Current Psychiatry

At a time when 28.4% of psychiatric practices provided no psychotherapy during a typical week,¹ applying psychotherapy principles to ethical judgments involving psychiatrists is inappropriate and should be abandoned (“Psychiatrist/patient boundaries: When it’s OK to stretch the line” Current Psychiatry, August 2008).

The American Psychiatric Association should revamp its ethics annotations by removing psychoanalytic references and terms such as “identification” to make these guidelines relevant to all psychiatric practices, even those without psychotherapy.

H. Berryman Edwards, MD
Bellevue, WA

At a time when 28.4% of psychiatric practices provided no psychotherapy during a typical week,¹ applying psychotherapy principles to ethical judgments involving psychiatrists is inappropriate and should be abandoned (“Psychiatrist/patient boundaries: When it’s OK to stretch the line” Current Psychiatry, August 2008).

The American Psychiatric Association should revamp its ethics annotations by removing psychoanalytic references and terms such as “identification” to make these guidelines relevant to all psychiatric practices, even those without psychotherapy.

H. Berryman Edwards, MD
Bellevue, WA

At a time when 28.4% of psychiatric practices provided no psychotherapy during a typical week,¹ applying psychotherapy principles to ethical judgments involving psychiatrists is inappropriate and should be abandoned (“Psychiatrist/patient boundaries: When it’s OK to stretch the line” Current Psychiatry, August 2008).

The American Psychiatric Association should revamp its ethics annotations by removing psychoanalytic references and terms such as “identification” to make these guidelines relevant to all psychiatric practices, even those without psychotherapy.

H. Berryman Edwards, MD
Bellevue, WA

The article “6 screening questions for military veterans” (Pearls, Current Psychiatry, September 2008) is a well written and thorough review for clinicians treating active duty personnel and military veterans. A very important and clinically relevant seventh question is to ask about traumatic brain injury (TBI) that may have occurred during training or deployment to Iraq or Afghanistan. One tool clinicians can use to screen these patients is the Brief Traumatic Brain Injury Screen developed by the Defense and Veterans Brain Injury Center.¹ Coupled with a clinical interview, this 3-question survey will assist clinicians in identifying possible TBI patients. Referring TBI patients to a polytrauma rehabilitative center through the Department of Veterans Affairs is part of a comprehensive treatment plan.²

Timothy Berigan, MD
Psychiatrist
Raymond W. Bliss Army Health Center
Fort Huachuca, AZ

1. Schwab KA, Baker G, Ivins B, et al. The Brief Traumatic Brain Injury Screen (BTBIS): investigating the validity of a self-report instrument for detecting traumatic brain injury (TBI) in troops returning from deployment in Afghanistan and Iraq. Neurology 2006;66(5)(suppl 2):A235.

2. Friedmann-Sánchez G, Sayer NA, Pickett T. Provider perspectives on rehabilitation of patients with polytrauma. Arch Phys Med Rehabil 2008;89(1):171-8.

Dr. Barry responds

Mild traumatic brain injury (mTBI) is one of the signature diagnoses of the conflicts in Iraq and Afghanistan, and we’re learning more about the pathogenesis, manifestations, sequelae, and treatment of this syndrome. Most veterans have basic knowledge about this condition because the military educates and screens all personnel for mTBI and employs aggressive, multi-disciplinary treatment plans for those in need. Thus, clinical interview, physical examination, and screening for mTBI—as suggested by Dr. Berigan—might yield useful information for clinicians treating veterans.

Matthew James Barry, DO
Chief of psychiatric services
U.S. Army’s Medical Department Activity
Fort Huachuca, AZ

The article “6 screening questions for military veterans” (Pearls, Current Psychiatry, September 2008) is a well written and thorough review for clinicians treating active duty personnel and military veterans. A very important and clinically relevant seventh question is to ask about traumatic brain injury (TBI) that may have occurred during training or deployment to Iraq or Afghanistan. One tool clinicians can use to screen these patients is the Brief Traumatic Brain Injury Screen developed by the Defense and Veterans Brain Injury Center.¹ Coupled with a clinical interview, this 3-question survey will assist clinicians in identifying possible TBI patients. Referring TBI patients to a polytrauma rehabilitative center through the Department of Veterans Affairs is part of a comprehensive treatment plan.²

Timothy Berigan, MD
Psychiatrist
Raymond W. Bliss Army Health Center
Fort Huachuca, AZ

1. Schwab KA, Baker G, Ivins B, et al. The Brief Traumatic Brain Injury Screen (BTBIS): investigating the validity of a self-report instrument for detecting traumatic brain injury (TBI) in troops returning from deployment in Afghanistan and Iraq. Neurology 2006;66(5)(suppl 2):A235.

2. Friedmann-Sánchez G, Sayer NA, Pickett T. Provider perspectives on rehabilitation of patients with polytrauma. Arch Phys Med Rehabil 2008;89(1):171-8.

Dr. Barry responds

Mild traumatic brain injury (mTBI) is one of the signature diagnoses of the conflicts in Iraq and Afghanistan, and we’re learning more about the pathogenesis, manifestations, sequelae, and treatment of this syndrome. Most veterans have basic knowledge about this condition because the military educates and screens all personnel for mTBI and employs aggressive, multi-disciplinary treatment plans for those in need. Thus, clinical interview, physical examination, and screening for mTBI—as suggested by Dr. Berigan—might yield useful information for clinicians treating veterans.

Matthew James Barry, DO
Chief of psychiatric services
U.S. Army’s Medical Department Activity
Fort Huachuca, AZ

The article “6 screening questions for military veterans” (Pearls, Current Psychiatry, September 2008) is a well written and thorough review for clinicians treating active duty personnel and military veterans. A very important and clinically relevant seventh question is to ask about traumatic brain injury (TBI) that may have occurred during training or deployment to Iraq or Afghanistan. One tool clinicians can use to screen these patients is the Brief Traumatic Brain Injury Screen developed by the Defense and Veterans Brain Injury Center.¹ Coupled with a clinical interview, this 3-question survey will assist clinicians in identifying possible TBI patients. Referring TBI patients to a polytrauma rehabilitative center through the Department of Veterans Affairs is part of a comprehensive treatment plan.²

Timothy Berigan, MD
Psychiatrist
Raymond W. Bliss Army Health Center
Fort Huachuca, AZ

1. Schwab KA, Baker G, Ivins B, et al. The Brief Traumatic Brain Injury Screen (BTBIS): investigating the validity of a self-report instrument for detecting traumatic brain injury (TBI) in troops returning from deployment in Afghanistan and Iraq. Neurology 2006;66(5)(suppl 2):A235.

2. Friedmann-Sánchez G, Sayer NA, Pickett T. Provider perspectives on rehabilitation of patients with polytrauma. Arch Phys Med Rehabil 2008;89(1):171-8.

Dr. Barry responds

Mild traumatic brain injury (mTBI) is one of the signature diagnoses of the conflicts in Iraq and Afghanistan, and we’re learning more about the pathogenesis, manifestations, sequelae, and treatment of this syndrome. Most veterans have basic knowledge about this condition because the military educates and screens all personnel for mTBI and employs aggressive, multi-disciplinary treatment plans for those in need. Thus, clinical interview, physical examination, and screening for mTBI—as suggested by Dr. Berigan—might yield useful information for clinicians treating veterans.

Matthew James Barry, DO
Chief of psychiatric services
U.S. Army’s Medical Department Activity
Fort Huachuca, AZ

Ever since Kraepelin used the label “dementia praecox” for the disorder Bleuler later renamed “schizophrenia,” psychiatrists have recognized cognitive impairment as a central feature of schizophrenia.¹ Cognitive deficits are an important component of many psychiatric disorders, but formal cognitive assessment still is not a part of standard clinical evaluation in psychiatric practice. It’s time that it becomes so.

Almost 2 decades ago, studies in my laboratory discovered that patients with bipolar disorder have significant deficits in cognition—including short-term memory and executive function—similar to those seen in schizophrenia.² This finding has been replicated extensively, and a book on the subject was published recently.³ Cognitive dysfunction also has been reported in unipolar depression,⁴ obsessive-compulsive disorder,⁵ posttraumatic stress disorder (PTSD),⁶ attention-deficit/hyperactivity disorder,⁷ and borderline personality disorder.⁸

This should not be surprising. Cognition is a major brain function, and mental illnesses are neurobiologic disorders in which cognitive domains can be moderately or seriously disrupted. Neurocognitive studies have established that specific cognitive dysfunctions correlate with brain pathology in specific regions. For example, because the hippocampus is a key brain region for memory, memory deficits are observed in any disorder that disrupts hippocampal structure, including Alzheimer’s disease, alcoholism, PTSD, schizophrenia, and depression.

Why assess cognition?

Cognitive measurement using validated test batteries as part of a thorough and systematic mental status examination is becoming essential—even required—in psychiatric practice. Formal cognitive assessment is useful for many clinical reasons:

• for diagnosis (the upcoming fifth edition of the Diagnostic and statistical manual of mental disorders [DSM-V] is sure to include cognitive performance in schizophrenia’s diagnostic criteria)
• to assess illness severity
• to localize dysfunctional neural pathways
• to formulate a reasonable prognosis
• to rule out possible mental retardation
• to tailor a pharmacologic treatment plan that does not further impair cognition but may enhance it
• to monitor response to treatment
• to assess cognitive side effects of pharmacotherapy
• to develop social and vocational rehabilitation programs that build on patients’ cognitive abilities.

‘Cognition enhancers’

The pharmaceutical industry’s recent surge of interest in developing cognition-enhancing (“nootropic”) drugs is timely, welcome, and supported by the National Institute of Mental Health. Initial targets of nootropic drug development are Alzheimer’s dementia and schizophrenia, but research is likely to extend to other psychiatric disorders.

