% (14.9 point drop on Beck Depression Inventory [BDI], P < .001) and self-esteem scores had increased significantly. In non-depressed participants, depression scores and self-esteem scores remained constant throughout the 12 weeks.

A pooled analysis reported results of 3 RCTs16,17 consisting of a total of 134 depressed patients, males and females age 18 to 65 who received BTA (n = 59) or pla­cebo (n = 74) into the glabellar region. At the end of 6 weeks, BDI scores in the depressed group had decreased by 47.4% (14.3 points) compared with a 16.2% decrease (5.1 points) in the placebo group. This corresponds to a 52.5% vs 8.0% response rate and a 42.4% vs 8.0% remission rate, respectively (Table 1,^{1,2,10,11,15-17}). There was no dif­ference between the 2 groups in sex, age, depression severity, and number of anti­depressants being taken. Females received 29 units and males received 10 to 11 units more to account for higher muscle mass (Figure).

Depression as measured by the physician-administered Hamilton Depression Rating Scale (HAM-D) and the Montgomery-Åsberg Depression Rating Scale showed similar reduction in overall scores (−45.7% vs −14.6%), response rates (54.2% vs 10.7%) and remission rates (30.5% vs 6.7%) with BTA.

Although these improvements in depres­sion scores do not reach those seen with electroconvulsive therapy,^19,20 they are com­parable to placebo-controlled studies of antidepressants.^21,22

Doesn’t this technique work because people who look better, feel better?
Aesthetic improvement alone is unlikely to explain the entire story. A recent study showed that improvement in wrinkle score did not correlate with improvement in mood.²³ Furthermore, some patients in RCTs did not like the cosmetic effects of BTA but still reported feeling less depressed after treatment.¹⁰

How might it work?
Several theories about the mechanism of action have been proposed:
• The facial feedback hypothesis dates to Charles Darwin in 1872: Facial move­ments influence emotional states. Numerous studies have confirmed this. Strack et al²⁴ found that patients asked to smile while reading comics found them to be funnier. Ekman et al²⁵ found that imitating angry facial expressions made body temperature and heart rate rise. Dialectical behavioral therapy expert Marsha Linehan recognized the importance of modifying facial expres­sions (from grimacing to smiling) and pos­ture (from clenched fists to open hands) when feeling distressed, because it is hard to feel “willful” when your “mind is going one way and your body is going another.”²⁶ Accordingly, for a person who continuously “looks” depressed or distressed, reducing the anguished facial expression using botu­linum toxin might diminish the entwined negative emotions.

• A more pleasant facial expression improves social interactions, which leads to improvement in self-esteem and mood. Social biologists argue that (1) we are attracted to those who have more pleas­ant facial expressions and (2) we steer clear of those who appear angry or depressed (a negative facial expression, such as a growl­ing dog, is perceived as a threat). Therefore, anyone who looks depressed might have less rewarding interpersonal interactions, which can contribute to a poor mood.

On a similar note, mirror neurons are regions in the brain that are activated by witnessing another person’s emotional cues. When our mirror neurons light up, we can feel an observed experience, which is why we often feel nervous around anxious peo­ple, or cringe when we see others get hurt, or why we might prefer engaging with people who appear happier. It is possible that, after BTA injection, a person’s social connectiv­ity is improved because of a more positive reciprocal firing of mirror neurons.  

• BTA leads to direct and indirect neu­rochemical changes in the brain that can reduce depression. Functional MRI studies have shown that after glabellar BTA injec­tions, the amygdala was less responsive to negative stimuli.^27,28 For example, patients who were treated with BTA and then shown pictures of angry people had an attenuated amygdala response to the photos.

This is an important finding, especially for patients who have been traumatized. After a traumatic event, the amygdala “remem­bers” what happened, which is good, in some ways (it prevents us from getting into a similar dangerous situation), but bad in others (the traumatized amygdala may falsely perceive a non-threatening stimuli as threatening). A hypervigilant amygdala can lead to an out-of-proportion fear response, depression, and anxiety. Therefore, quell­ing an overactive amygdala with BTA could improve emotional dysregulation and posttraumatic disorders.

Many of our patients reported that, after BTA injection, “traumatic events didn’t feel as traumatizing,” as one said. The emotional pain and rumination that often follow a life stressor “does not overstay its welcome” and patients are able to “move on” more quickly.

It is unknown why the amygdala is qui­eted after BTA; researchers hypothesize that BTA suppresses facial feedback signals from the forehead branch of the trigeminal nerve to the brain. Another hypothesis is that BTA is directly transported by the trigeminal nerve into the brain and exerts central phar­macological effects on the amygdala.²⁹ This theory has only been studied in rat models.³⁰

When does it start working? How long does it last?
From what we know, BTA for depression could start working as early as 2 weeks and could last as long as 6 months. In one RCT, the earliest follow-up was 2 weeks,10 at which time the depressed patients had responded to botulinum toxin (P ≤ .05). In the other 2 controlled trials, the earliest follow-up was 3 weeks, at which time a more robust response was seen (P < .001). Aesthetically, BTA usually lasts 3 months. It is unclear how long the antidepressant effects last but, in the longest trial,3 depression symptoms contin­ued to improve at 6 months, after cosmetic effects had worn off.

These findings raise a series of questions:
   • Do mood effects outlast cosmetic effects? If so, why?
   • Does botulinum toxin start to work sooner than 2 weeks?
   • Will adherence improve if a patient has to be treated only every 6 months?

In our clinical experience, depressed patients who responded to BTA injection report a slow resurfacing of depressive symptoms 4 to 6 months after treatment, at which point they usually return for “mainte­nance treatment” (same dosing, same injec­tion configuration).

Will psychiatrists administer the treatment?
Any physician or physician extender can, when properly trained, inject BTA. The ques­tion is: Do psychiatrists want to? Administrating botulinum toxin requires more labor and preparation than prescrib­ing a drug (Table 2,³¹) and requires placing hands on patients. Depending on the type of psychiatric practice, this may be a “deal-breaker” for some providers, such as those in a psychoanalytic practice who might worry about boundaries.

As a basis for comparison, despite sev­eral indications for BTA for headache and neurologic conditions, few neurologists have added botulinum toxin to their prac­tice. Dermatologists who are comfortable seeing psychiatric patients or family practi­tioners, who are already set up for injection procedures, could become custodians of this intervention.

Which patients are candidates for the treatment?
Patients with anxious or agitated depres­sion might be ideal candidates for BTA injection. A recent study looked at predic­tors of response: Patients with a high agitation score (as measured on item 9 of the HAM-D) were more likely to respond, with a sensitivity of 100%, a specificity of 56%, and an overall precision of 78%.³² So far, no other predictors of response have been clearly identified. Higher baseline wrinkle scores do not predict better response.23 Sex and age do not have any predictive value. The treatment appears to be equally effective in males and females; because only a handful of males have been treated (n = 14), however, these patients need to be studied further.

Is botulinum toxin better as monotherapy or augmentation strategy?
So far, it appears to be equally effective as monotherapy or augmentation strategy,¹⁶ but more studies are needed.

How expensive is it?
Estimates of patient cost include the cost of the product and the professional fee for injection. As a point of reference, for cos­metic purposes, depending on practice loca­tion, dermatologists charge $11 to $20 per unit of BTA. Therefore, 1 treatment of BTA for depression (29 to 40 units) can cost a patient $319 to $800.

When treating a patient with BTA for medical indications, such as tension head­ache, insurance often reimburses the physi­cian for the BTA at cost (paid with a J code: J0585) and pay an injection fee (a procedure code) of $150 to $200. A recent analysis of cost-effectiveness estimated that BTA for depression would cost a patient $1,200 to $1,600 annually.³³ Compared with the price of branded medications (eg, $500 to $2,000 annually)³³ plus weekly psychotherapy (eg, $2,000 to $5,000 annually), BTA may be a cost-effective option for patients who do not respond to conventional treatments. Of course, for patients who tolerate and respond to generic medications or have a therapist who charges on a sliding scale, BTA is not the most cost-effective option.

What about injecting other areas of the face?
We’ve thought about it but haven’t tried it.  There are several muscles around the mouth that allow us to smile and frown. BTA injections in the depressor anguli oris, a muscle around the mouth that is largely responsible for frowning, could treat depres­sion. However, if the mechanism of action is via amygdala desensitization through the trigeminal nerve, treating mouth frown muscles might not work.

Is it safe?
BTA in the glabella has an exceptionally good safety profile.^9,31,34 Adverse reactions, which include eyelid droop, pain, bruising, and redness at the injection site, are minor and temporary.⁹ In addition, BTA has few drug–drug interactions. The biggest com­plaint for most patients is discomfort upon injection, which often is described as feeling like “an ant bite.”

In the pooled analysis of RCTs, apart from local irritation immediately after injection, temporary headache was the only relevant, and possibly treatment-related, adverse event. Headache occurred in 13.6% (n = 8) of the BTA group and 9.3% (n = 7) of the placebo group (P = .44). Compared with antidepressants such as citalopram, where approximately 38.6% of patients report a moderate or severe side-effect burden,²¹ BTA is well tolerated.

Are other studies underway?
Larger studies are being conducted,³⁵ mainly to confirm what pilot studies have shown. It would be interesting to discover other predictors of response and if different dos­ing and injection configurations could strengthen the response rate and extend the duration of effect.

Because of the cosmetic effects of BTA, further studies are needed to address the problem of blinding. In earlier studies, raters were blinded during appointments because patients wore surgical caps that covered their glabellar region.^3,10 Patients did not know their treatment intervention, but 52% to 90% of patients guessed correctly.^3,10,11 Although unblinding is a common problem in “blinded” trials in which some research­ers have reported >75% of participants and raters guessed the intervention correctly,³⁶ it is a particularly sensitive area in studies that involve a change in appearance because it is almost impossible to prevent someone from looking in a mirror.

Summing up
Botulinum toxin for depression is not ready for prime time. The FDA has not approved its use for psychiatric indications, and Medicare and commercial insurance do not reimburse for this procedure as a treatment for depression. Patients who request BTA for depression must be informed that this use is off-label.

For now, we recommend psychotherapy or medication management, or both, for most patients with major depression. In addition, until larger studies are done, we recommend that patients who are interested in BTA for depression use it as an add-on to conventional treatment. However, if larger studies replicate the findings of the smaller studies we have described, botulinum toxin could become a novel therapeutic agent in the fight against depression.

Bottom Line
In pilot studies, botulinum toxin A (BTA) has shown efficacy in improving symptoms of depression. Although considered safe, BTA is not FDA-approved for psychiatric indications, and Medicare and commercial insurance do not reimburse for this procedure for depression. Larger studies are underway to determine if this novel treatment can be introduced into practice.

Related Resources
• Wollmer MA, Magid M, Kruger THC. Botulinum toxin treat­ment in depression. In: Bewley A, Taylor RE, Reichenberg JS, et al, eds. Practical psychodermatology. Hoboken, NJ: John Wiley & Sons; 2014:216-219.
• Botox for depression. www.botoxfordepression.com.
• Botox and depression. www.botoxanddepression.com.

Drug Brand Names
Botulinum toxin A • Botox
Citalopram • Celexa

Acknowledgments
We thank the Brain and Behavior Research Foundation for grant­ing Dr. Magid a young investigator award and for continuing to in­vest in innovative research ideas. We thank Dr. Eric Finzi, MD, PhD, Axel Wollmer, MD, and Tillmann Krüger, MD, for their continued collaboration in this area of research.

Disclosures
In July 2011, Dr. Magid received a young investigator award from the Brain and Behavior Research Foundation for her study on treating depression using botulinum toxin (Grant number 17648). In November 2012, after completion and as a result of the study on treating depression using botulinum toxin, Dr. Magid became a consultant with Allergan to discuss study findings. In September 2015, Dr. Magid became a speaker for IPSEN Innovation. Dr. Reichenberg is married to Dr. Magid. Dr. Reichenberg has no other conflicts of interest to disclose.

Psychiatry is experiencing a major paradigm shift.¹ No longer is depression a disease of norepinephrine and serotonin deficiency. Today, we are exploring inflam­mation, methylation, epigenetics, and neuroplasticity as major players; we are using innovative treatment interven­tions such as ketamine, magnets, psilocin, anti-inflammato­ries, and even botulinum toxin.

In 2006, dermatologist Eric Finzi, MD, PhD, reported a case series of 10 depressed patients who were given a single course of botulinum toxin A (BTA, onabotulinum-toxinA) injections in the forehead.² After 2 months, 9 out of the 10 patients were no longer depressed. The 10th patient, who reported improvement in symptoms but not remission, was the only patient with bipolar depression.

As a psychiatrist (M.M.) and a dermatologist (J.R.), we conducted a randomized controlled trial³ to challenge the difficult-to-swallow notion that a cosmetic intervention could help severely depressed patients. After reporting our positive findings and hearing numerous encourag­ing patient testimonials, we present a favorable review on the treatment of depression using BTA. We also present the top 10 questions we are asked at lectures about this novel treatment.

A deadly toxin used to treat medical conditions
Botulinum toxin is one of the deadliest sub­stance known to man.⁴ It was named after the gram-positive bacterium Clostridium botulinum, which causes so-called floppy baby syndrome in infants who eat contam­inated honey. Botulinum toxin prevents nerves from releasing acetylcholine, which causes muscle paralysis.

In the wrong hands, botulinum toxin can be exploited for chemical warfare.⁴ However, doctors are using it to treat >50 medical conditions, including migraine, cervical dystonia, strabismus, overac­tive bladder, urinary incontinence, exces­sive sweating, muscle spasm, and now depression.^5,6 In 2014, BTA was the top cos­metic treatment in the United States, with >3 million procedures performed, generat­ing more than 1 billion dollars in revenue.⁷

The most common site injected with BTA for cosmetic treatments is the glabel­lar region, which is the area directly above and in between the eyebrows (ie, the lower forehead). The glabella comprises 2 main muscles: the central procerus flanked by a pair of corrugators (Figure). When express­ing fear, anger, sadness, or anguish, these muscles contract, causing the appearance of 2 vertical wrinkles, referred to as the “11s.” The wrinkles also can form the shape of an upside-down “U,” known as the omega sign.⁸ BTA prevents contraction of these muscles and therefore prevents the appear­ance of a furrowed brow. During cosmetic procedures, approximately 20 to 50 units of BTA are spread out over 5 glabellar injection sites.⁹ A similar technique is being used in studies of BTA for depression^2,3,10,11 (Figure).

BTA for depression is new to the men­tal health world but, before psychiatrists caught on, dermatologists were aware that BTA could improve quality of life,¹² reduce negative emotions,¹³ and increase feelings of well-being.¹⁴

The evidence
To date, there have been 2 case series,^2,15 3 ran­domized control trials (RCTs),^3,10,11 1 pooled analysis,^16,17 and 1 meta-analysis¹⁸ looking at botulinum for depression (Table 1).^{2,10,11,15-17} In each trial, a single treatment of BTA (ie, 1 doctor’s visit; 29 to 40 units of BTA distributed into 5 glabellar injections sites), was the intervention studied.²

The first case series, by Finzi and Wasserman² is described above. A second case series, published in 2013, describes 50 female patients, one-half depressed and one-half non-depressed, all of whom received 20 units of BTA into the gla­bella.¹⁵ At 12 weeks, depression scores in the depressed group had decreased by 54% (14.9 point drop on Beck Depression Inventory [BDI], P < .001) and self-esteem scores had increased significantly. In non-depressed participants, depression scores and self-esteem scores remained constant throughout the 12 weeks.

A pooled analysis reported results of 3 RCTs16,17 consisting of a total of 134 depressed patients, males and females age 18 to 65 who received BTA (n = 59) or pla­cebo (n = 74) into the glabellar region. At the end of 6 weeks, BDI scores in the depressed group had decreased by 47.4% (14.3 points) compared with a 16.2% decrease (5.1 points) in the placebo group. This corresponds to a 52.5% vs 8.0% response rate and a 42.4% vs 8.0% remission rate, respectively (Table 1,^{1,2,10,11,15-17}). There was no dif­ference between the 2 groups in sex, age, depression severity, and number of anti­depressants being taken. Females received 29 units and males received 10 to 11 units more to account for higher muscle mass (Figure).

Depression as measured by the physician-administered Hamilton Depression Rating Scale (HAM-D) and the Montgomery-Åsberg Depression Rating Scale showed similar reduction in overall scores (−45.7% vs −14.6%), response rates (54.2% vs 10.7%) and remission rates (30.5% vs 6.7%) with BTA.

Although these improvements in depres­sion scores do not reach those seen with electroconvulsive therapy,^19,20 they are com­parable to placebo-controlled studies of antidepressants.^21,22

Doesn’t this technique work because people who look better, feel better?
Aesthetic improvement alone is unlikely to explain the entire story. A recent study showed that improvement in wrinkle score did not correlate with improvement in mood.²³ Furthermore, some patients in RCTs did not like the cosmetic effects of BTA but still reported feeling less depressed after treatment.¹⁰

How might it work?
Several theories about the mechanism of action have been proposed:
• The facial feedback hypothesis dates to Charles Darwin in 1872: Facial move­ments influence emotional states. Numerous studies have confirmed this. Strack et al²⁴ found that patients asked to smile while reading comics found them to be funnier. Ekman et al²⁵ found that imitating angry facial expressions made body temperature and heart rate rise. Dialectical behavioral therapy expert Marsha Linehan recognized the importance of modifying facial expres­sions (from grimacing to smiling) and pos­ture (from clenched fists to open hands) when feeling distressed, because it is hard to feel “willful” when your “mind is going one way and your body is going another.”²⁶ Accordingly, for a person who continuously “looks” depressed or distressed, reducing the anguished facial expression using botu­linum toxin might diminish the entwined negative emotions.

• A more pleasant facial expression improves social interactions, which leads to improvement in self-esteem and mood. Social biologists argue that (1) we are attracted to those who have more pleas­ant facial expressions and (2) we steer clear of those who appear angry or depressed (a negative facial expression, such as a growl­ing dog, is perceived as a threat). Therefore, anyone who looks depressed might have less rewarding interpersonal interactions, which can contribute to a poor mood.

On a similar note, mirror neurons are regions in the brain that are activated by witnessing another person’s emotional cues. When our mirror neurons light up, we can feel an observed experience, which is why we often feel nervous around anxious peo­ple, or cringe when we see others get hurt, or why we might prefer engaging with people who appear happier. It is possible that, after BTA injection, a person’s social connectiv­ity is improved because of a more positive reciprocal firing of mirror neurons.  

• BTA leads to direct and indirect neu­rochemical changes in the brain that can reduce depression. Functional MRI studies have shown that after glabellar BTA injec­tions, the amygdala was less responsive to negative stimuli.^27,28 For example, patients who were treated with BTA and then shown pictures of angry people had an attenuated amygdala response to the photos.

This is an important finding, especially for patients who have been traumatized. After a traumatic event, the amygdala “remem­bers” what happened, which is good, in some ways (it prevents us from getting into a similar dangerous situation), but bad in others (the traumatized amygdala may falsely perceive a non-threatening stimuli as threatening). A hypervigilant amygdala can lead to an out-of-proportion fear response, depression, and anxiety. Therefore, quell­ing an overactive amygdala with BTA could improve emotional dysregulation and posttraumatic disorders.

Many of our patients reported that, after BTA injection, “traumatic events didn’t feel as traumatizing,” as one said. The emotional pain and rumination that often follow a life stressor “does not overstay its welcome” and patients are able to “move on” more quickly.

It is unknown why the amygdala is qui­eted after BTA; researchers hypothesize that BTA suppresses facial feedback signals from the forehead branch of the trigeminal nerve to the brain. Another hypothesis is that BTA is directly transported by the trigeminal nerve into the brain and exerts central phar­macological effects on the amygdala.²⁹ This theory has only been studied in rat models.³⁰

When does it start working? How long does it last?
From what we know, BTA for depression could start working as early as 2 weeks and could last as long as 6 months. In one RCT, the earliest follow-up was 2 weeks,10 at which time the depressed patients had responded to botulinum toxin (P ≤ .05). In the other 2 controlled trials, the earliest follow-up was 3 weeks, at which time a more robust response was seen (P < .001). Aesthetically, BTA usually lasts 3 months. It is unclear how long the antidepressant effects last but, in the longest trial,3 depression symptoms contin­ued to improve at 6 months, after cosmetic effects had worn off.

These findings raise a series of questions:
   • Do mood effects outlast cosmetic effects? If so, why?
   • Does botulinum toxin start to work sooner than 2 weeks?
   • Will adherence improve if a patient has to be treated only every 6 months?

In our clinical experience, depressed patients who responded to BTA injection report a slow resurfacing of depressive symptoms 4 to 6 months after treatment, at which point they usually return for “mainte­nance treatment” (same dosing, same injec­tion configuration).

Will psychiatrists administer the treatment?
Any physician or physician extender can, when properly trained, inject BTA. The ques­tion is: Do psychiatrists want to? Administrating botulinum toxin requires more labor and preparation than prescrib­ing a drug (Table 2,³¹) and requires placing hands on patients. Depending on the type of psychiatric practice, this may be a “deal-breaker” for some providers, such as those in a psychoanalytic practice who might worry about boundaries.

As a basis for comparison, despite sev­eral indications for BTA for headache and neurologic conditions, few neurologists have added botulinum toxin to their prac­tice. Dermatologists who are comfortable seeing psychiatric patients or family practi­tioners, who are already set up for injection procedures, could become custodians of this intervention.

Which patients are candidates for the treatment?
Patients with anxious or agitated depres­sion might be ideal candidates for BTA injection. A recent study looked at predic­tors of response: Patients with a high agitation score (as measured on item 9 of the HAM-D) were more likely to respond, with a sensitivity of 100%, a specificity of 56%, and an overall precision of 78%.³² So far, no other predictors of response have been clearly identified. Higher baseline wrinkle scores do not predict better response.23 Sex and age do not have any predictive value. The treatment appears to be equally effective in males and females; because only a handful of males have been treated (n = 14), however, these patients need to be studied further.

Is botulinum toxin better as monotherapy or augmentation strategy?
So far, it appears to be equally effective as monotherapy or augmentation strategy,¹⁶ but more studies are needed.

How expensive is it?
Estimates of patient cost include the cost of the product and the professional fee for injection. As a point of reference, for cos­metic purposes, depending on practice loca­tion, dermatologists charge $11 to $20 per unit of BTA. Therefore, 1 treatment of BTA for depression (29 to 40 units) can cost a patient $319 to $800.

When treating a patient with BTA for medical indications, such as tension head­ache, insurance often reimburses the physi­cian for the BTA at cost (paid with a J code: J0585) and pay an injection fee (a procedure code) of $150 to $200. A recent analysis of cost-effectiveness estimated that BTA for depression would cost a patient $1,200 to $1,600 annually.³³ Compared with the price of branded medications (eg, $500 to $2,000 annually)³³ plus weekly psychotherapy (eg, $2,000 to $5,000 annually), BTA may be a cost-effective option for patients who do not respond to conventional treatments. Of course, for patients who tolerate and respond to generic medications or have a therapist who charges on a sliding scale, BTA is not the most cost-effective option.

What about injecting other areas of the face?
We’ve thought about it but haven’t tried it.  There are several muscles around the mouth that allow us to smile and frown. BTA injections in the depressor anguli oris, a muscle around the mouth that is largely responsible for frowning, could treat depres­sion. However, if the mechanism of action is via amygdala desensitization through the trigeminal nerve, treating mouth frown muscles might not work.

Is it safe?
BTA in the glabella has an exceptionally good safety profile.^9,31,34 Adverse reactions, which include eyelid droop, pain, bruising, and redness at the injection site, are minor and temporary.⁹ In addition, BTA has few drug–drug interactions. The biggest com­plaint for most patients is discomfort upon injection, which often is described as feeling like “an ant bite.”

In the pooled analysis of RCTs, apart from local irritation immediately after injection, temporary headache was the only relevant, and possibly treatment-related, adverse event. Headache occurred in 13.6% (n = 8) of the BTA group and 9.3% (n = 7) of the placebo group (P = .44). Compared with antidepressants such as citalopram, where approximately 38.6% of patients report a moderate or severe side-effect burden,²¹ BTA is well tolerated.

Are other studies underway?
Larger studies are being conducted,³⁵ mainly to confirm what pilot studies have shown. It would be interesting to discover other predictors of response and if different dos­ing and injection configurations could strengthen the response rate and extend the duration of effect.

Because of the cosmetic effects of BTA, further studies are needed to address the problem of blinding. In earlier studies, raters were blinded during appointments because patients wore surgical caps that covered their glabellar region.^3,10 Patients did not know their treatment intervention, but 52% to 90% of patients guessed correctly.^3,10,11 Although unblinding is a common problem in “blinded” trials in which some research­ers have reported >75% of participants and raters guessed the intervention correctly,³⁶ it is a particularly sensitive area in studies that involve a change in appearance because it is almost impossible to prevent someone from looking in a mirror.

Summing up
Botulinum toxin for depression is not ready for prime time. The FDA has not approved its use for psychiatric indications, and Medicare and commercial insurance do not reimburse for this procedure as a treatment for depression. Patients who request BTA for depression must be informed that this use is off-label.

For now, we recommend psychotherapy or medication management, or both, for most patients with major depression. In addition, until larger studies are done, we recommend that patients who are interested in BTA for depression use it as an add-on to conventional treatment. However, if larger studies replicate the findings of the smaller studies we have described, botulinum toxin could become a novel therapeutic agent in the fight against depression.

Bottom Line
In pilot studies, botulinum toxin A (BTA) has shown efficacy in improving symptoms of depression. Although considered safe, BTA is not FDA-approved for psychiatric indications, and Medicare and commercial insurance do not reimburse for this procedure for depression. Larger studies are underway to determine if this novel treatment can be introduced into practice.

Related Resources
• Wollmer MA, Magid M, Kruger THC. Botulinum toxin treat­ment in depression. In: Bewley A, Taylor RE, Reichenberg JS, et al, eds. Practical psychodermatology. Hoboken, NJ: John Wiley & Sons; 2014:216-219.
• Botox for depression. www.botoxfordepression.com.
• Botox and depression. www.botoxanddepression.com.

Drug Brand Names
Botulinum toxin A • Botox
Citalopram • Celexa

Acknowledgments
We thank the Brain and Behavior Research Foundation for grant­ing Dr. Magid a young investigator award and for continuing to in­vest in innovative research ideas. We thank Dr. Eric Finzi, MD, PhD, Axel Wollmer, MD, and Tillmann Krüger, MD, for their continued collaboration in this area of research.

Disclosures
In July 2011, Dr. Magid received a young investigator award from the Brain and Behavior Research Foundation for her study on treating depression using botulinum toxin (Grant number 17648). In November 2012, after completion and as a result of the study on treating depression using botulinum toxin, Dr. Magid became a consultant with Allergan to discuss study findings. In September 2015, Dr. Magid became a speaker for IPSEN Innovation. Dr. Reichenberg is married to Dr. Magid. Dr. Reichenberg has no other conflicts of interest to disclose.

Psychiatry is experiencing a major paradigm shift.¹ No longer is depression a disease of norepinephrine and serotonin deficiency. Today, we are exploring inflam­mation, methylation, epigenetics, and neuroplasticity as major players; we are using innovative treatment interven­tions such as ketamine, magnets, psilocin, anti-inflammato­ries, and even botulinum toxin.

In 2006, dermatologist Eric Finzi, MD, PhD, reported a case series of 10 depressed patients who were given a single course of botulinum toxin A (BTA, onabotulinum-toxinA) injections in the forehead.² After 2 months, 9 out of the 10 patients were no longer depressed. The 10th patient, who reported improvement in symptoms but not remission, was the only patient with bipolar depression.

As a psychiatrist (M.M.) and a dermatologist (J.R.), we conducted a randomized controlled trial³ to challenge the difficult-to-swallow notion that a cosmetic intervention could help severely depressed patients. After reporting our positive findings and hearing numerous encourag­ing patient testimonials, we present a favorable review on the treatment of depression using BTA. We also present the top 10 questions we are asked at lectures about this novel treatment.

A deadly toxin used to treat medical conditions
Botulinum toxin is one of the deadliest sub­stance known to man.⁴ It was named after the gram-positive bacterium Clostridium botulinum, which causes so-called floppy baby syndrome in infants who eat contam­inated honey. Botulinum toxin prevents nerves from releasing acetylcholine, which causes muscle paralysis.

In the wrong hands, botulinum toxin can be exploited for chemical warfare.⁴ However, doctors are using it to treat >50 medical conditions, including migraine, cervical dystonia, strabismus, overac­tive bladder, urinary incontinence, exces­sive sweating, muscle spasm, and now depression.^5,6 In 2014, BTA was the top cos­metic treatment in the United States, with >3 million procedures performed, generat­ing more than 1 billion dollars in revenue.⁷

The most common site injected with BTA for cosmetic treatments is the glabel­lar region, which is the area directly above and in between the eyebrows (ie, the lower forehead). The glabella comprises 2 main muscles: the central procerus flanked by a pair of corrugators (Figure). When express­ing fear, anger, sadness, or anguish, these muscles contract, causing the appearance of 2 vertical wrinkles, referred to as the “11s.” The wrinkles also can form the shape of an upside-down “U,” known as the omega sign.⁸ BTA prevents contraction of these muscles and therefore prevents the appear­ance of a furrowed brow. During cosmetic procedures, approximately 20 to 50 units of BTA are spread out over 5 glabellar injection sites.⁹ A similar technique is being used in studies of BTA for depression^2,3,10,11 (Figure).

BTA for depression is new to the men­tal health world but, before psychiatrists caught on, dermatologists were aware that BTA could improve quality of life,¹² reduce negative emotions,¹³ and increase feelings of well-being.¹⁴

The evidence
To date, there have been 2 case series,^2,15 3 ran­domized control trials (RCTs),^3,10,11 1 pooled analysis,^16,17 and 1 meta-analysis¹⁸ looking at botulinum for depression (Table 1).^{2,10,11,15-17} In each trial, a single treatment of BTA (ie, 1 doctor’s visit; 29 to 40 units of BTA distributed into 5 glabellar injections sites), was the intervention studied.²

The first case series, by Finzi and Wasserman² is described above. A second case series, published in 2013, describes 50 female patients, one-half depressed and one-half non-depressed, all of whom received 20 units of BTA into the gla­bella.¹⁵ At 12 weeks, depression scores in the depressed group had decreased by 54% (14.9 point drop on Beck Depression Inventory [BDI], P < .001) and self-esteem scores had increased significantly. In non-depressed participants, depression scores and self-esteem scores remained constant throughout the 12 weeks.

A pooled analysis reported results of 3 RCTs16,17 consisting of a total of 134 depressed patients, males and females age 18 to 65 who received BTA (n = 59) or pla­cebo (n = 74) into the glabellar region. At the end of 6 weeks, BDI scores in the depressed group had decreased by 47.4% (14.3 points) compared with a 16.2% decrease (5.1 points) in the placebo group. This corresponds to a 52.5% vs 8.0% response rate and a 42.4% vs 8.0% remission rate, respectively (Table 1,^{1,2,10,11,15-17}). There was no dif­ference between the 2 groups in sex, age, depression severity, and number of anti­depressants being taken. Females received 29 units and males received 10 to 11 units more to account for higher muscle mass (Figure).

Depression as measured by the physician-administered Hamilton Depression Rating Scale (HAM-D) and the Montgomery-Åsberg Depression Rating Scale showed similar reduction in overall scores (−45.7% vs −14.6%), response rates (54.2% vs 10.7%) and remission rates (30.5% vs 6.7%) with BTA.

Although these improvements in depres­sion scores do not reach those seen with electroconvulsive therapy,^19,20 they are com­parable to placebo-controlled studies of antidepressants.^21,22

Doesn’t this technique work because people who look better, feel better?
Aesthetic improvement alone is unlikely to explain the entire story. A recent study showed that improvement in wrinkle score did not correlate with improvement in mood.²³ Furthermore, some patients in RCTs did not like the cosmetic effects of BTA but still reported feeling less depressed after treatment.¹⁰

How might it work?
Several theories about the mechanism of action have been proposed:
• The facial feedback hypothesis dates to Charles Darwin in 1872: Facial move­ments influence emotional states. Numerous studies have confirmed this. Strack et al²⁴ found that patients asked to smile while reading comics found them to be funnier. Ekman et al²⁵ found that imitating angry facial expressions made body temperature and heart rate rise. Dialectical behavioral therapy expert Marsha Linehan recognized the importance of modifying facial expres­sions (from grimacing to smiling) and pos­ture (from clenched fists to open hands) when feeling distressed, because it is hard to feel “willful” when your “mind is going one way and your body is going another.”²⁶ Accordingly, for a person who continuously “looks” depressed or distressed, reducing the anguished facial expression using botu­linum toxin might diminish the entwined negative emotions.

• A more pleasant facial expression improves social interactions, which leads to improvement in self-esteem and mood. Social biologists argue that (1) we are attracted to those who have more pleas­ant facial expressions and (2) we steer clear of those who appear angry or depressed (a negative facial expression, such as a growl­ing dog, is perceived as a threat). Therefore, anyone who looks depressed might have less rewarding interpersonal interactions, which can contribute to a poor mood.

On a similar note, mirror neurons are regions in the brain that are activated by witnessing another person’s emotional cues. When our mirror neurons light up, we can feel an observed experience, which is why we often feel nervous around anxious peo­ple, or cringe when we see others get hurt, or why we might prefer engaging with people who appear happier. It is possible that, after BTA injection, a person’s social connectiv­ity is improved because of a more positive reciprocal firing of mirror neurons.  

• BTA leads to direct and indirect neu­rochemical changes in the brain that can reduce depression. Functional MRI studies have shown that after glabellar BTA injec­tions, the amygdala was less responsive to negative stimuli.^27,28 For example, patients who were treated with BTA and then shown pictures of angry people had an attenuated amygdala response to the photos.

This is an important finding, especially for patients who have been traumatized. After a traumatic event, the amygdala “remem­bers” what happened, which is good, in some ways (it prevents us from getting into a similar dangerous situation), but bad in others (the traumatized amygdala may falsely perceive a non-threatening stimuli as threatening). A hypervigilant amygdala can lead to an out-of-proportion fear response, depression, and anxiety. Therefore, quell­ing an overactive amygdala with BTA could improve emotional dysregulation and posttraumatic disorders.

Many of our patients reported that, after BTA injection, “traumatic events didn’t feel as traumatizing,” as one said. The emotional pain and rumination that often follow a life stressor “does not overstay its welcome” and patients are able to “move on” more quickly.

It is unknown why the amygdala is qui­eted after BTA; researchers hypothesize that BTA suppresses facial feedback signals from the forehead branch of the trigeminal nerve to the brain. Another hypothesis is that BTA is directly transported by the trigeminal nerve into the brain and exerts central phar­macological effects on the amygdala.²⁹ This theory has only been studied in rat models.³⁰

When does it start working? How long does it last?
From what we know, BTA for depression could start working as early as 2 weeks and could last as long as 6 months. In one RCT, the earliest follow-up was 2 weeks,10 at which time the depressed patients had responded to botulinum toxin (P ≤ .05). In the other 2 controlled trials, the earliest follow-up was 3 weeks, at which time a more robust response was seen (P < .001). Aesthetically, BTA usually lasts 3 months. It is unclear how long the antidepressant effects last but, in the longest trial,3 depression symptoms contin­ued to improve at 6 months, after cosmetic effects had worn off.

These findings raise a series of questions:
   • Do mood effects outlast cosmetic effects? If so, why?
   • Does botulinum toxin start to work sooner than 2 weeks?
   • Will adherence improve if a patient has to be treated only every 6 months?

In our clinical experience, depressed patients who responded to BTA injection report a slow resurfacing of depressive symptoms 4 to 6 months after treatment, at which point they usually return for “mainte­nance treatment” (same dosing, same injec­tion configuration).

Will psychiatrists administer the treatment?
Any physician or physician extender can, when properly trained, inject BTA. The ques­tion is: Do psychiatrists want to? Administrating botulinum toxin requires more labor and preparation than prescrib­ing a drug (Table 2,³¹) and requires placing hands on patients. Depending on the type of psychiatric practice, this may be a “deal-breaker” for some providers, such as those in a psychoanalytic practice who might worry about boundaries.

As a basis for comparison, despite sev­eral indications for BTA for headache and neurologic conditions, few neurologists have added botulinum toxin to their prac­tice. Dermatologists who are comfortable seeing psychiatric patients or family practi­tioners, who are already set up for injection procedures, could become custodians of this intervention.

Which patients are candidates for the treatment?
Patients with anxious or agitated depres­sion might be ideal candidates for BTA injection. A recent study looked at predic­tors of response: Patients with a high agitation score (as measured on item 9 of the HAM-D) were more likely to respond, with a sensitivity of 100%, a specificity of 56%, and an overall precision of 78%.³² So far, no other predictors of response have been clearly identified. Higher baseline wrinkle scores do not predict better response.23 Sex and age do not have any predictive value. The treatment appears to be equally effective in males and females; because only a handful of males have been treated (n = 14), however, these patients need to be studied further.

Is botulinum toxin better as monotherapy or augmentation strategy?
So far, it appears to be equally effective as monotherapy or augmentation strategy,¹⁶ but more studies are needed.

How expensive is it?
Estimates of patient cost include the cost of the product and the professional fee for injection. As a point of reference, for cos­metic purposes, depending on practice loca­tion, dermatologists charge $11 to $20 per unit of BTA. Therefore, 1 treatment of BTA for depression (29 to 40 units) can cost a patient $319 to $800.

When treating a patient with BTA for medical indications, such as tension head­ache, insurance often reimburses the physi­cian for the BTA at cost (paid with a J code: J0585) and pay an injection fee (a procedure code) of $150 to $200. A recent analysis of cost-effectiveness estimated that BTA for depression would cost a patient $1,200 to $1,600 annually.³³ Compared with the price of branded medications (eg, $500 to $2,000 annually)³³ plus weekly psychotherapy (eg, $2,000 to $5,000 annually), BTA may be a cost-effective option for patients who do not respond to conventional treatments. Of course, for patients who tolerate and respond to generic medications or have a therapist who charges on a sliding scale, BTA is not the most cost-effective option.

What about injecting other areas of the face?
We’ve thought about it but haven’t tried it.  There are several muscles around the mouth that allow us to smile and frown. BTA injections in the depressor anguli oris, a muscle around the mouth that is largely responsible for frowning, could treat depres­sion. However, if the mechanism of action is via amygdala desensitization through the trigeminal nerve, treating mouth frown muscles might not work.

Is it safe?
BTA in the glabella has an exceptionally good safety profile.^9,31,34 Adverse reactions, which include eyelid droop, pain, bruising, and redness at the injection site, are minor and temporary.⁹ In addition, BTA has few drug–drug interactions. The biggest com­plaint for most patients is discomfort upon injection, which often is described as feeling like “an ant bite.”

In the pooled analysis of RCTs, apart from local irritation immediately after injection, temporary headache was the only relevant, and possibly treatment-related, adverse event. Headache occurred in 13.6% (n = 8) of the BTA group and 9.3% (n = 7) of the placebo group (P = .44). Compared with antidepressants such as citalopram, where approximately 38.6% of patients report a moderate or severe side-effect burden,²¹ BTA is well tolerated.

Are other studies underway?
Larger studies are being conducted,³⁵ mainly to confirm what pilot studies have shown. It would be interesting to discover other predictors of response and if different dos­ing and injection configurations could strengthen the response rate and extend the duration of effect.

Because of the cosmetic effects of BTA, further studies are needed to address the problem of blinding. In earlier studies, raters were blinded during appointments because patients wore surgical caps that covered their glabellar region.^3,10 Patients did not know their treatment intervention, but 52% to 90% of patients guessed correctly.^3,10,11 Although unblinding is a common problem in “blinded” trials in which some research­ers have reported >75% of participants and raters guessed the intervention correctly,³⁶ it is a particularly sensitive area in studies that involve a change in appearance because it is almost impossible to prevent someone from looking in a mirror.

Summing up
Botulinum toxin for depression is not ready for prime time. The FDA has not approved its use for psychiatric indications, and Medicare and commercial insurance do not reimburse for this procedure as a treatment for depression. Patients who request BTA for depression must be informed that this use is off-label.

For now, we recommend psychotherapy or medication management, or both, for most patients with major depression. In addition, until larger studies are done, we recommend that patients who are interested in BTA for depression use it as an add-on to conventional treatment. However, if larger studies replicate the findings of the smaller studies we have described, botulinum toxin could become a novel therapeutic agent in the fight against depression.

Bottom Line
In pilot studies, botulinum toxin A (BTA) has shown efficacy in improving symptoms of depression. Although considered safe, BTA is not FDA-approved for psychiatric indications, and Medicare and commercial insurance do not reimburse for this procedure for depression. Larger studies are underway to determine if this novel treatment can be introduced into practice.

Related Resources
• Wollmer MA, Magid M, Kruger THC. Botulinum toxin treat­ment in depression. In: Bewley A, Taylor RE, Reichenberg JS, et al, eds. Practical psychodermatology. Hoboken, NJ: John Wiley & Sons; 2014:216-219.
• Botox for depression. www.botoxfordepression.com.
• Botox and depression. www.botoxanddepression.com.

Drug Brand Names
Botulinum toxin A • Botox
Citalopram • Celexa

Acknowledgments
We thank the Brain and Behavior Research Foundation for grant­ing Dr. Magid a young investigator award and for continuing to in­vest in innovative research ideas. We thank Dr. Eric Finzi, MD, PhD, Axel Wollmer, MD, and Tillmann Krüger, MD, for their continued collaboration in this area of research.

Disclosures
In July 2011, Dr. Magid received a young investigator award from the Brain and Behavior Research Foundation for her study on treating depression using botulinum toxin (Grant number 17648). In November 2012, after completion and as a result of the study on treating depression using botulinum toxin, Dr. Magid became a consultant with Allergan to discuss study findings. In September 2015, Dr. Magid became a speaker for IPSEN Innovation. Dr. Reichenberg is married to Dr. Magid. Dr. Reichenberg has no other conflicts of interest to disclose.

In its simplest form, elder abuse refers to the intentional infliction of injury or neglect of an older adult by a caregiver. The 5 primary types of elder abuse include neglect, physical, financial, psychological/emotional, and sexual, with a subtype of social abuse that falls under psychological/emotional abuse.

Differentiating abuse from the normal sequelae of physiologic aging can be difficult; therefore, early identification and awareness of risk factors is key, as well as detailed documentation of the patient encounter. As soon as abuse is suspected, clinicians should report it to Adult Protective Services (APS) for further investigation. In terms of prevention, regular screening for elder abuse is still up for debate, but as the incidence of elder abuse continues to rise so should research and preventive efforts to combat this growing public health concern.
What is ‘elder abuse’?According to the Elder Abuse Prevention, Identification, and Treatment Act of 1985, elder abuse is:
…willful infliction of injury, unreasonable confinement, intimidation or cruel punishment with resulting physical harm or pain or mental anguish or the willful deprivation by a caretaker of goods or services which are necessary to avoid physical harm, mental anguish or mental illness.¹

There are 2 main components to this definition:
   • an older adult has suffered injury or deprivation
   • another person is responsible for causing or failing to prevent it.²

Although definitions vary, it generally is accepted that, for elder abuse to occur, it must take place within a relationship of trust.³

An ‘older adult’ is a person age ≥65, representing 14% of the U.S. population.⁴
According to U.S. Census Bureau data, there were 40 million older adults in 20105;
recent data project that this number will rise to 90 million by 2060 as Baby Boomers age.⁶ Studies suggest that as many as 10% of older adults in the United States experience elder abuse each year²; one study estimated that 6% of older adults in the community experienced significant abuse in the past month.⁷

Although elder abuse can occur in any setting, it takes place most often in the community. A survey of state APSs in 2000 showed that 60.7% of abuse was domestic; only 8.3% of incidents occurred in institutional settings.⁸ The annual direct medical costs associated with elder abuse injury in the United States are estimated at $5.3 billion, which is likely to increase with anticipated growth of the geriatric population.⁹

Although the number of older adults and the incidence of elder abuse are on the rise, as few as 1 in 14 cases is reported to authorities¹⁰; health care providers are some of the least likely of involved parties to report suspected abuse. One study found that 63% of physicians never asked about elder abuse, and only 31% reported encountering it in the previous 12 months.¹¹ A busy clinician—ie, one who sees 20 to 40 geriatric patients a day—has a high likelihood of encountering at least 1 victim of elder abuse,² yet many cases go unrecognized.
Types of abuseElder abuse comprises 5 categories¹²:
   • neglect (58.5% of cases)
   • physical (15.7%)
   • financial (12.3%)
   • psychological and emotional (7.3%)
   • sexual (0.04%).

Social abuse is considered a subtype of psychological and emotional abuse. All “other” types of abuse constitute 5.1% cases; 0.06% are of unknown type.¹²

Neglect is (1) failure of the caregiver to provide life necessities or (2) the responsible person’s refusal to permit others to provide appropriate care.³ This is one of the most common types of elder abuse in residential facilities. Signs of neglect include dehydration, depression, fecal impaction, and malnutrition (Table).⁴ The prevalence of dehydration in nursing home patients is reported to be as high as 35%, which may be the result of passive or active withholding of liquids (ie, decreasing hydration to reduce the need to change the resident’s clothing or bedding).⁴ Other forms of neglect include medication misuse (overdosing or underdosing) and self-neglect, which occurs among people living alone and often is listed as a subtype of neglect, but is controversial because it does not involve another person.
 

Financial exploitation is the illegal or improper use, or mismanagement, of a person’s money, property, or financial resources³—often, to his (her) detriment. Estimates are that 1 of every 20 older adults has been subject to financial abuse at some point in their life.¹⁵ There should be a high index of suspicion for financial exploitation when one notices unexplained changes in power of attorney, wills, or other legal documents; missing checks, money, or belongings.¹⁶ In the past, adult children were most likely to be financial abusers; in recent years, however, the abuser is more often a spouse—especially a second spouse.¹⁷ Bankers, accountants, and other financial advisors are among those trained to identify risk factors for financial abuse; they are encouraged to caution clients about this possibility.¹⁸

Psychological and emotional abuse occurs when a caregiver inflicts mental stress on an older adult by actions and threats that cause fear, violence, isolation, deprivation, or feelings of shame and powerlessness.³ Examples are threatening to put the older adult in a nursing home or verbally abusing him (her). Suspect this type of abuse when a caregiver refuses to leave the older adult or speaks for him, or if the older adult expresses fear in the presence of the caregiver.⁴ This type of abuse also is prevalent in nursing homes and other long-term care facilities.¹⁹

Sexual abuse involves nonconsensual touching or sexual activities (rape, language, exploitive behavior) that are threatened or forced on an older adult.¹⁶ Sexual abuse is more common in frail or dependent persons.³ Physical exam findings—particularly dysuria, tender genitalia, and evidence of sexually transmitted infections⁴—are required to identify sexual abuse, along with signs of depression and display of fear.

Social abuse can be considered a subtype of psychological and emotional abuse, in which a caregiver denies an older adult contact with family and friends or deprives him from access to transportation. Other examples include not allowing the older adult to use the telephone, monitoring phone calls, and claiming that his friends or family are “interfering.”²⁰ Intentionally embarrassing an older adult in front of others also can be considered social abuse.

Technology, particularly smart phones and social media, can complicate and exacerbate elder abuse:
   • In July 2013, employees of a Wisconsin nursing home were found with videos and photographs of residents bathing and of a nude resident who had a bowel obstruction being mocked.²¹
   • In May 2014, employees of a nursing home in Massachusetts recorded themselves physically and verbally abusing several older adults with Alzheimer’s disease, including one episode of the employees “hitting the woman on her arms, flicking her ears and then pinching the woman’s nose closed.” The employees also possessed a photograph of her naked.²²
   • In June 2015, an employee of a nursing home in Indiana was accused of taking
photos of a resident naked and sharing them on the messaging application Snapchat,²³ in which images disappear 10 seconds after they are viewed.

As technology evolves, caregivers are finding more cunning ways to abuse older adults. Considering current events and trends in this area, technology as a gateway to elder abuse should be of growing concern.
Risk factorsA 2013 literature review on elder abuse reported that the most important risk factors are related to relationship (family disharmony, poor or conflicting relationships) and environment (a low level of social support),³ although other variables can play a role. Regardless of these findings, it is important to recognize that (1) elder abuse is not a necessary consequence in a family with many risk factors and (2) elder abuse can occur in the absence of any risk factors.

As a whole, women are at a higher risk of abuse, particularly when combined with loneliness, poor social support, cohabitation (especially family members), substance abuse, cognitive impairment, and dementia and other mental health problems.⁴ Other risk factors include functional deficiency, poor physical health or frailty, low income or wealth, and trauma or past abuse.³

Lower income, poor health, low social support, and belonging to a non-white racial group put an older adult at risk for neglect; female sex is a specific risk factor for sexual assault.¹⁵ One study found that, among older adults who suffered physical, mental, or cognitive impairment, 1 of every 4 was at risk of abuse.⁷

Mental illness. Dementia puts an older adult at higher risk because of increased
caregiver stress resulting from disruptive and aggressive behaviors²; the same is true when the older adult suffers another mental illness, such as anxiety, depression, schizophrenia or bipolar disorder. Presumably, older adults with any of these disorders are at risk of financial and psychological and emotional abuse because of their decreased social support, lack of independence, and inability to hold a job—leaving their caregiver to shoulder more responsibilities and with more opportunities to inflict abuse. In addition, an older adult suffering from depression can feel helpless and unworthy, possibly making him more susceptible to psychological and emotional abuse, and less likely to seek help.

More research is needed to establish racial and ethnic differences in the risk of abuse. Some research states that older adults who are a member of a minority are at greater risk of abuse; however, the difference dissipates after adjusting for variables such as income and social support.²⁴ Cultural confounders, such as varying interpretations of the same set of interactions between older adults, need to be examined further.

Sexual orientation. Identifying one’s self as a lesbian, gay, bisexual, or transgender (LGBT) person is an additional risk factor for elder abuse. In 1997, a report described a nursing home employee who refused to bathe a resident because he didn’t want to “touch the lesbian.”²⁵ Despite evolving attitudes in society toward support and acceptance of sexual orientation, fear of homophobia still prevents some LGBT older adults from seeking help when they have been abused because of their orientation—especially ones who have internalized that
homophobia and feel that they are unworthy of seeking help.²⁵

In addition, health care providers and nursing home staff members might neglect the particular care needs of LGBT older adults, intentionally or unintentionally. APS staff and providers must be cognizant of underlying biases and exhibit respect when assisting LGBT clients.

Approximately 75% of caregivers of older adults are family members; 70% are female²⁶; and most are adult children, spouses, and partners of those receiving care.²⁷ Male caregivers age ≥40 are more likely to be the abuser, however, especially when they possess any of these risk factors: fatigue, burnout, medical illness, mental illness, lack of financial and support services, family history of abusive behavior, and substance abuse.4 People who commit elder abuse also tend to be significantly dependent on the person they are abusing.² In some cases, and especially when the abuser is financially needy, caregivers turn to elder abuse to obtain resources from the victim.²

From your standpoint as a practitioner, it is important to determine the root cause of elder abuse. According to one review,²⁸ family members with mental illness or a history of substance abuse, or who are stressed by the burden of caregiving, abuse older adults at a higher rate than family members who are not affected in those ways. Depression in particular is a common characteristic of abusers,² often secondary to the stress of caring for an older adult.

Abuse caused by stress can be addressed by referral to a support group and counseling for the caregiver; psychiatric conditions, such as depression, might be better treated with pharmacotherapy. Evaluate for depression and posttraumatic stress disorder (PTSD) in both the abuser and the abused,²⁹ and for other mental health issues that might compound the situation. It is possible for you to have 2 patients: the older adult and his caregiver. Regardless of the challenge,keep in mind that the older adult’s safety is your priority.
Consequences for the abusedThe abused adult is at risk of a number of serious physical and psychological consequences.³⁰ They tend to have a shorter lifespan, after adjusting for other variables associated with increased mortality.

The reason for shortened lifespan is multifactorial³⁰:
   • Bruises, abrasions, and fractures may take longer to heal because of diminished skin and bone regeneration.
   • Diseases that affect the heart, lungs, and kidneys might prevent the person from bouncing back from major stressors caused by abuse, such as blood loss, severe injury, and pain.
   • Injury from abuse can exacerbate an underlying illness.
   • Elder abuse also is associated with increased emergency department use, hospitalization (including readmission within 30 days), and nursing home placement.³¹

Elder abuse can lead to depression, shame, and guilt; increased isolation; and
increased risk of alcohol abuse and substance use.³¹ A study found that victims of
elder abuse are significantly more depressed than non-victims.³²

In the same study, being a victim of abuse was found to be the second-strongest
predictor of depression, after the state of one’s health.³² Other potential psychiatric
consequences of abuse that need further study include increased risk of developing
fear and anxiety disorders; learned helplessness; and PTSD.³³ According to LoFaso,
“depression and anxiety can consume their days and leave them emotionally and
physically frail.”29 Such feelings make these older adults less likely to resolve abuse or break off relations with the abuser.³²

Because mental illness can be a risk factor for, and a consequence of, elder abuse,
be aware of such complications and address them appropriately. Keep in mind that older adults are more likely to visit a primary care practitioner than a psychiatrist for a routine health check-up or evaluation of initial cognition-related problems; however, they are more likely to see a psychiatrist for advanced neuropsychiatric problems such as dementia, paranoia, delusions, hallucinations, and insomnia. Adequate education on elder abuse should not be limited to a single medical specialty because it can present in several clinical settings.
Identifying abuseIdentification of elder abuse in the home poses a greater challenge to clinicians than abuse in an institutional setting because it is not directly observable. Compounding this is the lack of unified standards for identifying and dealing with elder abuse. It is first necessary for you to determine the likelihood that abuse or neglect occurred, which can be difficult because the signs of elder abuse and manifestations of normal aging often are similar. You also must establish whether (1) the abused person will accept intervention and (2) the abused person who refuses intervention has the capacity to make that decision. Both of these conditions will guide your approach to management.²

Obtain the history from several sources; review the records; and carefully examine patterns of injury, in particular assessing functional status and level of dependency on the caregiver. Explanations that do not match injuries signal the need for further investigation and examination.

To help differentiate elder abuse from normal physiologic aging, look at the skin for bruises, rashes that do not heal, and ulcers—all of which could be signs of abuse or neglect. Keep in mind that bruising generally is more common in older adults because of the slower turnover rate of epidermal cells; physiologic bruising tends to occur on dorsal aspects of the hands and arms.⁴ In contrast, bruising secondary to neglect or physical abuse can manifest as a subgaleal hematoma (caused by traumatic hair pulling), tracking in the peritoneum after genital trauma, Battle’s sign, and raccoon eyes, among other findings.⁴

In addition, larger bruises (>5 cm in diameter) are more likely the result of elder abuse.³⁴ To complicate matters, many older persons are taking anticoagulant therapy, making bruising more likely. In addition, be on the lookout for burns during the physical exam. Evidence suggests that at least 10% of burns caused by battery and assault occur in the context of elder abuse; most burn facilities do not have formal guidelines for screening for abuse and neglect, however.³⁵ According to one retrospective study, the most common causes of burns in older adults are hot water scalds and radiator contact, and the mortality rate of older burn patients in general is higher than among the overall population.³⁶

Falls and fractures are common among older adults, regardless of whether they are
abused, because of polypharmacy, underlying medical conditions, and functional
limitations. Many abusers, however, use these factors to cover up intentional injury
that might have resulted in the older person falling, including overmedication (a form of physical abuse) and withholding a necessary walking aid (a form of neglect). Maintain a high index of suspicion of elder abuse when (1) the caregiver’s and the older adult’s stories of an injury don’t add up and (2) physical findings that might have been caused by abuse are present.

A number of psychiatric and cognitive symptoms suggest other types of elder abuse. Take note of emotional upset, agitation, and unusual behaviors³⁷—especially if you can follow the patient over time to observe marked changes in the presentation. Likewise, be aware of proposed alterations in guardianship, which should be evaluated by a forensic psychiatrist with analysis of medical history, social attachments, home environment, self-care, and finances.³⁸ Such evaluation should provide clues to the motivation behind a change of guardianship and will help to determine if elder abuse should be suspected.

Brandi et al³⁷ provided an informative table that identifies pertinent signs, symptoms, and other findings that clinicians should be aware of to support a suspicion of elder abuse (Table).

Documentation is of utmost importance in evaluating potential elder abuse; keep in mind that the medical record might be used in an investigation of abuse by social workers, law enforcement, and prosecutors. Your records should be legible, clearly indicate who the main caregiver is and what his (her) responsibilities are, and specify who is present at your encounter with the patient.⁴ Document your observations of patient behavior, reactions to questions, and family dynamics and conflicts¹⁶; make note of warning signs such as fear, silence, and inability to interview the patient alone.

In addition to written documentation, take photographs of injuries, with a ruler in the image to record their size. Serial photographs are helpful; so are photographs from a variety of distances (close-up, regional, wholebody) to capture detail and place the wound in the context of a specific area of the body.⁴

Safety is paramount. Given the findings of the history and physical exam, it is necessary to determine whether it is safe for the patient to return home with the caregiver, or if alternate accommodations or resources, such as a social worker or a support group, are required. Include details of planned follow-up in your evaluation, and offer consideration of possible psychiatric disorders that can develop as a result of such abuse.
ReportingElder abuse is a criminal offense in all states.³⁹ A clinician who has reasonable suspicion that elder abuse occurred must report it, regardless of whether the proof of abuse is concrete.⁴⁰ At a point of reasonable suspicion, immediately contact APS, law enforcement, and a social worker. Adult Protective Services, modeled after Child Protective Services, is typically administered by local and state health
departments.⁴¹

After a report is filed with APS, an assigned social worker makes an in-person home visit to investigate the allegation and determine whether elder abuse is substantiated, partially substantiated, or unsubstantiated.¹⁶ In most states, elder abuse reporting is not anonymous because follow-up may be needed to provide additional evidence, especially if the report was made by a health care provider.¹⁶

No federal standard exists for states to follow when defining and addressing elder abuse, which can complicate identification and reporting of abuse. Laws governing elder abuse do not allow states to determine the fate of the older adult, who can decide for himself (herself) whether to use or waive protective services.⁴² Older adults might choose not to report abuse because of shame, intimidation, or fear,⁴³ or to protect a caregiver, who often is a family member.

Elder abuse reports can come from a variety of sources; convincing evidence is, as noted, unnecessary to report it. Health care providers are mandated reporters, but
it is believed that the number of clinicians who report elder abuse based on suspicion is far below what it should be. One study found that 94% of physicians said that they either were unable to prove that the abuse had occurred or decided not to report it.¹¹ Another study found that only 1.4% of elder abuse cases reported to APS come from physicians.⁴⁴

There are several possible reasons for underreporting elder abuse, including (1) the difficulty of distinguishing elder abuse and neglect from sequelae of normal aging and (2) the fact that cognitive and functional impairment of the abused person makes it difficult, even impossible, to establish the narrative of how the abuse happened. Nursing homes in particular provide a high level of oversight because residents have an average of ≥3 functional deficits.⁴ Other reasons for underreporting—some of which are difficult to understand, and excuse, in a clinician—are:
   • subtlety of signs
   • victim denial
   • ignorance of reporting procedures
   • inadequate training
   • lack of information about resources
   • concern about losing physician–patient rapport
   • concern about involvement in the legal system
   • time limitations
   • doubt about the effectiveness of APS.¹⁶
 

Assessing capacityThe older adult’s wishes must be respected unless a health care provider or the legal system determines that he lacks functional capacity to make decisions.¹⁶

How is capacity evaluated? A capacity evaluation has 3 components:
   • Comprehension is a person’s factual understanding of the situation, including
consequences and alternatives
   • Free choice is a person’s voluntary decision to accept or reject a proposed treatment, free of coercion (in this setting, free choice is the older adult’s decision whether to report the abuse)
   • Reliability is a person’s ability to provide a consistent choice over time.⁴⁵

Most capacity evaluations are conducted by clinical interview. No single, brief test is used universally, and there is the possibility of inter-rater variability.⁴⁵ Examples of tests used to assess capacity are the Folstein Mini-Mental Status Examination and the MacArthur Competence Assessment Tool-Treatment⁴⁵; the latter is a structured interview that incorporates information specific to the individual patient’s decision-making situation.⁴⁶ Regardless of the approach, the psychiatrist-evaluator ensures that the older adult has been given the appropriate information
to provide informed consent about the situation.⁴⁷

If the evaluator determines that a person lacks capacity to make decisions, efforts should be made to determine if the cause of that impairment is reversible.⁴⁷ Older adults who have dementia or other underlying psychiatric condition that impairs cognition might benefit from more education on their situation; ones who appear fearful of consequences should be introduced to a trusted advisor to assist in making competent judgments.⁴⁷

If the older adult is found to lack capacity, a substitute decision-maker must be sought.⁴⁷ Many states have statutes specifying the order in which family members are contacted.⁴⁸ The need to appoint an advisor can become knotty because the suspected abuser often is a family member; clinicians and others involved in identifying a decision-maker to speak on behalf of an older adult should choose carefully.
Prevention and screeningKey to reducing the prevalence of elder abuse in the community is formulating
strategies for prevention and screening. The American Medical Association recommends that clinicians “incorporate routine questions related to elder abuse and neglect into daily practice.”⁴⁹ Older adults might not admit to abuse or neglect unless they are asked; speak to patients at eye level, keep questions simple, direct, and nonjudgmental, and assure them (1) that discussions are confidential and (2) that their safety is your primary goal.^50,51

Comprehensive approaches to questioning patients are available and often recommended for screening for elder abuse.⁴ However, screening in the office setting often involves short, directly administered questionnaires.⁴⁹ For example, the Health and Safety Screen developed at the University of Maine comprises 6 questions⁵²:
   • Has anyone close to you called you names or insulted you recently?
   • Are you afraid of anyone in your life?
   • Are you able to use the telephone anytime you want to?
   • Has anyone forced you to do things you didn’t want to do?
   • Has anyone taken things or money that belong to you without your OK?
   • Has anyone close to you tried to hurt you or harm you recently?

Because of time constraints and lack of a universal standard, there is debate whether regular elder abuse screening is time-effective. It often is recommended, therefore, that clinicians in primary care (1) refer older adults with risk factors for abuse to geriatric medical teams trained in these measures and (2) perform periodic follow-up on such patients⁴ (Figure).

This Web Exclusive Table provides a 7-point summary reference guide for understanding and preventing elder abuse in your practice.

BOTTOM LINEIdentification of elder abuse can be difficult because signs and symptoms of abuse closely resemble physiologic aging. Older adults with identifiable risk factors should be screened for abuse; time constraints make universal screening impossible at this time. In the future, multidisciplinary approaches likely will make elder abuse more easily identifiable through the combined work of health care providers, law enforcement agencies, banks, and other institutions—with the ultimate goal of protecting older adults in the community from abuse.

Related Resources
• Frazão SL, Correia AM, Norton P, et al. Physical abuse against elderly persons in institutional settings. J Forensic Leg Med. 2015;36:54-60.
• Sorrentino R. Performing capacity evaluations: what’s expected from your consult. Current Psychiatry. 2014;13(1):41-44.

Disclosures
Ms. Hubert reports no financial relationships with any company whose products are mentioned in this article or with manufacturers of competing products. Dr. Gupta is a member of the speakers’ bureau of Alkermes, Allergan, Avanir Pharmaceuticals, Takeda Pharmaceutical, Lundbeck, Otsuka Pharmaceutical, and Sunovion Pharmaceuticals.

In its simplest form, elder abuse refers to the intentional infliction of injury or neglect of an older adult by a caregiver. The 5 primary types of elder abuse include neglect, physical, financial, psychological/emotional, and sexual, with a subtype of social abuse that falls under psychological/emotional abuse.

Differentiating abuse from the normal sequelae of physiologic aging can be difficult; therefore, early identification and awareness of risk factors is key, as well as detailed documentation of the patient encounter. As soon as abuse is suspected, clinicians should report it to Adult Protective Services (APS) for further investigation. In terms of prevention, regular screening for elder abuse is still up for debate, but as the incidence of elder abuse continues to rise so should research and preventive efforts to combat this growing public health concern.
What is ‘elder abuse’?According to the Elder Abuse Prevention, Identification, and Treatment Act of 1985, elder abuse is:
…willful infliction of injury, unreasonable confinement, intimidation or cruel punishment with resulting physical harm or pain or mental anguish or the willful deprivation by a caretaker of goods or services which are necessary to avoid physical harm, mental anguish or mental illness.¹

There are 2 main components to this definition:
   • an older adult has suffered injury or deprivation
   • another person is responsible for causing or failing to prevent it.²

Although definitions vary, it generally is accepted that, for elder abuse to occur, it must take place within a relationship of trust.³

An ‘older adult’ is a person age ≥65, representing 14% of the U.S. population.⁴
According to U.S. Census Bureau data, there were 40 million older adults in 20105;
recent data project that this number will rise to 90 million by 2060 as Baby Boomers age.⁶ Studies suggest that as many as 10% of older adults in the United States experience elder abuse each year²; one study estimated that 6% of older adults in the community experienced significant abuse in the past month.⁷

Although elder abuse can occur in any setting, it takes place most often in the community. A survey of state APSs in 2000 showed that 60.7% of abuse was domestic; only 8.3% of incidents occurred in institutional settings.⁸ The annual direct medical costs associated with elder abuse injury in the United States are estimated at $5.3 billion, which is likely to increase with anticipated growth of the geriatric population.⁹

Although the number of older adults and the incidence of elder abuse are on the rise, as few as 1 in 14 cases is reported to authorities¹⁰; health care providers are some of the least likely of involved parties to report suspected abuse. One study found that 63% of physicians never asked about elder abuse, and only 31% reported encountering it in the previous 12 months.¹¹ A busy clinician—ie, one who sees 20 to 40 geriatric patients a day—has a high likelihood of encountering at least 1 victim of elder abuse,² yet many cases go unrecognized.
Types of abuseElder abuse comprises 5 categories¹²:
   • neglect (58.5% of cases)
   • physical (15.7%)
   • financial (12.3%)
   • psychological and emotional (7.3%)
   • sexual (0.04%).

Social abuse is considered a subtype of psychological and emotional abuse. All “other” types of abuse constitute 5.1% cases; 0.06% are of unknown type.¹²

Neglect is (1) failure of the caregiver to provide life necessities or (2) the responsible person’s refusal to permit others to provide appropriate care.³ This is one of the most common types of elder abuse in residential facilities. Signs of neglect include dehydration, depression, fecal impaction, and malnutrition (Table).⁴ The prevalence of dehydration in nursing home patients is reported to be as high as 35%, which may be the result of passive or active withholding of liquids (ie, decreasing hydration to reduce the need to change the resident’s clothing or bedding).⁴ Other forms of neglect include medication misuse (overdosing or underdosing) and self-neglect, which occurs among people living alone and often is listed as a subtype of neglect, but is controversial because it does not involve another person.
 

Financial exploitation is the illegal or improper use, or mismanagement, of a person’s money, property, or financial resources³—often, to his (her) detriment. Estimates are that 1 of every 20 older adults has been subject to financial abuse at some point in their life.¹⁵ There should be a high index of suspicion for financial exploitation when one notices unexplained changes in power of attorney, wills, or other legal documents; missing checks, money, or belongings.¹⁶ In the past, adult children were most likely to be financial abusers; in recent years, however, the abuser is more often a spouse—especially a second spouse.¹⁷ Bankers, accountants, and other financial advisors are among those trained to identify risk factors for financial abuse; they are encouraged to caution clients about this possibility.¹⁸

Psychological and emotional abuse occurs when a caregiver inflicts mental stress on an older adult by actions and threats that cause fear, violence, isolation, deprivation, or feelings of shame and powerlessness.³ Examples are threatening to put the older adult in a nursing home or verbally abusing him (her). Suspect this type of abuse when a caregiver refuses to leave the older adult or speaks for him, or if the older adult expresses fear in the presence of the caregiver.⁴ This type of abuse also is prevalent in nursing homes and other long-term care facilities.¹⁹

Sexual abuse involves nonconsensual touching or sexual activities (rape, language, exploitive behavior) that are threatened or forced on an older adult.¹⁶ Sexual abuse is more common in frail or dependent persons.³ Physical exam findings—particularly dysuria, tender genitalia, and evidence of sexually transmitted infections⁴—are required to identify sexual abuse, along with signs of depression and display of fear.

Social abuse can be considered a subtype of psychological and emotional abuse, in which a caregiver denies an older adult contact with family and friends or deprives him from access to transportation. Other examples include not allowing the older adult to use the telephone, monitoring phone calls, and claiming that his friends or family are “interfering.”²⁰ Intentionally embarrassing an older adult in front of others also can be considered social abuse.

Technology, particularly smart phones and social media, can complicate and exacerbate elder abuse:
   • In July 2013, employees of a Wisconsin nursing home were found with videos and photographs of residents bathing and of a nude resident who had a bowel obstruction being mocked.²¹
   • In May 2014, employees of a nursing home in Massachusetts recorded themselves physically and verbally abusing several older adults with Alzheimer’s disease, including one episode of the employees “hitting the woman on her arms, flicking her ears and then pinching the woman’s nose closed.” The employees also possessed a photograph of her naked.²²
   • In June 2015, an employee of a nursing home in Indiana was accused of taking
photos of a resident naked and sharing them on the messaging application Snapchat,²³ in which images disappear 10 seconds after they are viewed.

As technology evolves, caregivers are finding more cunning ways to abuse older adults. Considering current events and trends in this area, technology as a gateway to elder abuse should be of growing concern.
Risk factorsA 2013 literature review on elder abuse reported that the most important risk factors are related to relationship (family disharmony, poor or conflicting relationships) and environment (a low level of social support),³ although other variables can play a role. Regardless of these findings, it is important to recognize that (1) elder abuse is not a necessary consequence in a family with many risk factors and (2) elder abuse can occur in the absence of any risk factors.

As a whole, women are at a higher risk of abuse, particularly when combined with loneliness, poor social support, cohabitation (especially family members), substance abuse, cognitive impairment, and dementia and other mental health problems.⁴ Other risk factors include functional deficiency, poor physical health or frailty, low income or wealth, and trauma or past abuse.³

Lower income, poor health, low social support, and belonging to a non-white racial group put an older adult at risk for neglect; female sex is a specific risk factor for sexual assault.¹⁵ One study found that, among older adults who suffered physical, mental, or cognitive impairment, 1 of every 4 was at risk of abuse.⁷

Mental illness. Dementia puts an older adult at higher risk because of increased
caregiver stress resulting from disruptive and aggressive behaviors²; the same is true when the older adult suffers another mental illness, such as anxiety, depression, schizophrenia or bipolar disorder. Presumably, older adults with any of these disorders are at risk of financial and psychological and emotional abuse because of their decreased social support, lack of independence, and inability to hold a job—leaving their caregiver to shoulder more responsibilities and with more opportunities to inflict abuse. In addition, an older adult suffering from depression can feel helpless and unworthy, possibly making him more susceptible to psychological and emotional abuse, and less likely to seek help.

More research is needed to establish racial and ethnic differences in the risk of abuse. Some research states that older adults who are a member of a minority are at greater risk of abuse; however, the difference dissipates after adjusting for variables such as income and social support.²⁴ Cultural confounders, such as varying interpretations of the same set of interactions between older adults, need to be examined further.

Sexual orientation. Identifying one’s self as a lesbian, gay, bisexual, or transgender (LGBT) person is an additional risk factor for elder abuse. In 1997, a report described a nursing home employee who refused to bathe a resident because he didn’t want to “touch the lesbian.”²⁵ Despite evolving attitudes in society toward support and acceptance of sexual orientation, fear of homophobia still prevents some LGBT older adults from seeking help when they have been abused because of their orientation—especially ones who have internalized that
homophobia and feel that they are unworthy of seeking help.²⁵

In addition, health care providers and nursing home staff members might neglect the particular care needs of LGBT older adults, intentionally or unintentionally. APS staff and providers must be cognizant of underlying biases and exhibit respect when assisting LGBT clients.

Approximately 75% of caregivers of older adults are family members; 70% are female²⁶; and most are adult children, spouses, and partners of those receiving care.²⁷ Male caregivers age ≥40 are more likely to be the abuser, however, especially when they possess any of these risk factors: fatigue, burnout, medical illness, mental illness, lack of financial and support services, family history of abusive behavior, and substance abuse.4 People who commit elder abuse also tend to be significantly dependent on the person they are abusing.² In some cases, and especially when the abuser is financially needy, caregivers turn to elder abuse to obtain resources from the victim.²

From your standpoint as a practitioner, it is important to determine the root cause of elder abuse. According to one review,²⁸ family members with mental illness or a history of substance abuse, or who are stressed by the burden of caregiving, abuse older adults at a higher rate than family members who are not affected in those ways. Depression in particular is a common characteristic of abusers,² often secondary to the stress of caring for an older adult.

Abuse caused by stress can be addressed by referral to a support group and counseling for the caregiver; psychiatric conditions, such as depression, might be better treated with pharmacotherapy. Evaluate for depression and posttraumatic stress disorder (PTSD) in both the abuser and the abused,²⁹ and for other mental health issues that might compound the situation. It is possible for you to have 2 patients: the older adult and his caregiver. Regardless of the challenge,keep in mind that the older adult’s safety is your priority.
Consequences for the abusedThe abused adult is at risk of a number of serious physical and psychological consequences.³⁰ They tend to have a shorter lifespan, after adjusting for other variables associated with increased mortality.

The reason for shortened lifespan is multifactorial³⁰:
   • Bruises, abrasions, and fractures may take longer to heal because of diminished skin and bone regeneration.
   • Diseases that affect the heart, lungs, and kidneys might prevent the person from bouncing back from major stressors caused by abuse, such as blood loss, severe injury, and pain.
   • Injury from abuse can exacerbate an underlying illness.
   • Elder abuse also is associated with increased emergency department use, hospitalization (including readmission within 30 days), and nursing home placement.³¹

Elder abuse can lead to depression, shame, and guilt; increased isolation; and
increased risk of alcohol abuse and substance use.³¹ A study found that victims of
elder abuse are significantly more depressed than non-victims.³²

In the same study, being a victim of abuse was found to be the second-strongest
predictor of depression, after the state of one’s health.³² Other potential psychiatric
consequences of abuse that need further study include increased risk of developing
fear and anxiety disorders; learned helplessness; and PTSD.³³ According to LoFaso,
“depression and anxiety can consume their days and leave them emotionally and
physically frail.”29 Such feelings make these older adults less likely to resolve abuse or break off relations with the abuser.³²

Because mental illness can be a risk factor for, and a consequence of, elder abuse,
be aware of such complications and address them appropriately. Keep in mind that older adults are more likely to visit a primary care practitioner than a psychiatrist for a routine health check-up or evaluation of initial cognition-related problems; however, they are more likely to see a psychiatrist for advanced neuropsychiatric problems such as dementia, paranoia, delusions, hallucinations, and insomnia. Adequate education on elder abuse should not be limited to a single medical specialty because it can present in several clinical settings.
Identifying abuseIdentification of elder abuse in the home poses a greater challenge to clinicians than abuse in an institutional setting because it is not directly observable. Compounding this is the lack of unified standards for identifying and dealing with elder abuse. It is first necessary for you to determine the likelihood that abuse or neglect occurred, which can be difficult because the signs of elder abuse and manifestations of normal aging often are similar. You also must establish whether (1) the abused person will accept intervention and (2) the abused person who refuses intervention has the capacity to make that decision. Both of these conditions will guide your approach to management.²

Obtain the history from several sources; review the records; and carefully examine patterns of injury, in particular assessing functional status and level of dependency on the caregiver. Explanations that do not match injuries signal the need for further investigation and examination.

To help differentiate elder abuse from normal physiologic aging, look at the skin for bruises, rashes that do not heal, and ulcers—all of which could be signs of abuse or neglect. Keep in mind that bruising generally is more common in older adults because of the slower turnover rate of epidermal cells; physiologic bruising tends to occur on dorsal aspects of the hands and arms.⁴ In contrast, bruising secondary to neglect or physical abuse can manifest as a subgaleal hematoma (caused by traumatic hair pulling), tracking in the peritoneum after genital trauma, Battle’s sign, and raccoon eyes, among other findings.⁴

In addition, larger bruises (>5 cm in diameter) are more likely the result of elder abuse.³⁴ To complicate matters, many older persons are taking anticoagulant therapy, making bruising more likely. In addition, be on the lookout for burns during the physical exam. Evidence suggests that at least 10% of burns caused by battery and assault occur in the context of elder abuse; most burn facilities do not have formal guidelines for screening for abuse and neglect, however.³⁵ According to one retrospective study, the most common causes of burns in older adults are hot water scalds and radiator contact, and the mortality rate of older burn patients in general is higher than among the overall population.³⁶

Falls and fractures are common among older adults, regardless of whether they are
abused, because of polypharmacy, underlying medical conditions, and functional
limitations. Many abusers, however, use these factors to cover up intentional injury
that might have resulted in the older person falling, including overmedication (a form of physical abuse) and withholding a necessary walking aid (a form of neglect). Maintain a high index of suspicion of elder abuse when (1) the caregiver’s and the older adult’s stories of an injury don’t add up and (2) physical findings that might have been caused by abuse are present.

A number of psychiatric and cognitive symptoms suggest other types of elder abuse. Take note of emotional upset, agitation, and unusual behaviors³⁷—especially if you can follow the patient over time to observe marked changes in the presentation. Likewise, be aware of proposed alterations in guardianship, which should be evaluated by a forensic psychiatrist with analysis of medical history, social attachments, home environment, self-care, and finances.³⁸ Such evaluation should provide clues to the motivation behind a change of guardianship and will help to determine if elder abuse should be suspected.

Brandi et al³⁷ provided an informative table that identifies pertinent signs, symptoms, and other findings that clinicians should be aware of to support a suspicion of elder abuse (Table).

Documentation is of utmost importance in evaluating potential elder abuse; keep in mind that the medical record might be used in an investigation of abuse by social workers, law enforcement, and prosecutors. Your records should be legible, clearly indicate who the main caregiver is and what his (her) responsibilities are, and specify who is present at your encounter with the patient.⁴ Document your observations of patient behavior, reactions to questions, and family dynamics and conflicts¹⁶; make note of warning signs such as fear, silence, and inability to interview the patient alone.

In addition to written documentation, take photographs of injuries, with a ruler in the image to record their size. Serial photographs are helpful; so are photographs from a variety of distances (close-up, regional, wholebody) to capture detail and place the wound in the context of a specific area of the body.⁴

Safety is paramount. Given the findings of the history and physical exam, it is necessary to determine whether it is safe for the patient to return home with the caregiver, or if alternate accommodations or resources, such as a social worker or a support group, are required. Include details of planned follow-up in your evaluation, and offer consideration of possible psychiatric disorders that can develop as a result of such abuse.
ReportingElder abuse is a criminal offense in all states.³⁹ A clinician who has reasonable suspicion that elder abuse occurred must report it, regardless of whether the proof of abuse is concrete.⁴⁰ At a point of reasonable suspicion, immediately contact APS, law enforcement, and a social worker. Adult Protective Services, modeled after Child Protective Services, is typically administered by local and state health
departments.⁴¹

After a report is filed with APS, an assigned social worker makes an in-person home visit to investigate the allegation and determine whether elder abuse is substantiated, partially substantiated, or unsubstantiated.¹⁶ In most states, elder abuse reporting is not anonymous because follow-up may be needed to provide additional evidence, especially if the report was made by a health care provider.¹⁶

No federal standard exists for states to follow when defining and addressing elder abuse, which can complicate identification and reporting of abuse. Laws governing elder abuse do not allow states to determine the fate of the older adult, who can decide for himself (herself) whether to use or waive protective services.⁴² Older adults might choose not to report abuse because of shame, intimidation, or fear,⁴³ or to protect a caregiver, who often is a family member.