When effective cognition-enhancing agents are developed and approved for use in dementia and schizophrenia, they undoubtedly will be tested in other neuropsychiatric disorders as well. They will be used as “add-on” medications to target cognitive deficits in many psychopathologic states.

Getting started

The time to vigorously assess and treat cognitive dysfunction is here. You can start—if you haven’t already—by incorporating into your practice a brief cognitive battery that measures performance on several key domains. One example is the Brief Assessment of Cognition in Schizophrenia (BACS)⁹ that was used in the Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) study. You not only will be ahead of the curve, but your patients will benefit from increased attention to cognition in their diagnosis and treatment.

Ever since Kraepelin used the label “dementia praecox” for the disorder Bleuler later renamed “schizophrenia,” psychiatrists have recognized cognitive impairment as a central feature of schizophrenia.¹ Cognitive deficits are an important component of many psychiatric disorders, but formal cognitive assessment still is not a part of standard clinical evaluation in psychiatric practice. It’s time that it becomes so.

Almost 2 decades ago, studies in my laboratory discovered that patients with bipolar disorder have significant deficits in cognition—including short-term memory and executive function—similar to those seen in schizophrenia.² This finding has been replicated extensively, and a book on the subject was published recently.³ Cognitive dysfunction also has been reported in unipolar depression,⁴ obsessive-compulsive disorder,⁵ posttraumatic stress disorder (PTSD),⁶ attention-deficit/hyperactivity disorder,⁷ and borderline personality disorder.⁸

This should not be surprising. Cognition is a major brain function, and mental illnesses are neurobiologic disorders in which cognitive domains can be moderately or seriously disrupted. Neurocognitive studies have established that specific cognitive dysfunctions correlate with brain pathology in specific regions. For example, because the hippocampus is a key brain region for memory, memory deficits are observed in any disorder that disrupts hippocampal structure, including Alzheimer’s disease, alcoholism, PTSD, schizophrenia, and depression.

Why assess cognition?

Cognitive measurement using validated test batteries as part of a thorough and systematic mental status examination is becoming essential—even required—in psychiatric practice. Formal cognitive assessment is useful for many clinical reasons:

• for diagnosis (the upcoming fifth edition of the Diagnostic and statistical manual of mental disorders [DSM-V] is sure to include cognitive performance in schizophrenia’s diagnostic criteria)
• to assess illness severity
• to localize dysfunctional neural pathways
• to formulate a reasonable prognosis
• to rule out possible mental retardation
• to tailor a pharmacologic treatment plan that does not further impair cognition but may enhance it
• to monitor response to treatment
• to assess cognitive side effects of pharmacotherapy
• to develop social and vocational rehabilitation programs that build on patients’ cognitive abilities.

‘Cognition enhancers’

The pharmaceutical industry’s recent surge of interest in developing cognition-enhancing (“nootropic”) drugs is timely, welcome, and supported by the National Institute of Mental Health. Initial targets of nootropic drug development are Alzheimer’s dementia and schizophrenia, but research is likely to extend to other psychiatric disorders.

When effective cognition-enhancing agents are developed and approved for use in dementia and schizophrenia, they undoubtedly will be tested in other neuropsychiatric disorders as well. They will be used as “add-on” medications to target cognitive deficits in many psychopathologic states.

Getting started

The time to vigorously assess and treat cognitive dysfunction is here. You can start—if you haven’t already—by incorporating into your practice a brief cognitive battery that measures performance on several key domains. One example is the Brief Assessment of Cognition in Schizophrenia (BACS)⁹ that was used in the Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) study. You not only will be ahead of the curve, but your patients will benefit from increased attention to cognition in their diagnosis and treatment.

Ever since Kraepelin used the label “dementia praecox” for the disorder Bleuler later renamed “schizophrenia,” psychiatrists have recognized cognitive impairment as a central feature of schizophrenia.¹ Cognitive deficits are an important component of many psychiatric disorders, but formal cognitive assessment still is not a part of standard clinical evaluation in psychiatric practice. It’s time that it becomes so.

Almost 2 decades ago, studies in my laboratory discovered that patients with bipolar disorder have significant deficits in cognition—including short-term memory and executive function—similar to those seen in schizophrenia.² This finding has been replicated extensively, and a book on the subject was published recently.³ Cognitive dysfunction also has been reported in unipolar depression,⁴ obsessive-compulsive disorder,⁵ posttraumatic stress disorder (PTSD),⁶ attention-deficit/hyperactivity disorder,⁷ and borderline personality disorder.⁸

This should not be surprising. Cognition is a major brain function, and mental illnesses are neurobiologic disorders in which cognitive domains can be moderately or seriously disrupted. Neurocognitive studies have established that specific cognitive dysfunctions correlate with brain pathology in specific regions. For example, because the hippocampus is a key brain region for memory, memory deficits are observed in any disorder that disrupts hippocampal structure, including Alzheimer’s disease, alcoholism, PTSD, schizophrenia, and depression.

Why assess cognition?

Cognitive measurement using validated test batteries as part of a thorough and systematic mental status examination is becoming essential—even required—in psychiatric practice. Formal cognitive assessment is useful for many clinical reasons:

• for diagnosis (the upcoming fifth edition of the Diagnostic and statistical manual of mental disorders [DSM-V] is sure to include cognitive performance in schizophrenia’s diagnostic criteria)
• to assess illness severity
• to localize dysfunctional neural pathways
• to formulate a reasonable prognosis
• to rule out possible mental retardation
• to tailor a pharmacologic treatment plan that does not further impair cognition but may enhance it
• to monitor response to treatment
• to assess cognitive side effects of pharmacotherapy
• to develop social and vocational rehabilitation programs that build on patients’ cognitive abilities.

‘Cognition enhancers’

The pharmaceutical industry’s recent surge of interest in developing cognition-enhancing (“nootropic”) drugs is timely, welcome, and supported by the National Institute of Mental Health. Initial targets of nootropic drug development are Alzheimer’s dementia and schizophrenia, but research is likely to extend to other psychiatric disorders.

When effective cognition-enhancing agents are developed and approved for use in dementia and schizophrenia, they undoubtedly will be tested in other neuropsychiatric disorders as well. They will be used as “add-on” medications to target cognitive deficits in many psychopathologic states.

Getting started

The time to vigorously assess and treat cognitive dysfunction is here. You can start—if you haven’t already—by incorporating into your practice a brief cognitive battery that measures performance on several key domains. One example is the Brief Assessment of Cognition in Schizophrenia (BACS)⁹ that was used in the Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) study. You not only will be ahead of the curve, but your patients will benefit from increased attention to cognition in their diagnosis and treatment.

CASE: Confused and suicidal

Mr. A, age 39, becomes disoriented while walking and approaches a suspension bridge. He borrows a passerby’s cell phone and calls his sister. His sister later states that he was confused and expressed his final goodbyes, saying, “I will see Mom in heaven.” He gives back the phone and leaps of the bridge. A nearby boat rescues him almost immediately.

Mr. A is brought to the trauma unit, where he is treated for a lacerated liver. After he is stabilized, Mr. A is awake and answering questions appropriately. He is placed on suicide precautions and direct 24-hour, one-to-one supervision. Our psychiatric team evaluates him.

Mr. A reports no history of diabetes, hypertension, cardiac disorders, or neurologic disorders, but does have a history of cognitive developmental delay. He has no history of psychiatric illness, suicide attempts, or self-injurious behavior. He denies a psychiatric family history or using alcohol, tobacco, or illicit drugs; drug screen is negative. He is unemployed, collects disability, and lives with his sister.

The authors’ observations

In our initial evaluation, we find no obvious reason for Mr. A’s confusion or suicide attempt. We decide to closely review Mr. A’s history in the days leading up to his jumping off the bridge.

HISTORY: Otitis media treatment

Mr. A has a history of chronic otitis media and sought treatment for ear pain at a local emergency room (ER) 10 days before his suicide attempt. He was prescribed amoxicillin, 500 mg tid for 10 days, and meclizine, 25 mg every 8 hours as needed for dizziness.

Immediately after his first dose of both drugs, the patient told his family he was feeling “weird,” but denied being dizzy. Thinking the unusual feeling was from meclizine, Mr. A stopped taking it but continued amoxicillin. On the second day of amoxicillin, he noticed bouts of confusion. He could perform his daily activities, but with difficulty. Mr. A’s niece said he had to ask for help with minor tasks, such as opening a can of soup.

On day 3, Mr. A developed prominent auditory hallucinations. He described hearing unrecognizable male and female voices chattering and mumbling throughout the day. The voices and confusion progressively worsened, but Mr. A continued taking the antibiotic and did not mention the voices to his family.

Mr. A’s sister reports that in a phone conversation with her brother on day 7, “he wasn’t himself…he was talking about my sister and mother but what he said didn’t make sense.” She asked a neighbor to check on Mr. A; he reported that Mr. A was “OK.” On the final day of amoxicillin—day 10—Mr. A became increasingly agitated. He says us that shortly before wandering onto the bridge and jumping, he was having a difficult time dealing with the voices and confusion.

We suspect amoxicillin might have been responsible for Mr. A’s psychotic symptoms.

The authors’ observations

Treatment modalities and pharmaceutical approaches used to treat infectious diseases carry many potential adverse effects. When a patient presents with new-onset psychiatric symptoms, explore whether they are related to an underlying mood disorder or medication side effects. Three important considerations are to:

• determine whether the condition is reversible by discontinuing a drug
• identify and characterize previously unrecognized adverse drug effects
• avoid inaccurate diagnosis that leads to nonindicated psychiatric treatment.¹

Antibiotic side effects vary, depending on the particular drug and its target bacteria. The most common are gastrointestinal, such as upset stomach and diarrhea. Antibiotics also can induce an anaphylactic reaction ranging from mild (pruritic rash or slight wheezing) to life-threatening (swelling of the throat, difficulty breathing, and hypotension).