Elder abuse reports can come from a variety of sources; convincing evidence is, as noted, unnecessary to report it. Health care providers are mandated reporters, but
it is believed that the number of clinicians who report elder abuse based on suspicion is far below what it should be. One study found that 94% of physicians said that they either were unable to prove that the abuse had occurred or decided not to report it.¹¹ Another study found that only 1.4% of elder abuse cases reported to APS come from physicians.⁴⁴

There are several possible reasons for underreporting elder abuse, including (1) the difficulty of distinguishing elder abuse and neglect from sequelae of normal aging and (2) the fact that cognitive and functional impairment of the abused person makes it difficult, even impossible, to establish the narrative of how the abuse happened. Nursing homes in particular provide a high level of oversight because residents have an average of ≥3 functional deficits.⁴ Other reasons for underreporting—some of which are difficult to understand, and excuse, in a clinician—are:
   • subtlety of signs
   • victim denial
   • ignorance of reporting procedures
   • inadequate training
   • lack of information about resources
   • concern about losing physician–patient rapport
   • concern about involvement in the legal system
   • time limitations
   • doubt about the effectiveness of APS.¹⁶
 

Assessing capacityThe older adult’s wishes must be respected unless a health care provider or the legal system determines that he lacks functional capacity to make decisions.¹⁶

How is capacity evaluated? A capacity evaluation has 3 components:
   • Comprehension is a person’s factual understanding of the situation, including
consequences and alternatives
   • Free choice is a person’s voluntary decision to accept or reject a proposed treatment, free of coercion (in this setting, free choice is the older adult’s decision whether to report the abuse)
   • Reliability is a person’s ability to provide a consistent choice over time.⁴⁵

Most capacity evaluations are conducted by clinical interview. No single, brief test is used universally, and there is the possibility of inter-rater variability.⁴⁵ Examples of tests used to assess capacity are the Folstein Mini-Mental Status Examination and the MacArthur Competence Assessment Tool-Treatment⁴⁵; the latter is a structured interview that incorporates information specific to the individual patient’s decision-making situation.⁴⁶ Regardless of the approach, the psychiatrist-evaluator ensures that the older adult has been given the appropriate information
to provide informed consent about the situation.⁴⁷

If the evaluator determines that a person lacks capacity to make decisions, efforts should be made to determine if the cause of that impairment is reversible.⁴⁷ Older adults who have dementia or other underlying psychiatric condition that impairs cognition might benefit from more education on their situation; ones who appear fearful of consequences should be introduced to a trusted advisor to assist in making competent judgments.⁴⁷

If the older adult is found to lack capacity, a substitute decision-maker must be sought.⁴⁷ Many states have statutes specifying the order in which family members are contacted.⁴⁸ The need to appoint an advisor can become knotty because the suspected abuser often is a family member; clinicians and others involved in identifying a decision-maker to speak on behalf of an older adult should choose carefully.
Prevention and screeningKey to reducing the prevalence of elder abuse in the community is formulating
strategies for prevention and screening. The American Medical Association recommends that clinicians “incorporate routine questions related to elder abuse and neglect into daily practice.”⁴⁹ Older adults might not admit to abuse or neglect unless they are asked; speak to patients at eye level, keep questions simple, direct, and nonjudgmental, and assure them (1) that discussions are confidential and (2) that their safety is your primary goal.^50,51

Comprehensive approaches to questioning patients are available and often recommended for screening for elder abuse.⁴ However, screening in the office setting often involves short, directly administered questionnaires.⁴⁹ For example, the Health and Safety Screen developed at the University of Maine comprises 6 questions⁵²:
   • Has anyone close to you called you names or insulted you recently?
   • Are you afraid of anyone in your life?
   • Are you able to use the telephone anytime you want to?
   • Has anyone forced you to do things you didn’t want to do?
   • Has anyone taken things or money that belong to you without your OK?
   • Has anyone close to you tried to hurt you or harm you recently?

Because of time constraints and lack of a universal standard, there is debate whether regular elder abuse screening is time-effective. It often is recommended, therefore, that clinicians in primary care (1) refer older adults with risk factors for abuse to geriatric medical teams trained in these measures and (2) perform periodic follow-up on such patients⁴ (Figure).

This Web Exclusive Table provides a 7-point summary reference guide for understanding and preventing elder abuse in your practice.

BOTTOM LINEIdentification of elder abuse can be difficult because signs and symptoms of abuse closely resemble physiologic aging. Older adults with identifiable risk factors should be screened for abuse; time constraints make universal screening impossible at this time. In the future, multidisciplinary approaches likely will make elder abuse more easily identifiable through the combined work of health care providers, law enforcement agencies, banks, and other institutions—with the ultimate goal of protecting older adults in the community from abuse.

Related Resources
• Frazão SL, Correia AM, Norton P, et al. Physical abuse against elderly persons in institutional settings. J Forensic Leg Med. 2015;36:54-60.
• Sorrentino R. Performing capacity evaluations: what’s expected from your consult. Current Psychiatry. 2014;13(1):41-44.

Disclosures
Ms. Hubert reports no financial relationships with any company whose products are mentioned in this article or with manufacturers of competing products. Dr. Gupta is a member of the speakers’ bureau of Alkermes, Allergan, Avanir Pharmaceuticals, Takeda Pharmaceutical, Lundbeck, Otsuka Pharmaceutical, and Sunovion Pharmaceuticals.

In its simplest form, elder abuse refers to the intentional infliction of injury or neglect of an older adult by a caregiver. The 5 primary types of elder abuse include neglect, physical, financial, psychological/emotional, and sexual, with a subtype of social abuse that falls under psychological/emotional abuse.

Differentiating abuse from the normal sequelae of physiologic aging can be difficult; therefore, early identification and awareness of risk factors is key, as well as detailed documentation of the patient encounter. As soon as abuse is suspected, clinicians should report it to Adult Protective Services (APS) for further investigation. In terms of prevention, regular screening for elder abuse is still up for debate, but as the incidence of elder abuse continues to rise so should research and preventive efforts to combat this growing public health concern.
What is ‘elder abuse’?According to the Elder Abuse Prevention, Identification, and Treatment Act of 1985, elder abuse is:
…willful infliction of injury, unreasonable confinement, intimidation or cruel punishment with resulting physical harm or pain or mental anguish or the willful deprivation by a caretaker of goods or services which are necessary to avoid physical harm, mental anguish or mental illness.¹

There are 2 main components to this definition:
   • an older adult has suffered injury or deprivation
   • another person is responsible for causing or failing to prevent it.²

Although definitions vary, it generally is accepted that, for elder abuse to occur, it must take place within a relationship of trust.³

An ‘older adult’ is a person age ≥65, representing 14% of the U.S. population.⁴
According to U.S. Census Bureau data, there were 40 million older adults in 20105;
recent data project that this number will rise to 90 million by 2060 as Baby Boomers age.⁶ Studies suggest that as many as 10% of older adults in the United States experience elder abuse each year²; one study estimated that 6% of older adults in the community experienced significant abuse in the past month.⁷

Although elder abuse can occur in any setting, it takes place most often in the community. A survey of state APSs in 2000 showed that 60.7% of abuse was domestic; only 8.3% of incidents occurred in institutional settings.⁸ The annual direct medical costs associated with elder abuse injury in the United States are estimated at $5.3 billion, which is likely to increase with anticipated growth of the geriatric population.⁹

Although the number of older adults and the incidence of elder abuse are on the rise, as few as 1 in 14 cases is reported to authorities¹⁰; health care providers are some of the least likely of involved parties to report suspected abuse. One study found that 63% of physicians never asked about elder abuse, and only 31% reported encountering it in the previous 12 months.¹¹ A busy clinician—ie, one who sees 20 to 40 geriatric patients a day—has a high likelihood of encountering at least 1 victim of elder abuse,² yet many cases go unrecognized.
Types of abuseElder abuse comprises 5 categories¹²:
   • neglect (58.5% of cases)
   • physical (15.7%)
   • financial (12.3%)
   • psychological and emotional (7.3%)
   • sexual (0.04%).

Social abuse is considered a subtype of psychological and emotional abuse. All “other” types of abuse constitute 5.1% cases; 0.06% are of unknown type.¹²

Neglect is (1) failure of the caregiver to provide life necessities or (2) the responsible person’s refusal to permit others to provide appropriate care.³ This is one of the most common types of elder abuse in residential facilities. Signs of neglect include dehydration, depression, fecal impaction, and malnutrition (Table).⁴ The prevalence of dehydration in nursing home patients is reported to be as high as 35%, which may be the result of passive or active withholding of liquids (ie, decreasing hydration to reduce the need to change the resident’s clothing or bedding).⁴ Other forms of neglect include medication misuse (overdosing or underdosing) and self-neglect, which occurs among people living alone and often is listed as a subtype of neglect, but is controversial because it does not involve another person.
 

Financial exploitation is the illegal or improper use, or mismanagement, of a person’s money, property, or financial resources³—often, to his (her) detriment. Estimates are that 1 of every 20 older adults has been subject to financial abuse at some point in their life.¹⁵ There should be a high index of suspicion for financial exploitation when one notices unexplained changes in power of attorney, wills, or other legal documents; missing checks, money, or belongings.¹⁶ In the past, adult children were most likely to be financial abusers; in recent years, however, the abuser is more often a spouse—especially a second spouse.¹⁷ Bankers, accountants, and other financial advisors are among those trained to identify risk factors for financial abuse; they are encouraged to caution clients about this possibility.¹⁸

Psychological and emotional abuse occurs when a caregiver inflicts mental stress on an older adult by actions and threats that cause fear, violence, isolation, deprivation, or feelings of shame and powerlessness.³ Examples are threatening to put the older adult in a nursing home or verbally abusing him (her). Suspect this type of abuse when a caregiver refuses to leave the older adult or speaks for him, or if the older adult expresses fear in the presence of the caregiver.⁴ This type of abuse also is prevalent in nursing homes and other long-term care facilities.¹⁹

Sexual abuse involves nonconsensual touching or sexual activities (rape, language, exploitive behavior) that are threatened or forced on an older adult.¹⁶ Sexual abuse is more common in frail or dependent persons.³ Physical exam findings—particularly dysuria, tender genitalia, and evidence of sexually transmitted infections⁴—are required to identify sexual abuse, along with signs of depression and display of fear.

Social abuse can be considered a subtype of psychological and emotional abuse, in which a caregiver denies an older adult contact with family and friends or deprives him from access to transportation. Other examples include not allowing the older adult to use the telephone, monitoring phone calls, and claiming that his friends or family are “interfering.”²⁰ Intentionally embarrassing an older adult in front of others also can be considered social abuse.

Technology, particularly smart phones and social media, can complicate and exacerbate elder abuse:
   • In July 2013, employees of a Wisconsin nursing home were found with videos and photographs of residents bathing and of a nude resident who had a bowel obstruction being mocked.²¹
   • In May 2014, employees of a nursing home in Massachusetts recorded themselves physically and verbally abusing several older adults with Alzheimer’s disease, including one episode of the employees “hitting the woman on her arms, flicking her ears and then pinching the woman’s nose closed.” The employees also possessed a photograph of her naked.²²
   • In June 2015, an employee of a nursing home in Indiana was accused of taking
photos of a resident naked and sharing them on the messaging application Snapchat,²³ in which images disappear 10 seconds after they are viewed.

As technology evolves, caregivers are finding more cunning ways to abuse older adults. Considering current events and trends in this area, technology as a gateway to elder abuse should be of growing concern.
Risk factorsA 2013 literature review on elder abuse reported that the most important risk factors are related to relationship (family disharmony, poor or conflicting relationships) and environment (a low level of social support),³ although other variables can play a role. Regardless of these findings, it is important to recognize that (1) elder abuse is not a necessary consequence in a family with many risk factors and (2) elder abuse can occur in the absence of any risk factors.

As a whole, women are at a higher risk of abuse, particularly when combined with loneliness, poor social support, cohabitation (especially family members), substance abuse, cognitive impairment, and dementia and other mental health problems.⁴ Other risk factors include functional deficiency, poor physical health or frailty, low income or wealth, and trauma or past abuse.³

Lower income, poor health, low social support, and belonging to a non-white racial group put an older adult at risk for neglect; female sex is a specific risk factor for sexual assault.¹⁵ One study found that, among older adults who suffered physical, mental, or cognitive impairment, 1 of every 4 was at risk of abuse.⁷

Mental illness. Dementia puts an older adult at higher risk because of increased
caregiver stress resulting from disruptive and aggressive behaviors²; the same is true when the older adult suffers another mental illness, such as anxiety, depression, schizophrenia or bipolar disorder. Presumably, older adults with any of these disorders are at risk of financial and psychological and emotional abuse because of their decreased social support, lack of independence, and inability to hold a job—leaving their caregiver to shoulder more responsibilities and with more opportunities to inflict abuse. In addition, an older adult suffering from depression can feel helpless and unworthy, possibly making him more susceptible to psychological and emotional abuse, and less likely to seek help.

More research is needed to establish racial and ethnic differences in the risk of abuse. Some research states that older adults who are a member of a minority are at greater risk of abuse; however, the difference dissipates after adjusting for variables such as income and social support.²⁴ Cultural confounders, such as varying interpretations of the same set of interactions between older adults, need to be examined further.

Sexual orientation. Identifying one’s self as a lesbian, gay, bisexual, or transgender (LGBT) person is an additional risk factor for elder abuse. In 1997, a report described a nursing home employee who refused to bathe a resident because he didn’t want to “touch the lesbian.”²⁵ Despite evolving attitudes in society toward support and acceptance of sexual orientation, fear of homophobia still prevents some LGBT older adults from seeking help when they have been abused because of their orientation—especially ones who have internalized that
homophobia and feel that they are unworthy of seeking help.²⁵

In addition, health care providers and nursing home staff members might neglect the particular care needs of LGBT older adults, intentionally or unintentionally. APS staff and providers must be cognizant of underlying biases and exhibit respect when assisting LGBT clients.

Approximately 75% of caregivers of older adults are family members; 70% are female²⁶; and most are adult children, spouses, and partners of those receiving care.²⁷ Male caregivers age ≥40 are more likely to be the abuser, however, especially when they possess any of these risk factors: fatigue, burnout, medical illness, mental illness, lack of financial and support services, family history of abusive behavior, and substance abuse.4 People who commit elder abuse also tend to be significantly dependent on the person they are abusing.² In some cases, and especially when the abuser is financially needy, caregivers turn to elder abuse to obtain resources from the victim.²

From your standpoint as a practitioner, it is important to determine the root cause of elder abuse. According to one review,²⁸ family members with mental illness or a history of substance abuse, or who are stressed by the burden of caregiving, abuse older adults at a higher rate than family members who are not affected in those ways. Depression in particular is a common characteristic of abusers,² often secondary to the stress of caring for an older adult.

Abuse caused by stress can be addressed by referral to a support group and counseling for the caregiver; psychiatric conditions, such as depression, might be better treated with pharmacotherapy. Evaluate for depression and posttraumatic stress disorder (PTSD) in both the abuser and the abused,²⁹ and for other mental health issues that might compound the situation. It is possible for you to have 2 patients: the older adult and his caregiver. Regardless of the challenge,keep in mind that the older adult’s safety is your priority.
Consequences for the abusedThe abused adult is at risk of a number of serious physical and psychological consequences.³⁰ They tend to have a shorter lifespan, after adjusting for other variables associated with increased mortality.

The reason for shortened lifespan is multifactorial³⁰:
   • Bruises, abrasions, and fractures may take longer to heal because of diminished skin and bone regeneration.
   • Diseases that affect the heart, lungs, and kidneys might prevent the person from bouncing back from major stressors caused by abuse, such as blood loss, severe injury, and pain.
   • Injury from abuse can exacerbate an underlying illness.
   • Elder abuse also is associated with increased emergency department use, hospitalization (including readmission within 30 days), and nursing home placement.³¹

Elder abuse can lead to depression, shame, and guilt; increased isolation; and
increased risk of alcohol abuse and substance use.³¹ A study found that victims of
elder abuse are significantly more depressed than non-victims.³²

In the same study, being a victim of abuse was found to be the second-strongest
predictor of depression, after the state of one’s health.³² Other potential psychiatric
consequences of abuse that need further study include increased risk of developing
fear and anxiety disorders; learned helplessness; and PTSD.³³ According to LoFaso,
“depression and anxiety can consume their days and leave them emotionally and
physically frail.”29 Such feelings make these older adults less likely to resolve abuse or break off relations with the abuser.³²

Because mental illness can be a risk factor for, and a consequence of, elder abuse,
be aware of such complications and address them appropriately. Keep in mind that older adults are more likely to visit a primary care practitioner than a psychiatrist for a routine health check-up or evaluation of initial cognition-related problems; however, they are more likely to see a psychiatrist for advanced neuropsychiatric problems such as dementia, paranoia, delusions, hallucinations, and insomnia. Adequate education on elder abuse should not be limited to a single medical specialty because it can present in several clinical settings.
Identifying abuseIdentification of elder abuse in the home poses a greater challenge to clinicians than abuse in an institutional setting because it is not directly observable. Compounding this is the lack of unified standards for identifying and dealing with elder abuse. It is first necessary for you to determine the likelihood that abuse or neglect occurred, which can be difficult because the signs of elder abuse and manifestations of normal aging often are similar. You also must establish whether (1) the abused person will accept intervention and (2) the abused person who refuses intervention has the capacity to make that decision. Both of these conditions will guide your approach to management.²

Obtain the history from several sources; review the records; and carefully examine patterns of injury, in particular assessing functional status and level of dependency on the caregiver. Explanations that do not match injuries signal the need for further investigation and examination.

To help differentiate elder abuse from normal physiologic aging, look at the skin for bruises, rashes that do not heal, and ulcers—all of which could be signs of abuse or neglect. Keep in mind that bruising generally is more common in older adults because of the slower turnover rate of epidermal cells; physiologic bruising tends to occur on dorsal aspects of the hands and arms.⁴ In contrast, bruising secondary to neglect or physical abuse can manifest as a subgaleal hematoma (caused by traumatic hair pulling), tracking in the peritoneum after genital trauma, Battle’s sign, and raccoon eyes, among other findings.⁴

In addition, larger bruises (>5 cm in diameter) are more likely the result of elder abuse.³⁴ To complicate matters, many older persons are taking anticoagulant therapy, making bruising more likely. In addition, be on the lookout for burns during the physical exam. Evidence suggests that at least 10% of burns caused by battery and assault occur in the context of elder abuse; most burn facilities do not have formal guidelines for screening for abuse and neglect, however.³⁵ According to one retrospective study, the most common causes of burns in older adults are hot water scalds and radiator contact, and the mortality rate of older burn patients in general is higher than among the overall population.³⁶

Falls and fractures are common among older adults, regardless of whether they are
abused, because of polypharmacy, underlying medical conditions, and functional
limitations. Many abusers, however, use these factors to cover up intentional injury
that might have resulted in the older person falling, including overmedication (a form of physical abuse) and withholding a necessary walking aid (a form of neglect). Maintain a high index of suspicion of elder abuse when (1) the caregiver’s and the older adult’s stories of an injury don’t add up and (2) physical findings that might have been caused by abuse are present.

A number of psychiatric and cognitive symptoms suggest other types of elder abuse. Take note of emotional upset, agitation, and unusual behaviors³⁷—especially if you can follow the patient over time to observe marked changes in the presentation. Likewise, be aware of proposed alterations in guardianship, which should be evaluated by a forensic psychiatrist with analysis of medical history, social attachments, home environment, self-care, and finances.³⁸ Such evaluation should provide clues to the motivation behind a change of guardianship and will help to determine if elder abuse should be suspected.

Brandi et al³⁷ provided an informative table that identifies pertinent signs, symptoms, and other findings that clinicians should be aware of to support a suspicion of elder abuse (Table).

Documentation is of utmost importance in evaluating potential elder abuse; keep in mind that the medical record might be used in an investigation of abuse by social workers, law enforcement, and prosecutors. Your records should be legible, clearly indicate who the main caregiver is and what his (her) responsibilities are, and specify who is present at your encounter with the patient.⁴ Document your observations of patient behavior, reactions to questions, and family dynamics and conflicts¹⁶; make note of warning signs such as fear, silence, and inability to interview the patient alone.

In addition to written documentation, take photographs of injuries, with a ruler in the image to record their size. Serial photographs are helpful; so are photographs from a variety of distances (close-up, regional, wholebody) to capture detail and place the wound in the context of a specific area of the body.⁴

Safety is paramount. Given the findings of the history and physical exam, it is necessary to determine whether it is safe for the patient to return home with the caregiver, or if alternate accommodations or resources, such as a social worker or a support group, are required. Include details of planned follow-up in your evaluation, and offer consideration of possible psychiatric disorders that can develop as a result of such abuse.
ReportingElder abuse is a criminal offense in all states.³⁹ A clinician who has reasonable suspicion that elder abuse occurred must report it, regardless of whether the proof of abuse is concrete.⁴⁰ At a point of reasonable suspicion, immediately contact APS, law enforcement, and a social worker. Adult Protective Services, modeled after Child Protective Services, is typically administered by local and state health
departments.⁴¹

After a report is filed with APS, an assigned social worker makes an in-person home visit to investigate the allegation and determine whether elder abuse is substantiated, partially substantiated, or unsubstantiated.¹⁶ In most states, elder abuse reporting is not anonymous because follow-up may be needed to provide additional evidence, especially if the report was made by a health care provider.¹⁶

No federal standard exists for states to follow when defining and addressing elder abuse, which can complicate identification and reporting of abuse. Laws governing elder abuse do not allow states to determine the fate of the older adult, who can decide for himself (herself) whether to use or waive protective services.⁴² Older adults might choose not to report abuse because of shame, intimidation, or fear,⁴³ or to protect a caregiver, who often is a family member.

Elder abuse reports can come from a variety of sources; convincing evidence is, as noted, unnecessary to report it. Health care providers are mandated reporters, but
it is believed that the number of clinicians who report elder abuse based on suspicion is far below what it should be. One study found that 94% of physicians said that they either were unable to prove that the abuse had occurred or decided not to report it.¹¹ Another study found that only 1.4% of elder abuse cases reported to APS come from physicians.⁴⁴

There are several possible reasons for underreporting elder abuse, including (1) the difficulty of distinguishing elder abuse and neglect from sequelae of normal aging and (2) the fact that cognitive and functional impairment of the abused person makes it difficult, even impossible, to establish the narrative of how the abuse happened. Nursing homes in particular provide a high level of oversight because residents have an average of ≥3 functional deficits.⁴ Other reasons for underreporting—some of which are difficult to understand, and excuse, in a clinician—are:
   • subtlety of signs
   • victim denial
   • ignorance of reporting procedures
   • inadequate training
   • lack of information about resources
   • concern about losing physician–patient rapport
   • concern about involvement in the legal system
   • time limitations
   • doubt about the effectiveness of APS.¹⁶
 

Assessing capacityThe older adult’s wishes must be respected unless a health care provider or the legal system determines that he lacks functional capacity to make decisions.¹⁶

How is capacity evaluated? A capacity evaluation has 3 components:
   • Comprehension is a person’s factual understanding of the situation, including
consequences and alternatives
   • Free choice is a person’s voluntary decision to accept or reject a proposed treatment, free of coercion (in this setting, free choice is the older adult’s decision whether to report the abuse)
   • Reliability is a person’s ability to provide a consistent choice over time.⁴⁵

Most capacity evaluations are conducted by clinical interview. No single, brief test is used universally, and there is the possibility of inter-rater variability.⁴⁵ Examples of tests used to assess capacity are the Folstein Mini-Mental Status Examination and the MacArthur Competence Assessment Tool-Treatment⁴⁵; the latter is a structured interview that incorporates information specific to the individual patient’s decision-making situation.⁴⁶ Regardless of the approach, the psychiatrist-evaluator ensures that the older adult has been given the appropriate information
to provide informed consent about the situation.⁴⁷

If the evaluator determines that a person lacks capacity to make decisions, efforts should be made to determine if the cause of that impairment is reversible.⁴⁷ Older adults who have dementia or other underlying psychiatric condition that impairs cognition might benefit from more education on their situation; ones who appear fearful of consequences should be introduced to a trusted advisor to assist in making competent judgments.⁴⁷

If the older adult is found to lack capacity, a substitute decision-maker must be sought.⁴⁷ Many states have statutes specifying the order in which family members are contacted.⁴⁸ The need to appoint an advisor can become knotty because the suspected abuser often is a family member; clinicians and others involved in identifying a decision-maker to speak on behalf of an older adult should choose carefully.
Prevention and screeningKey to reducing the prevalence of elder abuse in the community is formulating
strategies for prevention and screening. The American Medical Association recommends that clinicians “incorporate routine questions related to elder abuse and neglect into daily practice.”⁴⁹ Older adults might not admit to abuse or neglect unless they are asked; speak to patients at eye level, keep questions simple, direct, and nonjudgmental, and assure them (1) that discussions are confidential and (2) that their safety is your primary goal.^50,51

Comprehensive approaches to questioning patients are available and often recommended for screening for elder abuse.⁴ However, screening in the office setting often involves short, directly administered questionnaires.⁴⁹ For example, the Health and Safety Screen developed at the University of Maine comprises 6 questions⁵²:
   • Has anyone close to you called you names or insulted you recently?
   • Are you afraid of anyone in your life?
   • Are you able to use the telephone anytime you want to?
   • Has anyone forced you to do things you didn’t want to do?
   • Has anyone taken things or money that belong to you without your OK?
   • Has anyone close to you tried to hurt you or harm you recently?

Because of time constraints and lack of a universal standard, there is debate whether regular elder abuse screening is time-effective. It often is recommended, therefore, that clinicians in primary care (1) refer older adults with risk factors for abuse to geriatric medical teams trained in these measures and (2) perform periodic follow-up on such patients⁴ (Figure).

This Web Exclusive Table provides a 7-point summary reference guide for understanding and preventing elder abuse in your practice.

BOTTOM LINEIdentification of elder abuse can be difficult because signs and symptoms of abuse closely resemble physiologic aging. Older adults with identifiable risk factors should be screened for abuse; time constraints make universal screening impossible at this time. In the future, multidisciplinary approaches likely will make elder abuse more easily identifiable through the combined work of health care providers, law enforcement agencies, banks, and other institutions—with the ultimate goal of protecting older adults in the community from abuse.

Related Resources
• Frazão SL, Correia AM, Norton P, et al. Physical abuse against elderly persons in institutional settings. J Forensic Leg Med. 2015;36:54-60.
• Sorrentino R. Performing capacity evaluations: what’s expected from your consult. Current Psychiatry. 2014;13(1):41-44.

Disclosures
Ms. Hubert reports no financial relationships with any company whose products are mentioned in this article or with manufacturers of competing products. Dr. Gupta is a member of the speakers’ bureau of Alkermes, Allergan, Avanir Pharmaceuticals, Takeda Pharmaceutical, Lundbeck, Otsuka Pharmaceutical, and Sunovion Pharmaceuticals.

When a known depressed patient newly develops signs of mania or hypomania, a cascade of diagnostic and therapeu­tic questions ensues: Does the event “automatically” signify the presence of bipolar disorder (BD), or could manic symptoms be secondary to another underlying medical problem, a prescribed anti­depressant or non-psychotropic medication, or illicit substances?

Even more questions confront the clinician: If mania symptoms are nothing more than an adverse drug reaction, will they go away by stop­ping the presumed offending agent? Or do symptoms always indicate the unmasking of a bipolar diathesis? Should anti-manic medication be prescribed immediately? If so, which one(s) and for how long? How extensive a medical or neurologic workup is indicated?

And, how do you differentiate ambiguous hypomania symptoms (irritability, insomnia, agitation) from other phenomena, such as akathisia, anxiety, and overstimulation?

In this article, we present an overview of how to approach and answer these key questions, so that you can identify, comprehend, and manage manic symptoms that arise in the course of your patient’s treatment for depression (Box).

Does disease exist on a unipolar−bipolar continuum?
There has been a resurgence of interest in Kraepelin’s original notion of mania and depression as falling along a continuum, rather than being distinct categories of pathology. True bipolar mania has its own identifiable epidemiology, familiality, and treatment, but symptomatic shades of gray often pose a formidable diagnostic and therapeutic challenge.

For example, DSM-5 relaxed its defini­tion of “mixed” episodes of BD to include subsyndromal mania features in unipolar depression. When a patient with unipolar depression develops a full, unequivocal manic episode, there usually isn’t much ambiguity or confusion about initial man­agement: assure a safe environment, stop any antidepressants, rule out drug- or medically induced causes, and begin an acute anti-manic medication.

Next steps can, sometimes, be murkier:
   • formulate a definitive, overarching diagnosis
   • provide psycho-education
   • forecast return to work or school
   • discuss prognosis and likelihood of relapse
   • address necessary lifestyle modifica­tions (eg, sleep hygiene, elimination of alcohol and illicit drug use)
   • determine whether indefinite mainte­nance pharmacotherapy is indicated— and, if so, with which medication(s).

CASE A diagnostic formulation isn’t always black and white
Ms. J, age 56, a medically healthy woman, has a 10-year history of depression and anxiety that has been treated effectively for most of that time with venlafaxine, 225 mg/d. The mother of 4 grown children, Ms. J has worked steadily for >20 years as a flight attendant for an international airline.

Today, Ms. J is brought by ambulance from work to the emergency department in a par­anoid and agitated state. The admission fol­lows her having e-blasted airline corporate executives with a voluminous manifesto that she worked on around the clock the preced­ing week, in which she explained her bold ideas to revolutionize the airline industry, under her leadership.

Ms. J’s family history is unremarkable for psychiatric illness.

How does one approach a case such as Ms. J’s?
Stark examples of classical mania, as depicted in this case vignette, are easy to recognize but not necessarily straightfor­ward, nosologically. Consider the follow­ing not-so-straightforward elements of Ms. J’s case:
   • a first-lifetime episode of mania or hypomania is rare after age 50
   • Ms. J took a serotonin-norepinephrine reuptake inhibitor (SNRI) for many years without evidence of mood destabilization
   • years of repetitive chronobiological stress (including probable frequent time zone changes with likely sleep disruption) appar­ently did not trigger mood destabilization
   • none of Ms. J’s 4 pregnancies led to postpartum mood episodes
   • at least on the surface, there are no obvious features that point to likely causes of a secondary mania (eg, drug-induced, toxic, metabolic, or medical)

• Ms. J has no known family history of BD or any other mood disorder.

Approaching a case such as Ms. J’s must involve a systematic strategy that can best be broken into 2 segments: (1) a period of acute initial assessment and treatment and (2) later efforts focused on broader diagnostic evalu­ation and longer-term relapse prevention.

Initial assessment and treatment
Immediate assessment and manage­ment hinges on initial triage and form­ing a working diagnostic impression. Although full-blown mania usually is obvious (sometimes even without a formal interview), be alert to patients who might minimize or altogether disavow mania symptoms—often because of denial of ill­ness, misidentification of symptoms, or impaired insight about changes in think­ing, mood, or behavior.

Because florid mania, by definition, impairs psychosocial functioning, the con­text of an initial presentation often holds diagnostic relevance. Manic patients who display disruptive behaviors often are brought to treatment by a third party, whereas a less severely ill patient might be more inclined to seek treatment for herself (himself) when psychosis is absent and insight is less compromised or when the patient feels she (he) might be depressed.

It is not uncommon for a manic patient to report “depression” as the chief complaint or to omit elements related to psychomo­tor acceleration (such as racing thoughts or psychomotor agitation) in the description of symptoms. An accurate diagnosis often requires clinical probing and clarification of symptoms (eg, differentiating simple insom­nia with consequent next-day fatigue from loss of the need for sleep with intact or even enhanced next-day energy) or discriminating racing thoughts from anxious ruminations that might be more intrusive than rapid.

Presentations of frank mania also can come to light as a consequence of symp­toms, rather than as symptoms per se (eg, conflict in relationships, problems at work, financial reversals).

Particularly in patients who do not have a history of mania, avoid the temptation to begin or modify existing pharmacother­apy until you have performed a basic ini­tial evaluation. Immediate considerations for initial assessment and management include the following:

Provide containment. Ensure a safe setting, level of care, and frequency of monitoring. Evaluate suicide risk (particularly when mixed features are present), and risk of with­drawal from any psychoactive substances.

Engage significant others. Close family members can provide essential history, particularly when a patient’s insight about her illness and need for treatment are impaired. Family members and significant others also often play important roles in helping to restrict access to finances, fos­tering medication adherence, preventing access to weapons in the home, and shar­ing information with providers about sub­stance use or high-risk behavior.

Systematically assess for DSM-5 symp­toms of mania and depression. DSM-5 modified criteria for mania/hypomania to necessitate increased energy, in addition to change in mood, to make a syndromal diag­nosis. Useful during a clinical interview is the popular mnemonic DIGFAST to aid recognition of core mania symptoms^a:
   • Distractibility
   • Indiscretion/impulsivity
   • Grandiosity
   • Flight of ideas
   • Activity increase
   • Sleep deficit
   • Talkativeness.

^aAlso see: “Mnemonics in a mnutshell: 32 aids to psychiatric diagnosis,” in the October 2008 issue Current Psychiatry and in the archive at CurrentPsychiatry.com.

These symptoms should represent a departure from normal baseline character­istics; it often is helpful to ask a significant other or collateral historian how the pres­ent symptoms differ from the patient’s usual state.

Assess for unstable medical conditions or toxicity states. When evaluating an acute change in mental status, toxicology screening is relatively standard and the absence of illicit substances should sel­dom, if ever, be taken for granted—espe­cially because occult substance use can lead to identification of false-positive BD “cases.”¹

Stop any antidepressant. During a manic episode, continuing antidepressant medi­cation serves no purpose other than to contribute to or exacerbate mania symp­toms. Nonetheless, observational studies demonstrate that approximately 15% of syndromally manic patients continue to receive an antidepressant, often when a clinician perceives more severe depression during mania, multiple prior depressive episodes, current anxiety, or rapid cycling.²

Importantly, antidepressants have been shown to harm, rather than alleviate, pre­sentations that involve a mixed state,³ and have no demonstrated value in preventing post-manic depression. Mere elimination of an antidepressant might ease symptoms during a manic or mixed episode.⁴

In some cases, it might be advisable to taper, not abruptly stop, a short half-life serotonergic antidepressant, even in the setting of mania, to minimize the potential for aggravating autonomic dysregulation that can result from antidepressant discon­tinuation effects. 

Begin anti-manic pharmacotherapy. Initiation of an anti-manic mood stabilizer, such as lithium and divalproex, has been standard in the treatment of acute mania.

In the 1990s, protocols for oral loading of divalproex (20 to 30 mg/kg/d) gained popularity for achieving more rapid symp­tom improvement than might occur with lithium. In the current era, atypical anti­psychotics have all but replaced mood sta­bilizers as an initial intervention to contain mania symptoms quickly (and with less risk than first-generation antipsychotics for acute adverse motor effects from so-called rapid neuroleptization).

Because atypical antipsychotics often rapidly subdue mania, psychosis, and agitation, regardless of the underlying process, many practitioners might feel more comfortable initiating them than a mood stabilizer when the diagnosis is ambiguous or provisional, although their longer-term efficacy and safety, relative to traditional mood stabilizers, remains contested. Considerations for choosing from among feasible anti-manic pharma­cotherapies are summarized in Table 1.⁵ 

Normalize the sleep-wake cycle. Chronobiological and circadian vari­ables, such as irregular sleep patterns, are thought to contribute to the pathophysiol­ogy of affective switch in BD. Behavioral and pharmacotherapeutic efforts to impose a normal sleep−wake schedule are considered fundamental to stabilizing acute mania.

Facilitate next steps after acute stabili­zation. For inpatients, this might involve step-down to a partial hospitalization or intensive outpatient program, alongside taking steps to ensure continued treatment adherence and minimize relapse.

What medical and neurologic workup is appropriate?
Not every first lifetime presentation of mania requires extensive medical and neurologic workup, particularly among patients who have a history of depression and those whose presentation neatly fits the demographic and clinical profile of newly emergent BD. Basic assessment should determine whether any new medication has been started that could plausibly contribute to abnormal mental sta­tus (Table 2).

Nevertheless, evaluation of almost all first presentations of mania should include:
   • urine toxicology screen 
   • complete blood count
   • comprehensive metabolic panel
   • thyroid-stimulating hormone assay
   • serum vitamin B₁₂ level assay
   • serum folic acid level assay
   • rapid plasma reagin test.

Clinical features that usually lead a cli­nician to pursue a more detailed medical and neurologic evaluation of first-episode mania include:
   • onset age >40
   • absence of a family history of mood disorder
   • symptoms arising during a major medical illness
   • multiple medications
   • suspicion of a degenerative or heredi­tary neurologic disorder
   • altered state of consciousness
   • signs of cortical or diffuse subcorti­cal dysfunction (eg, cognitive deficits, motor deficits, tremor)
   • abnormal vital signs.

Depending on the presentation, addi­tional testing might include:
   • tests of HIV antibody, immune auto­antibodies, and Lyme disease antibody
   • heavy metal screening (when sug­gested by environmental exposure)
   • lumbar puncture (eg, in a setting of manic delirium or suspected central nervous system infection or paraneoplastic syndrome)
   • neuroimaging (note: MRI provides bet­ter visualization than CT of white matter pathology and small vessel cerebrovascular disease) electroencephalography.

Making an overarching diagnosis: Is mania always bipolar disorder?
Mania is considered a manifestation of BD when symptoms cannot be attributed to another psychiatric condition, another underlying medical or neurologic condi­tion, or a toxic-metabolic state (Table 3 and Table 4^6-9). Classification of mania that occurs soon after antidepressant exposure in patients without a known history of BD continues to be the subject of debate, vary­ing in its conceptualization across editions of DSM.

The National Institute of Mental Health (NIMH) Systematic Treatment Enhancement Program for Bipolar Disorder, or STEP-BD, observed a fairly low (approximately 10%) incidence of switch from depression to mania when an antidepressant is added to a mood stabilizer; the study authors con­cluded that much of what is presumed to be antidepressant-induced mania might simply be the natural course of illness.¹⁰

Notably, several reports suggest that antidepressants might pose a greater risk of mood destabilization in people with BD I than with either BD II or other sus­pected variants on the bipolar spectrum.

DSM-5 advises that a diagnosis of substance-induced mood disorder appro­priately describes symptoms that spontane­ously dissipate once an antidepressant has been discontinued, whereas a diagnosis of BD can be made when manic or hypomanic symptoms persist at a syndromal level after an antidepressant has been stopped and its physiological effects are no longer present. With respect to time course, the International Society of Bipolar Disorders proposes that, beyond 12 to 16 weeks after an antidepressant has been started or the dosage has been increased, it is unlikely that new-onset mania/hypomania can rea­sonably be attributed to “triggering” by an antidepressant¹¹ (although antidepressants should be stopped when symptoms of mania emerge).

Several clinical features have been linked in the literature with an increased suscepti­bility to BD after an initial depressive epi­sode, including:
   • early (pre-adolescent) age at onset of first mood disorder episode⁶
   • family history of BD, highly recurrent depression, or psychosis^12,13
   • psychosis when depressed.^7,14

A number of other characteristics of depressive illness—including seasonal depression, atypical depressive features, suicidality, irritability, anxiety or sub­stance use comorbidity, postpartum mood episodes, and brief recurrent depressive episodes—have been described in the lit­erature as potential correlates of a bipolar diathesis; none have proved to be robust or pathognomonic of a BD diagnosis, as opposed to a unipolar diagnosis.

Data from the NIMH Collaborative Depression Study suggest that recurrent mania/hypomania after an antidepressant-associated polarity switch is greater when a family history of BD is present; other clinical variables might hold less predictive value.¹⁵

In addition, although some practitioners consider a history of nonresponse to trials of multiple antidepressants suggestive of an underlying bipolar process, polarity is only one of many variables that must be considered in the differential diagnosis of antidepressant-resistant depression.^b Likewise, molecular genetic studies do not support a link between antidepressant nonresponse and the likelihood of a diag­nosis of BD.¹⁶

^bSee “A practical approach to subtyping depression among your patients” in the April 2014 issue of Current Psychiatry or in the archive at CurrentPsychiatry.com.

Indefinite pharmacotherapy for bipolar disorder?
An important but nagging issue when diag­nosing BD after a first manic (or hypomanic) episode is the implied need for indefinite pharmacotherapy to sustain remission and prevent relapse and recurrence.