Several classes of antibiotics have psychiatric side effects that range from minor confusion and irritability to severe encephalopathy and suicide (Table 1).² Case reports have described psychotic symptoms associated with cotrimoxazole,³ trimethoprim/sulfamethoxazole,⁴ and ciprofloxacin.⁵ An older review found that amoxicillin is among the top 10 most commonly prescribed medications associated with psychiatric side effects.¹

Table 1

Potential psychiatric effects of antibiotics

Amoxicillin is a penicillin-based, broad-spectrum antibiotic (Box).^1,6 Its potential psychiatric side effects include encephalopathy, irritability, sedation, anxiety, and hallucinations.² These symptoms usually are managed by reducing the dosage or discontinuing the medication. In some cases, antipsychotics may be used to control the symptoms.

Box

A literature search reveals 3 cases of amoxicillin-related psychosis (Table 2).^7-9 A 30-year-old woman with a urinary tract infection (UTI) developed “confusional manic symptoms” after 10 days of amoxicillin.⁷ The patient’s family reported she’d had a similar reaction 14 years earlier following 9 days of ampicillin for a perforated appendix; since then she had received non-aminopenicillins without incident. In both incidents, her psychotic symptoms resolved.

A 55-year-old man developed auditory, visual, and tactile hallucinations within hours of his first dose of amoxicillin for presumed pneumonia. The patient “was able to describe what he had experienced clearly with evidence of subjective terror.”⁸

Most recently, a 63-year-old woman taking amoxicillin, 250 mg tid, for a UTI developed sleep disturbance after 1 day and auditory and visual hallucinations after 4 days. She had a similar episode that required hospitalization 5 years earlier. In both episodes, psychotic symptoms resolved within 3 days of antibiotic discontinuation, with no psychotropic drug treatment.⁹

Table 2

Amoxicillin-triggered psychosis: 3 case reports

Mechanism of psychiatric effects

The mechanisms of antibiotic-related neuropsychiatric sequelae are uncertain and vary with drug class and patient factors.

Hoigné’s syndrome—an acute psychotic reaction to intramuscular procaine penicillin first reported around 1950—is characterized by psychiatric symptoms, predominantly anxiety and hallucinations, almost immediately following injection. Anxiety is marked by a fear of imminent death as well as autonomic hyperactivity. This “pseudoanaphylactic reaction” persists for 5 to 30 minutes and has been noted for its resemblance to temporal lobe and limbic seizures (perceptual disturbance, sympathetic hyperactivity, and “doom anxiety”).

The underlying pathophysiology remains unclear; the reaction was originally attributed to microembolization of procaine crystals to the lungs and brain, later to direct procaine neurotoxicity, and most recently to temporolimbic kindling—the appearance of physiologic and behavioral responses to repetition of a stimulus (procaine) that initially is without effect.¹⁰

A potential mechanism for amoxicillin’s neuropsychiatric effects is less clear. Because amoxicillin is an oral medication, hypotheses regarding Hoigné’s syndrome seem inapplicable. In addition, amoxicillin is largely excreted unchanged by the kidneys; the lack of significant P450 metabolism argues against mechanisms mediated by polypharmacy or altered metabolite levels. Furthermore, penicillins are polar molecules with poor CNS penetration.⁶ Penicillins demonstrate known neurotoxicity, however, most often causing convulsions or myelopathy. Identified risk factors for penicillin neurotoxicity include:

• intravenous/thecal administration
• high doses
• CNS disease
• renal insufficiency
• advanced age
• use of drugs that block antibiotic export from the CNS
• conditions that increase blood-brain barrier permeability.

One hypothesis focuses on penicillins’ inhibition of both the GABAA receptor-chloride ionophore complex and the benzodiazepine receptor, yielding CNS disinhibition and decreasing the seizure threshold. Notably, GABA antagonism is considered a primary facilitator of CNS kindling. Penicillin also has been reported to cause delirium related to allergy-mediated cerebral edema.¹¹ Beal et al⁷ argue for an immune-mediated cerebritis.

Psychiatric symptoms secondary to antibiotics—particularly penicillins—are likely multifactorial, suggesting certain individuals may be predisposed to “Hoigné’s syndrome” from amoxicillin. In the 3 case reports of amoxicillin-related psychosis, there is variation in duration of exposure until symptom onset, medical indication for the antibiotic, and patient age and gender. Any or all of these factors may be clinically significant. None of these patients, however, had a psychiatric history.

It is not clear whether a single 25-mg dose of meclizine—an H1-receptor antagonist—played a role in Mr. A’s psychotic symptoms. Meclizine overdose can cause extreme drowsiness, seizures, hallucinations, and decreased breathing. This anticholinergic has a half-life of only 6 hours and a duration of action of up to 24 hours, although anticholinergic toxicity from overdose can last for days.¹⁰ Mr. A ingested a single 25-mg dose of meclizine, however, and his auditory hallucinations persisted for 9 days. Furthermore, Mr. A’s previous well-tolerated meclizine use and lack of other signs and symptoms of anticholinergic toxicity do not support a substantial role for meclizine in his psychotic symptoms.

OUTCOME: Symptoms resolve

Mr. A’s confusion and auditory hallucinations resolve approximately 36 hours after he completed amoxicillin treatment. When transferred to the psychiatric unit, he denies auditory hallucinations or suicidal ideation. He also denies ear pain, tinnitus, vertigo, or ear tenderness; physical examination of the ear is unremarkable. Throughout the hospital admission, Mr. A experiences no confusion or changes in mental status and he continues to adamantly deny suicidal ideation.

He does not require treatment with anti-psychotics or other psychotropic medications and is discharged in stable condition.

Related resources

• Levenson JL, Schneider RK. Infectious diseases. In: Levenson JL, ed. The American Psychiatric Publishing textbook of psychosomatic medicine. Washington, DC: American Psychiatric Publishing; 2005:577-98.

Drug brand names

• Amoxicillin • Amoxil, Trimox, others
• Amphotericin B • Amphocin, Abelcet
• Ampicillin • Principen
• Chloramphenicol • Chloromycetin
• Ciprofloxacin • Cipro
• Clofazimine • Lamprene
• Cycloserine • Seromycin
• Ethionamide • Trecator
• Flucytosine • Ancobon
• Ganciclovir • Cytovene
• Griseofulvin • Fulvicin U/F, Grifulvin V
• Isoniazid • Nydrazid
• Ketoconazole • Nizoral
• Meclizine • Antivert, Bonine, others
• Rifampin • Rifadin, Rimactane
• Trimethoprim/sulfamethoxazole • Bactrim, Septra

Disclosure

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

CASE: Confused and suicidal

Mr. A, age 39, becomes disoriented while walking and approaches a suspension bridge. He borrows a passerby’s cell phone and calls his sister. His sister later states that he was confused and expressed his final goodbyes, saying, “I will see Mom in heaven.” He gives back the phone and leaps of the bridge. A nearby boat rescues him almost immediately.

Mr. A is brought to the trauma unit, where he is treated for a lacerated liver. After he is stabilized, Mr. A is awake and answering questions appropriately. He is placed on suicide precautions and direct 24-hour, one-to-one supervision. Our psychiatric team evaluates him.

Mr. A reports no history of diabetes, hypertension, cardiac disorders, or neurologic disorders, but does have a history of cognitive developmental delay. He has no history of psychiatric illness, suicide attempts, or self-injurious behavior. He denies a psychiatric family history or using alcohol, tobacco, or illicit drugs; drug screen is negative. He is unemployed, collects disability, and lives with his sister.

The authors’ observations

In our initial evaluation, we find no obvious reason for Mr. A’s confusion or suicide attempt. We decide to closely review Mr. A’s history in the days leading up to his jumping off the bridge.

HISTORY: Otitis media treatment

Mr. A has a history of chronic otitis media and sought treatment for ear pain at a local emergency room (ER) 10 days before his suicide attempt. He was prescribed amoxicillin, 500 mg tid for 10 days, and meclizine, 25 mg every 8 hours as needed for dizziness.

Immediately after his first dose of both drugs, the patient told his family he was feeling “weird,” but denied being dizzy. Thinking the unusual feeling was from meclizine, Mr. A stopped taking it but continued amoxicillin. On the second day of amoxicillin, he noticed bouts of confusion. He could perform his daily activities, but with difficulty. Mr. A’s niece said he had to ask for help with minor tasks, such as opening a can of soup.

On day 3, Mr. A developed prominent auditory hallucinations. He described hearing unrecognizable male and female voices chattering and mumbling throughout the day. The voices and confusion progressively worsened, but Mr. A continued taking the antibiotic and did not mention the voices to his family.

Mr. A’s sister reports that in a phone conversation with her brother on day 7, “he wasn’t himself…he was talking about my sister and mother but what he said didn’t make sense.” She asked a neighbor to check on Mr. A; he reported that Mr. A was “OK.” On the final day of amoxicillin—day 10—Mr. A became increasingly agitated. He says us that shortly before wandering onto the bridge and jumping, he was having a difficult time dealing with the voices and confusion.

We suspect amoxicillin might have been responsible for Mr. A’s psychotic symptoms.

The authors’ observations

Treatment modalities and pharmaceutical approaches used to treat infectious diseases carry many potential adverse effects. When a patient presents with new-onset psychiatric symptoms, explore whether they are related to an underlying mood disorder or medication side effects. Three important considerations are to:

• determine whether the condition is reversible by discontinuing a drug
• identify and characterize previously unrecognized adverse drug effects
• avoid inaccurate diagnosis that leads to nonindicated psychiatric treatment.¹

Antibiotic side effects vary, depending on the particular drug and its target bacteria. The most common are gastrointestinal, such as upset stomach and diarrhea. Antibiotics also can induce an anaphylactic reaction ranging from mild (pruritic rash or slight wheezing) to life-threatening (swelling of the throat, difficulty breathing, and hypotension).

Several classes of antibiotics have psychiatric side effects that range from minor confusion and irritability to severe encephalopathy and suicide (Table 1).² Case reports have described psychotic symptoms associated with cotrimoxazole,³ trimethoprim/sulfamethoxazole,⁴ and ciprofloxacin.⁵ An older review found that amoxicillin is among the top 10 most commonly prescribed medications associated with psychiatric side effects.¹

Table 1

Potential psychiatric effects of antibiotics

Amoxicillin is a penicillin-based, broad-spectrum antibiotic (Box).^1,6 Its potential psychiatric side effects include encephalopathy, irritability, sedation, anxiety, and hallucinations.² These symptoms usually are managed by reducing the dosage or discontinuing the medication. In some cases, antipsychotics may be used to control the symptoms.

Box

A literature search reveals 3 cases of amoxicillin-related psychosis (Table 2).^7-9 A 30-year-old woman with a urinary tract infection (UTI) developed “confusional manic symptoms” after 10 days of amoxicillin.⁷ The patient’s family reported she’d had a similar reaction 14 years earlier following 9 days of ampicillin for a perforated appendix; since then she had received non-aminopenicillins without incident. In both incidents, her psychotic symptoms resolved.

A 55-year-old man developed auditory, visual, and tactile hallucinations within hours of his first dose of amoxicillin for presumed pneumonia. The patient “was able to describe what he had experienced clearly with evidence of subjective terror.”⁸

Most recently, a 63-year-old woman taking amoxicillin, 250 mg tid, for a UTI developed sleep disturbance after 1 day and auditory and visual hallucinations after 4 days. She had a similar episode that required hospitalization 5 years earlier. In both episodes, psychotic symptoms resolved within 3 days of antibiotic discontinuation, with no psychotropic drug treatment.⁹

Table 2

Amoxicillin-triggered psychosis: 3 case reports

Mechanism of psychiatric effects

The mechanisms of antibiotic-related neuropsychiatric sequelae are uncertain and vary with drug class and patient factors.

Hoigné’s syndrome—an acute psychotic reaction to intramuscular procaine penicillin first reported around 1950—is characterized by psychiatric symptoms, predominantly anxiety and hallucinations, almost immediately following injection. Anxiety is marked by a fear of imminent death as well as autonomic hyperactivity. This “pseudoanaphylactic reaction” persists for 5 to 30 minutes and has been noted for its resemblance to temporal lobe and limbic seizures (perceptual disturbance, sympathetic hyperactivity, and “doom anxiety”).

The underlying pathophysiology remains unclear; the reaction was originally attributed to microembolization of procaine crystals to the lungs and brain, later to direct procaine neurotoxicity, and most recently to temporolimbic kindling—the appearance of physiologic and behavioral responses to repetition of a stimulus (procaine) that initially is without effect.¹⁰

A potential mechanism for amoxicillin’s neuropsychiatric effects is less clear. Because amoxicillin is an oral medication, hypotheses regarding Hoigné’s syndrome seem inapplicable. In addition, amoxicillin is largely excreted unchanged by the kidneys; the lack of significant P450 metabolism argues against mechanisms mediated by polypharmacy or altered metabolite levels. Furthermore, penicillins are polar molecules with poor CNS penetration.⁶ Penicillins demonstrate known neurotoxicity, however, most often causing convulsions or myelopathy. Identified risk factors for penicillin neurotoxicity include:

• intravenous/thecal administration
• high doses
• CNS disease
• renal insufficiency
• advanced age
• use of drugs that block antibiotic export from the CNS
• conditions that increase blood-brain barrier permeability.

One hypothesis focuses on penicillins’ inhibition of both the GABAA receptor-chloride ionophore complex and the benzodiazepine receptor, yielding CNS disinhibition and decreasing the seizure threshold. Notably, GABA antagonism is considered a primary facilitator of CNS kindling. Penicillin also has been reported to cause delirium related to allergy-mediated cerebral edema.¹¹ Beal et al⁷ argue for an immune-mediated cerebritis.

Psychiatric symptoms secondary to antibiotics—particularly penicillins—are likely multifactorial, suggesting certain individuals may be predisposed to “Hoigné’s syndrome” from amoxicillin. In the 3 case reports of amoxicillin-related psychosis, there is variation in duration of exposure until symptom onset, medical indication for the antibiotic, and patient age and gender. Any or all of these factors may be clinically significant. None of these patients, however, had a psychiatric history.

It is not clear whether a single 25-mg dose of meclizine—an H1-receptor antagonist—played a role in Mr. A’s psychotic symptoms. Meclizine overdose can cause extreme drowsiness, seizures, hallucinations, and decreased breathing. This anticholinergic has a half-life of only 6 hours and a duration of action of up to 24 hours, although anticholinergic toxicity from overdose can last for days.¹⁰ Mr. A ingested a single 25-mg dose of meclizine, however, and his auditory hallucinations persisted for 9 days. Furthermore, Mr. A’s previous well-tolerated meclizine use and lack of other signs and symptoms of anticholinergic toxicity do not support a substantial role for meclizine in his psychotic symptoms.

OUTCOME: Symptoms resolve

Mr. A’s confusion and auditory hallucinations resolve approximately 36 hours after he completed amoxicillin treatment. When transferred to the psychiatric unit, he denies auditory hallucinations or suicidal ideation. He also denies ear pain, tinnitus, vertigo, or ear tenderness; physical examination of the ear is unremarkable. Throughout the hospital admission, Mr. A experiences no confusion or changes in mental status and he continues to adamantly deny suicidal ideation.

He does not require treatment with anti-psychotics or other psychotropic medications and is discharged in stable condition.

Related resources

• Levenson JL, Schneider RK. Infectious diseases. In: Levenson JL, ed. The American Psychiatric Publishing textbook of psychosomatic medicine. Washington, DC: American Psychiatric Publishing; 2005:577-98.

Drug brand names

• Amoxicillin • Amoxil, Trimox, others
• Amphotericin B • Amphocin, Abelcet
• Ampicillin • Principen
• Chloramphenicol • Chloromycetin
• Ciprofloxacin • Cipro
• Clofazimine • Lamprene
• Cycloserine • Seromycin
• Ethionamide • Trecator
• Flucytosine • Ancobon
• Ganciclovir • Cytovene
• Griseofulvin • Fulvicin U/F, Grifulvin V
• Isoniazid • Nydrazid
• Ketoconazole • Nizoral
• Meclizine • Antivert, Bonine, others
• Rifampin • Rifadin, Rimactane
• Trimethoprim/sulfamethoxazole • Bactrim, Septra

Disclosure

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

CASE: Confused and suicidal

Mr. A, age 39, becomes disoriented while walking and approaches a suspension bridge. He borrows a passerby’s cell phone and calls his sister. His sister later states that he was confused and expressed his final goodbyes, saying, “I will see Mom in heaven.” He gives back the phone and leaps of the bridge. A nearby boat rescues him almost immediately.

Mr. A is brought to the trauma unit, where he is treated for a lacerated liver. After he is stabilized, Mr. A is awake and answering questions appropriately. He is placed on suicide precautions and direct 24-hour, one-to-one supervision. Our psychiatric team evaluates him.

Mr. A reports no history of diabetes, hypertension, cardiac disorders, or neurologic disorders, but does have a history of cognitive developmental delay. He has no history of psychiatric illness, suicide attempts, or self-injurious behavior. He denies a psychiatric family history or using alcohol, tobacco, or illicit drugs; drug screen is negative. He is unemployed, collects disability, and lives with his sister.

The authors’ observations

In our initial evaluation, we find no obvious reason for Mr. A’s confusion or suicide attempt. We decide to closely review Mr. A’s history in the days leading up to his jumping off the bridge.

HISTORY: Otitis media treatment

Mr. A has a history of chronic otitis media and sought treatment for ear pain at a local emergency room (ER) 10 days before his suicide attempt. He was prescribed amoxicillin, 500 mg tid for 10 days, and meclizine, 25 mg every 8 hours as needed for dizziness.

Immediately after his first dose of both drugs, the patient told his family he was feeling “weird,” but denied being dizzy. Thinking the unusual feeling was from meclizine, Mr. A stopped taking it but continued amoxicillin. On the second day of amoxicillin, he noticed bouts of confusion. He could perform his daily activities, but with difficulty. Mr. A’s niece said he had to ask for help with minor tasks, such as opening a can of soup.

On day 3, Mr. A developed prominent auditory hallucinations. He described hearing unrecognizable male and female voices chattering and mumbling throughout the day. The voices and confusion progressively worsened, but Mr. A continued taking the antibiotic and did not mention the voices to his family.

Mr. A’s sister reports that in a phone conversation with her brother on day 7, “he wasn’t himself…he was talking about my sister and mother but what he said didn’t make sense.” She asked a neighbor to check on Mr. A; he reported that Mr. A was “OK.” On the final day of amoxicillin—day 10—Mr. A became increasingly agitated. He says us that shortly before wandering onto the bridge and jumping, he was having a difficult time dealing with the voices and confusion.

We suspect amoxicillin might have been responsible for Mr. A’s psychotic symptoms.

The authors’ observations

Treatment modalities and pharmaceutical approaches used to treat infectious diseases carry many potential adverse effects. When a patient presents with new-onset psychiatric symptoms, explore whether they are related to an underlying mood disorder or medication side effects. Three important considerations are to:

• determine whether the condition is reversible by discontinuing a drug
• identify and characterize previously unrecognized adverse drug effects
• avoid inaccurate diagnosis that leads to nonindicated psychiatric treatment.¹

Antibiotic side effects vary, depending on the particular drug and its target bacteria. The most common are gastrointestinal, such as upset stomach and diarrhea. Antibiotics also can induce an anaphylactic reaction ranging from mild (pruritic rash or slight wheezing) to life-threatening (swelling of the throat, difficulty breathing, and hypotension).