The likelihood of subsequent depression or mania/hypomania remains high after an index manic/hypomanic episode, par­ticularly for 6 to 8 months after recovery.^8,17 Natural history data suggest that, during the year that follows a first lifetime mania, approximately 40% of patients experience a second manic episode.⁸ A second lifetime mania might be especially likely in patients whose index episode involved mood-congruent psychosis, low premorbid work functioning, and an initial manic episode, as opposed to a mixed episode¹⁷ or early age at onset.⁸

In the absence of randomized, placebo-controlled studies of maintenance phar­macotherapy after a first lifetime manic episode, clinical judgment often drives decisions about the duration of continuing pharmacotherapy after initial symptoms resolve. The Texas Medication Algorithm Project for BD advises that:

Similarly, in the most recent (2004) Expert Consensus Guideline Series for the Treatment of Bipolar Disorder,¹⁹ 84% of practitioner−respondents favored indefi­nite mood stabilizer therapy after a second lifetime manic episode. No recommen­dation was made about the duration of maintenance pharmacotherapy after a first lifetime manic/hypomanic episode.

Avoid or reintroduce an antidepressant if depression recurs after a first mania?
Controversies surrounding antidepressant use in BD are extensive; detailed discus­sion is beyond the scope of this review (Goldberg and Ghaemi provided a broader discussion of risks and benefits of antide­pressants in BD20). Although the main clinical concern regarding antidepres­sant use was, at one time, the potential to induce mania or accelerate the frequency of recurrent episodes, more recent, empiri­cal studies suggest that the greater risk of using antidepressants for BD is lack of efficacy.^10,21

If a careful longitudinal history and clin­ical evaluation reveal that an initial manic episode heralds the onset of BD, decisions about whether to avoid an antidepressant (as opposed to using other, more evidence-based interventions for bipolar depres­sion) depend on a number of variables, including establishing whether the index episode was manic or hypomanic; ruling out current subthreshold mixed features; and clarifying how recently mania devel­oped. Decisions about future antidepres­sant use (or avoidance) might be less clear if an index manic/hypomanic episode was brief and self-limited once the antidepres­sant was stopped.

Although some experts eschew antidepressant monotherapy after such occurrences, there is no body of litera­ture to inform decisions about the safety or efficacy of undertaking a future antidepressant trial in such patients. That said, reasonable judgment probably includes several considerations:
   • Re-exposure to the same antidepres­sant that was associated with an induction of mania is likely riskier than choosing a different antidepressant; in general, purely serotonergic antidepressants or bupropion are considered to pose less risk of mood destabilization than is seen with an SNRI or tricyclic antidepressant.
   • After a manic episode, a subsequent antidepressant trial generally shouldn’t be attempted without concurrent anti-manic medication.
   • Introducing any antidepressant is probably ill-advised in the recent (~2 months) aftermath of acute manic/ hypomanic symptoms.²²
   • Patients and their significant other should be apprised of the risk of emerg­ing symptoms of mania or hypomania, or mixed features, and should be familiar with key target symptoms to watch for. Prospective mood charting can be helpful.
   • Patients should be monitored closely both for an exacerbation of depression and recurrence of mania/hypomania symptoms.
   • Any antidepressant should be discon­tinued promptly at the first sign of psy­chomotor acceleration or the emergence of mixed features, as defined by DSM-5.

Psychoeducation and forecasting
Functional recovery from a manic episode can lag behind symptomatic recovery. Subsyndromal symptoms often persist after a full episode subsides.

Mania often is followed by a depres­sive episode, and questions inevitably arise about how to prevent and treat these epi­sodes. Because the median duration of a manic episode is approximately 13 weeks,²³ it is crucial for patients and their immedi­ate family to recognize that recovery might be gradual, and that it will likely take time before she (he) can resume full-time respon­sibilities at work or school or in the home.

Today, a patient who is hospitalized for severe acute mania (as Ms. J was, in the case vignette) seldom remains an inpa­tient long enough to achieve remission of symptoms; sometimes, she (he) might con­tinue to manifest significant symptoms, even though decisions about the “medical necessity” of ongoing inpatient care tend to be governed mainly by issues of safety and imminent danger. (This web exclusive Table^20,24,25 provides considerations when making the transition from the acute phase to the continuation phase of treatment.)

To minimize risk of relapse, psycho-education should include discussion of:
   • psychiatrically deleterious effects of alcohol and illicit drug use
   • suicide risk, including what to do in an emergency
   • protecting a regular sleep schedule and avoiding sleep deprivation
   • the potential for poor medication adherence and management of side effects
   • the need for periodic laboratory moni­toring, as needed
   • the role of adjunctive psychotherapy and effective stress management
   • familiarity with symptoms that serve as warning signs, and how to monitor their onset.

Bottom Line
When a patient being treated for depression develops signs of mania or hypomania, stop any antidepressant and consider initiating a mood stabilizer, antipsychotic, or both, to contain and stabilize symptoms. Entertain medical and substance-related causes of mania symptoms, and evaluate and treat as suggested by the patient’s presentation. Long-term drug therapy to prevent recurrence of mania/hypomania, as well as risks and benefits of future exposure to antidepressants, should be decided case by case.

Related Resources
• Proudfoot J, Whitton A, Parker G, et al. Triggers of mania and depression in young adults with bipolar disorder. J Affect Disord. 2012;143(1-3):196-202.
• Stange JP, Sylvia LG, Magalhães PV, et al. Extreme at­tributions predict transition from depression to ma­nia or hypomania in bipolar disorder. J Psychiatr Res. 2013;47(10):1329-1336.

Drug Brand Names
Albuterol • Proventil, Ventolin
Anastrozole • Arimidex
Aripiprazole • Abilify
Bupropion • Wellbutrin
Carbamazepine • Tegretol
Chloroquine • Aralen
Ciprofloxacin • Cipro
Clarithromycin • Biaxin
Clomiphene • Clomid
Digoxin • Digox, Lanoxin
Divalproex • Depakote
5-Fluorouracil • Carac, Efudex
Human chorionic gonadotropin • Novarel, Pregnyl
Ifosfamide • Ifex
Isoniazid • Nydrazid
Lamotrigine • Lamictal
Letrozole • Femara
Lithium • Eskalith, Lithobid
Lurasidone • Latuda
Mefloquine • Lariam
Olanzapine • Zyprexa
Olanzapine/fluoxetine combination • Symbyax
ramipexole • Mirapex
Procarbazine • Matulane
Quetiapine • Seroquel
Ropinirole • Requip
Rotigotine • Neupro
Venlafaxine • Effexor
Zidovudine • Retrovir

Disclosures
Dr. Goldberg is a consultant to Merck & Co. and Sunovion. He is a member of the speakers’ bureau of AstraZeneca, Janssen, Merck & Co., Takeda and Lundbeck, and Sunovion.

Dr. Ernst reports no financial relationships with any company whose products are mentioned in this article or with manufacturers of competing products.

When a known depressed patient newly develops signs of mania or hypomania, a cascade of diagnostic and therapeu­tic questions ensues: Does the event “automatically” signify the presence of bipolar disorder (BD), or could manic symptoms be secondary to another underlying medical problem, a prescribed anti­depressant or non-psychotropic medication, or illicit substances?

Even more questions confront the clinician: If mania symptoms are nothing more than an adverse drug reaction, will they go away by stop­ping the presumed offending agent? Or do symptoms always indicate the unmasking of a bipolar diathesis? Should anti-manic medication be prescribed immediately? If so, which one(s) and for how long? How extensive a medical or neurologic workup is indicated?

And, how do you differentiate ambiguous hypomania symptoms (irritability, insomnia, agitation) from other phenomena, such as akathisia, anxiety, and overstimulation?

In this article, we present an overview of how to approach and answer these key questions, so that you can identify, comprehend, and manage manic symptoms that arise in the course of your patient’s treatment for depression (Box).

Does disease exist on a unipolar−bipolar continuum?
There has been a resurgence of interest in Kraepelin’s original notion of mania and depression as falling along a continuum, rather than being distinct categories of pathology. True bipolar mania has its own identifiable epidemiology, familiality, and treatment, but symptomatic shades of gray often pose a formidable diagnostic and therapeutic challenge.

For example, DSM-5 relaxed its defini­tion of “mixed” episodes of BD to include subsyndromal mania features in unipolar depression. When a patient with unipolar depression develops a full, unequivocal manic episode, there usually isn’t much ambiguity or confusion about initial man­agement: assure a safe environment, stop any antidepressants, rule out drug- or medically induced causes, and begin an acute anti-manic medication.

Next steps can, sometimes, be murkier:
   • formulate a definitive, overarching diagnosis
   • provide psycho-education
   • forecast return to work or school
   • discuss prognosis and likelihood of relapse
   • address necessary lifestyle modifica­tions (eg, sleep hygiene, elimination of alcohol and illicit drug use)
   • determine whether indefinite mainte­nance pharmacotherapy is indicated— and, if so, with which medication(s).

CASE A diagnostic formulation isn’t always black and white
Ms. J, age 56, a medically healthy woman, has a 10-year history of depression and anxiety that has been treated effectively for most of that time with venlafaxine, 225 mg/d. The mother of 4 grown children, Ms. J has worked steadily for >20 years as a flight attendant for an international airline.

Today, Ms. J is brought by ambulance from work to the emergency department in a par­anoid and agitated state. The admission fol­lows her having e-blasted airline corporate executives with a voluminous manifesto that she worked on around the clock the preced­ing week, in which she explained her bold ideas to revolutionize the airline industry, under her leadership.

Ms. J’s family history is unremarkable for psychiatric illness.

How does one approach a case such as Ms. J’s?
Stark examples of classical mania, as depicted in this case vignette, are easy to recognize but not necessarily straightfor­ward, nosologically. Consider the follow­ing not-so-straightforward elements of Ms. J’s case:
   • a first-lifetime episode of mania or hypomania is rare after age 50
   • Ms. J took a serotonin-norepinephrine reuptake inhibitor (SNRI) for many years without evidence of mood destabilization
   • years of repetitive chronobiological stress (including probable frequent time zone changes with likely sleep disruption) appar­ently did not trigger mood destabilization
   • none of Ms. J’s 4 pregnancies led to postpartum mood episodes
   • at least on the surface, there are no obvious features that point to likely causes of a secondary mania (eg, drug-induced, toxic, metabolic, or medical)

• Ms. J has no known family history of BD or any other mood disorder.

Approaching a case such as Ms. J’s must involve a systematic strategy that can best be broken into 2 segments: (1) a period of acute initial assessment and treatment and (2) later efforts focused on broader diagnostic evalu­ation and longer-term relapse prevention.

Initial assessment and treatment
Immediate assessment and manage­ment hinges on initial triage and form­ing a working diagnostic impression. Although full-blown mania usually is obvious (sometimes even without a formal interview), be alert to patients who might minimize or altogether disavow mania symptoms—often because of denial of ill­ness, misidentification of symptoms, or impaired insight about changes in think­ing, mood, or behavior.

Because florid mania, by definition, impairs psychosocial functioning, the con­text of an initial presentation often holds diagnostic relevance. Manic patients who display disruptive behaviors often are brought to treatment by a third party, whereas a less severely ill patient might be more inclined to seek treatment for herself (himself) when psychosis is absent and insight is less compromised or when the patient feels she (he) might be depressed.

It is not uncommon for a manic patient to report “depression” as the chief complaint or to omit elements related to psychomo­tor acceleration (such as racing thoughts or psychomotor agitation) in the description of symptoms. An accurate diagnosis often requires clinical probing and clarification of symptoms (eg, differentiating simple insom­nia with consequent next-day fatigue from loss of the need for sleep with intact or even enhanced next-day energy) or discriminating racing thoughts from anxious ruminations that might be more intrusive than rapid.

Presentations of frank mania also can come to light as a consequence of symp­toms, rather than as symptoms per se (eg, conflict in relationships, problems at work, financial reversals).

Particularly in patients who do not have a history of mania, avoid the temptation to begin or modify existing pharmacother­apy until you have performed a basic ini­tial evaluation. Immediate considerations for initial assessment and management include the following:

Provide containment. Ensure a safe setting, level of care, and frequency of monitoring. Evaluate suicide risk (particularly when mixed features are present), and risk of with­drawal from any psychoactive substances.

Engage significant others. Close family members can provide essential history, particularly when a patient’s insight about her illness and need for treatment are impaired. Family members and significant others also often play important roles in helping to restrict access to finances, fos­tering medication adherence, preventing access to weapons in the home, and shar­ing information with providers about sub­stance use or high-risk behavior.

Systematically assess for DSM-5 symp­toms of mania and depression. DSM-5 modified criteria for mania/hypomania to necessitate increased energy, in addition to change in mood, to make a syndromal diag­nosis. Useful during a clinical interview is the popular mnemonic DIGFAST to aid recognition of core mania symptoms^a:
   • Distractibility
   • Indiscretion/impulsivity
   • Grandiosity
   • Flight of ideas
   • Activity increase
   • Sleep deficit
   • Talkativeness.

^aAlso see: “Mnemonics in a mnutshell: 32 aids to psychiatric diagnosis,” in the October 2008 issue Current Psychiatry and in the archive at CurrentPsychiatry.com.

These symptoms should represent a departure from normal baseline character­istics; it often is helpful to ask a significant other or collateral historian how the pres­ent symptoms differ from the patient’s usual state.

Assess for unstable medical conditions or toxicity states. When evaluating an acute change in mental status, toxicology screening is relatively standard and the absence of illicit substances should sel­dom, if ever, be taken for granted—espe­cially because occult substance use can lead to identification of false-positive BD “cases.”¹

Stop any antidepressant. During a manic episode, continuing antidepressant medi­cation serves no purpose other than to contribute to or exacerbate mania symp­toms. Nonetheless, observational studies demonstrate that approximately 15% of syndromally manic patients continue to receive an antidepressant, often when a clinician perceives more severe depression during mania, multiple prior depressive episodes, current anxiety, or rapid cycling.²

Importantly, antidepressants have been shown to harm, rather than alleviate, pre­sentations that involve a mixed state,³ and have no demonstrated value in preventing post-manic depression. Mere elimination of an antidepressant might ease symptoms during a manic or mixed episode.⁴

In some cases, it might be advisable to taper, not abruptly stop, a short half-life serotonergic antidepressant, even in the setting of mania, to minimize the potential for aggravating autonomic dysregulation that can result from antidepressant discon­tinuation effects. 

Begin anti-manic pharmacotherapy. Initiation of an anti-manic mood stabilizer, such as lithium and divalproex, has been standard in the treatment of acute mania.

In the 1990s, protocols for oral loading of divalproex (20 to 30 mg/kg/d) gained popularity for achieving more rapid symp­tom improvement than might occur with lithium. In the current era, atypical anti­psychotics have all but replaced mood sta­bilizers as an initial intervention to contain mania symptoms quickly (and with less risk than first-generation antipsychotics for acute adverse motor effects from so-called rapid neuroleptization).

Because atypical antipsychotics often rapidly subdue mania, psychosis, and agitation, regardless of the underlying process, many practitioners might feel more comfortable initiating them than a mood stabilizer when the diagnosis is ambiguous or provisional, although their longer-term efficacy and safety, relative to traditional mood stabilizers, remains contested. Considerations for choosing from among feasible anti-manic pharma­cotherapies are summarized in Table 1.⁵ 

Normalize the sleep-wake cycle. Chronobiological and circadian vari­ables, such as irregular sleep patterns, are thought to contribute to the pathophysiol­ogy of affective switch in BD. Behavioral and pharmacotherapeutic efforts to impose a normal sleep−wake schedule are considered fundamental to stabilizing acute mania.

Facilitate next steps after acute stabili­zation. For inpatients, this might involve step-down to a partial hospitalization or intensive outpatient program, alongside taking steps to ensure continued treatment adherence and minimize relapse.

What medical and neurologic workup is appropriate?
Not every first lifetime presentation of mania requires extensive medical and neurologic workup, particularly among patients who have a history of depression and those whose presentation neatly fits the demographic and clinical profile of newly emergent BD. Basic assessment should determine whether any new medication has been started that could plausibly contribute to abnormal mental sta­tus (Table 2).

Nevertheless, evaluation of almost all first presentations of mania should include:
   • urine toxicology screen 
   • complete blood count
   • comprehensive metabolic panel
   • thyroid-stimulating hormone assay
   • serum vitamin B₁₂ level assay
   • serum folic acid level assay
   • rapid plasma reagin test.

Clinical features that usually lead a cli­nician to pursue a more detailed medical and neurologic evaluation of first-episode mania include:
   • onset age >40
   • absence of a family history of mood disorder
   • symptoms arising during a major medical illness
   • multiple medications
   • suspicion of a degenerative or heredi­tary neurologic disorder
   • altered state of consciousness
   • signs of cortical or diffuse subcorti­cal dysfunction (eg, cognitive deficits, motor deficits, tremor)
   • abnormal vital signs.

Depending on the presentation, addi­tional testing might include:
   • tests of HIV antibody, immune auto­antibodies, and Lyme disease antibody
   • heavy metal screening (when sug­gested by environmental exposure)
   • lumbar puncture (eg, in a setting of manic delirium or suspected central nervous system infection or paraneoplastic syndrome)
   • neuroimaging (note: MRI provides bet­ter visualization than CT of white matter pathology and small vessel cerebrovascular disease) electroencephalography.

Making an overarching diagnosis: Is mania always bipolar disorder?
Mania is considered a manifestation of BD when symptoms cannot be attributed to another psychiatric condition, another underlying medical or neurologic condi­tion, or a toxic-metabolic state (Table 3 and Table 4^6-9). Classification of mania that occurs soon after antidepressant exposure in patients without a known history of BD continues to be the subject of debate, vary­ing in its conceptualization across editions of DSM.

The National Institute of Mental Health (NIMH) Systematic Treatment Enhancement Program for Bipolar Disorder, or STEP-BD, observed a fairly low (approximately 10%) incidence of switch from depression to mania when an antidepressant is added to a mood stabilizer; the study authors con­cluded that much of what is presumed to be antidepressant-induced mania might simply be the natural course of illness.¹⁰

Notably, several reports suggest that antidepressants might pose a greater risk of mood destabilization in people with BD I than with either BD II or other sus­pected variants on the bipolar spectrum.

DSM-5 advises that a diagnosis of substance-induced mood disorder appro­priately describes symptoms that spontane­ously dissipate once an antidepressant has been discontinued, whereas a diagnosis of BD can be made when manic or hypomanic symptoms persist at a syndromal level after an antidepressant has been stopped and its physiological effects are no longer present. With respect to time course, the International Society of Bipolar Disorders proposes that, beyond 12 to 16 weeks after an antidepressant has been started or the dosage has been increased, it is unlikely that new-onset mania/hypomania can rea­sonably be attributed to “triggering” by an antidepressant¹¹ (although antidepressants should be stopped when symptoms of mania emerge).

Several clinical features have been linked in the literature with an increased suscepti­bility to BD after an initial depressive epi­sode, including:
   • early (pre-adolescent) age at onset of first mood disorder episode⁶
   • family history of BD, highly recurrent depression, or psychosis^12,13
   • psychosis when depressed.^7,14

A number of other characteristics of depressive illness—including seasonal depression, atypical depressive features, suicidality, irritability, anxiety or sub­stance use comorbidity, postpartum mood episodes, and brief recurrent depressive episodes—have been described in the lit­erature as potential correlates of a bipolar diathesis; none have proved to be robust or pathognomonic of a BD diagnosis, as opposed to a unipolar diagnosis.

Data from the NIMH Collaborative Depression Study suggest that recurrent mania/hypomania after an antidepressant-associated polarity switch is greater when a family history of BD is present; other clinical variables might hold less predictive value.¹⁵

In addition, although some practitioners consider a history of nonresponse to trials of multiple antidepressants suggestive of an underlying bipolar process, polarity is only one of many variables that must be considered in the differential diagnosis of antidepressant-resistant depression.^b Likewise, molecular genetic studies do not support a link between antidepressant nonresponse and the likelihood of a diag­nosis of BD.¹⁶

^bSee “A practical approach to subtyping depression among your patients” in the April 2014 issue of Current Psychiatry or in the archive at CurrentPsychiatry.com.

Indefinite pharmacotherapy for bipolar disorder?
An important but nagging issue when diag­nosing BD after a first manic (or hypomanic) episode is the implied need for indefinite pharmacotherapy to sustain remission and prevent relapse and recurrence.

The likelihood of subsequent depression or mania/hypomania remains high after an index manic/hypomanic episode, par­ticularly for 6 to 8 months after recovery.^8,17 Natural history data suggest that, during the year that follows a first lifetime mania, approximately 40% of patients experience a second manic episode.⁸ A second lifetime mania might be especially likely in patients whose index episode involved mood-congruent psychosis, low premorbid work functioning, and an initial manic episode, as opposed to a mixed episode¹⁷ or early age at onset.⁸

In the absence of randomized, placebo-controlled studies of maintenance phar­macotherapy after a first lifetime manic episode, clinical judgment often drives decisions about the duration of continuing pharmacotherapy after initial symptoms resolve. The Texas Medication Algorithm Project for BD advises that:

Similarly, in the most recent (2004) Expert Consensus Guideline Series for the Treatment of Bipolar Disorder,¹⁹ 84% of practitioner−respondents favored indefi­nite mood stabilizer therapy after a second lifetime manic episode. No recommen­dation was made about the duration of maintenance pharmacotherapy after a first lifetime manic/hypomanic episode.

Avoid or reintroduce an antidepressant if depression recurs after a first mania?
Controversies surrounding antidepressant use in BD are extensive; detailed discus­sion is beyond the scope of this review (Goldberg and Ghaemi provided a broader discussion of risks and benefits of antide­pressants in BD20). Although the main clinical concern regarding antidepres­sant use was, at one time, the potential to induce mania or accelerate the frequency of recurrent episodes, more recent, empiri­cal studies suggest that the greater risk of using antidepressants for BD is lack of efficacy.^10,21

If a careful longitudinal history and clin­ical evaluation reveal that an initial manic episode heralds the onset of BD, decisions about whether to avoid an antidepressant (as opposed to using other, more evidence-based interventions for bipolar depres­sion) depend on a number of variables, including establishing whether the index episode was manic or hypomanic; ruling out current subthreshold mixed features; and clarifying how recently mania devel­oped. Decisions about future antidepres­sant use (or avoidance) might be less clear if an index manic/hypomanic episode was brief and self-limited once the antidepres­sant was stopped.

Although some experts eschew antidepressant monotherapy after such occurrences, there is no body of litera­ture to inform decisions about the safety or efficacy of undertaking a future antidepressant trial in such patients. That said, reasonable judgment probably includes several considerations:
   • Re-exposure to the same antidepres­sant that was associated with an induction of mania is likely riskier than choosing a different antidepressant; in general, purely serotonergic antidepressants or bupropion are considered to pose less risk of mood destabilization than is seen with an SNRI or tricyclic antidepressant.
   • After a manic episode, a subsequent antidepressant trial generally shouldn’t be attempted without concurrent anti-manic medication.
   • Introducing any antidepressant is probably ill-advised in the recent (~2 months) aftermath of acute manic/ hypomanic symptoms.²²
   • Patients and their significant other should be apprised of the risk of emerg­ing symptoms of mania or hypomania, or mixed features, and should be familiar with key target symptoms to watch for. Prospective mood charting can be helpful.
   • Patients should be monitored closely both for an exacerbation of depression and recurrence of mania/hypomania symptoms.
   • Any antidepressant should be discon­tinued promptly at the first sign of psy­chomotor acceleration or the emergence of mixed features, as defined by DSM-5.

Psychoeducation and forecasting
Functional recovery from a manic episode can lag behind symptomatic recovery. Subsyndromal symptoms often persist after a full episode subsides.

Mania often is followed by a depres­sive episode, and questions inevitably arise about how to prevent and treat these epi­sodes. Because the median duration of a manic episode is approximately 13 weeks,²³ it is crucial for patients and their immedi­ate family to recognize that recovery might be gradual, and that it will likely take time before she (he) can resume full-time respon­sibilities at work or school or in the home.

Today, a patient who is hospitalized for severe acute mania (as Ms. J was, in the case vignette) seldom remains an inpa­tient long enough to achieve remission of symptoms; sometimes, she (he) might con­tinue to manifest significant symptoms, even though decisions about the “medical necessity” of ongoing inpatient care tend to be governed mainly by issues of safety and imminent danger. (This web exclusive Table^20,24,25 provides considerations when making the transition from the acute phase to the continuation phase of treatment.)

To minimize risk of relapse, psycho-education should include discussion of:
   • psychiatrically deleterious effects of alcohol and illicit drug use
   • suicide risk, including what to do in an emergency
   • protecting a regular sleep schedule and avoiding sleep deprivation
   • the potential for poor medication adherence and management of side effects
   • the need for periodic laboratory moni­toring, as needed
   • the role of adjunctive psychotherapy and effective stress management
   • familiarity with symptoms that serve as warning signs, and how to monitor their onset.

Bottom Line
When a patient being treated for depression develops signs of mania or hypomania, stop any antidepressant and consider initiating a mood stabilizer, antipsychotic, or both, to contain and stabilize symptoms. Entertain medical and substance-related causes of mania symptoms, and evaluate and treat as suggested by the patient’s presentation. Long-term drug therapy to prevent recurrence of mania/hypomania, as well as risks and benefits of future exposure to antidepressants, should be decided case by case.

Related Resources
• Proudfoot J, Whitton A, Parker G, et al. Triggers of mania and depression in young adults with bipolar disorder. J Affect Disord. 2012;143(1-3):196-202.
• Stange JP, Sylvia LG, Magalhães PV, et al. Extreme at­tributions predict transition from depression to ma­nia or hypomania in bipolar disorder. J Psychiatr Res. 2013;47(10):1329-1336.

Drug Brand Names
Albuterol • Proventil, Ventolin
Anastrozole • Arimidex
Aripiprazole • Abilify
Bupropion • Wellbutrin
Carbamazepine • Tegretol
Chloroquine • Aralen
Ciprofloxacin • Cipro
Clarithromycin • Biaxin
Clomiphene • Clomid
Digoxin • Digox, Lanoxin
Divalproex • Depakote
5-Fluorouracil • Carac, Efudex
Human chorionic gonadotropin • Novarel, Pregnyl
Ifosfamide • Ifex
Isoniazid • Nydrazid
Lamotrigine • Lamictal
Letrozole • Femara
Lithium • Eskalith, Lithobid
Lurasidone • Latuda
Mefloquine • Lariam
Olanzapine • Zyprexa
Olanzapine/fluoxetine combination • Symbyax
ramipexole • Mirapex
Procarbazine • Matulane
Quetiapine • Seroquel
Ropinirole • Requip
Rotigotine • Neupro
Venlafaxine • Effexor
Zidovudine • Retrovir

Disclosures
Dr. Goldberg is a consultant to Merck & Co. and Sunovion. He is a member of the speakers’ bureau of AstraZeneca, Janssen, Merck & Co., Takeda and Lundbeck, and Sunovion.

Dr. Ernst reports no financial relationships with any company whose products are mentioned in this article or with manufacturers of competing products.

When a known depressed patient newly develops signs of mania or hypomania, a cascade of diagnostic and therapeu­tic questions ensues: Does the event “automatically” signify the presence of bipolar disorder (BD), or could manic symptoms be secondary to another underlying medical problem, a prescribed anti­depressant or non-psychotropic medication, or illicit substances?

Even more questions confront the clinician: If mania symptoms are nothing more than an adverse drug reaction, will they go away by stop­ping the presumed offending agent? Or do symptoms always indicate the unmasking of a bipolar diathesis? Should anti-manic medication be prescribed immediately? If so, which one(s) and for how long? How extensive a medical or neurologic workup is indicated?

And, how do you differentiate ambiguous hypomania symptoms (irritability, insomnia, agitation) from other phenomena, such as akathisia, anxiety, and overstimulation?

In this article, we present an overview of how to approach and answer these key questions, so that you can identify, comprehend, and manage manic symptoms that arise in the course of your patient’s treatment for depression (Box).

Does disease exist on a unipolar−bipolar continuum?
There has been a resurgence of interest in Kraepelin’s original notion of mania and depression as falling along a continuum, rather than being distinct categories of pathology. True bipolar mania has its own identifiable epidemiology, familiality, and treatment, but symptomatic shades of gray often pose a formidable diagnostic and therapeutic challenge.

For example, DSM-5 relaxed its defini­tion of “mixed” episodes of BD to include subsyndromal mania features in unipolar depression. When a patient with unipolar depression develops a full, unequivocal manic episode, there usually isn’t much ambiguity or confusion about initial man­agement: assure a safe environment, stop any antidepressants, rule out drug- or medically induced causes, and begin an acute anti-manic medication.

Next steps can, sometimes, be murkier:
   • formulate a definitive, overarching diagnosis
   • provide psycho-education
   • forecast return to work or school
   • discuss prognosis and likelihood of relapse
   • address necessary lifestyle modifica­tions (eg, sleep hygiene, elimination of alcohol and illicit drug use)
   • determine whether indefinite mainte­nance pharmacotherapy is indicated— and, if so, with which medication(s).

CASE A diagnostic formulation isn’t always black and white
Ms. J, age 56, a medically healthy woman, has a 10-year history of depression and anxiety that has been treated effectively for most of that time with venlafaxine, 225 mg/d. The mother of 4 grown children, Ms. J has worked steadily for >20 years as a flight attendant for an international airline.

Today, Ms. J is brought by ambulance from work to the emergency department in a par­anoid and agitated state. The admission fol­lows her having e-blasted airline corporate executives with a voluminous manifesto that she worked on around the clock the preced­ing week, in which she explained her bold ideas to revolutionize the airline industry, under her leadership.

Ms. J’s family history is unremarkable for psychiatric illness.

How does one approach a case such as Ms. J’s?
Stark examples of classical mania, as depicted in this case vignette, are easy to recognize but not necessarily straightfor­ward, nosologically. Consider the follow­ing not-so-straightforward elements of Ms. J’s case:
   • a first-lifetime episode of mania or hypomania is rare after age 50
   • Ms. J took a serotonin-norepinephrine reuptake inhibitor (SNRI) for many years without evidence of mood destabilization
   • years of repetitive chronobiological stress (including probable frequent time zone changes with likely sleep disruption) appar­ently did not trigger mood destabilization
   • none of Ms. J’s 4 pregnancies led to postpartum mood episodes
   • at least on the surface, there are no obvious features that point to likely causes of a secondary mania (eg, drug-induced, toxic, metabolic, or medical)

• Ms. J has no known family history of BD or any other mood disorder.

Approaching a case such as Ms. J’s must involve a systematic strategy that can best be broken into 2 segments: (1) a period of acute initial assessment and treatment and (2) later efforts focused on broader diagnostic evalu­ation and longer-term relapse prevention.

Initial assessment and treatment
Immediate assessment and manage­ment hinges on initial triage and form­ing a working diagnostic impression. Although full-blown mania usually is obvious (sometimes even without a formal interview), be alert to patients who might minimize or altogether disavow mania symptoms—often because of denial of ill­ness, misidentification of symptoms, or impaired insight about changes in think­ing, mood, or behavior.

Because florid mania, by definition, impairs psychosocial functioning, the con­text of an initial presentation often holds diagnostic relevance. Manic patients who display disruptive behaviors often are brought to treatment by a third party, whereas a less severely ill patient might be more inclined to seek treatment for herself (himself) when psychosis is absent and insight is less compromised or when the patient feels she (he) might be depressed.

It is not uncommon for a manic patient to report “depression” as the chief complaint or to omit elements related to psychomo­tor acceleration (such as racing thoughts or psychomotor agitation) in the description of symptoms. An accurate diagnosis often requires clinical probing and clarification of symptoms (eg, differentiating simple insom­nia with consequent next-day fatigue from loss of the need for sleep with intact or even enhanced next-day energy) or discriminating racing thoughts from anxious ruminations that might be more intrusive than rapid.

Presentations of frank mania also can come to light as a consequence of symp­toms, rather than as symptoms per se (eg, conflict in relationships, problems at work, financial reversals).

Particularly in patients who do not have a history of mania, avoid the temptation to begin or modify existing pharmacother­apy until you have performed a basic ini­tial evaluation. Immediate considerations for initial assessment and management include the following:

Provide containment. Ensure a safe setting, level of care, and frequency of monitoring. Evaluate suicide risk (particularly when mixed features are present), and risk of with­drawal from any psychoactive substances.

Engage significant others. Close family members can provide essential history, particularly when a patient’s insight about her illness and need for treatment are impaired. Family members and significant others also often play important roles in helping to restrict access to finances, fos­tering medication adherence, preventing access to weapons in the home, and shar­ing information with providers about sub­stance use or high-risk behavior.

Systematically assess for DSM-5 symp­toms of mania and depression. DSM-5 modified criteria for mania/hypomania to necessitate increased energy, in addition to change in mood, to make a syndromal diag­nosis. Useful during a clinical interview is the popular mnemonic DIGFAST to aid recognition of core mania symptoms^a:
   • Distractibility
   • Indiscretion/impulsivity
   • Grandiosity
   • Flight of ideas
   • Activity increase
   • Sleep deficit
   • Talkativeness.

^aAlso see: “Mnemonics in a mnutshell: 32 aids to psychiatric diagnosis,” in the October 2008 issue Current Psychiatry and in the archive at CurrentPsychiatry.com.

These symptoms should represent a departure from normal baseline character­istics; it often is helpful to ask a significant other or collateral historian how the pres­ent symptoms differ from the patient’s usual state.

Assess for unstable medical conditions or toxicity states. When evaluating an acute change in mental status, toxicology screening is relatively standard and the absence of illicit substances should sel­dom, if ever, be taken for granted—espe­cially because occult substance use can lead to identification of false-positive BD “cases.”¹

Stop any antidepressant. During a manic episode, continuing antidepressant medi­cation serves no purpose other than to contribute to or exacerbate mania symp­toms. Nonetheless, observational studies demonstrate that approximately 15% of syndromally manic patients continue to receive an antidepressant, often when a clinician perceives more severe depression during mania, multiple prior depressive episodes, current anxiety, or rapid cycling.²

Importantly, antidepressants have been shown to harm, rather than alleviate, pre­sentations that involve a mixed state,³ and have no demonstrated value in preventing post-manic depression. Mere elimination of an antidepressant might ease symptoms during a manic or mixed episode.⁴

In some cases, it might be advisable to taper, not abruptly stop, a short half-life serotonergic antidepressant, even in the setting of mania, to minimize the potential for aggravating autonomic dysregulation that can result from antidepressant discon­tinuation effects. 

Begin anti-manic pharmacotherapy. Initiation of an anti-manic mood stabilizer, such as lithium and divalproex, has been standard in the treatment of acute mania.

In the 1990s, protocols for oral loading of divalproex (20 to 30 mg/kg/d) gained popularity for achieving more rapid symp­tom improvement than might occur with lithium. In the current era, atypical anti­psychotics have all but replaced mood sta­bilizers as an initial intervention to contain mania symptoms quickly (and with less risk than first-generation antipsychotics for acute adverse motor effects from so-called rapid neuroleptization).

Because atypical antipsychotics often rapidly subdue mania, psychosis, and agitation, regardless of the underlying process, many practitioners might feel more comfortable initiating them than a mood stabilizer when the diagnosis is ambiguous or provisional, although their longer-term efficacy and safety, relative to traditional mood stabilizers, remains contested. Considerations for choosing from among feasible anti-manic pharma­cotherapies are summarized in Table 1.⁵ 

Normalize the sleep-wake cycle. Chronobiological and circadian vari­ables, such as irregular sleep patterns, are thought to contribute to the pathophysiol­ogy of affective switch in BD. Behavioral and pharmacotherapeutic efforts to impose a normal sleep−wake schedule are considered fundamental to stabilizing acute mania.

Facilitate next steps after acute stabili­zation. For inpatients, this might involve step-down to a partial hospitalization or intensive outpatient program, alongside taking steps to ensure continued treatment adherence and minimize relapse.

What medical and neurologic workup is appropriate?
Not every first lifetime presentation of mania requires extensive medical and neurologic workup, particularly among patients who have a history of depression and those whose presentation neatly fits the demographic and clinical profile of newly emergent BD. Basic assessment should determine whether any new medication has been started that could plausibly contribute to abnormal mental sta­tus (Table 2).

Nevertheless, evaluation of almost all first presentations of mania should include:
   • urine toxicology screen 
   • complete blood count
   • comprehensive metabolic panel
   • thyroid-stimulating hormone assay
   • serum vitamin B₁₂ level assay
   • serum folic acid level assay
   • rapid plasma reagin test.

Clinical features that usually lead a cli­nician to pursue a more detailed medical and neurologic evaluation of first-episode mania include:
   • onset age >40
   • absence of a family history of mood disorder
   • symptoms arising during a major medical illness
   • multiple medications
   • suspicion of a degenerative or heredi­tary neurologic disorder
   • altered state of consciousness
   • signs of cortical or diffuse subcorti­cal dysfunction (eg, cognitive deficits, motor deficits, tremor)
   • abnormal vital signs.

Depending on the presentation, addi­tional testing might include:
   • tests of HIV antibody, immune auto­antibodies, and Lyme disease antibody
   • heavy metal screening (when sug­gested by environmental exposure)
   • lumbar puncture (eg, in a setting of manic delirium or suspected central nervous system infection or paraneoplastic syndrome)
   • neuroimaging (note: MRI provides bet­ter visualization than CT of white matter pathology and small vessel cerebrovascular disease) electroencephalography.

Making an overarching diagnosis: Is mania always bipolar disorder?
Mania is considered a manifestation of BD when symptoms cannot be attributed to another psychiatric condition, another underlying medical or neurologic condi­tion, or a toxic-metabolic state (Table 3 and Table 4^6-9). Classification of mania that occurs soon after antidepressant exposure in patients without a known history of BD continues to be the subject of debate, vary­ing in its conceptualization across editions of DSM.

The National Institute of Mental Health (NIMH) Systematic Treatment Enhancement Program for Bipolar Disorder, or STEP-BD, observed a fairly low (approximately 10%) incidence of switch from depression to mania when an antidepressant is added to a mood stabilizer; the study authors con­cluded that much of what is presumed to be antidepressant-induced mania might simply be the natural course of illness.¹⁰

Notably, several reports suggest that antidepressants might pose a greater risk of mood destabilization in people with BD I than with either BD II or other sus­pected variants on the bipolar spectrum.

DSM-5 advises that a diagnosis of substance-induced mood disorder appro­priately describes symptoms that spontane­ously dissipate once an antidepressant has been discontinued, whereas a diagnosis of BD can be made when manic or hypomanic symptoms persist at a syndromal level after an antidepressant has been stopped and its physiological effects are no longer present. With respect to time course, the International Society of Bipolar Disorders proposes that, beyond 12 to 16 weeks after an antidepressant has been started or the dosage has been increased, it is unlikely that new-onset mania/hypomania can rea­sonably be attributed to “triggering” by an antidepressant¹¹ (although antidepressants should be stopped when symptoms of mania emerge).

Several clinical features have been linked in the literature with an increased suscepti­bility to BD after an initial depressive epi­sode, including:
   • early (pre-adolescent) age at onset of first mood disorder episode⁶
   • family history of BD, highly recurrent depression, or psychosis^12,13
   • psychosis when depressed.^7,14

A number of other characteristics of depressive illness—including seasonal depression, atypical depressive features, suicidality, irritability, anxiety or sub­stance use comorbidity, postpartum mood episodes, and brief recurrent depressive episodes—have been described in the lit­erature as potential correlates of a bipolar diathesis; none have proved to be robust or pathognomonic of a BD diagnosis, as opposed to a unipolar diagnosis.

Data from the NIMH Collaborative Depression Study suggest that recurrent mania/hypomania after an antidepressant-associated polarity switch is greater when a family history of BD is present; other clinical variables might hold less predictive value.¹⁵

In addition, although some practitioners consider a history of nonresponse to trials of multiple antidepressants suggestive of an underlying bipolar process, polarity is only one of many variables that must be considered in the differential diagnosis of antidepressant-resistant depression.^b Likewise, molecular genetic studies do not support a link between antidepressant nonresponse and the likelihood of a diag­nosis of BD.¹⁶

^bSee “A practical approach to subtyping depression among your patients” in the April 2014 issue of Current Psychiatry or in the archive at CurrentPsychiatry.com.