Several classes of antibiotics have psychiatric side effects that range from minor confusion and irritability to severe encephalopathy and suicide (Table 1).² Case reports have described psychotic symptoms associated with cotrimoxazole,³ trimethoprim/sulfamethoxazole,⁴ and ciprofloxacin.⁵ An older review found that amoxicillin is among the top 10 most commonly prescribed medications associated with psychiatric side effects.¹

Table 1

Potential psychiatric effects of antibiotics

Amoxicillin is a penicillin-based, broad-spectrum antibiotic (Box).^1,6 Its potential psychiatric side effects include encephalopathy, irritability, sedation, anxiety, and hallucinations.² These symptoms usually are managed by reducing the dosage or discontinuing the medication. In some cases, antipsychotics may be used to control the symptoms.

Box

A literature search reveals 3 cases of amoxicillin-related psychosis (Table 2).^7-9 A 30-year-old woman with a urinary tract infection (UTI) developed “confusional manic symptoms” after 10 days of amoxicillin.⁷ The patient’s family reported she’d had a similar reaction 14 years earlier following 9 days of ampicillin for a perforated appendix; since then she had received non-aminopenicillins without incident. In both incidents, her psychotic symptoms resolved.

A 55-year-old man developed auditory, visual, and tactile hallucinations within hours of his first dose of amoxicillin for presumed pneumonia. The patient “was able to describe what he had experienced clearly with evidence of subjective terror.”⁸

Most recently, a 63-year-old woman taking amoxicillin, 250 mg tid, for a UTI developed sleep disturbance after 1 day and auditory and visual hallucinations after 4 days. She had a similar episode that required hospitalization 5 years earlier. In both episodes, psychotic symptoms resolved within 3 days of antibiotic discontinuation, with no psychotropic drug treatment.⁹

Table 2

Amoxicillin-triggered psychosis: 3 case reports

Mechanism of psychiatric effects

The mechanisms of antibiotic-related neuropsychiatric sequelae are uncertain and vary with drug class and patient factors.

Hoigné’s syndrome—an acute psychotic reaction to intramuscular procaine penicillin first reported around 1950—is characterized by psychiatric symptoms, predominantly anxiety and hallucinations, almost immediately following injection. Anxiety is marked by a fear of imminent death as well as autonomic hyperactivity. This “pseudoanaphylactic reaction” persists for 5 to 30 minutes and has been noted for its resemblance to temporal lobe and limbic seizures (perceptual disturbance, sympathetic hyperactivity, and “doom anxiety”).

The underlying pathophysiology remains unclear; the reaction was originally attributed to microembolization of procaine crystals to the lungs and brain, later to direct procaine neurotoxicity, and most recently to temporolimbic kindling—the appearance of physiologic and behavioral responses to repetition of a stimulus (procaine) that initially is without effect.¹⁰

A potential mechanism for amoxicillin’s neuropsychiatric effects is less clear. Because amoxicillin is an oral medication, hypotheses regarding Hoigné’s syndrome seem inapplicable. In addition, amoxicillin is largely excreted unchanged by the kidneys; the lack of significant P450 metabolism argues against mechanisms mediated by polypharmacy or altered metabolite levels. Furthermore, penicillins are polar molecules with poor CNS penetration.⁶ Penicillins demonstrate known neurotoxicity, however, most often causing convulsions or myelopathy. Identified risk factors for penicillin neurotoxicity include:

• intravenous/thecal administration
• high doses
• CNS disease
• renal insufficiency
• advanced age
• use of drugs that block antibiotic export from the CNS
• conditions that increase blood-brain barrier permeability.

One hypothesis focuses on penicillins’ inhibition of both the GABAA receptor-chloride ionophore complex and the benzodiazepine receptor, yielding CNS disinhibition and decreasing the seizure threshold. Notably, GABA antagonism is considered a primary facilitator of CNS kindling. Penicillin also has been reported to cause delirium related to allergy-mediated cerebral edema.¹¹ Beal et al⁷ argue for an immune-mediated cerebritis.

Psychiatric symptoms secondary to antibiotics—particularly penicillins—are likely multifactorial, suggesting certain individuals may be predisposed to “Hoigné’s syndrome” from amoxicillin. In the 3 case reports of amoxicillin-related psychosis, there is variation in duration of exposure until symptom onset, medical indication for the antibiotic, and patient age and gender. Any or all of these factors may be clinically significant. None of these patients, however, had a psychiatric history.

It is not clear whether a single 25-mg dose of meclizine—an H1-receptor antagonist—played a role in Mr. A’s psychotic symptoms. Meclizine overdose can cause extreme drowsiness, seizures, hallucinations, and decreased breathing. This anticholinergic has a half-life of only 6 hours and a duration of action of up to 24 hours, although anticholinergic toxicity from overdose can last for days.¹⁰ Mr. A ingested a single 25-mg dose of meclizine, however, and his auditory hallucinations persisted for 9 days. Furthermore, Mr. A’s previous well-tolerated meclizine use and lack of other signs and symptoms of anticholinergic toxicity do not support a substantial role for meclizine in his psychotic symptoms.

OUTCOME: Symptoms resolve

Mr. A’s confusion and auditory hallucinations resolve approximately 36 hours after he completed amoxicillin treatment. When transferred to the psychiatric unit, he denies auditory hallucinations or suicidal ideation. He also denies ear pain, tinnitus, vertigo, or ear tenderness; physical examination of the ear is unremarkable. Throughout the hospital admission, Mr. A experiences no confusion or changes in mental status and he continues to adamantly deny suicidal ideation.

He does not require treatment with anti-psychotics or other psychotropic medications and is discharged in stable condition.

Related resources

• Levenson JL, Schneider RK. Infectious diseases. In: Levenson JL, ed. The American Psychiatric Publishing textbook of psychosomatic medicine. Washington, DC: American Psychiatric Publishing; 2005:577-98.

Drug brand names

• Amoxicillin • Amoxil, Trimox, others
• Amphotericin B • Amphocin, Abelcet
• Ampicillin • Principen
• Chloramphenicol • Chloromycetin
• Ciprofloxacin • Cipro
• Clofazimine • Lamprene
• Cycloserine • Seromycin
• Ethionamide • Trecator
• Flucytosine • Ancobon
• Ganciclovir • Cytovene
• Griseofulvin • Fulvicin U/F, Grifulvin V
• Isoniazid • Nydrazid
• Ketoconazole • Nizoral
• Meclizine • Antivert, Bonine, others
• Rifampin • Rifadin, Rimactane
• Trimethoprim/sulfamethoxazole • Bactrim, Septra

Disclosure

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

Serum prolactin increases in children and adolescents when risperidone therapy begins, then decreases over time in many patients. When prolactin levels remain elevated, evidence suggests that children may experience adverse effects such as delayed sexual maturation or reduced bone growth because of hypothalamic-pituitary-gonadal axis (HPG) dysfunction.

To help you make informed prescribing decisions, we discuss what the evidence says about the effects of elevated prolactin in children and adolescents. We then suggest clinical steps to help you manage hyperprolactinemia when prescribing risperidone.

Pediatric indications

Based on short-term clinical trials of efficacy and tolerability, risperidone is FDA-approved for 3 pediatric indications:

• short-term treatment of acute mania or mixed episodes associated with bipolar I disorder in patients age 10 to 17
• schizophrenia treatment in patients age 13 to 17
• treatment of irritability (including aggression, self-injury, temper tantrums, and mood swings) associated with autistic disorder in patients age 5 to 16.

Recommended risperidone dosages are lower for children and adolescents than for adults (Table 1). Off-label pediatric uses described in case reports include psychotic, mood, disruptive, movement, and pervasive developmental disorders.

Table 1

Recommended risperidone dosing for pediatric indications*

Prolactin physiology

Prolactin’s primary physiologic function is to cause breast enlargement during pregnancy and milk secretion during lactation.¹ A polypeptide hormone, prolactin is secreted by lactotroph cells in the anterior pituitary, under the complex control of stimulatory and inhibitory factors (Table 2). Its pulsatile secretion peaks 13 to 14 times daily, with approximately 95 minutes between pulses.

Serum prolactin levels show marked circadian variation.² The reference value for serum prolactin is 1 to 25 ng/mL for women and 1 to 20 ng/mL for men. The higher prolactin levels seen in women begin after puberty and presumably are caused by estrogen’s stimulatory effect.³ Age- and sex-specific normal prolactin ranges vary widely and from lab to lab (Table 3).

Risperidone is a strong dopamine D2 and serotonin 5HT-2A antagonist with low affinity for alpha-1 and alpha-2 adrenergic receptors and histamine H1 receptors.⁴ Antagonism of these receptors is thought to explain the drug’s therapeutic effects and many of its side effects, including hyperprolactinemia.⁵ Prolactin release is also influenced by thyrotropin-releasing hormone.⁶ A rare association between pituitary tumors and atypical antipsychotics has been proposed as a probable cause of sustained prolactin elevation.⁷

Pituitary prolactin secretion is regulated by neuroendocrine neurons in the hypothalamus, specifically in the tuberoinfundibular tract that extends from the arcuate nucleus of the mediobasal hypothalamus (tuberal region) and projects to the median eminence (infundibular region). Neurosecretory dopamine neurons of the arcuate nucleus inhibit prolactin secretion. Hence, prolactin secretion increases when antipsychotic therapy results in dopamine receptor blockade.