Indefinite pharmacotherapy for bipolar disorder?
An important but nagging issue when diag­nosing BD after a first manic (or hypomanic) episode is the implied need for indefinite pharmacotherapy to sustain remission and prevent relapse and recurrence.

The likelihood of subsequent depression or mania/hypomania remains high after an index manic/hypomanic episode, par­ticularly for 6 to 8 months after recovery.^8,17 Natural history data suggest that, during the year that follows a first lifetime mania, approximately 40% of patients experience a second manic episode.⁸ A second lifetime mania might be especially likely in patients whose index episode involved mood-congruent psychosis, low premorbid work functioning, and an initial manic episode, as opposed to a mixed episode¹⁷ or early age at onset.⁸

In the absence of randomized, placebo-controlled studies of maintenance phar­macotherapy after a first lifetime manic episode, clinical judgment often drives decisions about the duration of continuing pharmacotherapy after initial symptoms resolve. The Texas Medication Algorithm Project for BD advises that:

Similarly, in the most recent (2004) Expert Consensus Guideline Series for the Treatment of Bipolar Disorder,¹⁹ 84% of practitioner−respondents favored indefi­nite mood stabilizer therapy after a second lifetime manic episode. No recommen­dation was made about the duration of maintenance pharmacotherapy after a first lifetime manic/hypomanic episode.

Avoid or reintroduce an antidepressant if depression recurs after a first mania?
Controversies surrounding antidepressant use in BD are extensive; detailed discus­sion is beyond the scope of this review (Goldberg and Ghaemi provided a broader discussion of risks and benefits of antide­pressants in BD20). Although the main clinical concern regarding antidepres­sant use was, at one time, the potential to induce mania or accelerate the frequency of recurrent episodes, more recent, empiri­cal studies suggest that the greater risk of using antidepressants for BD is lack of efficacy.^10,21

If a careful longitudinal history and clin­ical evaluation reveal that an initial manic episode heralds the onset of BD, decisions about whether to avoid an antidepressant (as opposed to using other, more evidence-based interventions for bipolar depres­sion) depend on a number of variables, including establishing whether the index episode was manic or hypomanic; ruling out current subthreshold mixed features; and clarifying how recently mania devel­oped. Decisions about future antidepres­sant use (or avoidance) might be less clear if an index manic/hypomanic episode was brief and self-limited once the antidepres­sant was stopped.

Although some experts eschew antidepressant monotherapy after such occurrences, there is no body of litera­ture to inform decisions about the safety or efficacy of undertaking a future antidepressant trial in such patients. That said, reasonable judgment probably includes several considerations:
   • Re-exposure to the same antidepres­sant that was associated with an induction of mania is likely riskier than choosing a different antidepressant; in general, purely serotonergic antidepressants or bupropion are considered to pose less risk of mood destabilization than is seen with an SNRI or tricyclic antidepressant.
   • After a manic episode, a subsequent antidepressant trial generally shouldn’t be attempted without concurrent anti-manic medication.
   • Introducing any antidepressant is probably ill-advised in the recent (~2 months) aftermath of acute manic/ hypomanic symptoms.²²
   • Patients and their significant other should be apprised of the risk of emerg­ing symptoms of mania or hypomania, or mixed features, and should be familiar with key target symptoms to watch for. Prospective mood charting can be helpful.
   • Patients should be monitored closely both for an exacerbation of depression and recurrence of mania/hypomania symptoms.
   • Any antidepressant should be discon­tinued promptly at the first sign of psy­chomotor acceleration or the emergence of mixed features, as defined by DSM-5.

Psychoeducation and forecasting
Functional recovery from a manic episode can lag behind symptomatic recovery. Subsyndromal symptoms often persist after a full episode subsides.

Mania often is followed by a depres­sive episode, and questions inevitably arise about how to prevent and treat these epi­sodes. Because the median duration of a manic episode is approximately 13 weeks,²³ it is crucial for patients and their immedi­ate family to recognize that recovery might be gradual, and that it will likely take time before she (he) can resume full-time respon­sibilities at work or school or in the home.

Today, a patient who is hospitalized for severe acute mania (as Ms. J was, in the case vignette) seldom remains an inpa­tient long enough to achieve remission of symptoms; sometimes, she (he) might con­tinue to manifest significant symptoms, even though decisions about the “medical necessity” of ongoing inpatient care tend to be governed mainly by issues of safety and imminent danger. (This web exclusive Table^20,24,25 provides considerations when making the transition from the acute phase to the continuation phase of treatment.)

To minimize risk of relapse, psycho-education should include discussion of:
   • psychiatrically deleterious effects of alcohol and illicit drug use
   • suicide risk, including what to do in an emergency
   • protecting a regular sleep schedule and avoiding sleep deprivation
   • the potential for poor medication adherence and management of side effects
   • the need for periodic laboratory moni­toring, as needed
   • the role of adjunctive psychotherapy and effective stress management
   • familiarity with symptoms that serve as warning signs, and how to monitor their onset.

Bottom Line
When a patient being treated for depression develops signs of mania or hypomania, stop any antidepressant and consider initiating a mood stabilizer, antipsychotic, or both, to contain and stabilize symptoms. Entertain medical and substance-related causes of mania symptoms, and evaluate and treat as suggested by the patient’s presentation. Long-term drug therapy to prevent recurrence of mania/hypomania, as well as risks and benefits of future exposure to antidepressants, should be decided case by case.

Related Resources
• Proudfoot J, Whitton A, Parker G, et al. Triggers of mania and depression in young adults with bipolar disorder. J Affect Disord. 2012;143(1-3):196-202.
• Stange JP, Sylvia LG, Magalhães PV, et al. Extreme at­tributions predict transition from depression to ma­nia or hypomania in bipolar disorder. J Psychiatr Res. 2013;47(10):1329-1336.

Drug Brand Names
Albuterol • Proventil, Ventolin
Anastrozole • Arimidex
Aripiprazole • Abilify
Bupropion • Wellbutrin
Carbamazepine • Tegretol
Chloroquine • Aralen
Ciprofloxacin • Cipro
Clarithromycin • Biaxin
Clomiphene • Clomid
Digoxin • Digox, Lanoxin
Divalproex • Depakote
5-Fluorouracil • Carac, Efudex
Human chorionic gonadotropin • Novarel, Pregnyl
Ifosfamide • Ifex
Isoniazid • Nydrazid
Lamotrigine • Lamictal
Letrozole • Femara
Lithium • Eskalith, Lithobid
Lurasidone • Latuda
Mefloquine • Lariam
Olanzapine • Zyprexa
Olanzapine/fluoxetine combination • Symbyax
ramipexole • Mirapex
Procarbazine • Matulane
Quetiapine • Seroquel
Ropinirole • Requip
Rotigotine • Neupro
Venlafaxine • Effexor
Zidovudine • Retrovir

Disclosures
Dr. Goldberg is a consultant to Merck & Co. and Sunovion. He is a member of the speakers’ bureau of AstraZeneca, Janssen, Merck & Co., Takeda and Lundbeck, and Sunovion.

Dr. Ernst reports no financial relationships with any company whose products are mentioned in this article or with manufacturers of competing products.

Attention-deficit/hyperactivity disorder (ADHD) is one of the most common pediatric psychiatric dis­orders, occurring in approximately 5% of children.¹ The disorder persists into adulthood in about one-half of those who are affected in childhood.² In adults and children, diagnosis continues to be based on the examiner’s subjec­tive assessment. (Box 1^3-9 describes how ADHD presents a compli­cated, moving target for the diagnostician.)

Patients who have ADHD are rarely studied with imag­ing; there are no established imaging findings associated with an ADHD diagnosis. Over the past 20 years, however, significant research has shown that molecular alterations along the dopaminergic−frontostriatal pathways occur in association with the behavioral constellation of ADHD symptoms—suggesting a pathophysiologic mechanism for this disorder.

In this article, we describe molecular findings from nuclear medicine imaging in ADHD. We also summarize imaging evidence for dysfunction of the dopaminergic-frontostriatal neural circuits as central in the pathophysiol­ogy of ADHD, with special focus on the dopamine reuptake transporter (DaT). Box 2^10,11 reviews our key observations and looks at the future of imaging in the management of ADHD.

Dopaminergic theory of ADHD
The executive functions that are disordered in ADHD (impulse control, judgment, main­taining attention) are thought to be centered in the infraorbital, dorsolateral, and medial frontal lobes. Neurotransmitters that have been implicated in the pathophysiology of ADHD include norepinephrine¹² and dopa­mine¹³; medications that selectively block reuptake of these neurotransmitters are used to treat ADHD.^14,15 Only the dopamine system has been extensively evaluated with molecular imaging techniques.

Because methylphenidate, a potent selec­tive dopamine reuptake inhibitor, has been shown to reduce disordered executive func­tional behaviors in ADHD, considerable imaging research has focused on the dopa­minergic neural circuits in the frontostriatal regions of the brain. The dopaminergic the­ory of ADHD is based on the hypothesis that alterations in the density or function of these circuits are responsible for behaviors that constitute ADHD.

Despite decades of efforts to delineate the underlying pathophysiology and neu­rochemistry of ADHD, no single unifying theory accounts for all imaging findings in all patients. This might be in part because of imprecision inherent in psychiatric diag­noses that are based on subjective observa­tions. The behavioral criteria for ADHD can manifest in several disorders. For example, anxiety-related symptoms seen in post­traumatic stress disorder, social anxiety disorder, and panic disorder also present as behaviors similar to those in ADHD diag­nostic criteria.

Molecular imaging might provide a window into the underlying pathophysiol­ogy of ADHD and, by identifying objective findings, (1) allow for patient stratification based on underlying physiologic subtypes, (2) refine diagnostic criteria, and (3) predict treatment response.

Nuclear medicine findings
In general, nuclear medicine investiga­tions of ADHD can be divided into studies of changes in regional cerebral blood flow (rCBF) or glucose metabolism (rCGM) and those that have assessed the concentration of synaptic structures, using highly specific radiolabeled ligands. Both kinds of studies provide limited anatomic resolution, unless co-registered with MRI or CT scans and either single photon emission computed tomography (SPECT) or positron emission tomography (PET).

Synaptic imaging using radiolabeled ligands with high biologic specificity for synaptic structures has high molecular resolution—that is, radiolabeled ligands used for selective imaging of the dopamine transporter or receptor do not identify sero­tonin transporters or receptors, and vice versa. (Details of SPECT and PET tech­niques are beyond the scope of this article but can be found in standard nuclear medi­cine textbooks.)

SPECT and PET of rCBF
Early investigations of rCBF in ADHD were performed using inhaled radioactive xenon-133 gas.¹⁶ Later, rCBF was assessed using fat-soluble radiolabeled ligands that rapidly distribute in the brain in proportion to blood flow by crossing the blood−brain barrier. Labeled with radioactive 99m-technetium, these ligands cross rapidly into brain cells after IV injection. Once intracellular, cova­lent bonds within the ligands cleave into 2 charged particles that do not easily recross the cell membrane. There is little redistribu­tion of tracer after initial uptake.

The imaging data set that results can be reconstructed as (1) surface images, on which defects indicate areas of reduced rCBF, or (2) tomographic slices on which color scales indicate relative rCBF values (Figure 1). Because of the minimal redistribution of the tracer, SPECT images obtained 1 or 2 hours after injection provide a snapshot of rCBF at the time tracer is injected. Patients can be injected under various conditions, such as at rest with eyes and ears open in a dimly lit, quiet room, and then under cognitive stress (Figure 2), such as performing a computer-based attention and impulse con­trol task, or during stimulant treatment.

Numerous investigators have found reduced frontal or striatal rCBF, or both, in patients with ADHD, unilaterally on the right¹⁷ or left,^18,19 or bilaterally.²⁰ Additionally, with stimulant therapy, normalization of striatal and frontal rCBF has been demon­strated^14,19—changes that correlate with reso­lution of behavioral symptoms of ADHD with stimulant treatment.²¹

SPECT of 32 boys with previously untreated ADHD. Kim et al²¹ found that the presence of reduced right or left, or both, frontal rCBF, which normal­ized with 8 weeks of stimulant therapy, predicted symptom improvement in 85% of patients. Absence of improve­ment of reduced frontal rCBF had a 75% negative predictive value for treatment response. (Additionally, hyperperfusion of the somatosensory cortex has been demonstrated in children with ADHD,^16,22 suggesting increased responsiveness to extraneous environmental input.)

SPECT of 40 untreated pediatric patients compared with 17 age-matched controls. Using SPECT, Lee et al²³ reported rCBF reductions in the orbitofrontal cortex and the medial temporal gyrus of participants; reductions corresponded to areas of motor and impulsivity control. The researchers also demonstrated increased rCBF in the somato­sensory area.

After methylphenidate treatment, blood flow to these areas normalized, and rCBF to higher visual and superior prefrontal areas decreased. Substantial clinical improve­ment occurred in 64% of patients—suggest­ing methylphenidate treatment of ADHD works by (1) increasing function of areas of the brain that control impulses, motor activ­ity, and attention, and (2) reducing function to sensory areas that lead to distraction by extraneous environmental sensory input.

O-15-labeled water PET of 10 adults with ADHD. Schweitzer et al²⁴ found that participants who demonstrated improve­ment in behavioral symptoms with chronic stimulant therapy had reduced rCBF in the striata at baseline—again, suggesting that baseline hypometabolism in the striata is associated with ADHD.

PET of regional cerebral glucose metabolism
Cerebral metabolism requires a constant supply of glucose; regional differences in cerebral glucose metabolism can be assessed directly with positron-emitting F-18-fluoro-2-deoxyglucose. Although metabolically inert, this agent is transported intracellularly simi­lar to glucose; once phosphorylated within brain cells, however, it can no longer undergo further metabolism or redistribution.

Studies using PET to assess rCGM were some of the earliest molecular imaging appli­cations in ADHD. Zametkin et al²⁵ reported low global cerebral glucose utilization in adults, but not adolescents,²⁶ with ADHD. However, further study, with normalization of the PET data, confirmed reduced rCGM in the left prefrontal cortex in both adolescents²⁶ and adults,²⁷ indicating hypometabolism of cortical areas associated with impulse con­trol and attention in ADHD. In adolescents, symptom severity was inversely related to rCBF in the left anterior frontal cortex.

Synaptic imaging
Nuclear imaging has been used to study several components of the striatal dopami­nergic synapse, including:
   • dopamine substrates, using fluorine- 18-labeled dopa or carbon-11-labeled dopa
   • dopamine receptors, using carbon- 11-labeled raclopride or iodine-123 iodobenzamide
   • the tDaT, using iodine-123 ioflupane, 99m-technetium TRODAT, or carbon-11 cocaine (Figure 3).

All of these synaptic imaging agents were used mainly as research tools until 2011, when the FDA approved the SPECT imag­ing agent iodine-123 ioflupane (DaTscan) for clinical use in assessment of Parkinson’s disease.²⁸ This commercially available agent has high specificity for the DaT, with little background activity noted on SPECT imag­ing (Figure 4).

Dopamine transporter imaging
Because the site of action of methylpheni­date is the DaT, imaging this component of the striatal dopaminergic synapse has been an area of intense investigation in ADHD. Located almost exclusively in the striata, DaT reduces synaptic concentrations of dopamine by means of reuptake channels in the cell membrane.²⁹ By reversibly bind­ing to, and occupying sites on, the DaT, methylphenidate impedes dopamine reup­take, which results in increased availability of dopamine at the synapse.³⁰

By demonstrating an increase in stria­tal DaT density in patients with ADHD— first reported by Dougherty et al³¹ using iodine-123 altropane (a dopaminergic uptake inhibitor) in 6 adults with ADHD—inves­tigators have hypothesized that excessive expression of the DaT protein in the striata, which may result from genetic or environ­mental factors, is a central causative agent of ADHD.³² Subsequent studies, however, have yielded contradictory findings: Hesse et al,³³ using SPECT imaging, and Volkow et al,³⁴ using carbon-11 cocaine PET imaging, found reduced DaT density in, respectively, 9 and 26 patients with ADHD.

To clarify the role of DaT levels in the eti­ology of ADHD and to explain discrepant results, Fusar-Poli et al³⁵ performed a meta-analysis of 9 published papers that reported the results of DaT imaging in a total of 169 ADHD patients and 129 controls. They noted that these studies included 6 differ­ent imaging agents and protocols. Patients were stimulant therapy-naïve (n = 137) or drug-free (refrained from stimulant therapy for a time [n = 32]). The team found that the degree of elevation of the striatal DaT con­centration correlated with a history of stimu­lant exposure, and that the drug-naïve group had a reduced DaT level.

Fusar-Poli’s hypothesis? Elevated DaT lev­els result from up-regulation in the presence of chronic methylphenidate therapy, which accounts for early reports that demonstrated increased striatal DaT density. Clinically, up-regulation might explain the lack of sustained relief of behavioral symptoms with stimulant therapy in 20% of patients with ADHD who showed clinical improvement initially.³⁶

Only limited conclusions can be drawn about the role of DaT levels in ADHD, given the small number of patients studied in pub­lished reports. In addition, the Fusar-Poli meta-analysis has come under strong criti­cism because of methodological errors with improper patient inclusion and characteriza­tion of treatment status,³⁷ calling into question the investigators’ conclusions.

Does the DaT level hold promise for practice? Despite a lack of clarity about the significance of DaT level in the etiology of ADHD, knowledge of a patient’s level might prove useful in predicting which patients will respond to methylphenidate. Namely, several researchers have found that:
   • an elevated baseline level of DaT (before stimulant therapy) correlates with robust clinical response
   • absence of an elevated baseline DaT level suggests that symptomatic improvement with stimulant therapy in unlikely.^38-40

Dresel et al³⁸ evaluated 17 drug-naïve adults, newly diagnosed with ADHD, using 99m-technetium TRODAT SPECT before and after methylphenidate therapy. They found a 15% increase in specific DaT binding in patients with ADHD, compared with con­trols, at baseline. After treatment, the research­ers observed a 28% reduction in specific DaT binding—a significant change from baseline that correlated with behavioral response.

Study: SPECT in 18 adults with ADHD given methylphenidate. Krause³⁹ used the same SPECT agent to study 18 adults before they received methylphenidate and 10 weeks after treatment. Participants were categorized as responders or nonresponders based on clinical assessment of ADHD symptoms after those 10 weeks. All 12 responders had an elevated striatal DaT concentration at baseline. Of the 6 nonresponders, 5 had a normal level of striatal DaT compared with age-matched controls.

Study: 22 Adult ADHD patients evalu­ated with 99m-technetium TRODAT SPECT. The same group of investigators⁴⁰ presented imaging findings in 22 additional adult patients. Seventeen had an elevated striatal DaT level, 16 of whom responded to stimulant therapy. The remaining 5 patients had reduced striatal DaT at base­line; none had a good clinical response to methylphenidate.

The positive clinical response to methylphenidate in 67%³⁷ and 77%⁴⁰ of patients is in good agreement with results from larger studies, which reported that approximately 75% of patients with ADHD show prompt clinical improvement with stimulants.⁴¹ Improvement might be related to an increase in functioning of the frontostriatal dopami­nergic circuit that is seen with stimulant therapy. Increased availability of dopamine at the synapse, resulting from stimulant blockade of the dopamine reuptake trans­porter, produces increased dopamine neu­rotransmission and increased activation of frontostriatal circuits.

In another study, rCBF in frontostriatal circuits was determined to be inversely proportional to DaT density; rCBF normalized with stimulant therapy.⁴²

Will imaging pave the way for therapeu­tic stratification? Baseline determinations of striatal DaT concentration with SPECT imaging might make it possible to stratify patients with ADHD symptoms into those likely to show significant behavioral symp­tom response to methylphenidate and those who are not likely to respond. There might be an objective imaging finding—striatal DaT density—that allows clinicians to dis­tinguish stimulant-responsive ADHD from stimulant-unresponsive ADHD.

Dopamine substrate imaging
Radiolabeled dopa (carbon-11 or fluorine-18) is transported into presynaptic dopaminer­gic neurons in the striatum, where it is decar­boxylated, converted to radio-dopamine, and stored within vesicles until released in response to neuronal excitation. Semi-quantitative assessment is achieved with calculation of specific (striatal) to nonspecific (background) uptake ratios. Increased val­ues are thought to indicate increased density of dopaminergic neurons.⁴³

Ernst et al⁴⁴ reported a 50% decrease in specific fluorine-18 dopa uptake in the left prefrontal cortex in 17 drug-naïve adults with ADHD, compared with 23 controls. The same team reported increased midbrain fluorine-18 dopa levels in 10 adolescents with ADHD—48% higher, overall, than what was seen in 10 controls.⁴³ They hypoth­esized that these opposite results were the results of a reduction in the dopaminergic neuronal density in adults, which might be part of the natural history of ADHD, or a normal age-related reduction in neuronal density, or both. Increased dopa levels in the team’s adolescent group were hypothesized to reflect up-regulation in dopamine syn­thesis due to low synaptic dopamine con­centrations that might result from increased dopamine reuptake.

Dopamine-receptor imaging
The 5 distinct dopamine receptors (D1, D2, D3, D4, and D5) can be grouped into 2 subtypes, based on their coupling with G proteins. D1 and D5 constitute a group; D2, D3, and D4, a second group.

The D1 receptor is the most common dopamine receptor in the brain and is widely distributed in the striatum and pre­frontal cerebral cortex. D1 receptor knock­out mice demonstrate hyperactivity and poorer performance on learning tasks and are used as an animal model for ADHD.⁴⁵ D1 has been imaged using C-11 SCH 23390 PET⁴⁶ in rats, but its role in ADHD has yet to be evaluated. D5 is the most recently cloned and most widely distributed of the known dopamine receptors; however, there are no imaging studies of the D5 receptor.¹³

D2 receptors are present in presynaptic and postsynaptic neurons⁴⁷ in the neocor­tex, substantia nigra, nucleus accumbens, and olfactory tubercle, as well as in other structures.⁴⁸ Presynaptic D2 receptors act as autoregulators, inhibiting dopaminergic synthesis, firing rate, and release.⁴⁹

Using C-11 raclopride PET imaging, Lou et al⁵⁰ reported high D2/3 receptor availabil­ity in adolescents who had a history of peri­natal cerebral ischemia. They found that this availability is associated with an increase in the severity of ADHD symptoms. They pro­posed that the increase in “empty” receptor density might have been caused by perinatal ischemia-induced presynaptic dopaminer­gic neuronal loss or an increase in presynap­tic dopamine reuptake (Figure 5⁵⁰). Either mechanism could result in up-regulation in postsynaptic D2/3 receptors.

Volkow et al⁵¹ reported that D2 receptor density correlated with methylphenidate-induced changes in rCBF in frontal and tem­poral lobes in humans. They postulated that the variable therapeutic effects of methyl­phenidate seen in ADHD patients might be related to variations in baseline D2 receptor availability.

Lou et al⁵⁰ reported elevated D2 recep­tor density, demonstrated using carbon-11 raclopride, in children with ADHD, com­pared with normal adults.

Further support for a relationship between D2-receptor density and symptom­atic improvement with methylphenidate in ADHD was presented by Ilgin et al⁵² using iodine-123 iodobenzamide SPECT. They found elevated D2 receptor levels in 9 drug-naïve children with ADHD, which is 20% to 60% above what is seen in unaffected chil­dren. They noted that these patients showed improvement in hyperactivity when treated with methylphenidate.

In a similar study of 20 drug-naïve adults, Volkow et al⁵³ found that durable symptom­atic improvement with methylphenidate therapy was associated with increased D2 receptor availability.

Summing up
Striatal DaT is the most likely synaptic tar­get for stratifying patients with ADHD, now that a dopamine transporter imaging agent is available commercially. Stratification might allow for refinement in the diagnostic categorization of ADHD, with introduction of stimulant-responsive and stimulant-unresponsive subtypes that are based on DaT imaging findings.

Bottom Line
Given recent advances showing molecular alterations in the dopaminergic-frontostriatal pathway as central to attention-deficit/hyperactivity disorder, molecular imaging might be useful as an objective study for diagnosis.

Related Resources
• Schweitzer JB, Lee DO, Hanford RB, et al. A positron emis­sion tomography study of methylphenidate in adults with ADHD: alterations in resting blood flow and predict­ing treatment response. Neuropsychopharmacology. 2003;28(5):967-973.
• Raz A. Brain imaging data of ADHD. Psychiatric Times. http://www.psychiatrictimes.com/adhd/brain-imaging-data-adhd.

Drug Brand Names
Iodine-123 ioflupane • Methylphenidate • Ritalin DaTscan

Acknowledgment
Kylee M. L. Unsdorfer, a medical student at Northeast Ohio Medical University, helped prepare the manuscript of this article.

Disclosures
Dr. Thacker reports no financial relationships with any company whose products are mentioned in this article or with manufacturers of competing products.
Dr. Binkovitz received 4 doses of ioflupane I123I (DaTscan) from General Electric for investigator-initiated research, used for animal imaging in 2012.

Attention-deficit/hyperactivity disorder (ADHD) is one of the most common pediatric psychiatric dis­orders, occurring in approximately 5% of children.¹ The disorder persists into adulthood in about one-half of those who are affected in childhood.² In adults and children, diagnosis continues to be based on the examiner’s subjec­tive assessment. (Box 1^3-9 describes how ADHD presents a compli­cated, moving target for the diagnostician.)

Patients who have ADHD are rarely studied with imag­ing; there are no established imaging findings associated with an ADHD diagnosis. Over the past 20 years, however, significant research has shown that molecular alterations along the dopaminergic−frontostriatal pathways occur in association with the behavioral constellation of ADHD symptoms—suggesting a pathophysiologic mechanism for this disorder.

In this article, we describe molecular findings from nuclear medicine imaging in ADHD. We also summarize imaging evidence for dysfunction of the dopaminergic-frontostriatal neural circuits as central in the pathophysiol­ogy of ADHD, with special focus on the dopamine reuptake transporter (DaT). Box 2^10,11 reviews our key observations and looks at the future of imaging in the management of ADHD.

Dopaminergic theory of ADHD
The executive functions that are disordered in ADHD (impulse control, judgment, main­taining attention) are thought to be centered in the infraorbital, dorsolateral, and medial frontal lobes. Neurotransmitters that have been implicated in the pathophysiology of ADHD include norepinephrine¹² and dopa­mine¹³; medications that selectively block reuptake of these neurotransmitters are used to treat ADHD.^14,15 Only the dopamine system has been extensively evaluated with molecular imaging techniques.

Because methylphenidate, a potent selec­tive dopamine reuptake inhibitor, has been shown to reduce disordered executive func­tional behaviors in ADHD, considerable imaging research has focused on the dopa­minergic neural circuits in the frontostriatal regions of the brain. The dopaminergic the­ory of ADHD is based on the hypothesis that alterations in the density or function of these circuits are responsible for behaviors that constitute ADHD.

Despite decades of efforts to delineate the underlying pathophysiology and neu­rochemistry of ADHD, no single unifying theory accounts for all imaging findings in all patients. This might be in part because of imprecision inherent in psychiatric diag­noses that are based on subjective observa­tions. The behavioral criteria for ADHD can manifest in several disorders. For example, anxiety-related symptoms seen in post­traumatic stress disorder, social anxiety disorder, and panic disorder also present as behaviors similar to those in ADHD diag­nostic criteria.

Molecular imaging might provide a window into the underlying pathophysiol­ogy of ADHD and, by identifying objective findings, (1) allow for patient stratification based on underlying physiologic subtypes, (2) refine diagnostic criteria, and (3) predict treatment response.

Nuclear medicine findings
In general, nuclear medicine investiga­tions of ADHD can be divided into studies of changes in regional cerebral blood flow (rCBF) or glucose metabolism (rCGM) and those that have assessed the concentration of synaptic structures, using highly specific radiolabeled ligands. Both kinds of studies provide limited anatomic resolution, unless co-registered with MRI or CT scans and either single photon emission computed tomography (SPECT) or positron emission tomography (PET).

Synaptic imaging using radiolabeled ligands with high biologic specificity for synaptic structures has high molecular resolution—that is, radiolabeled ligands used for selective imaging of the dopamine transporter or receptor do not identify sero­tonin transporters or receptors, and vice versa. (Details of SPECT and PET tech­niques are beyond the scope of this article but can be found in standard nuclear medi­cine textbooks.)

SPECT and PET of rCBF
Early investigations of rCBF in ADHD were performed using inhaled radioactive xenon-133 gas.¹⁶ Later, rCBF was assessed using fat-soluble radiolabeled ligands that rapidly distribute in the brain in proportion to blood flow by crossing the blood−brain barrier. Labeled with radioactive 99m-technetium, these ligands cross rapidly into brain cells after IV injection. Once intracellular, cova­lent bonds within the ligands cleave into 2 charged particles that do not easily recross the cell membrane. There is little redistribu­tion of tracer after initial uptake.

The imaging data set that results can be reconstructed as (1) surface images, on which defects indicate areas of reduced rCBF, or (2) tomographic slices on which color scales indicate relative rCBF values (Figure 1). Because of the minimal redistribution of the tracer, SPECT images obtained 1 or 2 hours after injection provide a snapshot of rCBF at the time tracer is injected. Patients can be injected under various conditions, such as at rest with eyes and ears open in a dimly lit, quiet room, and then under cognitive stress (Figure 2), such as performing a computer-based attention and impulse con­trol task, or during stimulant treatment.

Numerous investigators have found reduced frontal or striatal rCBF, or both, in patients with ADHD, unilaterally on the right¹⁷ or left,^18,19 or bilaterally.²⁰ Additionally, with stimulant therapy, normalization of striatal and frontal rCBF has been demon­strated^14,19—changes that correlate with reso­lution of behavioral symptoms of ADHD with stimulant treatment.²¹

SPECT of 32 boys with previously untreated ADHD. Kim et al²¹ found that the presence of reduced right or left, or both, frontal rCBF, which normal­ized with 8 weeks of stimulant therapy, predicted symptom improvement in 85% of patients. Absence of improve­ment of reduced frontal rCBF had a 75% negative predictive value for treatment response. (Additionally, hyperperfusion of the somatosensory cortex has been demonstrated in children with ADHD,^16,22 suggesting increased responsiveness to extraneous environmental input.)

SPECT of 40 untreated pediatric patients compared with 17 age-matched controls. Using SPECT, Lee et al²³ reported rCBF reductions in the orbitofrontal cortex and the medial temporal gyrus of participants; reductions corresponded to areas of motor and impulsivity control. The researchers also demonstrated increased rCBF in the somato­sensory area.

After methylphenidate treatment, blood flow to these areas normalized, and rCBF to higher visual and superior prefrontal areas decreased. Substantial clinical improve­ment occurred in 64% of patients—suggest­ing methylphenidate treatment of ADHD works by (1) increasing function of areas of the brain that control impulses, motor activ­ity, and attention, and (2) reducing function to sensory areas that lead to distraction by extraneous environmental sensory input.

O-15-labeled water PET of 10 adults with ADHD. Schweitzer et al²⁴ found that participants who demonstrated improve­ment in behavioral symptoms with chronic stimulant therapy had reduced rCBF in the striata at baseline—again, suggesting that baseline hypometabolism in the striata is associated with ADHD.

PET of regional cerebral glucose metabolism
Cerebral metabolism requires a constant supply of glucose; regional differences in cerebral glucose metabolism can be assessed directly with positron-emitting F-18-fluoro-2-deoxyglucose. Although metabolically inert, this agent is transported intracellularly simi­lar to glucose; once phosphorylated within brain cells, however, it can no longer undergo further metabolism or redistribution.

Studies using PET to assess rCGM were some of the earliest molecular imaging appli­cations in ADHD. Zametkin et al²⁵ reported low global cerebral glucose utilization in adults, but not adolescents,²⁶ with ADHD. However, further study, with normalization of the PET data, confirmed reduced rCGM in the left prefrontal cortex in both adolescents²⁶ and adults,²⁷ indicating hypometabolism of cortical areas associated with impulse con­trol and attention in ADHD. In adolescents, symptom severity was inversely related to rCBF in the left anterior frontal cortex.

Synaptic imaging
Nuclear imaging has been used to study several components of the striatal dopami­nergic synapse, including:
   • dopamine substrates, using fluorine- 18-labeled dopa or carbon-11-labeled dopa
   • dopamine receptors, using carbon- 11-labeled raclopride or iodine-123 iodobenzamide
   • the tDaT, using iodine-123 ioflupane, 99m-technetium TRODAT, or carbon-11 cocaine (Figure 3).

All of these synaptic imaging agents were used mainly as research tools until 2011, when the FDA approved the SPECT imag­ing agent iodine-123 ioflupane (DaTscan) for clinical use in assessment of Parkinson’s disease.²⁸ This commercially available agent has high specificity for the DaT, with little background activity noted on SPECT imag­ing (Figure 4).

Dopamine transporter imaging
Because the site of action of methylpheni­date is the DaT, imaging this component of the striatal dopaminergic synapse has been an area of intense investigation in ADHD. Located almost exclusively in the striata, DaT reduces synaptic concentrations of dopamine by means of reuptake channels in the cell membrane.²⁹ By reversibly bind­ing to, and occupying sites on, the DaT, methylphenidate impedes dopamine reup­take, which results in increased availability of dopamine at the synapse.³⁰

By demonstrating an increase in stria­tal DaT density in patients with ADHD— first reported by Dougherty et al³¹ using iodine-123 altropane (a dopaminergic uptake inhibitor) in 6 adults with ADHD—inves­tigators have hypothesized that excessive expression of the DaT protein in the striata, which may result from genetic or environ­mental factors, is a central causative agent of ADHD.³² Subsequent studies, however, have yielded contradictory findings: Hesse et al,³³ using SPECT imaging, and Volkow et al,³⁴ using carbon-11 cocaine PET imaging, found reduced DaT density in, respectively, 9 and 26 patients with ADHD.

To clarify the role of DaT levels in the eti­ology of ADHD and to explain discrepant results, Fusar-Poli et al³⁵ performed a meta-analysis of 9 published papers that reported the results of DaT imaging in a total of 169 ADHD patients and 129 controls. They noted that these studies included 6 differ­ent imaging agents and protocols. Patients were stimulant therapy-naïve (n = 137) or drug-free (refrained from stimulant therapy for a time [n = 32]). The team found that the degree of elevation of the striatal DaT con­centration correlated with a history of stimu­lant exposure, and that the drug-naïve group had a reduced DaT level.

Fusar-Poli’s hypothesis? Elevated DaT lev­els result from up-regulation in the presence of chronic methylphenidate therapy, which accounts for early reports that demonstrated increased striatal DaT density. Clinically, up-regulation might explain the lack of sustained relief of behavioral symptoms with stimulant therapy in 20% of patients with ADHD who showed clinical improvement initially.³⁶

Only limited conclusions can be drawn about the role of DaT levels in ADHD, given the small number of patients studied in pub­lished reports. In addition, the Fusar-Poli meta-analysis has come under strong criti­cism because of methodological errors with improper patient inclusion and characteriza­tion of treatment status,³⁷ calling into question the investigators’ conclusions.

Does the DaT level hold promise for practice? Despite a lack of clarity about the significance of DaT level in the etiology of ADHD, knowledge of a patient’s level might prove useful in predicting which patients will respond to methylphenidate. Namely, several researchers have found that:
   • an elevated baseline level of DaT (before stimulant therapy) correlates with robust clinical response
   • absence of an elevated baseline DaT level suggests that symptomatic improvement with stimulant therapy in unlikely.^38-40

Dresel et al³⁸ evaluated 17 drug-naïve adults, newly diagnosed with ADHD, using 99m-technetium TRODAT SPECT before and after methylphenidate therapy. They found a 15% increase in specific DaT binding in patients with ADHD, compared with con­trols, at baseline. After treatment, the research­ers observed a 28% reduction in specific DaT binding—a significant change from baseline that correlated with behavioral response.

Study: SPECT in 18 adults with ADHD given methylphenidate. Krause³⁹ used the same SPECT agent to study 18 adults before they received methylphenidate and 10 weeks after treatment. Participants were categorized as responders or nonresponders based on clinical assessment of ADHD symptoms after those 10 weeks. All 12 responders had an elevated striatal DaT concentration at baseline. Of the 6 nonresponders, 5 had a normal level of striatal DaT compared with age-matched controls.

Study: 22 Adult ADHD patients evalu­ated with 99m-technetium TRODAT SPECT. The same group of investigators⁴⁰ presented imaging findings in 22 additional adult patients. Seventeen had an elevated striatal DaT level, 16 of whom responded to stimulant therapy. The remaining 5 patients had reduced striatal DaT at base­line; none had a good clinical response to methylphenidate.

The positive clinical response to methylphenidate in 67%³⁷ and 77%⁴⁰ of patients is in good agreement with results from larger studies, which reported that approximately 75% of patients with ADHD show prompt clinical improvement with stimulants.⁴¹ Improvement might be related to an increase in functioning of the frontostriatal dopami­nergic circuit that is seen with stimulant therapy. Increased availability of dopamine at the synapse, resulting from stimulant blockade of the dopamine reuptake trans­porter, produces increased dopamine neu­rotransmission and increased activation of frontostriatal circuits.

In another study, rCBF in frontostriatal circuits was determined to be inversely proportional to DaT density; rCBF normalized with stimulant therapy.⁴²

Will imaging pave the way for therapeu­tic stratification? Baseline determinations of striatal DaT concentration with SPECT imaging might make it possible to stratify patients with ADHD symptoms into those likely to show significant behavioral symp­tom response to methylphenidate and those who are not likely to respond. There might be an objective imaging finding—striatal DaT density—that allows clinicians to dis­tinguish stimulant-responsive ADHD from stimulant-unresponsive ADHD.

Dopamine substrate imaging
Radiolabeled dopa (carbon-11 or fluorine-18) is transported into presynaptic dopaminer­gic neurons in the striatum, where it is decar­boxylated, converted to radio-dopamine, and stored within vesicles until released in response to neuronal excitation. Semi-quantitative assessment is achieved with calculation of specific (striatal) to nonspecific (background) uptake ratios. Increased val­ues are thought to indicate increased density of dopaminergic neurons.⁴³

Ernst et al⁴⁴ reported a 50% decrease in specific fluorine-18 dopa uptake in the left prefrontal cortex in 17 drug-naïve adults with ADHD, compared with 23 controls. The same team reported increased midbrain fluorine-18 dopa levels in 10 adolescents with ADHD—48% higher, overall, than what was seen in 10 controls.⁴³ They hypoth­esized that these opposite results were the results of a reduction in the dopaminergic neuronal density in adults, which might be part of the natural history of ADHD, or a normal age-related reduction in neuronal density, or both. Increased dopa levels in the team’s adolescent group were hypothesized to reflect up-regulation in dopamine syn­thesis due to low synaptic dopamine con­centrations that might result from increased dopamine reuptake.

Dopamine-receptor imaging
The 5 distinct dopamine receptors (D1, D2, D3, D4, and D5) can be grouped into 2 subtypes, based on their coupling with G proteins. D1 and D5 constitute a group; D2, D3, and D4, a second group.

The D1 receptor is the most common dopamine receptor in the brain and is widely distributed in the striatum and pre­frontal cerebral cortex. D1 receptor knock­out mice demonstrate hyperactivity and poorer performance on learning tasks and are used as an animal model for ADHD.⁴⁵ D1 has been imaged using C-11 SCH 23390 PET⁴⁶ in rats, but its role in ADHD has yet to be evaluated. D5 is the most recently cloned and most widely distributed of the known dopamine receptors; however, there are no imaging studies of the D5 receptor.¹³

D2 receptors are present in presynaptic and postsynaptic neurons⁴⁷ in the neocor­tex, substantia nigra, nucleus accumbens, and olfactory tubercle, as well as in other structures.⁴⁸ Presynaptic D2 receptors act as autoregulators, inhibiting dopaminergic synthesis, firing rate, and release.⁴⁹

Using C-11 raclopride PET imaging, Lou et al⁵⁰ reported high D2/3 receptor availabil­ity in adolescents who had a history of peri­natal cerebral ischemia. They found that this availability is associated with an increase in the severity of ADHD symptoms. They pro­posed that the increase in “empty” receptor density might have been caused by perinatal ischemia-induced presynaptic dopaminer­gic neuronal loss or an increase in presynap­tic dopamine reuptake (Figure 5⁵⁰). Either mechanism could result in up-regulation in postsynaptic D2/3 receptors.