Antipsychotics vary in affinity for the D2 dopamine receptor, rate of dissociation from the receptor, and ability to act on the receptor as both a dopamine agonist (which lowers serum prolactin) and a dopamine antagonist (which increases serum prolactin). Based on adult and pediatric data, the relative potency of antipsychotic drugs in inducing hyperprolactinemia is roughly risperidone > haloperidol > olanzapine > ziprasidone > quetiapine > clozapine > aripiprazole.⁸ Even though risperidone ranks highest in the hierarchy to cause hyperprolactinemia, it is accepted as the first-line antipsychotic in children and adolescents. This is probably because risperidone:

• has been in clinical use longer than other atypical antipsychotics except clozapine
• has received FDA approval for 3 pediatric indications.

Table 2

Factors that regulate prolactin secretion

Table 3

Sample age- and sex-specific reference ranges for serum prolactin (ng/mL)*

Prolactin and the HPG axis

Elevated serum prolactin inhibits the hypothalamus’ pulsatile release of gonadotrophin-releasing hormone (GnRH), which in turn decreases the pituitary’s secretion of follicle-stimulating hormone (FSH) and luteinizing hormone (LH). In women, prolactin also blocks the feedback effect of estradiol on LH secretion (Figure). The prolactin level that triggers gonadal hypofunction appears to vary substantially among individuals.⁹

Symptoms of elevated prolactin can occur as a direct result of prolactin’s physiologic effect on breast tissue or indirectly through hypogonadism related to decreased FSH and LH. Symptoms of hyperprolactinemia—which can be seen more readily in sexually mature adolescents than in children—include:

• amenorrhea or oligorrhea
• breast enlargement or engorgement in females and males
• galactorrhea (females > males)
• decreased libido
• erectile dysfunction.

Although evidence is inconclusive, other problems may be associated with increased prolactin in children and adolescents. These include failure to enter or progress through puberty,⁸ increased risk of benign breast tumors,²² and reduced bone density.¹⁰

Bone changes. Decreased estrogen related to hyperprolactinemia may inhibit bone mineralization, causing osteopenia, osteoporosis, and increased fracture risk.¹⁰ The mechanism of bone density loss may be estrogen’s osteoclast activating and osteoblast inhibiting action. The level and duration of prolactin elevation that can hamper bone growth has not been defined, although evidence suggests a pervasive effect:

• 65% of a group of 38 premenstrual patients developed osteoporosis or osteopenia when taking risperidone or typical antipsychotics for schizophrenia for a mean of 8 years.¹¹
• Bone loss has persisted 2 years after prolactin normalized in adolescents with prolactinomas.¹²

Prolactin secretion is controlled by stimulatory and inhibitory influences (A). Antipsychotic blockade of dopamine’s inhibitory influence (B) increases serum prolactin and its effect on mammary tissue (C). In the brain, hyperprolactinemia inhibits the release of gonadotropin-releasing hormone (GnRH) by the hypothalamus, which results in decreased follicle-stimulating hormone (FSH) and luteinizing hormone (LH) secretion by the pituitary (D). FSH and LH are important determinants of male and female gonadal maturation by their direct action on testes and ovaries within the hypothalamic-pituitary-gonadal axis.
Source: Developed by Manpreet Khemka, MBBS, and Jeffrey Ali, MD, MSc. Current Psychiatry Illustration by Rob Flewell

Hyperprolactinemia in children and adolescents

We suggest that children and adolescents receiving prolonged risperidone treatment can present with symptoms similar to those associated with hyperprolactinemia secondary to other causes, including:

• prolactinomas (the most common cause)¹³
• thyrotropin-releasing hormone stimulation in primary hypothyroidism
• hypoglycemia
• inherited endocrine syndromes
• physiologic stress
• medications.

The most common presenting symptoms of prolactinomas are headache, amenorrhea, and galactorrhea. A few patients have delayed puberty.¹³ In a review of hyperprolactinemia in children, Massart and Saggese¹⁴ proposed a correlation between elevated serum prolactin and underlying pathology:

• >100 ng/mL usually suggests organic pathology and requires MRI or CT confirmation
• <100 ng/mL usually indicates functional pathology.

What the evidence says. Using the key words “risperidone and hyperprolactinemia in children and adolescents” in a PubMed search, we identified 7 prospective, cross-sectional, and retrospective studies.^14-20 We then analyzed these studies in terms of subjects’ age, sex, primary psychiatric disorder, dosage of risperidone used, prolactin elevation pattern, reported clinical consequences, and interventions used to ameliorate asymptomatic or symptomatic hyperprolactinemia.

Prolactin and antipsychotics. In a cross-sectional study, Staller¹⁵ compared serum prolactin at baseline and after 6 months in 50 children treated with atypical antipsychotics. Patients taking risperidone showed greater increases in prolactin than those taking quetiapine or olanzapine. Saito et al¹⁶ reached a similar conclusion in a prospective study of 40 children treated with atypical antipsychotics for 4 to 15 weeks.

Ups and downs. Prolactin levels increase sharply in the first weeks of risperidone treatment, peak at around 6 to 8 weeks, and then trend downward toward normal.¹⁷ In a post hoc analysis of pooled data from 5 clinical trials totaling 700 patients age 5 to 15, Findling et al¹⁷ reported that mean serum prolactin:

• peaked in the first 1 to 2 months of patients’ starting risperidone, 0.02 to 0.06 mg/kg/d
• returned to within or close to normal range by 3 to 5 months.

No correlation was seen between prolactin elevation and side effects that could be attributed to prolactin.

In a 2-part study, Anderson et al¹⁸ examined the short- and long-term effects of risperidone treatment on prolactin in children age 5 to 17 with autism. In the initial double-blind, placebo-controlled trial, 101 children were randomly assigned to risperidone 1.8 mg/d, or placebo. After 8 weeks, 63 children continued with open-label risperidone, mean dose 1.96 mg/d, for up to a total 22 months. Serum prolactin was measured at baseline (9.3±7.5 ng/ mL) and then at 2, 6, and 22 months.

Serum prolactin increased sharply in children treated with risperidone in the placebo-controlled trial. After 8 weeks, prolactin levels were 4 times higher with risperidone treatment than with placebo (39.0±19.2 ng/mL vs 10.1±8.8 ng/mL [P< 0.0001]). In the open-label risperidone continuation trial, prolactin levels remained significantly higher than at baseline but decreased over time to:

• 32.4±17.8 ng/mL in 43 children who remained in the study at 6 months (P< 0.0001)
• 25.3±15.6 ng/mL in 30 children who remained at 22 months (P< 0.0001).

In this study, a sharp rise in serum prolactin in the first 2 months trended down to the upper limit of normal at 22 months. None of the children showed known clinical manifestations of elevated prolactin, including gynecomastia, galactorrhea, or menstrual disturbance.

A double-blind, placebo-controlled trial by Hellings et al¹⁸ examined risperidone’s effect on aggression and self-injury in children, adolescents, and adults with mental retardation and pervasive developmental disorders. In a subset of 10 children and adolescents whose serum prolactin was measured during the trial, prolactin remained elevated during at least 26 weeks of risperidone treatment. Mean prolactin levels were:

• 13.2±8.6 ng/mL at baseline
• 31.0±11.6 ng/mL during acute risperidone therapy
• 37.9±10.4 ng/mL during maintenance therapy.

Clinical features. Higher risperidone dosages—rather than longer duration of use—appear more likely to cause symptomatic elevated serum prolactin. A case series of 3 adolescents with symptomatic prolactin elevation showed:

• gynecomastia and galactorrhea in 2 adolescent males age 17 and 18, receiving risperidone, 4 mg/d and 5 mg/d, respectively
• amenorrhea within 2 to 6 weeks of starting treatment in a female patient age 15 receiving risperidone, 6 mg/d.²⁰

Holzer et al²¹ described 5 adolescents who showed symptoms of elevated prolactin after 3 to 15 months while taking risperidone, 2 to 6 mg/d, as treatment for psychosis.

Long-term health risks?

Some evidence suggests an association between elevated serum prolactin and carcinogenesis and infertility in adults. No studies have examined these long-term risks in children and adolescents who develop hyperprolactinemia from risperidone treatment. A link may be possible, however, if prolactin elevation affects postpubertal and HPG axis development.

Breast cancer. Halbriech et al²² reviewed mammograms and charts of 275 female psychiatric hospital patients age >40 and 928 women of similar age at a hospital radiology clinic. The incidence of breast cancer among psychiatric patients was:

• >3.5 times higher than among radiology clinic patients
• 9.5 times higher than in the general population.

The authors speculated that the observed increased breast cancer incidence in psychiatric patients could be associated with medications, although high rates of cigarette smoking and alcohol consumption also might have played a role.

A case-control study by Hankinson et al²³ of blood samples collected from women in the Nurses’ Health Study found a statistically significant association between hyperprolactinemia and breast cancer. This analysis included 306 postmenopausal women diagnosed with breast cancer and 448 controls matched for age, postmenopausal hormone use, and time of day and month when blood samples were drawn.

Several putative mechanisms have been proposed to explain a possible role of prolactin in breast carcinoma. Breast tissue—whether normal or cancerous—expresses the prolactin receptor, but the density of prolactin receptors is higher in tumor tissue. In several mouse models, prolactin induces tumor formation and increases tumor growth rates.^23,24

Infertility. As noted, hyperprolactinemia can cause HPG axis dysfunction.⁹ A retrospective review by Sigman and Jarow²⁵ linked endocrine disorders with infertility in 10% of 1,035 consecutive men attending 2 infertility centers. Hyperprolactinemia accounted for infertility in 0.4% of that population. No studies have associated hyperprolactinemia with female infertility.

Targeting hyperprolactinemia

When prescribing risperidone, consider obtaining a baseline serum prolactin level, especially in sexually mature patients. Repeat after 2 months, and ask the patient about menstruation, nipple discharge, sexual functioning, and pubertal development. Sexual side effects may be difficult to ascertain in patients receiving antipsychotics because of psychiatric comorbidities in this population.