Volkow et al⁵¹ reported that D2 receptor density correlated with methylphenidate-induced changes in rCBF in frontal and tem­poral lobes in humans. They postulated that the variable therapeutic effects of methyl­phenidate seen in ADHD patients might be related to variations in baseline D2 receptor availability.

Lou et al⁵⁰ reported elevated D2 recep­tor density, demonstrated using carbon-11 raclopride, in children with ADHD, com­pared with normal adults.

Further support for a relationship between D2-receptor density and symptom­atic improvement with methylphenidate in ADHD was presented by Ilgin et al⁵² using iodine-123 iodobenzamide SPECT. They found elevated D2 receptor levels in 9 drug-naïve children with ADHD, which is 20% to 60% above what is seen in unaffected chil­dren. They noted that these patients showed improvement in hyperactivity when treated with methylphenidate.

In a similar study of 20 drug-naïve adults, Volkow et al⁵³ found that durable symptom­atic improvement with methylphenidate therapy was associated with increased D2 receptor availability.

Summing up
Striatal DaT is the most likely synaptic tar­get for stratifying patients with ADHD, now that a dopamine transporter imaging agent is available commercially. Stratification might allow for refinement in the diagnostic categorization of ADHD, with introduction of stimulant-responsive and stimulant-unresponsive subtypes that are based on DaT imaging findings.

Bottom Line
Given recent advances showing molecular alterations in the dopaminergic-frontostriatal pathway as central to attention-deficit/hyperactivity disorder, molecular imaging might be useful as an objective study for diagnosis.

Related Resources
• Schweitzer JB, Lee DO, Hanford RB, et al. A positron emis­sion tomography study of methylphenidate in adults with ADHD: alterations in resting blood flow and predict­ing treatment response. Neuropsychopharmacology. 2003;28(5):967-973.
• Raz A. Brain imaging data of ADHD. Psychiatric Times. http://www.psychiatrictimes.com/adhd/brain-imaging-data-adhd.

Drug Brand Names
Iodine-123 ioflupane • Methylphenidate • Ritalin DaTscan

Acknowledgment
Kylee M. L. Unsdorfer, a medical student at Northeast Ohio Medical University, helped prepare the manuscript of this article.

Disclosures
Dr. Thacker reports no financial relationships with any company whose products are mentioned in this article or with manufacturers of competing products.
Dr. Binkovitz received 4 doses of ioflupane I123I (DaTscan) from General Electric for investigator-initiated research, used for animal imaging in 2012.

Attention-deficit/hyperactivity disorder (ADHD) is one of the most common pediatric psychiatric dis­orders, occurring in approximately 5% of children.¹ The disorder persists into adulthood in about one-half of those who are affected in childhood.² In adults and children, diagnosis continues to be based on the examiner’s subjec­tive assessment. (Box 1^3-9 describes how ADHD presents a compli­cated, moving target for the diagnostician.)

Patients who have ADHD are rarely studied with imag­ing; there are no established imaging findings associated with an ADHD diagnosis. Over the past 20 years, however, significant research has shown that molecular alterations along the dopaminergic−frontostriatal pathways occur in association with the behavioral constellation of ADHD symptoms—suggesting a pathophysiologic mechanism for this disorder.

In this article, we describe molecular findings from nuclear medicine imaging in ADHD. We also summarize imaging evidence for dysfunction of the dopaminergic-frontostriatal neural circuits as central in the pathophysiol­ogy of ADHD, with special focus on the dopamine reuptake transporter (DaT). Box 2^10,11 reviews our key observations and looks at the future of imaging in the management of ADHD.

Dopaminergic theory of ADHD
The executive functions that are disordered in ADHD (impulse control, judgment, main­taining attention) are thought to be centered in the infraorbital, dorsolateral, and medial frontal lobes. Neurotransmitters that have been implicated in the pathophysiology of ADHD include norepinephrine¹² and dopa­mine¹³; medications that selectively block reuptake of these neurotransmitters are used to treat ADHD.^14,15 Only the dopamine system has been extensively evaluated with molecular imaging techniques.

Because methylphenidate, a potent selec­tive dopamine reuptake inhibitor, has been shown to reduce disordered executive func­tional behaviors in ADHD, considerable imaging research has focused on the dopa­minergic neural circuits in the frontostriatal regions of the brain. The dopaminergic the­ory of ADHD is based on the hypothesis that alterations in the density or function of these circuits are responsible for behaviors that constitute ADHD.

Despite decades of efforts to delineate the underlying pathophysiology and neu­rochemistry of ADHD, no single unifying theory accounts for all imaging findings in all patients. This might be in part because of imprecision inherent in psychiatric diag­noses that are based on subjective observa­tions. The behavioral criteria for ADHD can manifest in several disorders. For example, anxiety-related symptoms seen in post­traumatic stress disorder, social anxiety disorder, and panic disorder also present as behaviors similar to those in ADHD diag­nostic criteria.

Molecular imaging might provide a window into the underlying pathophysiol­ogy of ADHD and, by identifying objective findings, (1) allow for patient stratification based on underlying physiologic subtypes, (2) refine diagnostic criteria, and (3) predict treatment response.

Nuclear medicine findings
In general, nuclear medicine investiga­tions of ADHD can be divided into studies of changes in regional cerebral blood flow (rCBF) or glucose metabolism (rCGM) and those that have assessed the concentration of synaptic structures, using highly specific radiolabeled ligands. Both kinds of studies provide limited anatomic resolution, unless co-registered with MRI or CT scans and either single photon emission computed tomography (SPECT) or positron emission tomography (PET).

Synaptic imaging using radiolabeled ligands with high biologic specificity for synaptic structures has high molecular resolution—that is, radiolabeled ligands used for selective imaging of the dopamine transporter or receptor do not identify sero­tonin transporters or receptors, and vice versa. (Details of SPECT and PET tech­niques are beyond the scope of this article but can be found in standard nuclear medi­cine textbooks.)

SPECT and PET of rCBF
Early investigations of rCBF in ADHD were performed using inhaled radioactive xenon-133 gas.¹⁶ Later, rCBF was assessed using fat-soluble radiolabeled ligands that rapidly distribute in the brain in proportion to blood flow by crossing the blood−brain barrier. Labeled with radioactive 99m-technetium, these ligands cross rapidly into brain cells after IV injection. Once intracellular, cova­lent bonds within the ligands cleave into 2 charged particles that do not easily recross the cell membrane. There is little redistribu­tion of tracer after initial uptake.

The imaging data set that results can be reconstructed as (1) surface images, on which defects indicate areas of reduced rCBF, or (2) tomographic slices on which color scales indicate relative rCBF values (Figure 1). Because of the minimal redistribution of the tracer, SPECT images obtained 1 or 2 hours after injection provide a snapshot of rCBF at the time tracer is injected. Patients can be injected under various conditions, such as at rest with eyes and ears open in a dimly lit, quiet room, and then under cognitive stress (Figure 2), such as performing a computer-based attention and impulse con­trol task, or during stimulant treatment.

Numerous investigators have found reduced frontal or striatal rCBF, or both, in patients with ADHD, unilaterally on the right¹⁷ or left,^18,19 or bilaterally.²⁰ Additionally, with stimulant therapy, normalization of striatal and frontal rCBF has been demon­strated^14,19—changes that correlate with reso­lution of behavioral symptoms of ADHD with stimulant treatment.²¹

SPECT of 32 boys with previously untreated ADHD. Kim et al²¹ found that the presence of reduced right or left, or both, frontal rCBF, which normal­ized with 8 weeks of stimulant therapy, predicted symptom improvement in 85% of patients. Absence of improve­ment of reduced frontal rCBF had a 75% negative predictive value for treatment response. (Additionally, hyperperfusion of the somatosensory cortex has been demonstrated in children with ADHD,^16,22 suggesting increased responsiveness to extraneous environmental input.)

SPECT of 40 untreated pediatric patients compared with 17 age-matched controls. Using SPECT, Lee et al²³ reported rCBF reductions in the orbitofrontal cortex and the medial temporal gyrus of participants; reductions corresponded to areas of motor and impulsivity control. The researchers also demonstrated increased rCBF in the somato­sensory area.

After methylphenidate treatment, blood flow to these areas normalized, and rCBF to higher visual and superior prefrontal areas decreased. Substantial clinical improve­ment occurred in 64% of patients—suggest­ing methylphenidate treatment of ADHD works by (1) increasing function of areas of the brain that control impulses, motor activ­ity, and attention, and (2) reducing function to sensory areas that lead to distraction by extraneous environmental sensory input.

O-15-labeled water PET of 10 adults with ADHD. Schweitzer et al²⁴ found that participants who demonstrated improve­ment in behavioral symptoms with chronic stimulant therapy had reduced rCBF in the striata at baseline—again, suggesting that baseline hypometabolism in the striata is associated with ADHD.

PET of regional cerebral glucose metabolism
Cerebral metabolism requires a constant supply of glucose; regional differences in cerebral glucose metabolism can be assessed directly with positron-emitting F-18-fluoro-2-deoxyglucose. Although metabolically inert, this agent is transported intracellularly simi­lar to glucose; once phosphorylated within brain cells, however, it can no longer undergo further metabolism or redistribution.

Studies using PET to assess rCGM were some of the earliest molecular imaging appli­cations in ADHD. Zametkin et al²⁵ reported low global cerebral glucose utilization in adults, but not adolescents,²⁶ with ADHD. However, further study, with normalization of the PET data, confirmed reduced rCGM in the left prefrontal cortex in both adolescents²⁶ and adults,²⁷ indicating hypometabolism of cortical areas associated with impulse con­trol and attention in ADHD. In adolescents, symptom severity was inversely related to rCBF in the left anterior frontal cortex.

Synaptic imaging
Nuclear imaging has been used to study several components of the striatal dopami­nergic synapse, including:
   • dopamine substrates, using fluorine- 18-labeled dopa or carbon-11-labeled dopa
   • dopamine receptors, using carbon- 11-labeled raclopride or iodine-123 iodobenzamide
   • the tDaT, using iodine-123 ioflupane, 99m-technetium TRODAT, or carbon-11 cocaine (Figure 3).

All of these synaptic imaging agents were used mainly as research tools until 2011, when the FDA approved the SPECT imag­ing agent iodine-123 ioflupane (DaTscan) for clinical use in assessment of Parkinson’s disease.²⁸ This commercially available agent has high specificity for the DaT, with little background activity noted on SPECT imag­ing (Figure 4).

Dopamine transporter imaging
Because the site of action of methylpheni­date is the DaT, imaging this component of the striatal dopaminergic synapse has been an area of intense investigation in ADHD. Located almost exclusively in the striata, DaT reduces synaptic concentrations of dopamine by means of reuptake channels in the cell membrane.²⁹ By reversibly bind­ing to, and occupying sites on, the DaT, methylphenidate impedes dopamine reup­take, which results in increased availability of dopamine at the synapse.³⁰

By demonstrating an increase in stria­tal DaT density in patients with ADHD— first reported by Dougherty et al³¹ using iodine-123 altropane (a dopaminergic uptake inhibitor) in 6 adults with ADHD—inves­tigators have hypothesized that excessive expression of the DaT protein in the striata, which may result from genetic or environ­mental factors, is a central causative agent of ADHD.³² Subsequent studies, however, have yielded contradictory findings: Hesse et al,³³ using SPECT imaging, and Volkow et al,³⁴ using carbon-11 cocaine PET imaging, found reduced DaT density in, respectively, 9 and 26 patients with ADHD.

To clarify the role of DaT levels in the eti­ology of ADHD and to explain discrepant results, Fusar-Poli et al³⁵ performed a meta-analysis of 9 published papers that reported the results of DaT imaging in a total of 169 ADHD patients and 129 controls. They noted that these studies included 6 differ­ent imaging agents and protocols. Patients were stimulant therapy-naïve (n = 137) or drug-free (refrained from stimulant therapy for a time [n = 32]). The team found that the degree of elevation of the striatal DaT con­centration correlated with a history of stimu­lant exposure, and that the drug-naïve group had a reduced DaT level.

Fusar-Poli’s hypothesis? Elevated DaT lev­els result from up-regulation in the presence of chronic methylphenidate therapy, which accounts for early reports that demonstrated increased striatal DaT density. Clinically, up-regulation might explain the lack of sustained relief of behavioral symptoms with stimulant therapy in 20% of patients with ADHD who showed clinical improvement initially.³⁶

Only limited conclusions can be drawn about the role of DaT levels in ADHD, given the small number of patients studied in pub­lished reports. In addition, the Fusar-Poli meta-analysis has come under strong criti­cism because of methodological errors with improper patient inclusion and characteriza­tion of treatment status,³⁷ calling into question the investigators’ conclusions.

Does the DaT level hold promise for practice? Despite a lack of clarity about the significance of DaT level in the etiology of ADHD, knowledge of a patient’s level might prove useful in predicting which patients will respond to methylphenidate. Namely, several researchers have found that:
   • an elevated baseline level of DaT (before stimulant therapy) correlates with robust clinical response
   • absence of an elevated baseline DaT level suggests that symptomatic improvement with stimulant therapy in unlikely.^38-40

Dresel et al³⁸ evaluated 17 drug-naïve adults, newly diagnosed with ADHD, using 99m-technetium TRODAT SPECT before and after methylphenidate therapy. They found a 15% increase in specific DaT binding in patients with ADHD, compared with con­trols, at baseline. After treatment, the research­ers observed a 28% reduction in specific DaT binding—a significant change from baseline that correlated with behavioral response.

Study: SPECT in 18 adults with ADHD given methylphenidate. Krause³⁹ used the same SPECT agent to study 18 adults before they received methylphenidate and 10 weeks after treatment. Participants were categorized as responders or nonresponders based on clinical assessment of ADHD symptoms after those 10 weeks. All 12 responders had an elevated striatal DaT concentration at baseline. Of the 6 nonresponders, 5 had a normal level of striatal DaT compared with age-matched controls.

Study: 22 Adult ADHD patients evalu­ated with 99m-technetium TRODAT SPECT. The same group of investigators⁴⁰ presented imaging findings in 22 additional adult patients. Seventeen had an elevated striatal DaT level, 16 of whom responded to stimulant therapy. The remaining 5 patients had reduced striatal DaT at base­line; none had a good clinical response to methylphenidate.

The positive clinical response to methylphenidate in 67%³⁷ and 77%⁴⁰ of patients is in good agreement with results from larger studies, which reported that approximately 75% of patients with ADHD show prompt clinical improvement with stimulants.⁴¹ Improvement might be related to an increase in functioning of the frontostriatal dopami­nergic circuit that is seen with stimulant therapy. Increased availability of dopamine at the synapse, resulting from stimulant blockade of the dopamine reuptake trans­porter, produces increased dopamine neu­rotransmission and increased activation of frontostriatal circuits.

In another study, rCBF in frontostriatal circuits was determined to be inversely proportional to DaT density; rCBF normalized with stimulant therapy.⁴²

Will imaging pave the way for therapeu­tic stratification? Baseline determinations of striatal DaT concentration with SPECT imaging might make it possible to stratify patients with ADHD symptoms into those likely to show significant behavioral symp­tom response to methylphenidate and those who are not likely to respond. There might be an objective imaging finding—striatal DaT density—that allows clinicians to dis­tinguish stimulant-responsive ADHD from stimulant-unresponsive ADHD.

Dopamine substrate imaging
Radiolabeled dopa (carbon-11 or fluorine-18) is transported into presynaptic dopaminer­gic neurons in the striatum, where it is decar­boxylated, converted to radio-dopamine, and stored within vesicles until released in response to neuronal excitation. Semi-quantitative assessment is achieved with calculation of specific (striatal) to nonspecific (background) uptake ratios. Increased val­ues are thought to indicate increased density of dopaminergic neurons.⁴³

Ernst et al⁴⁴ reported a 50% decrease in specific fluorine-18 dopa uptake in the left prefrontal cortex in 17 drug-naïve adults with ADHD, compared with 23 controls. The same team reported increased midbrain fluorine-18 dopa levels in 10 adolescents with ADHD—48% higher, overall, than what was seen in 10 controls.⁴³ They hypoth­esized that these opposite results were the results of a reduction in the dopaminergic neuronal density in adults, which might be part of the natural history of ADHD, or a normal age-related reduction in neuronal density, or both. Increased dopa levels in the team’s adolescent group were hypothesized to reflect up-regulation in dopamine syn­thesis due to low synaptic dopamine con­centrations that might result from increased dopamine reuptake.

Dopamine-receptor imaging
The 5 distinct dopamine receptors (D1, D2, D3, D4, and D5) can be grouped into 2 subtypes, based on their coupling with G proteins. D1 and D5 constitute a group; D2, D3, and D4, a second group.

The D1 receptor is the most common dopamine receptor in the brain and is widely distributed in the striatum and pre­frontal cerebral cortex. D1 receptor knock­out mice demonstrate hyperactivity and poorer performance on learning tasks and are used as an animal model for ADHD.⁴⁵ D1 has been imaged using C-11 SCH 23390 PET⁴⁶ in rats, but its role in ADHD has yet to be evaluated. D5 is the most recently cloned and most widely distributed of the known dopamine receptors; however, there are no imaging studies of the D5 receptor.¹³

D2 receptors are present in presynaptic and postsynaptic neurons⁴⁷ in the neocor­tex, substantia nigra, nucleus accumbens, and olfactory tubercle, as well as in other structures.⁴⁸ Presynaptic D2 receptors act as autoregulators, inhibiting dopaminergic synthesis, firing rate, and release.⁴⁹

Using C-11 raclopride PET imaging, Lou et al⁵⁰ reported high D2/3 receptor availabil­ity in adolescents who had a history of peri­natal cerebral ischemia. They found that this availability is associated with an increase in the severity of ADHD symptoms. They pro­posed that the increase in “empty” receptor density might have been caused by perinatal ischemia-induced presynaptic dopaminer­gic neuronal loss or an increase in presynap­tic dopamine reuptake (Figure 5⁵⁰). Either mechanism could result in up-regulation in postsynaptic D2/3 receptors.

Volkow et al⁵¹ reported that D2 receptor density correlated with methylphenidate-induced changes in rCBF in frontal and tem­poral lobes in humans. They postulated that the variable therapeutic effects of methyl­phenidate seen in ADHD patients might be related to variations in baseline D2 receptor availability.

Lou et al⁵⁰ reported elevated D2 recep­tor density, demonstrated using carbon-11 raclopride, in children with ADHD, com­pared with normal adults.

Further support for a relationship between D2-receptor density and symptom­atic improvement with methylphenidate in ADHD was presented by Ilgin et al⁵² using iodine-123 iodobenzamide SPECT. They found elevated D2 receptor levels in 9 drug-naïve children with ADHD, which is 20% to 60% above what is seen in unaffected chil­dren. They noted that these patients showed improvement in hyperactivity when treated with methylphenidate.

In a similar study of 20 drug-naïve adults, Volkow et al⁵³ found that durable symptom­atic improvement with methylphenidate therapy was associated with increased D2 receptor availability.

Summing up
Striatal DaT is the most likely synaptic tar­get for stratifying patients with ADHD, now that a dopamine transporter imaging agent is available commercially. Stratification might allow for refinement in the diagnostic categorization of ADHD, with introduction of stimulant-responsive and stimulant-unresponsive subtypes that are based on DaT imaging findings.

Bottom Line
Given recent advances showing molecular alterations in the dopaminergic-frontostriatal pathway as central to attention-deficit/hyperactivity disorder, molecular imaging might be useful as an objective study for diagnosis.

Related Resources
• Schweitzer JB, Lee DO, Hanford RB, et al. A positron emis­sion tomography study of methylphenidate in adults with ADHD: alterations in resting blood flow and predict­ing treatment response. Neuropsychopharmacology. 2003;28(5):967-973.
• Raz A. Brain imaging data of ADHD. Psychiatric Times. http://www.psychiatrictimes.com/adhd/brain-imaging-data-adhd.

Drug Brand Names
Iodine-123 ioflupane • Methylphenidate • Ritalin DaTscan

Acknowledgment
Kylee M. L. Unsdorfer, a medical student at Northeast Ohio Medical University, helped prepare the manuscript of this article.

Disclosures
Dr. Thacker reports no financial relationships with any company whose products are mentioned in this article or with manufacturers of competing products.
Dr. Binkovitz received 4 doses of ioflupane I123I (DaTscan) from General Electric for investigator-initiated research, used for animal imaging in 2012.

Many college students suffer from mental illness (Table 1),¹ which can have a negative impact on academic performance. Although psychotropic medications are an important part of treat­ment for many college students, the potential for misuse always is present. Drug misuse occurs when patients use medications for reasons inconsistent with legal or medical guidelines.² For example, patients may take a medica­tion that has not been prescribed for them or in a manner that is inconsis­tent with the prescriber’s instructions, including administration with other substances.³

Misuse of psychotropic drugs is prevalent among college students. A study of 14,175 students from 26 campuses reported that 14.7% of students taking a psychotropic are doing so without a prescription, including stimu­lants (52.6%), anxiolytics (38.4%), and antidepressants (17.4%).⁴ Another study states that more than one-third of responders reported misuse of >1 class of medication.⁵

Psychotropic misuse is concerning because it increases the risk of adverse events. Nearly one-half of medication errors are associated with writing and dispensing the prescription, which means that prescribers can work to reduce these errors.⁶ However, nonadherence, prescription misuse, and failure to disclose use of over-the-counter drugs, illicit drugs, and herbal products makes preventing most adverse events difficult, if not impossible, for prescribers.^7,8

Psychotropic drug misuse among college students is highly variable and unpredictable. Students misuse medications, including stimulants, ben­zodiazepines, and antidepressants, for a variety of reasons, such as study enhancement, experimentation, intoxication, self-medication, relaxation, and stress management.⁸ One survey reported that >70% of students taking a psychotropic medication took it with alcohol or another illicit drug.⁹

However, <20% of those using a psy­chotropic medication with alcohol or other illicit drugs told their health care provider(s),⁹ making it impossible for cli­nicians to predict a patient’s risk of drug− drug interactions and subsequent adverse events. Additionally, additive effects could occur¹⁰ and changes in a patient’s presenta­tion could be caused by a reaction to a com­bination of medications, rather than a new symptom of mental illness.

This article will examine common issues associated with drug misuse among col­lege-age students and review prevention strategies (Table 2).

Stimulants
Stimulants have the highest rate of diver­sion; 61.7% of college students prescribed stimulants have shared or sold their medi­cation.¹¹ A survey of 115 students from 2 universities reported that the most common reason for stimulant misuse was to enhance academic performance.¹² The same survey showed that some students take stimulants with Cannabis (17%) and alcohol (30%).¹² As a result, in addition to lowering grade point average (GPA) and other academic difficul­ties,¹³ students misusing stimulants are at risk of drug interactions.¹⁴

It is critical to ascertain the route of drug administration, because non-oral routes, including crushing then snorting or inject­ing, are associated with additional health concerns, such as accidental death or blood-borne illnesses.^15,16 Cardiac adverse effects of stimulants include hypertension, vasospasm, tachycardia, and dysrhythmia; psychiatric and other effects include sero­tonin syndrome, hallucinations, anxiety, paranoia, seizures, tics, hyperthermia, and tremor.¹⁷ Health care providers prescrib­ing or caring for people taking a stimulant should monitor for these potential effects.

The risk of switch to mania might not be apparent to those who prescribe stimulants or to young people who take non-prescribed stimulants for academic enhancement or to achieve medication-induced euphoria. Adolescent stimulant use is associated with symptoms of early-onset bipolar disorder in patients who have attention-deficit/ hyperactivity disorder (ADHD) and undi­agnosed bipolarity.¹⁸

The cardiovascular risk associated with stimulant use is debatable. Although sev­eral studies have been conducted,^19-21 meth­odological factors limit their applicability. To minimize potential risks, several precau­tions should be taken before prescribing a stimulant to treat ADHD.

First, obtain a detailed personal and family medical history, asking about pos­sible cardiovascular disease. Second, care­fully scrutinize the patient’s cardiovascular system during the physical exam. Third, consider additional testing, such as an elec­trocardiogram, if the patient’s history or physical exam indicates possible risk.²²

As a prescriber, you should be aware of the prevalence of stimulant use among students with and without ADHD, includ­ing those who could be feigning ADHD symptoms.¹⁵ Diversion could occur through sharing medications or selling them to friends and family.¹¹ It also is possible that these medications may be used with other illicit substances, such as Cannabis, ecstasy, cocaine, and opiates.²³ Students also could misuse stimulants by taking more than the prescribed dosage.²⁴

Risk factors for misuse of stimulants include: heavy alcohol use, previous illicit drug use, white race, fraternity or sorority membership, low GPA, increased hyperac­tivity symptoms, and attendance at a com­petitive college or university.^25-27

Benzodiazepines
Misuse of benzodiazepine is a significant component of prescription drug abuse and often occurs with other medications and alcohol.²⁸ Additional methods of mis­use include increased dosage and non-oral routes of administration.²⁹

A 2001 national survey reported that 7.8% of college students have misused benzodiaz­epines.²³ Common characteristics of benzo­diazepine abusers include young age, male sex, personality characteristics of impulsiv­ity and hopelessness, and abuse of other drugs, including cocaine and methadone.^28,29

Benzodiazepines are prescribed for their anxiolytic and hypnotic properties and stu­dents could use these drugs with other agents to augment the euphoric effects or diminish withdrawal symptoms.³⁰ Patients taking ben­zodiazepines for anxiety might self-medicate with alcohol, which increases sedation and depression, and can contribute to the risk for respiratory depression.¹⁰ Misuse of benzodi­azepines can result in cognitive and psycho­motor impairment and increase the risk of accidents and overdose.^29,31

Although overdose with monotherapy is rare, the risk increases when a benzodiaz­epine is used with alcohol¹⁰ or another respi­ratory depressants, such as opioids, because combination use can produce additive effects.²⁸ You should therefore avoid prescrib­ing benzodiazepines to patients who have a history of significant substance abuse and consider using alternative, non-addictive agents, such as selective serotonin reuptake inhibitors, or non-pharmaceutical treatment when such patients present with an anxiety disorder. The risk of adverse effects of ben­zodiazepines can be reduced by limiting the dosing and the duration of the treatment, and by using longer-acting rather than the more addictive, shorter-acting, agents.

Antidepressants
Health care providers should be aware that, despite the relative absence of physically addictive properties, antidepressants from most classes are abusable agents sought by young people for non-medical use. In particular, the literature highlights mono­amine oxidase inhibitors (MAOIs), tricyclic antidepressants, serotonin-norepinephrine reuptake inhibitors, and bupropion as the antidepressants most likely to be misused for their amphetamine-like euphoric effects or serotonin-induced dissociative effects.³² However, compared with other drug classes discussed in this article, the rate of antide­pressant misuse is relatively low among col­lege students.

Regardless of the antidepressant selected, clinicians should be concerned about alcohol use among college-age patients. Persons with depression are at increased risk of alcohol­ism compared with the general population.³³ This combination can increase depressive symptoms and sedation, and decrease coor­dination, judgment, and reaction time.³³

Excessive alcohol use can increase the risk of seizures in patients taking antidepressants such as buproprion.³⁴ Employ caution when prescribing bupropion to patients who have a predisposing clini­cal factor that increases seizure risk, such as excessive alcohol use and abrupt ces­sation, use of other medications that may lower seizure threshold (eg, theophylline, amphetamines, phenothiazines), and a his­tory of head trauma.³⁴

To minimize the risk of seizures with bupropion, titrate up the dosage slowly. Furthermore, using a low dosage during dual therapy for antidepressant augmenta­tion further decreases the risk of seizure.³⁵ For these reasons, we recommend that you avoid bupropion in patients who are at risk of binge drinking, and give careful consid­eration to providing alternative therapies for them.

Prescribers and patients should also keep in mind that hypertensive crisis could occur if MAOIs are combined with certain types of alcoholic beverages containing tyramine, including some wines and draft beer.³³

How you can identify and prevent misuse
Careful communication between health care provider and patient that is necessary to minimize the risk of adverse drug events with psychotropic medications often is lack­ing. For example, 24% of study college-age participants did not remember if their physi­cian provided a diagnosis and 28.8% could not recall being informed about side effects and, perhaps as a result, many students did not take their medications as prescribed.⁹

Further, prescribers should ask college-age patients who are undergoing stimulant treatment if they believe that they are being adequately treated. They should inquire about how they are taking their medica­tions.¹¹ These questions can lead to discus­sion of the need for these medications and reevaluation of their perceived indication.¹¹

Remind patients to take their medication only as directed.³⁶ Highlight the need to:
   • store medications in a discreet location
   • properly dispose of unused medications
   • keep tabs on the quantity of pills
   • know how to resist requests for diver­sion from peers.

The Substance Abuse and Mental Health Services Administration offers additional useful strategies,³⁷ and pharmacists also can be partners in substance use education and prevention.³⁸ These are examples of how health care providers can take an active role in providing patients with a thorough and detailed understanding of (1) their condi­tions and (2) their prescribed medications to improve efficacy and safety while pre­venting misuse.⁸

A study found that the most common method of obtaining these medications without a prescription is acquiring them from peers; 54% of undergraduate patients with stimulant prescriptions have been approached by peers to give, trade, or sell their drugs.²⁵ Other methods include pur­chasing medications online or faking pre­scriptions.³⁹ Health care providers should remind patients of the legal ramifications of sharing or selling their prescribed medi­cations. Finally, providers must be vigilant for students who may feign symptoms to obtain a prescription:
   • be wary if symptom presentation sounds too “textbook”
   • seek collateral history from family. Adults with ADHD should have shown symptoms during childhood
   • use external verification such as neuropsychological testing for ADHD. A neuropsychologist can detect deception by analyzing the pattern of responses to questions.

Patient assessment is a key step to in pre­venting abuse of psychotropic medications. Gentle inquiry about school-related stress and other risk factors for misuse can help practitioners determine if students are at risk of diversion and if additional screening is necessary.

In response to these issues, Stone and Merlo⁸ have suggested that, in addition to the educational programs held on col­lege campuses on alcohol, illicit drugs, and prescription painkillers, patients should be better informed on the appropriate use of prescription psychiatric medications, instructed to avoid sharing with family and friends, and assessed for abuse risk at regu­lar intervals.

To further protect patients from adverse outcomes during treatment, you can employ conservative and safe prescribing techniques. One strategy might be to keep a personal formulary that lists key medica­tions you use in everyday practice, includ­ing knowledge about each drug’s dosage, potential adverse effects, key warnings, and drug−drug interactions.⁴⁰

Furthermore, maintain healthy caution about newly approved medications and carefully consider how they measure up to existing agents—in other words, prac­tice evidence-based medicine, particu­larly when students request a particular agent.^40,41 Prescribers should evaluate the risk of abuse before prescribing and attempt to prevent misuse by limiting quantities and minimizing polypharmacy.

Last, pharmacists can be key allies for consultation and appropriate medication selection.

 
Bottom Line
Psychotropic medications are necessary to treat the variety of conditions—anxiety, attention-deficit/hyperactivity disorder, depression, and panic disorder—common among college students. However, students are at risk of combining their prescribed medications with other medications, drugs, and alcohol or could sell or share their medication with peers. Proper counseling and identification of risk factors can be important tools for preventing such events.

Related Resources
• American College Health Association-National College Health Assessment. www.acha-ncha.org.
• Schwartz VI. College mental health: How to provide care for students in need. Current Psychiatry. 2011;10(12):22-29.

Drug Brand Names
Bupropion • Wellbutrin, Zyban
Methadone • Methadose, Dolophine
Theophylline • Theo-24, Theolair, Uniphyl

Disclosures
The authors report no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.

Many college students suffer from mental illness (Table 1),¹ which can have a negative impact on academic performance. Although psychotropic medications are an important part of treat­ment for many college students, the potential for misuse always is present. Drug misuse occurs when patients use medications for reasons inconsistent with legal or medical guidelines.² For example, patients may take a medica­tion that has not been prescribed for them or in a manner that is inconsis­tent with the prescriber’s instructions, including administration with other substances.³

Misuse of psychotropic drugs is prevalent among college students. A study of 14,175 students from 26 campuses reported that 14.7% of students taking a psychotropic are doing so without a prescription, including stimu­lants (52.6%), anxiolytics (38.4%), and antidepressants (17.4%).⁴ Another study states that more than one-third of responders reported misuse of >1 class of medication.⁵

Psychotropic misuse is concerning because it increases the risk of adverse events. Nearly one-half of medication errors are associated with writing and dispensing the prescription, which means that prescribers can work to reduce these errors.⁶ However, nonadherence, prescription misuse, and failure to disclose use of over-the-counter drugs, illicit drugs, and herbal products makes preventing most adverse events difficult, if not impossible, for prescribers.^7,8

Psychotropic drug misuse among college students is highly variable and unpredictable. Students misuse medications, including stimulants, ben­zodiazepines, and antidepressants, for a variety of reasons, such as study enhancement, experimentation, intoxication, self-medication, relaxation, and stress management.⁸ One survey reported that >70% of students taking a psychotropic medication took it with alcohol or another illicit drug.⁹

However, <20% of those using a psy­chotropic medication with alcohol or other illicit drugs told their health care provider(s),⁹ making it impossible for cli­nicians to predict a patient’s risk of drug− drug interactions and subsequent adverse events. Additionally, additive effects could occur¹⁰ and changes in a patient’s presenta­tion could be caused by a reaction to a com­bination of medications, rather than a new symptom of mental illness.

This article will examine common issues associated with drug misuse among col­lege-age students and review prevention strategies (Table 2).

Stimulants
Stimulants have the highest rate of diver­sion; 61.7% of college students prescribed stimulants have shared or sold their medi­cation.¹¹ A survey of 115 students from 2 universities reported that the most common reason for stimulant misuse was to enhance academic performance.¹² The same survey showed that some students take stimulants with Cannabis (17%) and alcohol (30%).¹² As a result, in addition to lowering grade point average (GPA) and other academic difficul­ties,¹³ students misusing stimulants are at risk of drug interactions.¹⁴

It is critical to ascertain the route of drug administration, because non-oral routes, including crushing then snorting or inject­ing, are associated with additional health concerns, such as accidental death or blood-borne illnesses.^15,16 Cardiac adverse effects of stimulants include hypertension, vasospasm, tachycardia, and dysrhythmia; psychiatric and other effects include sero­tonin syndrome, hallucinations, anxiety, paranoia, seizures, tics, hyperthermia, and tremor.¹⁷ Health care providers prescrib­ing or caring for people taking a stimulant should monitor for these potential effects.

The risk of switch to mania might not be apparent to those who prescribe stimulants or to young people who take non-prescribed stimulants for academic enhancement or to achieve medication-induced euphoria. Adolescent stimulant use is associated with symptoms of early-onset bipolar disorder in patients who have attention-deficit/ hyperactivity disorder (ADHD) and undi­agnosed bipolarity.¹⁸

The cardiovascular risk associated with stimulant use is debatable. Although sev­eral studies have been conducted,^19-21 meth­odological factors limit their applicability. To minimize potential risks, several precau­tions should be taken before prescribing a stimulant to treat ADHD.

First, obtain a detailed personal and family medical history, asking about pos­sible cardiovascular disease. Second, care­fully scrutinize the patient’s cardiovascular system during the physical exam. Third, consider additional testing, such as an elec­trocardiogram, if the patient’s history or physical exam indicates possible risk.²²

As a prescriber, you should be aware of the prevalence of stimulant use among students with and without ADHD, includ­ing those who could be feigning ADHD symptoms.¹⁵ Diversion could occur through sharing medications or selling them to friends and family.¹¹ It also is possible that these medications may be used with other illicit substances, such as Cannabis, ecstasy, cocaine, and opiates.²³ Students also could misuse stimulants by taking more than the prescribed dosage.²⁴

Risk factors for misuse of stimulants include: heavy alcohol use, previous illicit drug use, white race, fraternity or sorority membership, low GPA, increased hyperac­tivity symptoms, and attendance at a com­petitive college or university.^25-27

Benzodiazepines
Misuse of benzodiazepine is a significant component of prescription drug abuse and often occurs with other medications and alcohol.²⁸ Additional methods of mis­use include increased dosage and non-oral routes of administration.²⁹

A 2001 national survey reported that 7.8% of college students have misused benzodiaz­epines.²³ Common characteristics of benzo­diazepine abusers include young age, male sex, personality characteristics of impulsiv­ity and hopelessness, and abuse of other drugs, including cocaine and methadone.^28,29

Benzodiazepines are prescribed for their anxiolytic and hypnotic properties and stu­dents could use these drugs with other agents to augment the euphoric effects or diminish withdrawal symptoms.³⁰ Patients taking ben­zodiazepines for anxiety might self-medicate with alcohol, which increases sedation and depression, and can contribute to the risk for respiratory depression.¹⁰ Misuse of benzodi­azepines can result in cognitive and psycho­motor impairment and increase the risk of accidents and overdose.^29,31

Although overdose with monotherapy is rare, the risk increases when a benzodiaz­epine is used with alcohol¹⁰ or another respi­ratory depressants, such as opioids, because combination use can produce additive effects.²⁸ You should therefore avoid prescrib­ing benzodiazepines to patients who have a history of significant substance abuse and consider using alternative, non-addictive agents, such as selective serotonin reuptake inhibitors, or non-pharmaceutical treatment when such patients present with an anxiety disorder. The risk of adverse effects of ben­zodiazepines can be reduced by limiting the dosing and the duration of the treatment, and by using longer-acting rather than the more addictive, shorter-acting, agents.

Antidepressants
Health care providers should be aware that, despite the relative absence of physically addictive properties, antidepressants from most classes are abusable agents sought by young people for non-medical use. In particular, the literature highlights mono­amine oxidase inhibitors (MAOIs), tricyclic antidepressants, serotonin-norepinephrine reuptake inhibitors, and bupropion as the antidepressants most likely to be misused for their amphetamine-like euphoric effects or serotonin-induced dissociative effects.³² However, compared with other drug classes discussed in this article, the rate of antide­pressant misuse is relatively low among col­lege students.

Regardless of the antidepressant selected, clinicians should be concerned about alcohol use among college-age patients. Persons with depression are at increased risk of alcohol­ism compared with the general population.³³ This combination can increase depressive symptoms and sedation, and decrease coor­dination, judgment, and reaction time.³³

Excessive alcohol use can increase the risk of seizures in patients taking antidepressants such as buproprion.³⁴ Employ caution when prescribing bupropion to patients who have a predisposing clini­cal factor that increases seizure risk, such as excessive alcohol use and abrupt ces­sation, use of other medications that may lower seizure threshold (eg, theophylline, amphetamines, phenothiazines), and a his­tory of head trauma.³⁴

To minimize the risk of seizures with bupropion, titrate up the dosage slowly. Furthermore, using a low dosage during dual therapy for antidepressant augmenta­tion further decreases the risk of seizure.³⁵ For these reasons, we recommend that you avoid bupropion in patients who are at risk of binge drinking, and give careful consid­eration to providing alternative therapies for them.

Prescribers and patients should also keep in mind that hypertensive crisis could occur if MAOIs are combined with certain types of alcoholic beverages containing tyramine, including some wines and draft beer.³³

How you can identify and prevent misuse
Careful communication between health care provider and patient that is necessary to minimize the risk of adverse drug events with psychotropic medications often is lack­ing. For example, 24% of study college-age participants did not remember if their physi­cian provided a diagnosis and 28.8% could not recall being informed about side effects and, perhaps as a result, many students did not take their medications as prescribed.⁹

Further, prescribers should ask college-age patients who are undergoing stimulant treatment if they believe that they are being adequately treated. They should inquire about how they are taking their medica­tions.¹¹ These questions can lead to discus­sion of the need for these medications and reevaluation of their perceived indication.¹¹

Remind patients to take their medication only as directed.³⁶ Highlight the need to:
   • store medications in a discreet location
   • properly dispose of unused medications
   • keep tabs on the quantity of pills
   • know how to resist requests for diver­sion from peers.

The Substance Abuse and Mental Health Services Administration offers additional useful strategies,³⁷ and pharmacists also can be partners in substance use education and prevention.³⁸ These are examples of how health care providers can take an active role in providing patients with a thorough and detailed understanding of (1) their condi­tions and (2) their prescribed medications to improve efficacy and safety while pre­venting misuse.⁸

A study found that the most common method of obtaining these medications without a prescription is acquiring them from peers; 54% of undergraduate patients with stimulant prescriptions have been approached by peers to give, trade, or sell their drugs.²⁵ Other methods include pur­chasing medications online or faking pre­scriptions.³⁹ Health care providers should remind patients of the legal ramifications of sharing or selling their prescribed medi­cations. Finally, providers must be vigilant for students who may feign symptoms to obtain a prescription:
   • be wary if symptom presentation sounds too “textbook”
   • seek collateral history from family. Adults with ADHD should have shown symptoms during childhood
   • use external verification such as neuropsychological testing for ADHD. A neuropsychologist can detect deception by analyzing the pattern of responses to questions.

Patient assessment is a key step to in pre­venting abuse of psychotropic medications. Gentle inquiry about school-related stress and other risk factors for misuse can help practitioners determine if students are at risk of diversion and if additional screening is necessary.

In response to these issues, Stone and Merlo⁸ have suggested that, in addition to the educational programs held on col­lege campuses on alcohol, illicit drugs, and prescription painkillers, patients should be better informed on the appropriate use of prescription psychiatric medications, instructed to avoid sharing with family and friends, and assessed for abuse risk at regu­lar intervals.

To further protect patients from adverse outcomes during treatment, you can employ conservative and safe prescribing techniques. One strategy might be to keep a personal formulary that lists key medica­tions you use in everyday practice, includ­ing knowledge about each drug’s dosage, potential adverse effects, key warnings, and drug−drug interactions.⁴⁰

Furthermore, maintain healthy caution about newly approved medications and carefully consider how they measure up to existing agents—in other words, prac­tice evidence-based medicine, particu­larly when students request a particular agent.^40,41 Prescribers should evaluate the risk of abuse before prescribing and attempt to prevent misuse by limiting quantities and minimizing polypharmacy.

Last, pharmacists can be key allies for consultation and appropriate medication selection.

 
Bottom Line
Psychotropic medications are necessary to treat the variety of conditions—anxiety, attention-deficit/hyperactivity disorder, depression, and panic disorder—common among college students. However, students are at risk of combining their prescribed medications with other medications, drugs, and alcohol or could sell or share their medication with peers. Proper counseling and identification of risk factors can be important tools for preventing such events.

Related Resources
• American College Health Association-National College Health Assessment. www.acha-ncha.org.
• Schwartz VI. College mental health: How to provide care for students in need. Current Psychiatry. 2011;10(12):22-29.

Drug Brand Names
Bupropion • Wellbutrin, Zyban
Methadone • Methadose, Dolophine
Theophylline • Theo-24, Theolair, Uniphyl

Disclosures
The authors report no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.

Many college students suffer from mental illness (Table 1),¹ which can have a negative impact on academic performance. Although psychotropic medications are an important part of treat­ment for many college students, the potential for misuse always is present. Drug misuse occurs when patients use medications for reasons inconsistent with legal or medical guidelines.² For example, patients may take a medica­tion that has not been prescribed for them or in a manner that is inconsis­tent with the prescriber’s instructions, including administration with other substances.³

Misuse of psychotropic drugs is prevalent among college students. A study of 14,175 students from 26 campuses reported that 14.7% of students taking a psychotropic are doing so without a prescription, including stimu­lants (52.6%), anxiolytics (38.4%), and antidepressants (17.4%).⁴ Another study states that more than one-third of responders reported misuse of >1 class of medication.⁵

Psychotropic misuse is concerning because it increases the risk of adverse events. Nearly one-half of medication errors are associated with writing and dispensing the prescription, which means that prescribers can work to reduce these errors.⁶ However, nonadherence, prescription misuse, and failure to disclose use of over-the-counter drugs, illicit drugs, and herbal products makes preventing most adverse events difficult, if not impossible, for prescribers.^7,8

Psychotropic drug misuse among college students is highly variable and unpredictable. Students misuse medications, including stimulants, ben­zodiazepines, and antidepressants, for a variety of reasons, such as study enhancement, experimentation, intoxication, self-medication, relaxation, and stress management.⁸ One survey reported that >70% of students taking a psychotropic medication took it with alcohol or another illicit drug.⁹

However, <20% of those using a psy­chotropic medication with alcohol or other illicit drugs told their health care provider(s),⁹ making it impossible for cli­nicians to predict a patient’s risk of drug− drug interactions and subsequent adverse events. Additionally, additive effects could occur¹⁰ and changes in a patient’s presenta­tion could be caused by a reaction to a com­bination of medications, rather than a new symptom of mental illness.

This article will examine common issues associated with drug misuse among col­lege-age students and review prevention strategies (Table 2).

Stimulants
Stimulants have the highest rate of diver­sion; 61.7% of college students prescribed stimulants have shared or sold their medi­cation.¹¹ A survey of 115 students from 2 universities reported that the most common reason for stimulant misuse was to enhance academic performance.¹² The same survey showed that some students take stimulants with Cannabis (17%) and alcohol (30%).¹² As a result, in addition to lowering grade point average (GPA) and other academic difficul­ties,¹³ students misusing stimulants are at risk of drug interactions.¹⁴

It is critical to ascertain the route of drug administration, because non-oral routes, including crushing then snorting or inject­ing, are associated with additional health concerns, such as accidental death or blood-borne illnesses.^15,16 Cardiac adverse effects of stimulants include hypertension, vasospasm, tachycardia, and dysrhythmia; psychiatric and other effects include sero­tonin syndrome, hallucinations, anxiety, paranoia, seizures, tics, hyperthermia, and tremor.¹⁷ Health care providers prescrib­ing or caring for people taking a stimulant should monitor for these potential effects.

The risk of switch to mania might not be apparent to those who prescribe stimulants or to young people who take non-prescribed stimulants for academic enhancement or to achieve medication-induced euphoria. Adolescent stimulant use is associated with symptoms of early-onset bipolar disorder in patients who have attention-deficit/ hyperactivity disorder (ADHD) and undi­agnosed bipolarity.¹⁸

The cardiovascular risk associated with stimulant use is debatable. Although sev­eral studies have been conducted,^19-21 meth­odological factors limit their applicability. To minimize potential risks, several precau­tions should be taken before prescribing a stimulant to treat ADHD.

First, obtain a detailed personal and family medical history, asking about pos­sible cardiovascular disease. Second, care­fully scrutinize the patient’s cardiovascular system during the physical exam. Third, consider additional testing, such as an elec­trocardiogram, if the patient’s history or physical exam indicates possible risk.²²

As a prescriber, you should be aware of the prevalence of stimulant use among students with and without ADHD, includ­ing those who could be feigning ADHD symptoms.¹⁵ Diversion could occur through sharing medications or selling them to friends and family.¹¹ It also is possible that these medications may be used with other illicit substances, such as Cannabis, ecstasy, cocaine, and opiates.²³ Students also could misuse stimulants by taking more than the prescribed dosage.²⁴

Risk factors for misuse of stimulants include: heavy alcohol use, previous illicit drug use, white race, fraternity or sorority membership, low GPA, increased hyperac­tivity symptoms, and attendance at a com­petitive college or university.^25-27

Benzodiazepines
Misuse of benzodiazepine is a significant component of prescription drug abuse and often occurs with other medications and alcohol.²⁸ Additional methods of mis­use include increased dosage and non-oral routes of administration.²⁹

A 2001 national survey reported that 7.8% of college students have misused benzodiaz­epines.²³ Common characteristics of benzo­diazepine abusers include young age, male sex, personality characteristics of impulsiv­ity and hopelessness, and abuse of other drugs, including cocaine and methadone.^28,29

Benzodiazepines are prescribed for their anxiolytic and hypnotic properties and stu­dents could use these drugs with other agents to augment the euphoric effects or diminish withdrawal symptoms.³⁰ Patients taking ben­zodiazepines for anxiety might self-medicate with alcohol, which increases sedation and depression, and can contribute to the risk for respiratory depression.¹⁰ Misuse of benzodi­azepines can result in cognitive and psycho­motor impairment and increase the risk of accidents and overdose.^29,31

Although overdose with monotherapy is rare, the risk increases when a benzodiaz­epine is used with alcohol¹⁰ or another respi­ratory depressants, such as opioids, because combination use can produce additive effects.²⁸ You should therefore avoid prescrib­ing benzodiazepines to patients who have a history of significant substance abuse and consider using alternative, non-addictive agents, such as selective serotonin reuptake inhibitors, or non-pharmaceutical treatment when such patients present with an anxiety disorder. The risk of adverse effects of ben­zodiazepines can be reduced by limiting the dosing and the duration of the treatment, and by using longer-acting rather than the more addictive, shorter-acting, agents.

Antidepressants
Health care providers should be aware that, despite the relative absence of physically addictive properties, antidepressants from most classes are abusable agents sought by young people for non-medical use. In particular, the literature highlights mono­amine oxidase inhibitors (MAOIs), tricyclic antidepressants, serotonin-norepinephrine reuptake inhibitors, and bupropion as the antidepressants most likely to be misused for their amphetamine-like euphoric effects or serotonin-induced dissociative effects.³² However, compared with other drug classes discussed in this article, the rate of antide­pressant misuse is relatively low among col­lege students.

Regardless of the antidepressant selected, clinicians should be concerned about alcohol use among college-age patients. Persons with depression are at increased risk of alcohol­ism compared with the general population.³³ This combination can increase depressive symptoms and sedation, and decrease coor­dination, judgment, and reaction time.³³

Excessive alcohol use can increase the risk of seizures in patients taking antidepressants such as buproprion.³⁴ Employ caution when prescribing bupropion to patients who have a predisposing clini­cal factor that increases seizure risk, such as excessive alcohol use and abrupt ces­sation, use of other medications that may lower seizure threshold (eg, theophylline, amphetamines, phenothiazines), and a his­tory of head trauma.³⁴

To minimize the risk of seizures with bupropion, titrate up the dosage slowly. Furthermore, using a low dosage during dual therapy for antidepressant augmenta­tion further decreases the risk of seizure.³⁵ For these reasons, we recommend that you avoid bupropion in patients who are at risk of binge drinking, and give careful consid­eration to providing alternative therapies for them.

Prescribers and patients should also keep in mind that hypertensive crisis could occur if MAOIs are combined with certain types of alcoholic beverages containing tyramine, including some wines and draft beer.³³

How you can identify and prevent misuse
Careful communication between health care provider and patient that is necessary to minimize the risk of adverse drug events with psychotropic medications often is lack­ing. For example, 24% of study college-age participants did not remember if their physi­cian provided a diagnosis and 28.8% could not recall being informed about side effects and, perhaps as a result, many students did not take their medications as prescribed.⁹

Further, prescribers should ask college-age patients who are undergoing stimulant treatment if they believe that they are being adequately treated. They should inquire about how they are taking their medica­tions.¹¹ These questions can lead to discus­sion of the need for these medications and reevaluation of their perceived indication.¹¹

Remind patients to take their medication only as directed.³⁶ Highlight the need to:
   • store medications in a discreet location
   • properly dispose of unused medications
   • keep tabs on the quantity of pills
   • know how to resist requests for diver­sion from peers.

The Substance Abuse and Mental Health Services Administration offers additional useful strategies,³⁷ and pharmacists also can be partners in substance use education and prevention.³⁸ These are examples of how health care providers can take an active role in providing patients with a thorough and detailed understanding of (1) their condi­tions and (2) their prescribed medications to improve efficacy and safety while pre­venting misuse.⁸

A study found that the most common method of obtaining these medications without a prescription is acquiring them from peers; 54% of undergraduate patients with stimulant prescriptions have been approached by peers to give, trade, or sell their drugs.²⁵ Other methods include pur­chasing medications online or faking pre­scriptions.³⁹ Health care providers should remind patients of the legal ramifications of sharing or selling their prescribed medi­cations. Finally, providers must be vigilant for students who may feign symptoms to obtain a prescription:
   • be wary if symptom presentation sounds too “textbook”
   • seek collateral history from family. Adults with ADHD should have shown symptoms during childhood
   • use external verification such as neuropsychological testing for ADHD. A neuropsychologist can detect deception by analyzing the pattern of responses to questions.

Patient assessment is a key step to in pre­venting abuse of psychotropic medications. Gentle inquiry about school-related stress and other risk factors for misuse can help practitioners determine if students are at risk of diversion and if additional screening is necessary.

In response to these issues, Stone and Merlo⁸ have suggested that, in addition to the educational programs held on col­lege campuses on alcohol, illicit drugs, and prescription painkillers, patients should be better informed on the appropriate use of prescription psychiatric medications, instructed to avoid sharing with family and friends, and assessed for abuse risk at regu­lar intervals.

To further protect patients from adverse outcomes during treatment, you can employ conservative and safe prescribing techniques. One strategy might be to keep a personal formulary that lists key medica­tions you use in everyday practice, includ­ing knowledge about each drug’s dosage, potential adverse effects, key warnings, and drug−drug interactions.⁴⁰

Furthermore, maintain healthy caution about newly approved medications and carefully consider how they measure up to existing agents—in other words, prac­tice evidence-based medicine, particu­larly when students request a particular agent.^40,41 Prescribers should evaluate the risk of abuse before prescribing and attempt to prevent misuse by limiting quantities and minimizing polypharmacy.

Last, pharmacists can be key allies for consultation and appropriate medication selection.

 
Bottom Line
Psychotropic medications are necessary to treat the variety of conditions—anxiety, attention-deficit/hyperactivity disorder, depression, and panic disorder—common among college students. However, students are at risk of combining their prescribed medications with other medications, drugs, and alcohol or could sell or share their medication with peers. Proper counseling and identification of risk factors can be important tools for preventing such events.

Related Resources
• American College Health Association-National College Health Assessment. www.acha-ncha.org.
• Schwartz VI. College mental health: How to provide care for students in need. Current Psychiatry. 2011;10(12):22-29.

Drug Brand Names
Bupropion • Wellbutrin, Zyban
Methadone • Methadose, Dolophine
Theophylline • Theo-24, Theolair, Uniphyl

Disclosures
The authors report no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.

Little is known about people who write to criminals incar­cerated for a violent crime. However, existence of Web sites such as WriteAPrisoner.com, Meet-An-Inmate.com, and PrisonPenPals.com suggests some appetite among the public for corresponding with the incarcerated. Writers of letters might be drawn to the “bad boy” image of prison­ers. Furthermore, much has been written of the willingness of some battered women to remain in an abusive domestic relationship, leading them to correspond with their abusers even after those abusers are incarcerated.^1,2

To our knowledge, no examination of letters written to a mass murderer has been published. Therefore, we catego­rized and analyzed 784 letters sent to a high-profile male mass murderer whose crime was committed during the past decade. Here is a description of the study and what we found, as well as discussion of how our findings might offer utility in a psychiatric practice.

Goals of the study
We hypothesized that a large percentage of those letters could be classified as “Romantic,” given the lay percep­tion that it is women who write to mass murderers. We also sought to evaluate follow-up letters sent by these writers to test the assumption that their individual goals would be con­stant over time.

We performed this study in the hope that the research could assist psychiatric practitioners in treating patients who seek to associate with a violent person (see “Treatment considerations,”). We thought it might be helpful for practitioners to get a better understanding of the nature of people who write to a violent offender or express a desire to do so.

Methods of study
Two authors (R.S.J. and D.P.G.) evaluated 819 letters that had been written by non-incarcerated, non-family adults to 1 mass murderer. The initial letter and follow-up letters written by each unique writer (n = 333) were categorized as follows:
   • state or country from which the letter was sent
   • age
   • sex
   • number of letters sent by each writer
   • whether a photograph was enclosed
   • whether additional items were enclosed (eg, gifts, drawings)
   • whether the letter was rejected by prison authorities
   • the writer’s purpose.

The study was approved by the insti­tutional review board of Baylor College of Medicine.

Letters were assigned to 1 of 5 categories:

Acquaintance letters sought ongoing cor­respondence relationship with the murderer. They focused largely on conveying informa­tion about the writer.

Show of support letters also sought an ongoing correspondence relationship with the murderer, but instead focused on him, not the writer.

Romance letters used words that conveyed romantic or non-platonic affection.

Spiritual letters gave advice to the mur­derer with a religious tone.

Words of wisdom letters offered advice but lacked a religious tone.

Given the nonstandardized nature of categorization and the lack of a formal questionnaire, we were unable to perform an exploratory factor analysis on our cat­egorizations. Inter-rater reliability of letter categorization was 0.79.

Results: Writer profiles, purpose for writing
In all, we reviewed 819 letters:
   • Thirty-five letters were excluded because they were written by family mem­bers, children, or other prisoners
   • Of the remaining 784 letters, there were 333 unique writers
   • Two-hundred sixty letters were writ­ten by women, 61 by men; 2 were co-written by both sexes; sex could not be determined for 10.

Women were more likely than men to write a letter (P = .014) and to write ≥3 letters (P = .001). The age of the writer was deter­mined for 117 (35.1%) letters; mean age was 27.8 (± 8.9) years (range, 18 to 59 years).

The purpose of the letters differed by sex (P < .001) but not by the writer’s age (P = .058). Women were more likely than men to write letters categorized as “Acquaintance,” “Romance,” and “Show of support”; in con­trast, men were more likely than women to write a letter categorized as “Spiritual”  (Table 1). Approximately 95% of let­ters were handwritten. Letters averaged 3 pages (range, 1 to 16 pages).

Two-hundred sixteen writers wrote a single letter; 53 wrote 2 letters; 18 wrote 3 let­ters; 11 wrote 4 letters; 30 wrote 5 to 10 let­ters; and 9 wrote 11 to 43 letters. The purpose of follow-up letters was associated with the age of the writer (P < .001) and with the writ­er’s sex (P < .001). Women were more likely to write “Show of support” and “Romance” follow-up letters; men were more likely to write “Spiritual” follow-up letters (Table 2).

Results suggested that the purpose of the initial letter was a reasonable predictor of the purpose of follow-up letters (P < .001) (Table 3). The murderer never responded to any letters. Letters were most often writ­ten from his state of incarceration; next, from contiguous states; then, from non-contiguous states; and, last, from interna­tional locations (P < .001).

Of the initial letters from writers who wrote ≥10, 60% were categorized as “Acquaintance” and 20% as “Romance.” The writer who wrote the most letters (43) moved during the course of her letter-writing to live in the same state as the murderer; she stated in her letters that she did so to be closer to him and to be able to attend his court hearings. Four other writers, each of whom wrote >5 letters, stated that they had traveled to the murderer’s state of incarcera­tion to attend some of his hearings in person.

Composite examples of more common categories of letters
Names and other pertinent identifying information have been changed.

Acquaintance. Hi, Steve. I’ve been follow­ing your case and just wanted to write you so that maybe we could be friends or keep in touch since you’re probably pretty bored. I’m a 27-year-old college student studying market­ing and working at Applebee’s as a waitress (for now) until I can land my dream job. I’ve enclosed a picture of me and my dachshund along with a photo of my favorite beach in the world. Write me back if you want. Jenny.

Show of support. Steve: I’ve been really wor­ried about you since first seeing you on TV. You look different lately and I hope they’re treating you OK and feeding you decent food. In case they’re not, I’ve enclosed a little something to buy yourself a treat. Just know that there are many of us that care about you and are really pulling for you to be strong in this tough situ­ation you’re in. Yours truly, Karen.

Romance. Dearest Steven: My mind has been filled with thoughts of you and of us since I last saw you in my dreams! Be strong, because you are going to beat this once they understand that you are not responsible for what happened! Don’t you see, sweetie, the system failed you, and now you’re caught up in something that you will soon overcome. When I think of the day that you get released, and how we’ll be able to settle down some­where together, it gets me incredibly excited. You and I are meant to be together, because I understand you and can help you get better. I love you, Steven! Please write me back so that I know we’re on the same page about our plans for the future. Love, ♥ Your sweetie, Rachel.

Spiritual. Dear Child of God: The Lord has a plan for you. I know that things right now might be confusing, and you’re in a black place, but He is there right beside you. If you need some reading materials to give you com­fort, just let me know and I can get a Bible to you along with some other books to give you solace and strengthen your walk with Him. God forgives you and he loves you so much! Much love in Christ, Mary.

Discussion
Given that the mass murderer in this study was a young man, it is not surpris­ing that 78% of writers of initial letters were women. However, it is interesting that, among women’s initial letters, 44% were “Acquaintance” letters and only 15% were categorized as “Romance.”

Given the severity of the murderer’s crime, it is remarkable that he received only 1 “Hate mail” letter.

Initial “Spiritual” letters were more likely to be followed by letters of the same category than any other category; “Romance” letters were a close second. This demonstrates the consistent efforts of writers in these 2 categories. Highly persis­tent writers (≥10 letters) were most likely to fall into “Acquaintance” and “Romance” categories. The persistence of these writers is remarkable, in view of the fact that none of their letters were answered. We hypoth­esize that the killer did not reply because he had no interest in correspondence.

Similarities to stalking. Given that 9 writ­ers wrote >10 letters each and 2 wrote >20 each, elements of their behavior are not unlike what is seen in stalkers.³ Consistent with the stalking literature and Mullen et al4 stalker typology, many writers in this study appeared to seek intimacy with the perpetrator through “Romance” or “Show of support” letters, and might be akin to Mullen’s so-called intimacy-seeking stalker. Such stalkers’ behavior arises out of loneliness, with a strong desire for a rela­tionship with the target; a significant per­centage of such stalkers suffer a delusional disorder.

Mullen’s so-called incompetent suitor stalker is similar to the intimacy-seeking type but, instead, has an interest in a short-term relationship and is far less persistent in his (her) stalking behavior⁴; this type might apply to the writers in this study who wrote >1 but <10 letters.

Two additional observations also are notable when trying to characterize people who write letters: (1) A high percentage of people who stalk a celebrity suffer a psy­chotic disorder^5,6; (2) 4 letter-writers trav­eled, and 1 relocated, to the murderer’s state of incarceration to attend his hearings and be closer to him.

This study has limitations:
   • categorization of letters is inherently subjective and the categories themselves were created by the researchers
   • the nature and categorization of such letters might vary considerably with the age and sex of the violent criminal; our findings in this case are not generalizable.

Last, researchers who plan to study writers of letters to incarcerated criminals should consider sending a personality test and other questionnaires to those writers to understand this population better.

Treatment considerations
Psychiatrists treating patients who seek a romantic attachment with a violent person should consider psychotherapy as a means of treating possible character pathology. The desire for romance with a violent crimi­nal was greater among repeat writers (20%) than in initial letters (15%), suggesting that people who have a strong inclination to associate with a violent person might benefit from exploring romantic feelings in therapy. Specifically, therapists would be wise to explore with such patients the possibility that they experienced violence or verbal abuse in childhood or adulthood.

To the extent that evidence of prior abuse exists, a diagnosis of posttraumatic stress disorder (PTSD) might be appro­priate; specialized therapy for men and women with a history of abuse might be indicated. It is important to provide vali­dation for patients who are victims when they describe their abuse, and to stress that they did nothing to provoke the violence. Furthermore, investigation of why the patient feels drawn romantically toward a violent criminal is helpful, as well as an examination of how such behavior is self-defeating.

There might be value in having patients keep a journal in lieu of actually sending letters; there is evidence that “journaling” can reduce substance use recidivism.⁷ This work can be performed in conjunction with group or individual psychotherapy that addresses any history of abuse and subse­quent PTSD.

Many patients are reluctant to discuss their romantic feelings toward a violent criminal until the psychiatrist has estab­lished a strong doctor−patient relationship. Last, clinicians should not hesitate to refer these patients to a therapist who specializes in domestic violence.

Related Resource
• Marazziti D, Falaschi V, Lombardi A, et al. Stalking: a neuro­biological perspective. Riv Psichiatr. 2015;50(1):12-18.

Disclosures
The authors report no financial relationships with any company whose products are mentioned in this article or with manufacturers of competing products.

Little is known about people who write to criminals incar­cerated for a violent crime. However, existence of Web sites such as WriteAPrisoner.com, Meet-An-Inmate.com, and PrisonPenPals.com suggests some appetite among the public for corresponding with the incarcerated. Writers of letters might be drawn to the “bad boy” image of prison­ers. Furthermore, much has been written of the willingness of some battered women to remain in an abusive domestic relationship, leading them to correspond with their abusers even after those abusers are incarcerated.^1,2

To our knowledge, no examination of letters written to a mass murderer has been published. Therefore, we catego­rized and analyzed 784 letters sent to a high-profile male mass murderer whose crime was committed during the past decade. Here is a description of the study and what we found, as well as discussion of how our findings might offer utility in a psychiatric practice.

Goals of the study
We hypothesized that a large percentage of those letters could be classified as “Romantic,” given the lay percep­tion that it is women who write to mass murderers. We also sought to evaluate follow-up letters sent by these writers to test the assumption that their individual goals would be con­stant over time.

We performed this study in the hope that the research could assist psychiatric practitioners in treating patients who seek to associate with a violent person (see “Treatment considerations,”). We thought it might be helpful for practitioners to get a better understanding of the nature of people who write to a violent offender or express a desire to do so.

Methods of study
Two authors (R.S.J. and D.P.G.) evaluated 819 letters that had been written by non-incarcerated, non-family adults to 1 mass murderer. The initial letter and follow-up letters written by each unique writer (n = 333) were categorized as follows:
   • state or country from which the letter was sent
   • age
   • sex
   • number of letters sent by each writer
   • whether a photograph was enclosed
   • whether additional items were enclosed (eg, gifts, drawings)
   • whether the letter was rejected by prison authorities
   • the writer’s purpose.

The study was approved by the insti­tutional review board of Baylor College of Medicine.

Letters were assigned to 1 of 5 categories:

Acquaintance letters sought ongoing cor­respondence relationship with the murderer. They focused largely on conveying informa­tion about the writer.

Show of support letters also sought an ongoing correspondence relationship with the murderer, but instead focused on him, not the writer.

Romance letters used words that conveyed romantic or non-platonic affection.

Spiritual letters gave advice to the mur­derer with a religious tone.

Words of wisdom letters offered advice but lacked a religious tone.

Given the nonstandardized nature of categorization and the lack of a formal questionnaire, we were unable to perform an exploratory factor analysis on our cat­egorizations. Inter-rater reliability of letter categorization was 0.79.

Results: Writer profiles, purpose for writing
In all, we reviewed 819 letters:
   • Thirty-five letters were excluded because they were written by family mem­bers, children, or other prisoners
   • Of the remaining 784 letters, there were 333 unique writers
   • Two-hundred sixty letters were writ­ten by women, 61 by men; 2 were co-written by both sexes; sex could not be determined for 10.

Women were more likely than men to write a letter (P = .014) and to write ≥3 letters (P = .001). The age of the writer was deter­mined for 117 (35.1%) letters; mean age was 27.8 (± 8.9) years (range, 18 to 59 years).

The purpose of the letters differed by sex (P < .001) but not by the writer’s age (P = .058). Women were more likely than men to write letters categorized as “Acquaintance,” “Romance,” and “Show of support”; in con­trast, men were more likely than women to write a letter categorized as “Spiritual”  (Table 1). Approximately 95% of let­ters were handwritten. Letters averaged 3 pages (range, 1 to 16 pages).

Two-hundred sixteen writers wrote a single letter; 53 wrote 2 letters; 18 wrote 3 let­ters; 11 wrote 4 letters; 30 wrote 5 to 10 let­ters; and 9 wrote 11 to 43 letters. The purpose of follow-up letters was associated with the age of the writer (P < .001) and with the writ­er’s sex (P < .001). Women were more likely to write “Show of support” and “Romance” follow-up letters; men were more likely to write “Spiritual” follow-up letters (Table 2).

Results suggested that the purpose of the initial letter was a reasonable predictor of the purpose of follow-up letters (P < .001) (Table 3). The murderer never responded to any letters. Letters were most often writ­ten from his state of incarceration; next, from contiguous states; then, from non-contiguous states; and, last, from interna­tional locations (P < .001).

Of the initial letters from writers who wrote ≥10, 60% were categorized as “Acquaintance” and 20% as “Romance.” The writer who wrote the most letters (43) moved during the course of her letter-writing to live in the same state as the murderer; she stated in her letters that she did so to be closer to him and to be able to attend his court hearings. Four other writers, each of whom wrote >5 letters, stated that they had traveled to the murderer’s state of incarcera­tion to attend some of his hearings in person.

Composite examples of more common categories of letters
Names and other pertinent identifying information have been changed.

Acquaintance. Hi, Steve. I’ve been follow­ing your case and just wanted to write you so that maybe we could be friends or keep in touch since you’re probably pretty bored. I’m a 27-year-old college student studying market­ing and working at Applebee’s as a waitress (for now) until I can land my dream job. I’ve enclosed a picture of me and my dachshund along with a photo of my favorite beach in the world. Write me back if you want. Jenny.

Show of support. Steve: I’ve been really wor­ried about you since first seeing you on TV. You look different lately and I hope they’re treating you OK and feeding you decent food. In case they’re not, I’ve enclosed a little something to buy yourself a treat. Just know that there are many of us that care about you and are really pulling for you to be strong in this tough situ­ation you’re in. Yours truly, Karen.

Romance. Dearest Steven: My mind has been filled with thoughts of you and of us since I last saw you in my dreams! Be strong, because you are going to beat this once they understand that you are not responsible for what happened! Don’t you see, sweetie, the system failed you, and now you’re caught up in something that you will soon overcome. When I think of the day that you get released, and how we’ll be able to settle down some­where together, it gets me incredibly excited. You and I are meant to be together, because I understand you and can help you get better. I love you, Steven! Please write me back so that I know we’re on the same page about our plans for the future. Love, ♥ Your sweetie, Rachel.

Spiritual. Dear Child of God: The Lord has a plan for you. I know that things right now might be confusing, and you’re in a black place, but He is there right beside you. If you need some reading materials to give you com­fort, just let me know and I can get a Bible to you along with some other books to give you solace and strengthen your walk with Him. God forgives you and he loves you so much! Much love in Christ, Mary.

Discussion
Given that the mass murderer in this study was a young man, it is not surpris­ing that 78% of writers of initial letters were women. However, it is interesting that, among women’s initial letters, 44% were “Acquaintance” letters and only 15% were categorized as “Romance.”

Given the severity of the murderer’s crime, it is remarkable that he received only 1 “Hate mail” letter.

Initial “Spiritual” letters were more likely to be followed by letters of the same category than any other category; “Romance” letters were a close second. This demonstrates the consistent efforts of writers in these 2 categories. Highly persis­tent writers (≥10 letters) were most likely to fall into “Acquaintance” and “Romance” categories. The persistence of these writers is remarkable, in view of the fact that none of their letters were answered. We hypoth­esize that the killer did not reply because he had no interest in correspondence.

Similarities to stalking. Given that 9 writ­ers wrote >10 letters each and 2 wrote >20 each, elements of their behavior are not unlike what is seen in stalkers.³ Consistent with the stalking literature and Mullen et al4 stalker typology, many writers in this study appeared to seek intimacy with the perpetrator through “Romance” or “Show of support” letters, and might be akin to Mullen’s so-called intimacy-seeking stalker. Such stalkers’ behavior arises out of loneliness, with a strong desire for a rela­tionship with the target; a significant per­centage of such stalkers suffer a delusional disorder.

Mullen’s so-called incompetent suitor stalker is similar to the intimacy-seeking type but, instead, has an interest in a short-term relationship and is far less persistent in his (her) stalking behavior⁴; this type might apply to the writers in this study who wrote >1 but <10 letters.

Two additional observations also are notable when trying to characterize people who write letters: (1) A high percentage of people who stalk a celebrity suffer a psy­chotic disorder^5,6; (2) 4 letter-writers trav­eled, and 1 relocated, to the murderer’s state of incarceration to attend his hearings and be closer to him.

This study has limitations:
   • categorization of letters is inherently subjective and the categories themselves were created by the researchers
   • the nature and categorization of such letters might vary considerably with the age and sex of the violent criminal; our findings in this case are not generalizable.

Last, researchers who plan to study writers of letters to incarcerated criminals should consider sending a personality test and other questionnaires to those writers to understand this population better.

Treatment considerations
Psychiatrists treating patients who seek a romantic attachment with a violent person should consider psychotherapy as a means of treating possible character pathology. The desire for romance with a violent crimi­nal was greater among repeat writers (20%) than in initial letters (15%), suggesting that people who have a strong inclination to associate with a violent person might benefit from exploring romantic feelings in therapy. Specifically, therapists would be wise to explore with such patients the possibility that they experienced violence or verbal abuse in childhood or adulthood.

To the extent that evidence of prior abuse exists, a diagnosis of posttraumatic stress disorder (PTSD) might be appro­priate; specialized therapy for men and women with a history of abuse might be indicated. It is important to provide vali­dation for patients who are victims when they describe their abuse, and to stress that they did nothing to provoke the violence. Furthermore, investigation of why the patient feels drawn romantically toward a violent criminal is helpful, as well as an examination of how such behavior is self-defeating.

There might be value in having patients keep a journal in lieu of actually sending letters; there is evidence that “journaling” can reduce substance use recidivism.⁷ This work can be performed in conjunction with group or individual psychotherapy that addresses any history of abuse and subse­quent PTSD.

Many patients are reluctant to discuss their romantic feelings toward a violent criminal until the psychiatrist has estab­lished a strong doctor−patient relationship. Last, clinicians should not hesitate to refer these patients to a therapist who specializes in domestic violence.

Related Resource
• Marazziti D, Falaschi V, Lombardi A, et al. Stalking: a neuro­biological perspective. Riv Psichiatr. 2015;50(1):12-18.

Disclosures
The authors report no financial relationships with any company whose products are mentioned in this article or with manufacturers of competing products.

Little is known about people who write to criminals incar­cerated for a violent crime. However, existence of Web sites such as WriteAPrisoner.com, Meet-An-Inmate.com, and PrisonPenPals.com suggests some appetite among the public for corresponding with the incarcerated. Writers of letters might be drawn to the “bad boy” image of prison­ers. Furthermore, much has been written of the willingness of some battered women to remain in an abusive domestic relationship, leading them to correspond with their abusers even after those abusers are incarcerated.^1,2

To our knowledge, no examination of letters written to a mass murderer has been published. Therefore, we catego­rized and analyzed 784 letters sent to a high-profile male mass murderer whose crime was committed during the past decade. Here is a description of the study and what we found, as well as discussion of how our findings might offer utility in a psychiatric practice.

Goals of the study
We hypothesized that a large percentage of those letters could be classified as “Romantic,” given the lay percep­tion that it is women who write to mass murderers. We also sought to evaluate follow-up letters sent by these writers to test the assumption that their individual goals would be con­stant over time.

We performed this study in the hope that the research could assist psychiatric practitioners in treating patients who seek to associate with a violent person (see “Treatment considerations,”). We thought it might be helpful for practitioners to get a better understanding of the nature of people who write to a violent offender or express a desire to do so.

Methods of study
Two authors (R.S.J. and D.P.G.) evaluated 819 letters that had been written by non-incarcerated, non-family adults to 1 mass murderer. The initial letter and follow-up letters written by each unique writer (n = 333) were categorized as follows:
   • state or country from which the letter was sent
   • age
   • sex
   • number of letters sent by each writer
   • whether a photograph was enclosed
   • whether additional items were enclosed (eg, gifts, drawings)
   • whether the letter was rejected by prison authorities
   • the writer’s purpose.

The study was approved by the insti­tutional review board of Baylor College of Medicine.

Letters were assigned to 1 of 5 categories:

Acquaintance letters sought ongoing cor­respondence relationship with the murderer. They focused largely on conveying informa­tion about the writer.

Show of support letters also sought an ongoing correspondence relationship with the murderer, but instead focused on him, not the writer.

Romance letters used words that conveyed romantic or non-platonic affection.

Spiritual letters gave advice to the mur­derer with a religious tone.

Words of wisdom letters offered advice but lacked a religious tone.

Given the nonstandardized nature of categorization and the lack of a formal questionnaire, we were unable to perform an exploratory factor analysis on our cat­egorizations. Inter-rater reliability of letter categorization was 0.79.

Results: Writer profiles, purpose for writing
In all, we reviewed 819 letters:
   • Thirty-five letters were excluded because they were written by family mem­bers, children, or other prisoners
   • Of the remaining 784 letters, there were 333 unique writers
   • Two-hundred sixty letters were writ­ten by women, 61 by men; 2 were co-written by both sexes; sex could not be determined for 10.

Women were more likely than men to write a letter (P = .014) and to write ≥3 letters (P = .001). The age of the writer was deter­mined for 117 (35.1%) letters; mean age was 27.8 (± 8.9) years (range, 18 to 59 years).

The purpose of the letters differed by sex (P < .001) but not by the writer’s age (P = .058). Women were more likely than men to write letters categorized as “Acquaintance,” “Romance,” and “Show of support”; in con­trast, men were more likely than women to write a letter categorized as “Spiritual”  (Table 1). Approximately 95% of let­ters were handwritten. Letters averaged 3 pages (range, 1 to 16 pages).

Two-hundred sixteen writers wrote a single letter; 53 wrote 2 letters; 18 wrote 3 let­ters; 11 wrote 4 letters; 30 wrote 5 to 10 let­ters; and 9 wrote 11 to 43 letters. The purpose of follow-up letters was associated with the age of the writer (P < .001) and with the writ­er’s sex (P < .001). Women were more likely to write “Show of support” and “Romance” follow-up letters; men were more likely to write “Spiritual” follow-up letters (Table 2).

Results suggested that the purpose of the initial letter was a reasonable predictor of the purpose of follow-up letters (P < .001) (Table 3). The murderer never responded to any letters. Letters were most often writ­ten from his state of incarceration; next, from contiguous states; then, from non-contiguous states; and, last, from interna­tional locations (P < .001).

Of the initial letters from writers who wrote ≥10, 60% were categorized as “Acquaintance” and 20% as “Romance.” The writer who wrote the most letters (43) moved during the course of her letter-writing to live in the same state as the murderer; she stated in her letters that she did so to be closer to him and to be able to attend his court hearings. Four other writers, each of whom wrote >5 letters, stated that they had traveled to the murderer’s state of incarcera­tion to attend some of his hearings in person.

Composite examples of more common categories of letters
Names and other pertinent identifying information have been changed.

Acquaintance. Hi, Steve. I’ve been follow­ing your case and just wanted to write you so that maybe we could be friends or keep in touch since you’re probably pretty bored. I’m a 27-year-old college student studying market­ing and working at Applebee’s as a waitress (for now) until I can land my dream job. I’ve enclosed a picture of me and my dachshund along with a photo of my favorite beach in the world. Write me back if you want. Jenny.

Show of support. Steve: I’ve been really wor­ried about you since first seeing you on TV. You look different lately and I hope they’re treating you OK and feeding you decent food. In case they’re not, I’ve enclosed a little something to buy yourself a treat. Just know that there are many of us that care about you and are really pulling for you to be strong in this tough situ­ation you’re in. Yours truly, Karen.

Romance. Dearest Steven: My mind has been filled with thoughts of you and of us since I last saw you in my dreams! Be strong, because you are going to beat this once they understand that you are not responsible for what happened! Don’t you see, sweetie, the system failed you, and now you’re caught up in something that you will soon overcome. When I think of the day that you get released, and how we’ll be able to settle down some­where together, it gets me incredibly excited. You and I are meant to be together, because I understand you and can help you get better. I love you, Steven! Please write me back so that I know we’re on the same page about our plans for the future. Love, ♥ Your sweetie, Rachel.

Spiritual. Dear Child of God: The Lord has a plan for you. I know that things right now might be confusing, and you’re in a black place, but He is there right beside you. If you need some reading materials to give you com­fort, just let me know and I can get a Bible to you along with some other books to give you solace and strengthen your walk with Him. God forgives you and he loves you so much! Much love in Christ, Mary.

Discussion
Given that the mass murderer in this study was a young man, it is not surpris­ing that 78% of writers of initial letters were women. However, it is interesting that, among women’s initial letters, 44% were “Acquaintance” letters and only 15% were categorized as “Romance.”

Given the severity of the murderer’s crime, it is remarkable that he received only 1 “Hate mail” letter.