Elevation without symptoms. If serum prolactin is elevated after 2 months but the patient has no clinical symptoms, repeat evaluation after another 2 months without altering the risperidone dosage. As discussed, serum prolactin tends to decline and may normalize with continued antipsychotic therapy in adults and children. A reasonable approach may be to wait 6 to 12 months for symptoms to resolve and hyperprolactinemia to diminish in patients who benefit from risperidone and have no or mild prolactin-related symptoms.⁸

Elevation with symptoms. Intervene if serum prolactin is elevated and your patient has clinical symptoms of hyperprolactinemia. Consider gradually tapering risperidone over 2 weeks and switching to a prolactin-sparing antipsychotic such as aripiprazole. If serum prolactin is >200 ng/mL or is persistently elevated despite switching to a prolactin-sparing antipsychotic, obtain an MRI of the sella turcica to look for a pituitary adenoma or parasellar tumor.

Use of dopamine agonists. Few studies have evaluated the safety and efficacy of using dopamine agonists such as cabergoline or amantadine to resolve the effects of hyperprolactinemia.^26,27 Further research is warranted before this approach can be recommended.

Related resources

• Masi G, Cosenza A. Prolactin levels in young children with pervasive developmental disorders during risperidone treatment. J Child Adolesc Psychopharmacol 2001;11:389-94.
• Cheng-Shannon J, McGough JJ, Pataki C, McCracken JT. Second-generation antipsychotic medications in children and adolescents. J Child Adolesc Psychopharmacol 2004;14:372-94.

Drug brand names

• Amantadine • Symmetrel
• Aripiprazole • Abilify
• Cabergoline • Dostinex
• Clozapine • Clozaril
• Haloperidol • Haldol
• Olanzapine • Zyprexa
• Quetiapine • Seroquel
• Risperidone • Risperdal
• Ziprasidone • Geodon

Disclosure

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

Serum prolactin increases in children and adolescents when risperidone therapy begins, then decreases over time in many patients. When prolactin levels remain elevated, evidence suggests that children may experience adverse effects such as delayed sexual maturation or reduced bone growth because of hypothalamic-pituitary-gonadal axis (HPG) dysfunction.

To help you make informed prescribing decisions, we discuss what the evidence says about the effects of elevated prolactin in children and adolescents. We then suggest clinical steps to help you manage hyperprolactinemia when prescribing risperidone.

Pediatric indications

Based on short-term clinical trials of efficacy and tolerability, risperidone is FDA-approved for 3 pediatric indications:

• short-term treatment of acute mania or mixed episodes associated with bipolar I disorder in patients age 10 to 17
• schizophrenia treatment in patients age 13 to 17
• treatment of irritability (including aggression, self-injury, temper tantrums, and mood swings) associated with autistic disorder in patients age 5 to 16.

Recommended risperidone dosages are lower for children and adolescents than for adults (Table 1). Off-label pediatric uses described in case reports include psychotic, mood, disruptive, movement, and pervasive developmental disorders.

Table 1

Recommended risperidone dosing for pediatric indications*

Prolactin physiology

Prolactin’s primary physiologic function is to cause breast enlargement during pregnancy and milk secretion during lactation.¹ A polypeptide hormone, prolactin is secreted by lactotroph cells in the anterior pituitary, under the complex control of stimulatory and inhibitory factors (Table 2). Its pulsatile secretion peaks 13 to 14 times daily, with approximately 95 minutes between pulses.

Serum prolactin levels show marked circadian variation.² The reference value for serum prolactin is 1 to 25 ng/mL for women and 1 to 20 ng/mL for men. The higher prolactin levels seen in women begin after puberty and presumably are caused by estrogen’s stimulatory effect.³ Age- and sex-specific normal prolactin ranges vary widely and from lab to lab (Table 3).

Risperidone is a strong dopamine D2 and serotonin 5HT-2A antagonist with low affinity for alpha-1 and alpha-2 adrenergic receptors and histamine H1 receptors.⁴ Antagonism of these receptors is thought to explain the drug’s therapeutic effects and many of its side effects, including hyperprolactinemia.⁵ Prolactin release is also influenced by thyrotropin-releasing hormone.⁶ A rare association between pituitary tumors and atypical antipsychotics has been proposed as a probable cause of sustained prolactin elevation.⁷

Pituitary prolactin secretion is regulated by neuroendocrine neurons in the hypothalamus, specifically in the tuberoinfundibular tract that extends from the arcuate nucleus of the mediobasal hypothalamus (tuberal region) and projects to the median eminence (infundibular region). Neurosecretory dopamine neurons of the arcuate nucleus inhibit prolactin secretion. Hence, prolactin secretion increases when antipsychotic therapy results in dopamine receptor blockade.

Antipsychotics vary in affinity for the D2 dopamine receptor, rate of dissociation from the receptor, and ability to act on the receptor as both a dopamine agonist (which lowers serum prolactin) and a dopamine antagonist (which increases serum prolactin). Based on adult and pediatric data, the relative potency of antipsychotic drugs in inducing hyperprolactinemia is roughly risperidone > haloperidol > olanzapine > ziprasidone > quetiapine > clozapine > aripiprazole.⁸ Even though risperidone ranks highest in the hierarchy to cause hyperprolactinemia, it is accepted as the first-line antipsychotic in children and adolescents. This is probably because risperidone:

• has been in clinical use longer than other atypical antipsychotics except clozapine
• has received FDA approval for 3 pediatric indications.

Table 2

Factors that regulate prolactin secretion

Table 3

Sample age- and sex-specific reference ranges for serum prolactin (ng/mL)*

Prolactin and the HPG axis

Elevated serum prolactin inhibits the hypothalamus’ pulsatile release of gonadotrophin-releasing hormone (GnRH), which in turn decreases the pituitary’s secretion of follicle-stimulating hormone (FSH) and luteinizing hormone (LH). In women, prolactin also blocks the feedback effect of estradiol on LH secretion (Figure). The prolactin level that triggers gonadal hypofunction appears to vary substantially among individuals.⁹

Symptoms of elevated prolactin can occur as a direct result of prolactin’s physiologic effect on breast tissue or indirectly through hypogonadism related to decreased FSH and LH. Symptoms of hyperprolactinemia—which can be seen more readily in sexually mature adolescents than in children—include:

• amenorrhea or oligorrhea
• breast enlargement or engorgement in females and males
• galactorrhea (females > males)
• decreased libido
• erectile dysfunction.

Although evidence is inconclusive, other problems may be associated with increased prolactin in children and adolescents. These include failure to enter or progress through puberty,⁸ increased risk of benign breast tumors,²² and reduced bone density.¹⁰

Bone changes. Decreased estrogen related to hyperprolactinemia may inhibit bone mineralization, causing osteopenia, osteoporosis, and increased fracture risk.¹⁰ The mechanism of bone density loss may be estrogen’s osteoclast activating and osteoblast inhibiting action. The level and duration of prolactin elevation that can hamper bone growth has not been defined, although evidence suggests a pervasive effect:

• 65% of a group of 38 premenstrual patients developed osteoporosis or osteopenia when taking risperidone or typical antipsychotics for schizophrenia for a mean of 8 years.¹¹
• Bone loss has persisted 2 years after prolactin normalized in adolescents with prolactinomas.¹²

Prolactin secretion is controlled by stimulatory and inhibitory influences (A). Antipsychotic blockade of dopamine’s inhibitory influence (B) increases serum prolactin and its effect on mammary tissue (C). In the brain, hyperprolactinemia inhibits the release of gonadotropin-releasing hormone (GnRH) by the hypothalamus, which results in decreased follicle-stimulating hormone (FSH) and luteinizing hormone (LH) secretion by the pituitary (D). FSH and LH are important determinants of male and female gonadal maturation by their direct action on testes and ovaries within the hypothalamic-pituitary-gonadal axis.
Source: Developed by Manpreet Khemka, MBBS, and Jeffrey Ali, MD, MSc. Current Psychiatry Illustration by Rob Flewell

Hyperprolactinemia in children and adolescents

We suggest that children and adolescents receiving prolonged risperidone treatment can present with symptoms similar to those associated with hyperprolactinemia secondary to other causes, including:

• prolactinomas (the most common cause)¹³
• thyrotropin-releasing hormone stimulation in primary hypothyroidism
• hypoglycemia
• inherited endocrine syndromes
• physiologic stress
• medications.

The most common presenting symptoms of prolactinomas are headache, amenorrhea, and galactorrhea. A few patients have delayed puberty.¹³ In a review of hyperprolactinemia in children, Massart and Saggese¹⁴ proposed a correlation between elevated serum prolactin and underlying pathology:

• >100 ng/mL usually suggests organic pathology and requires MRI or CT confirmation
• <100 ng/mL usually indicates functional pathology.

What the evidence says. Using the key words “risperidone and hyperprolactinemia in children and adolescents” in a PubMed search, we identified 7 prospective, cross-sectional, and retrospective studies.^14-20 We then analyzed these studies in terms of subjects’ age, sex, primary psychiatric disorder, dosage of risperidone used, prolactin elevation pattern, reported clinical consequences, and interventions used to ameliorate asymptomatic or symptomatic hyperprolactinemia.

Prolactin and antipsychotics. In a cross-sectional study, Staller¹⁵ compared serum prolactin at baseline and after 6 months in 50 children treated with atypical antipsychotics. Patients taking risperidone showed greater increases in prolactin than those taking quetiapine or olanzapine. Saito et al¹⁶ reached a similar conclusion in a prospective study of 40 children treated with atypical antipsychotics for 4 to 15 weeks.

Ups and downs. Prolactin levels increase sharply in the first weeks of risperidone treatment, peak at around 6 to 8 weeks, and then trend downward toward normal.¹⁷ In a post hoc analysis of pooled data from 5 clinical trials totaling 700 patients age 5 to 15, Findling et al¹⁷ reported that mean serum prolactin:

• peaked in the first 1 to 2 months of patients’ starting risperidone, 0.02 to 0.06 mg/kg/d
• returned to within or close to normal range by 3 to 5 months.