Initial “Spiritual” letters were more likely to be followed by letters of the same category than any other category; “Romance” letters were a close second. This demonstrates the consistent efforts of writers in these 2 categories. Highly persis­tent writers (≥10 letters) were most likely to fall into “Acquaintance” and “Romance” categories. The persistence of these writers is remarkable, in view of the fact that none of their letters were answered. We hypoth­esize that the killer did not reply because he had no interest in correspondence.

Similarities to stalking. Given that 9 writ­ers wrote >10 letters each and 2 wrote >20 each, elements of their behavior are not unlike what is seen in stalkers.³ Consistent with the stalking literature and Mullen et al4 stalker typology, many writers in this study appeared to seek intimacy with the perpetrator through “Romance” or “Show of support” letters, and might be akin to Mullen’s so-called intimacy-seeking stalker. Such stalkers’ behavior arises out of loneliness, with a strong desire for a rela­tionship with the target; a significant per­centage of such stalkers suffer a delusional disorder.

Mullen’s so-called incompetent suitor stalker is similar to the intimacy-seeking type but, instead, has an interest in a short-term relationship and is far less persistent in his (her) stalking behavior⁴; this type might apply to the writers in this study who wrote >1 but <10 letters.

Two additional observations also are notable when trying to characterize people who write letters: (1) A high percentage of people who stalk a celebrity suffer a psy­chotic disorder^5,6; (2) 4 letter-writers trav­eled, and 1 relocated, to the murderer’s state of incarceration to attend his hearings and be closer to him.

This study has limitations:
   • categorization of letters is inherently subjective and the categories themselves were created by the researchers
   • the nature and categorization of such letters might vary considerably with the age and sex of the violent criminal; our findings in this case are not generalizable.

Last, researchers who plan to study writers of letters to incarcerated criminals should consider sending a personality test and other questionnaires to those writers to understand this population better.

Treatment considerations
Psychiatrists treating patients who seek a romantic attachment with a violent person should consider psychotherapy as a means of treating possible character pathology. The desire for romance with a violent crimi­nal was greater among repeat writers (20%) than in initial letters (15%), suggesting that people who have a strong inclination to associate with a violent person might benefit from exploring romantic feelings in therapy. Specifically, therapists would be wise to explore with such patients the possibility that they experienced violence or verbal abuse in childhood or adulthood.

To the extent that evidence of prior abuse exists, a diagnosis of posttraumatic stress disorder (PTSD) might be appro­priate; specialized therapy for men and women with a history of abuse might be indicated. It is important to provide vali­dation for patients who are victims when they describe their abuse, and to stress that they did nothing to provoke the violence. Furthermore, investigation of why the patient feels drawn romantically toward a violent criminal is helpful, as well as an examination of how such behavior is self-defeating.

There might be value in having patients keep a journal in lieu of actually sending letters; there is evidence that “journaling” can reduce substance use recidivism.⁷ This work can be performed in conjunction with group or individual psychotherapy that addresses any history of abuse and subse­quent PTSD.

Many patients are reluctant to discuss their romantic feelings toward a violent criminal until the psychiatrist has estab­lished a strong doctor−patient relationship. Last, clinicians should not hesitate to refer these patients to a therapist who specializes in domestic violence.

Related Resource
• Marazziti D, Falaschi V, Lombardi A, et al. Stalking: a neuro­biological perspective. Riv Psichiatr. 2015;50(1):12-18.

Disclosures
The authors report no financial relationships with any company whose products are mentioned in this article or with manufacturers of competing products.

First-episode psychosis (FEP) in schizophrenia is char­acterized by high response rates to antipsychotic therapy, followed by frequent antipsychotic discon­tinuation and elevated relapse rates soon after mainte­nance treatment begins.^1,2 With subsequent episodes, time to response progressively increases and likelihood of response decreases.^3,4

To address these issues, this article—the second of 2 parts⁵—describes the rationale and evidence for using non­standard first-line antipsychotic therapies to manage FEP. Specifically, we discuss when clinicians might consider mono­therapy exceeding FDA-approved maximum dosages, combi­nation therapy, long-acting injectable antipsychotics (LAIA), or clozapine.

Monotherapy beyond FDA-approved dosages
Treatment guidelines for FEP recommend oral antipsy­chotic dosages in the lower half of the treatment range and lower than those that are required for multi-episode schizo­phrenia.^6-16 Ultimately, clinicians prescribe individualized dosages for their patients based on symptom improvement and tolerability. The optimal dosage at which to achieve a favorable D2 receptor occupancy likely will vary from patient to patient.¹⁷

To control symptoms, higher dosages may be needed than those used in FEP clinical tri­als, recommended by guidelines for FEP or multi-episode patients, or approved by the FDA. Patients seen in everyday practice may be more complicated (eg, have a comorbid condition or history of nonresponse) than study populations. Higher dosages also may be reasonable to overcome drug−drug interactions (eg, cigarette smoking-mediated cytochrome P450 1A2 induction, resulting in increased olanzapine metabolism),¹⁸ or to establish antipsychotic failure if adequate trials at lower dosages have resulted in a suboptimal response and the patient is not experiencing tolerability or safety concerns.

In a study of low-, full-, and high-dosage antipsychotic therapy in FEP, an additional 15% of patients responded to higher dos­ages of olanzapine and risperidone after failing to respond to a standard dosage.¹⁹ A study of data from the Recovery After an Initial Schizophrenia Episode Project’s Early Treatment Program (RAISE-ETP) found that, of participants identified who may benefit from therapy modification, 8.8% were pre­scribed an antipsychotic (often, olanzapine, risperidone, and haloperidol) at a higher-than-recommended dosage.²⁰ Of note, only olanzapine was prescribed at higher than FDA-approved dosages.

Antipsychotic combination therapy
Prescribing combinations of antipsychot­ics—antipsychotic polypharmacy (APP)— has a negative connotation because of limited efficacy and safety data,²¹ and limited endorsement in schizophrenia treatment guidelines.^9,13 Caution with APP is war­ranted; a complex medication regimen may increase the potential for adverse effects, poorer adherence, and adverse drug-drug interactions.⁹ APP has been shown to inde­pendently predict both shorter treatment duration and discontinuation before 1 year.²²

Nonetheless, the clinician and patient may share the decision to implement APP and observe whether benefits outweigh risks in situations such as:
   • to optimize neuroreceptor occupancy and targets (eg, attempting to achieve ade­quate D2 receptor blockade while minimiz­ing side effects secondary to binding other receptors)
   • to manage co-existing symptom domains (eg, mood changes, aggression, negative symptoms, disorganization, and cognitive deficits)
   • to mitigate antipsychotic-induced side effects (eg, initiating aripiprazole to treat hyperprolactinemia induced by another anti­psychotic to which the patient has achieved a favorable response).²³

Clinicians report using APP to treat as many as 50% of patients with a history of multiple psychotic episodes.²³ For FEP patients, 23% of participants in the RAISE-ETP trial who were identified as possibly benefiting from therapy modification were prescribed APP.²⁰ Regrettably, research­ers have not found evidence to support a reported rationale for using APP—that lower dosages of individual antipsychotics when used in combination may avoid high-dosage prescriptions.²⁴

Before implementing APP, thoroughly explore and manage reasons for a patient’s suboptimal response to monotherapy.25 An adequate trial with any antipsychotic should be at the highest tolerated dosage for 12 to 16 weeks. Be mindful that response to an APP trial may be the result of additional time on the original antipsychotic.

Long-acting injectable antipsychotics in FEP
Guideline recommendations. Most older guidelines for schizophrenia treat­ment suggest LAIA after multiple relapses related to medication nonadherence or when a patient prefers injected medica­tion (Table 1).^6-13 Expert consensus guide­lines also recommend considering LAIA in patients who lack insight into their illness. The Texas Medication Algorithm Project (TMAP) guidelines⁷ state LAIA can be con­sidered for inadequate adherence at any stage, whereas the 2010 British Association for Psychopharmacology (BAP) guide­lines⁹ express uncertainty about their use in FEP, because of limited evidence. Both the BAP and National Institute for Health and Care Excellence guidelines¹³ urge cli­nicians to consider LAIA when avoiding nonadherence is a treatment priority.

Barriers to LAIA use include:
   • slow dosage titration and increased time to reach steady state drug level
   • oral supplementation for some (eg, risperidone microspheres and aripiprazole long-acting injectable)
   • logistical challenges for some (eg, 3-hour post-injection monitoring for delir­ium sedation syndrome with olanzapine pamoate)
   • additional planning to coordinate care for scheduled injections
   • higher expenses up front
   • local injection site reactions
   • dosage adjustment difficulties if adverse effects occur.^28,29

Adoption rates of LAIA are low, especially for FEP.³⁰ Most surveys indicate that (1) physi­cians believe LAIA treatment is ineffective for FEP31 and (2) patients do not prefer injectable to oral antipsychotics,³² despite evidence to the contrary.^33,34 A survey of 198 psychiatrists identified 3 factors that influenced their deci­sions against using LAIA patients with FEP:
   • limited availability of SGA depot formu­lations (4, to date, in the United States)
   • frequent rejection by the patient when LAIA is offered without adequate expla­nation or encouragement
   • skepticism of FEP patients (and their family) who lack experience with relapse.³⁵

In reality, when SGA depots were intro­duced in the United Kingdom, prescribing rates of LAIA did not increase. As for patient rejection being a major reason for not pre­scribing LAIA, few patients (5% to 36%) are offered depot injections, particularly in FEP.²⁹ Most patients using LAIA are chronic, multi-episode, violent people who are receiving medications involuntarily.²⁹ Interestingly, this survey did not find 2 factors to be influential in psychiatrists’ decision not to use LAIA in FEP:
   • guidelines do not explicitly recommend depot treatment in FEP
   • treatment in FEP may be limited to 1 year, therefore depot administration is not worthwhile.³⁵

Preliminary evidence. At least a dozen stud­ies have explored LAIA treatment for FEP, with the use of fluphenazine decanoate,³⁶ per­phenazine enanthate³⁷ (discontinued), and risperidone microspheres.^37-48 The research demonstrates the efficacy and safety of LAIA in FEP as measured by these endpoints:
   • improved symptom control^{38,40-43,46,48}
   • adherence^43,44,48
   • reduced relapse rates^37,43 and rehospitalizations^37,47
   • lesser reductions in white matter brain volume⁴⁵
   • no differences in extrapyramidal side effects or prolactin-associated adverse effects.⁴⁸

A few small studies demonstrate signifi­cant differences in outcomes between ris­peridone LAIA and oral comparator groups (Table 3).^43-45 Ongoing studies of LAIA use in FEP are comparing paliperidone palmitate with risperidone microspheres and other oral antipsychotics.^49-51 No stud­ies are examining olanzapine pamoate in FEP, likely because several guidelines do not recommended its use. No studies have been published regarding aripiprazole long-acting injectable in FEP. This LAIA formulation was approved in February 2013, and robust studies of the oral formulation in FEP are limited.⁵²

Relapses begin within a few months of illness stabilization after FEP, and >50% of patients relapse within 1 or 2 years²—the recommended minimum treatment dura­tion for FEP.^8,9,13 The use of LAIA is advis­able in any patient with schizophrenia for whom long-term antipsychotic therapy is indicated.⁵⁶ LAIA administration require­ments objectively track medication adher­ence, which allows clinicians to be proactive in relapse prevention. Not using an inter­vention in FEP that improves adherence and decreases relapse rates contradicts our goal of instituting early, effective treatment to improve long-term functional outcomes (Figure 2).²⁹

Considering clozapine in FEP
Guideline recommendations. Schizo-phrenia treatment guidelines and FDA labeling⁵⁷ reserve clozapine for third-line treatment of refractory schizophrenia after 2 adequate antipsychotic trials have failed despite optimal dosing (Table 1).^6-13 Some guidelines specify 1 of the 2 failed anti­psychotic trials must include an SGA.^{6,7,10,11,13-16} Most say clozapine may be considered in patients with chronic aggression or hostility,^7-9,14,16 or suicidal thoughts and behav­iors.^6-8,14,16 TMAP guidelines recommend a clozapine trial with concomitant substance abuse, persistent positive symptoms during 2 years of consistent medication treatment, and after 5 years of inadequate response (“treatment resistance”), regardless of the number of antipsychotic trials.⁷ ­

Rationale and concerns. Clozapine is a superior choice for treatment-refractory delusions or hallucinations of schizophrenia, because it markedly enhances the response rate to antipsychotic therapy.⁵⁸ Researchers therefore have investigated whether clozap­ine, compared with other antipsychotics, would yield more favorable initial and long-term outcomes when used first-line in FEP.

Preliminary evidence. Five studies have explored the use of clozapine as first-line therapy in FEP (Table 4).^59-63 Interpreting the results is difficult because clozapine trials may be brief (mostly, 12 to 52 weeks); lack a comparator arm; suffer from a high attrition rate; enroll few patients; and lack potentially important outcome measures such as nega­tive symptoms, suicidality, and functional assessment.

At present, clozapine has not been shown superior to other antipsychotics as a first-line treatment for FEP. Research does underscore the importance of a clozapine trial as third-line treatment for FEP patients who have not responded well to 2 SGA trials.⁶³ Many of these nonresponders (77%) have demon­strated a favorable response when promptly switched to clozapine.⁶⁴

Discussion and recommendations. The limited evidence argues against using clo­zapine earlier than as third-line treatment in FEP. Perhaps the high treatment response that characterizes FEP creates a ceiling effect that obscures differences in antipsychotic efficacy at this stage.⁶⁵ Clozapine use as first-line treatment should be re-evaluated with more robust methodology. One approach could be to assess its benefit in FEP by the duration of untreated psychosis.

The odds of achieving remission have been shown to decrease by 15% for each year that psychosis has not been treated.⁵⁹ Studies exploring the use of clozapine as a second-line agent for FEP also are warranted, as anti­psychotic response during subsequent trials is substantially reduced. In fact, the Scottish Intercollegiate Guidelines Network guide­lines recommend this as an area for future research.¹¹

For now, clozapine should continue to be reserved as second- or third-line treatment in a patient with FEP. The risks of clozap­ine’s potentially serious adverse effects (eg, agranulocytosis, seizures, obesity, diabe­tes, dyslipidemia, myocarditis, pancreatitis, hypotension, sialorrhea, severe sedation, ileus) can be justified only in the treatment of severe and persistent psychotic symptoms.⁵⁷

Bottom Line
Nonstandard use of antipsychotic monotherapy dosages beyond the approved FDA limit and combination antipsychotic therapy may be reasonable for select first-episode psychosis (FEP) patients. Strongly consider long-acting injectable antipsychotics in FEP to proactively combat the high relapse rate and more easily identify antipsychotic failure. Continue to use clozapine as second- or third-line therapy in FEP: Studies have not found that it is more efficacious than other antipsychotics for first-line use.

Related Resource
• Recovery After an Initial Schizophrenia Episode (RAISE) Project Early Treatment Program. National Institute of Mental Health. http://raiseetp.org.

Drug Brand Names
Aripiprazole • Abilify, Abilify Maintena      
Chlorpromazine • Thorazine                     
Clozapine • Clozaril                              
Fluphenazine decanoate • Prolixin-D      
Haloperidol • Haldol                                      
Haloperidol decanoate • Haldol-D       
Olanzapine • Zyprexa
Olanzapine pamoate • Zyprexa Relprevv
Paliperidone palmitate • Invega Sustenna
Quetiapine • Seroquel
Risperidone • Risperdal
Risperidone microspheres • Risperdal Consta

Disclosures
Dr. Gardner reports no financial relationships with any companies whose products are mentioned in this article or with manufacturers of competing products. Dr. Nasrallah is a consultant to Acadia, Alkermes, Lundbeck, Janssen, Merck, Otsuka, and Sunovion, and is a speaker for Alkermes, Lundbeck, Janssen, Otsuka, and Sunovion.

First-episode psychosis (FEP) in schizophrenia is char­acterized by high response rates to antipsychotic therapy, followed by frequent antipsychotic discon­tinuation and elevated relapse rates soon after mainte­nance treatment begins.^1,2 With subsequent episodes, time to response progressively increases and likelihood of response decreases.^3,4

To address these issues, this article—the second of 2 parts⁵—describes the rationale and evidence for using non­standard first-line antipsychotic therapies to manage FEP. Specifically, we discuss when clinicians might consider mono­therapy exceeding FDA-approved maximum dosages, combi­nation therapy, long-acting injectable antipsychotics (LAIA), or clozapine.

Monotherapy beyond FDA-approved dosages
Treatment guidelines for FEP recommend oral antipsy­chotic dosages in the lower half of the treatment range and lower than those that are required for multi-episode schizo­phrenia.^6-16 Ultimately, clinicians prescribe individualized dosages for their patients based on symptom improvement and tolerability. The optimal dosage at which to achieve a favorable D2 receptor occupancy likely will vary from patient to patient.¹⁷

To control symptoms, higher dosages may be needed than those used in FEP clinical tri­als, recommended by guidelines for FEP or multi-episode patients, or approved by the FDA. Patients seen in everyday practice may be more complicated (eg, have a comorbid condition or history of nonresponse) than study populations. Higher dosages also may be reasonable to overcome drug−drug interactions (eg, cigarette smoking-mediated cytochrome P450 1A2 induction, resulting in increased olanzapine metabolism),¹⁸ or to establish antipsychotic failure if adequate trials at lower dosages have resulted in a suboptimal response and the patient is not experiencing tolerability or safety concerns.

In a study of low-, full-, and high-dosage antipsychotic therapy in FEP, an additional 15% of patients responded to higher dos­ages of olanzapine and risperidone after failing to respond to a standard dosage.¹⁹ A study of data from the Recovery After an Initial Schizophrenia Episode Project’s Early Treatment Program (RAISE-ETP) found that, of participants identified who may benefit from therapy modification, 8.8% were pre­scribed an antipsychotic (often, olanzapine, risperidone, and haloperidol) at a higher-than-recommended dosage.²⁰ Of note, only olanzapine was prescribed at higher than FDA-approved dosages.

Antipsychotic combination therapy
Prescribing combinations of antipsychot­ics—antipsychotic polypharmacy (APP)— has a negative connotation because of limited efficacy and safety data,²¹ and limited endorsement in schizophrenia treatment guidelines.^9,13 Caution with APP is war­ranted; a complex medication regimen may increase the potential for adverse effects, poorer adherence, and adverse drug-drug interactions.⁹ APP has been shown to inde­pendently predict both shorter treatment duration and discontinuation before 1 year.²²

Nonetheless, the clinician and patient may share the decision to implement APP and observe whether benefits outweigh risks in situations such as:
   • to optimize neuroreceptor occupancy and targets (eg, attempting to achieve ade­quate D2 receptor blockade while minimiz­ing side effects secondary to binding other receptors)
   • to manage co-existing symptom domains (eg, mood changes, aggression, negative symptoms, disorganization, and cognitive deficits)
   • to mitigate antipsychotic-induced side effects (eg, initiating aripiprazole to treat hyperprolactinemia induced by another anti­psychotic to which the patient has achieved a favorable response).²³

Clinicians report using APP to treat as many as 50% of patients with a history of multiple psychotic episodes.²³ For FEP patients, 23% of participants in the RAISE-ETP trial who were identified as possibly benefiting from therapy modification were prescribed APP.²⁰ Regrettably, research­ers have not found evidence to support a reported rationale for using APP—that lower dosages of individual antipsychotics when used in combination may avoid high-dosage prescriptions.²⁴

Before implementing APP, thoroughly explore and manage reasons for a patient’s suboptimal response to monotherapy.25 An adequate trial with any antipsychotic should be at the highest tolerated dosage for 12 to 16 weeks. Be mindful that response to an APP trial may be the result of additional time on the original antipsychotic.

Long-acting injectable antipsychotics in FEP
Guideline recommendations. Most older guidelines for schizophrenia treat­ment suggest LAIA after multiple relapses related to medication nonadherence or when a patient prefers injected medica­tion (Table 1).^6-13 Expert consensus guide­lines also recommend considering LAIA in patients who lack insight into their illness. The Texas Medication Algorithm Project (TMAP) guidelines⁷ state LAIA can be con­sidered for inadequate adherence at any stage, whereas the 2010 British Association for Psychopharmacology (BAP) guide­lines⁹ express uncertainty about their use in FEP, because of limited evidence. Both the BAP and National Institute for Health and Care Excellence guidelines¹³ urge cli­nicians to consider LAIA when avoiding nonadherence is a treatment priority.

Barriers to LAIA use include:
   • slow dosage titration and increased time to reach steady state drug level
   • oral supplementation for some (eg, risperidone microspheres and aripiprazole long-acting injectable)
   • logistical challenges for some (eg, 3-hour post-injection monitoring for delir­ium sedation syndrome with olanzapine pamoate)
   • additional planning to coordinate care for scheduled injections
   • higher expenses up front
   • local injection site reactions
   • dosage adjustment difficulties if adverse effects occur.^28,29

Adoption rates of LAIA are low, especially for FEP.³⁰ Most surveys indicate that (1) physi­cians believe LAIA treatment is ineffective for FEP31 and (2) patients do not prefer injectable to oral antipsychotics,³² despite evidence to the contrary.^33,34 A survey of 198 psychiatrists identified 3 factors that influenced their deci­sions against using LAIA patients with FEP:
   • limited availability of SGA depot formu­lations (4, to date, in the United States)
   • frequent rejection by the patient when LAIA is offered without adequate expla­nation or encouragement
   • skepticism of FEP patients (and their family) who lack experience with relapse.³⁵

In reality, when SGA depots were intro­duced in the United Kingdom, prescribing rates of LAIA did not increase. As for patient rejection being a major reason for not pre­scribing LAIA, few patients (5% to 36%) are offered depot injections, particularly in FEP.²⁹ Most patients using LAIA are chronic, multi-episode, violent people who are receiving medications involuntarily.²⁹ Interestingly, this survey did not find 2 factors to be influential in psychiatrists’ decision not to use LAIA in FEP:
   • guidelines do not explicitly recommend depot treatment in FEP
   • treatment in FEP may be limited to 1 year, therefore depot administration is not worthwhile.³⁵

Preliminary evidence. At least a dozen stud­ies have explored LAIA treatment for FEP, with the use of fluphenazine decanoate,³⁶ per­phenazine enanthate³⁷ (discontinued), and risperidone microspheres.^37-48 The research demonstrates the efficacy and safety of LAIA in FEP as measured by these endpoints:
   • improved symptom control^{38,40-43,46,48}
   • adherence^43,44,48
   • reduced relapse rates^37,43 and rehospitalizations^37,47
   • lesser reductions in white matter brain volume⁴⁵
   • no differences in extrapyramidal side effects or prolactin-associated adverse effects.⁴⁸

A few small studies demonstrate signifi­cant differences in outcomes between ris­peridone LAIA and oral comparator groups (Table 3).^43-45 Ongoing studies of LAIA use in FEP are comparing paliperidone palmitate with risperidone microspheres and other oral antipsychotics.^49-51 No stud­ies are examining olanzapine pamoate in FEP, likely because several guidelines do not recommended its use. No studies have been published regarding aripiprazole long-acting injectable in FEP. This LAIA formulation was approved in February 2013, and robust studies of the oral formulation in FEP are limited.⁵²

Relapses begin within a few months of illness stabilization after FEP, and >50% of patients relapse within 1 or 2 years²—the recommended minimum treatment dura­tion for FEP.^8,9,13 The use of LAIA is advis­able in any patient with schizophrenia for whom long-term antipsychotic therapy is indicated.⁵⁶ LAIA administration require­ments objectively track medication adher­ence, which allows clinicians to be proactive in relapse prevention. Not using an inter­vention in FEP that improves adherence and decreases relapse rates contradicts our goal of instituting early, effective treatment to improve long-term functional outcomes (Figure 2).²⁹

Considering clozapine in FEP
Guideline recommendations. Schizo-phrenia treatment guidelines and FDA labeling⁵⁷ reserve clozapine for third-line treatment of refractory schizophrenia after 2 adequate antipsychotic trials have failed despite optimal dosing (Table 1).^6-13 Some guidelines specify 1 of the 2 failed anti­psychotic trials must include an SGA.^{6,7,10,11,13-16} Most say clozapine may be considered in patients with chronic aggression or hostility,^7-9,14,16 or suicidal thoughts and behav­iors.^6-8,14,16 TMAP guidelines recommend a clozapine trial with concomitant substance abuse, persistent positive symptoms during 2 years of consistent medication treatment, and after 5 years of inadequate response (“treatment resistance”), regardless of the number of antipsychotic trials.⁷ ­

Rationale and concerns. Clozapine is a superior choice for treatment-refractory delusions or hallucinations of schizophrenia, because it markedly enhances the response rate to antipsychotic therapy.⁵⁸ Researchers therefore have investigated whether clozap­ine, compared with other antipsychotics, would yield more favorable initial and long-term outcomes when used first-line in FEP.

Preliminary evidence. Five studies have explored the use of clozapine as first-line therapy in FEP (Table 4).^59-63 Interpreting the results is difficult because clozapine trials may be brief (mostly, 12 to 52 weeks); lack a comparator arm; suffer from a high attrition rate; enroll few patients; and lack potentially important outcome measures such as nega­tive symptoms, suicidality, and functional assessment.

At present, clozapine has not been shown superior to other antipsychotics as a first-line treatment for FEP. Research does underscore the importance of a clozapine trial as third-line treatment for FEP patients who have not responded well to 2 SGA trials.⁶³ Many of these nonresponders (77%) have demon­strated a favorable response when promptly switched to clozapine.⁶⁴

Discussion and recommendations. The limited evidence argues against using clo­zapine earlier than as third-line treatment in FEP. Perhaps the high treatment response that characterizes FEP creates a ceiling effect that obscures differences in antipsychotic efficacy at this stage.⁶⁵ Clozapine use as first-line treatment should be re-evaluated with more robust methodology. One approach could be to assess its benefit in FEP by the duration of untreated psychosis.

The odds of achieving remission have been shown to decrease by 15% for each year that psychosis has not been treated.⁵⁹ Studies exploring the use of clozapine as a second-line agent for FEP also are warranted, as anti­psychotic response during subsequent trials is substantially reduced. In fact, the Scottish Intercollegiate Guidelines Network guide­lines recommend this as an area for future research.¹¹

For now, clozapine should continue to be reserved as second- or third-line treatment in a patient with FEP. The risks of clozap­ine’s potentially serious adverse effects (eg, agranulocytosis, seizures, obesity, diabe­tes, dyslipidemia, myocarditis, pancreatitis, hypotension, sialorrhea, severe sedation, ileus) can be justified only in the treatment of severe and persistent psychotic symptoms.⁵⁷

Bottom Line
Nonstandard use of antipsychotic monotherapy dosages beyond the approved FDA limit and combination antipsychotic therapy may be reasonable for select first-episode psychosis (FEP) patients. Strongly consider long-acting injectable antipsychotics in FEP to proactively combat the high relapse rate and more easily identify antipsychotic failure. Continue to use clozapine as second- or third-line therapy in FEP: Studies have not found that it is more efficacious than other antipsychotics for first-line use.

Related Resource
• Recovery After an Initial Schizophrenia Episode (RAISE) Project Early Treatment Program. National Institute of Mental Health. http://raiseetp.org.

Drug Brand Names
Aripiprazole • Abilify, Abilify Maintena      
Chlorpromazine • Thorazine                     
Clozapine • Clozaril                              
Fluphenazine decanoate • Prolixin-D      
Haloperidol • Haldol                                      
Haloperidol decanoate • Haldol-D       
Olanzapine • Zyprexa
Olanzapine pamoate • Zyprexa Relprevv
Paliperidone palmitate • Invega Sustenna
Quetiapine • Seroquel
Risperidone • Risperdal
Risperidone microspheres • Risperdal Consta

Disclosures
Dr. Gardner reports no financial relationships with any companies whose products are mentioned in this article or with manufacturers of competing products. Dr. Nasrallah is a consultant to Acadia, Alkermes, Lundbeck, Janssen, Merck, Otsuka, and Sunovion, and is a speaker for Alkermes, Lundbeck, Janssen, Otsuka, and Sunovion.

First-episode psychosis (FEP) in schizophrenia is char­acterized by high response rates to antipsychotic therapy, followed by frequent antipsychotic discon­tinuation and elevated relapse rates soon after mainte­nance treatment begins.^1,2 With subsequent episodes, time to response progressively increases and likelihood of response decreases.^3,4

To address these issues, this article—the second of 2 parts⁵—describes the rationale and evidence for using non­standard first-line antipsychotic therapies to manage FEP. Specifically, we discuss when clinicians might consider mono­therapy exceeding FDA-approved maximum dosages, combi­nation therapy, long-acting injectable antipsychotics (LAIA), or clozapine.

Monotherapy beyond FDA-approved dosages
Treatment guidelines for FEP recommend oral antipsy­chotic dosages in the lower half of the treatment range and lower than those that are required for multi-episode schizo­phrenia.^6-16 Ultimately, clinicians prescribe individualized dosages for their patients based on symptom improvement and tolerability. The optimal dosage at which to achieve a favorable D2 receptor occupancy likely will vary from patient to patient.¹⁷

To control symptoms, higher dosages may be needed than those used in FEP clinical tri­als, recommended by guidelines for FEP or multi-episode patients, or approved by the FDA. Patients seen in everyday practice may be more complicated (eg, have a comorbid condition or history of nonresponse) than study populations. Higher dosages also may be reasonable to overcome drug−drug interactions (eg, cigarette smoking-mediated cytochrome P450 1A2 induction, resulting in increased olanzapine metabolism),¹⁸ or to establish antipsychotic failure if adequate trials at lower dosages have resulted in a suboptimal response and the patient is not experiencing tolerability or safety concerns.

In a study of low-, full-, and high-dosage antipsychotic therapy in FEP, an additional 15% of patients responded to higher dos­ages of olanzapine and risperidone after failing to respond to a standard dosage.¹⁹ A study of data from the Recovery After an Initial Schizophrenia Episode Project’s Early Treatment Program (RAISE-ETP) found that, of participants identified who may benefit from therapy modification, 8.8% were pre­scribed an antipsychotic (often, olanzapine, risperidone, and haloperidol) at a higher-than-recommended dosage.²⁰ Of note, only olanzapine was prescribed at higher than FDA-approved dosages.

Antipsychotic combination therapy
Prescribing combinations of antipsychot­ics—antipsychotic polypharmacy (APP)— has a negative connotation because of limited efficacy and safety data,²¹ and limited endorsement in schizophrenia treatment guidelines.^9,13 Caution with APP is war­ranted; a complex medication regimen may increase the potential for adverse effects, poorer adherence, and adverse drug-drug interactions.⁹ APP has been shown to inde­pendently predict both shorter treatment duration and discontinuation before 1 year.²²

Nonetheless, the clinician and patient may share the decision to implement APP and observe whether benefits outweigh risks in situations such as:
   • to optimize neuroreceptor occupancy and targets (eg, attempting to achieve ade­quate D2 receptor blockade while minimiz­ing side effects secondary to binding other receptors)
   • to manage co-existing symptom domains (eg, mood changes, aggression, negative symptoms, disorganization, and cognitive deficits)
   • to mitigate antipsychotic-induced side effects (eg, initiating aripiprazole to treat hyperprolactinemia induced by another anti­psychotic to which the patient has achieved a favorable response).²³

Clinicians report using APP to treat as many as 50% of patients with a history of multiple psychotic episodes.²³ For FEP patients, 23% of participants in the RAISE-ETP trial who were identified as possibly benefiting from therapy modification were prescribed APP.²⁰ Regrettably, research­ers have not found evidence to support a reported rationale for using APP—that lower dosages of individual antipsychotics when used in combination may avoid high-dosage prescriptions.²⁴

Before implementing APP, thoroughly explore and manage reasons for a patient’s suboptimal response to monotherapy.25 An adequate trial with any antipsychotic should be at the highest tolerated dosage for 12 to 16 weeks. Be mindful that response to an APP trial may be the result of additional time on the original antipsychotic.

Long-acting injectable antipsychotics in FEP
Guideline recommendations. Most older guidelines for schizophrenia treat­ment suggest LAIA after multiple relapses related to medication nonadherence or when a patient prefers injected medica­tion (Table 1).^6-13 Expert consensus guide­lines also recommend considering LAIA in patients who lack insight into their illness. The Texas Medication Algorithm Project (TMAP) guidelines⁷ state LAIA can be con­sidered for inadequate adherence at any stage, whereas the 2010 British Association for Psychopharmacology (BAP) guide­lines⁹ express uncertainty about their use in FEP, because of limited evidence. Both the BAP and National Institute for Health and Care Excellence guidelines¹³ urge cli­nicians to consider LAIA when avoiding nonadherence is a treatment priority.

Barriers to LAIA use include:
   • slow dosage titration and increased time to reach steady state drug level
   • oral supplementation for some (eg, risperidone microspheres and aripiprazole long-acting injectable)
   • logistical challenges for some (eg, 3-hour post-injection monitoring for delir­ium sedation syndrome with olanzapine pamoate)
   • additional planning to coordinate care for scheduled injections
   • higher expenses up front
   • local injection site reactions
   • dosage adjustment difficulties if adverse effects occur.^28,29

Adoption rates of LAIA are low, especially for FEP.³⁰ Most surveys indicate that (1) physi­cians believe LAIA treatment is ineffective for FEP31 and (2) patients do not prefer injectable to oral antipsychotics,³² despite evidence to the contrary.^33,34 A survey of 198 psychiatrists identified 3 factors that influenced their deci­sions against using LAIA patients with FEP:
   • limited availability of SGA depot formu­lations (4, to date, in the United States)
   • frequent rejection by the patient when LAIA is offered without adequate expla­nation or encouragement
   • skepticism of FEP patients (and their family) who lack experience with relapse.³⁵

In reality, when SGA depots were intro­duced in the United Kingdom, prescribing rates of LAIA did not increase. As for patient rejection being a major reason for not pre­scribing LAIA, few patients (5% to 36%) are offered depot injections, particularly in FEP.²⁹ Most patients using LAIA are chronic, multi-episode, violent people who are receiving medications involuntarily.²⁹ Interestingly, this survey did not find 2 factors to be influential in psychiatrists’ decision not to use LAIA in FEP:
   • guidelines do not explicitly recommend depot treatment in FEP
   • treatment in FEP may be limited to 1 year, therefore depot administration is not worthwhile.³⁵

Preliminary evidence. At least a dozen stud­ies have explored LAIA treatment for FEP, with the use of fluphenazine decanoate,³⁶ per­phenazine enanthate³⁷ (discontinued), and risperidone microspheres.^37-48 The research demonstrates the efficacy and safety of LAIA in FEP as measured by these endpoints:
   • improved symptom control^{38,40-43,46,48}
   • adherence^43,44,48
   • reduced relapse rates^37,43 and rehospitalizations^37,47
   • lesser reductions in white matter brain volume⁴⁵
   • no differences in extrapyramidal side effects or prolactin-associated adverse effects.⁴⁸

A few small studies demonstrate signifi­cant differences in outcomes between ris­peridone LAIA and oral comparator groups (Table 3).^43-45 Ongoing studies of LAIA use in FEP are comparing paliperidone palmitate with risperidone microspheres and other oral antipsychotics.^49-51 No stud­ies are examining olanzapine pamoate in FEP, likely because several guidelines do not recommended its use. No studies have been published regarding aripiprazole long-acting injectable in FEP. This LAIA formulation was approved in February 2013, and robust studies of the oral formulation in FEP are limited.⁵²

Relapses begin within a few months of illness stabilization after FEP, and >50% of patients relapse within 1 or 2 years²—the recommended minimum treatment dura­tion for FEP.^8,9,13 The use of LAIA is advis­able in any patient with schizophrenia for whom long-term antipsychotic therapy is indicated.⁵⁶ LAIA administration require­ments objectively track medication adher­ence, which allows clinicians to be proactive in relapse prevention. Not using an inter­vention in FEP that improves adherence and decreases relapse rates contradicts our goal of instituting early, effective treatment to improve long-term functional outcomes (Figure 2).²⁹

Considering clozapine in FEP
Guideline recommendations. Schizo-phrenia treatment guidelines and FDA labeling⁵⁷ reserve clozapine for third-line treatment of refractory schizophrenia after 2 adequate antipsychotic trials have failed despite optimal dosing (Table 1).^6-13 Some guidelines specify 1 of the 2 failed anti­psychotic trials must include an SGA.^{6,7,10,11,13-16} Most say clozapine may be considered in patients with chronic aggression or hostility,^7-9,14,16 or suicidal thoughts and behav­iors.^6-8,14,16 TMAP guidelines recommend a clozapine trial with concomitant substance abuse, persistent positive symptoms during 2 years of consistent medication treatment, and after 5 years of inadequate response (“treatment resistance”), regardless of the number of antipsychotic trials.⁷ ­

Rationale and concerns. Clozapine is a superior choice for treatment-refractory delusions or hallucinations of schizophrenia, because it markedly enhances the response rate to antipsychotic therapy.⁵⁸ Researchers therefore have investigated whether clozap­ine, compared with other antipsychotics, would yield more favorable initial and long-term outcomes when used first-line in FEP.

Preliminary evidence. Five studies have explored the use of clozapine as first-line therapy in FEP (Table 4).^59-63 Interpreting the results is difficult because clozapine trials may be brief (mostly, 12 to 52 weeks); lack a comparator arm; suffer from a high attrition rate; enroll few patients; and lack potentially important outcome measures such as nega­tive symptoms, suicidality, and functional assessment.

At present, clozapine has not been shown superior to other antipsychotics as a first-line treatment for FEP. Research does underscore the importance of a clozapine trial as third-line treatment for FEP patients who have not responded well to 2 SGA trials.⁶³ Many of these nonresponders (77%) have demon­strated a favorable response when promptly switched to clozapine.⁶⁴

Discussion and recommendations. The limited evidence argues against using clo­zapine earlier than as third-line treatment in FEP. Perhaps the high treatment response that characterizes FEP creates a ceiling effect that obscures differences in antipsychotic efficacy at this stage.⁶⁵ Clozapine use as first-line treatment should be re-evaluated with more robust methodology. One approach could be to assess its benefit in FEP by the duration of untreated psychosis.

The odds of achieving remission have been shown to decrease by 15% for each year that psychosis has not been treated.⁵⁹ Studies exploring the use of clozapine as a second-line agent for FEP also are warranted, as anti­psychotic response during subsequent trials is substantially reduced. In fact, the Scottish Intercollegiate Guidelines Network guide­lines recommend this as an area for future research.¹¹

For now, clozapine should continue to be reserved as second- or third-line treatment in a patient with FEP. The risks of clozap­ine’s potentially serious adverse effects (eg, agranulocytosis, seizures, obesity, diabe­tes, dyslipidemia, myocarditis, pancreatitis, hypotension, sialorrhea, severe sedation, ileus) can be justified only in the treatment of severe and persistent psychotic symptoms.⁵⁷

Bottom Line
Nonstandard use of antipsychotic monotherapy dosages beyond the approved FDA limit and combination antipsychotic therapy may be reasonable for select first-episode psychosis (FEP) patients. Strongly consider long-acting injectable antipsychotics in FEP to proactively combat the high relapse rate and more easily identify antipsychotic failure. Continue to use clozapine as second- or third-line therapy in FEP: Studies have not found that it is more efficacious than other antipsychotics for first-line use.

Related Resource
• Recovery After an Initial Schizophrenia Episode (RAISE) Project Early Treatment Program. National Institute of Mental Health. http://raiseetp.org.

Drug Brand Names
Aripiprazole • Abilify, Abilify Maintena      
Chlorpromazine • Thorazine                     
Clozapine • Clozaril                              
Fluphenazine decanoate • Prolixin-D      
Haloperidol • Haldol                                      
Haloperidol decanoate • Haldol-D       
Olanzapine • Zyprexa
Olanzapine pamoate • Zyprexa Relprevv
Paliperidone palmitate • Invega Sustenna
Quetiapine • Seroquel
Risperidone • Risperdal
Risperidone microspheres • Risperdal Consta

Disclosures
Dr. Gardner reports no financial relationships with any companies whose products are mentioned in this article or with manufacturers of competing products. Dr. Nasrallah is a consultant to Acadia, Alkermes, Lundbeck, Janssen, Merck, Otsuka, and Sunovion, and is a speaker for Alkermes, Lundbeck, Janssen, Otsuka, and Sunovion.