No correlation was seen between prolactin elevation and side effects that could be attributed to prolactin.

In a 2-part study, Anderson et al¹⁸ examined the short- and long-term effects of risperidone treatment on prolactin in children age 5 to 17 with autism. In the initial double-blind, placebo-controlled trial, 101 children were randomly assigned to risperidone 1.8 mg/d, or placebo. After 8 weeks, 63 children continued with open-label risperidone, mean dose 1.96 mg/d, for up to a total 22 months. Serum prolactin was measured at baseline (9.3±7.5 ng/ mL) and then at 2, 6, and 22 months.

Serum prolactin increased sharply in children treated with risperidone in the placebo-controlled trial. After 8 weeks, prolactin levels were 4 times higher with risperidone treatment than with placebo (39.0±19.2 ng/mL vs 10.1±8.8 ng/mL [P< 0.0001]). In the open-label risperidone continuation trial, prolactin levels remained significantly higher than at baseline but decreased over time to:

• 32.4±17.8 ng/mL in 43 children who remained in the study at 6 months (P< 0.0001)
• 25.3±15.6 ng/mL in 30 children who remained at 22 months (P< 0.0001).

In this study, a sharp rise in serum prolactin in the first 2 months trended down to the upper limit of normal at 22 months. None of the children showed known clinical manifestations of elevated prolactin, including gynecomastia, galactorrhea, or menstrual disturbance.

A double-blind, placebo-controlled trial by Hellings et al¹⁸ examined risperidone’s effect on aggression and self-injury in children, adolescents, and adults with mental retardation and pervasive developmental disorders. In a subset of 10 children and adolescents whose serum prolactin was measured during the trial, prolactin remained elevated during at least 26 weeks of risperidone treatment. Mean prolactin levels were:

• 13.2±8.6 ng/mL at baseline
• 31.0±11.6 ng/mL during acute risperidone therapy
• 37.9±10.4 ng/mL during maintenance therapy.

Clinical features. Higher risperidone dosages—rather than longer duration of use—appear more likely to cause symptomatic elevated serum prolactin. A case series of 3 adolescents with symptomatic prolactin elevation showed:

• gynecomastia and galactorrhea in 2 adolescent males age 17 and 18, receiving risperidone, 4 mg/d and 5 mg/d, respectively
• amenorrhea within 2 to 6 weeks of starting treatment in a female patient age 15 receiving risperidone, 6 mg/d.²⁰

Holzer et al²¹ described 5 adolescents who showed symptoms of elevated prolactin after 3 to 15 months while taking risperidone, 2 to 6 mg/d, as treatment for psychosis.

Long-term health risks?

Some evidence suggests an association between elevated serum prolactin and carcinogenesis and infertility in adults. No studies have examined these long-term risks in children and adolescents who develop hyperprolactinemia from risperidone treatment. A link may be possible, however, if prolactin elevation affects postpubertal and HPG axis development.

Breast cancer. Halbriech et al²² reviewed mammograms and charts of 275 female psychiatric hospital patients age >40 and 928 women of similar age at a hospital radiology clinic. The incidence of breast cancer among psychiatric patients was:

• >3.5 times higher than among radiology clinic patients
• 9.5 times higher than in the general population.

The authors speculated that the observed increased breast cancer incidence in psychiatric patients could be associated with medications, although high rates of cigarette smoking and alcohol consumption also might have played a role.

A case-control study by Hankinson et al²³ of blood samples collected from women in the Nurses’ Health Study found a statistically significant association between hyperprolactinemia and breast cancer. This analysis included 306 postmenopausal women diagnosed with breast cancer and 448 controls matched for age, postmenopausal hormone use, and time of day and month when blood samples were drawn.

Several putative mechanisms have been proposed to explain a possible role of prolactin in breast carcinoma. Breast tissue—whether normal or cancerous—expresses the prolactin receptor, but the density of prolactin receptors is higher in tumor tissue. In several mouse models, prolactin induces tumor formation and increases tumor growth rates.^23,24

Infertility. As noted, hyperprolactinemia can cause HPG axis dysfunction.⁹ A retrospective review by Sigman and Jarow²⁵ linked endocrine disorders with infertility in 10% of 1,035 consecutive men attending 2 infertility centers. Hyperprolactinemia accounted for infertility in 0.4% of that population. No studies have associated hyperprolactinemia with female infertility.

Targeting hyperprolactinemia

When prescribing risperidone, consider obtaining a baseline serum prolactin level, especially in sexually mature patients. Repeat after 2 months, and ask the patient about menstruation, nipple discharge, sexual functioning, and pubertal development. Sexual side effects may be difficult to ascertain in patients receiving antipsychotics because of psychiatric comorbidities in this population.

Elevation without symptoms. If serum prolactin is elevated after 2 months but the patient has no clinical symptoms, repeat evaluation after another 2 months without altering the risperidone dosage. As discussed, serum prolactin tends to decline and may normalize with continued antipsychotic therapy in adults and children. A reasonable approach may be to wait 6 to 12 months for symptoms to resolve and hyperprolactinemia to diminish in patients who benefit from risperidone and have no or mild prolactin-related symptoms.⁸

Elevation with symptoms. Intervene if serum prolactin is elevated and your patient has clinical symptoms of hyperprolactinemia. Consider gradually tapering risperidone over 2 weeks and switching to a prolactin-sparing antipsychotic such as aripiprazole. If serum prolactin is >200 ng/mL or is persistently elevated despite switching to a prolactin-sparing antipsychotic, obtain an MRI of the sella turcica to look for a pituitary adenoma or parasellar tumor.

Use of dopamine agonists. Few studies have evaluated the safety and efficacy of using dopamine agonists such as cabergoline or amantadine to resolve the effects of hyperprolactinemia.^26,27 Further research is warranted before this approach can be recommended.

Related resources

• Masi G, Cosenza A. Prolactin levels in young children with pervasive developmental disorders during risperidone treatment. J Child Adolesc Psychopharmacol 2001;11:389-94.
• Cheng-Shannon J, McGough JJ, Pataki C, McCracken JT. Second-generation antipsychotic medications in children and adolescents. J Child Adolesc Psychopharmacol 2004;14:372-94.

Drug brand names

• Amantadine • Symmetrel
• Aripiprazole • Abilify
• Cabergoline • Dostinex
• Clozapine • Clozaril
• Haloperidol • Haldol
• Olanzapine • Zyprexa
• Quetiapine • Seroquel
• Risperidone • Risperdal
• Ziprasidone • Geodon

Disclosure

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

Serum prolactin increases in children and adolescents when risperidone therapy begins, then decreases over time in many patients. When prolactin levels remain elevated, evidence suggests that children may experience adverse effects such as delayed sexual maturation or reduced bone growth because of hypothalamic-pituitary-gonadal axis (HPG) dysfunction.

To help you make informed prescribing decisions, we discuss what the evidence says about the effects of elevated prolactin in children and adolescents. We then suggest clinical steps to help you manage hyperprolactinemia when prescribing risperidone.

Pediatric indications

Based on short-term clinical trials of efficacy and tolerability, risperidone is FDA-approved for 3 pediatric indications:

• short-term treatment of acute mania or mixed episodes associated with bipolar I disorder in patients age 10 to 17
• schizophrenia treatment in patients age 13 to 17
• treatment of irritability (including aggression, self-injury, temper tantrums, and mood swings) associated with autistic disorder in patients age 5 to 16.

Recommended risperidone dosages are lower for children and adolescents than for adults (Table 1). Off-label pediatric uses described in case reports include psychotic, mood, disruptive, movement, and pervasive developmental disorders.

Table 1

Recommended risperidone dosing for pediatric indications*

Prolactin physiology

Prolactin’s primary physiologic function is to cause breast enlargement during pregnancy and milk secretion during lactation.¹ A polypeptide hormone, prolactin is secreted by lactotroph cells in the anterior pituitary, under the complex control of stimulatory and inhibitory factors (Table 2). Its pulsatile secretion peaks 13 to 14 times daily, with approximately 95 minutes between pulses.

Serum prolactin levels show marked circadian variation.² The reference value for serum prolactin is 1 to 25 ng/mL for women and 1 to 20 ng/mL for men. The higher prolactin levels seen in women begin after puberty and presumably are caused by estrogen’s stimulatory effect.³ Age- and sex-specific normal prolactin ranges vary widely and from lab to lab (Table 3).

Risperidone is a strong dopamine D2 and serotonin 5HT-2A antagonist with low affinity for alpha-1 and alpha-2 adrenergic receptors and histamine H1 receptors.⁴ Antagonism of these receptors is thought to explain the drug’s therapeutic effects and many of its side effects, including hyperprolactinemia.⁵ Prolactin release is also influenced by thyrotropin-releasing hormone.⁶ A rare association between pituitary tumors and atypical antipsychotics has been proposed as a probable cause of sustained prolactin elevation.⁷

Pituitary prolactin secretion is regulated by neuroendocrine neurons in the hypothalamus, specifically in the tuberoinfundibular tract that extends from the arcuate nucleus of the mediobasal hypothalamus (tuberal region) and projects to the median eminence (infundibular region). Neurosecretory dopamine neurons of the arcuate nucleus inhibit prolactin secretion. Hence, prolactin secretion increases when antipsychotic therapy results in dopamine receptor blockade.

Antipsychotics vary in affinity for the D2 dopamine receptor, rate of dissociation from the receptor, and ability to act on the receptor as both a dopamine agonist (which lowers serum prolactin) and a dopamine antagonist (which increases serum prolactin). Based on adult and pediatric data, the relative potency of antipsychotic drugs in inducing hyperprolactinemia is roughly risperidone > haloperidol > olanzapine > ziprasidone > quetiapine > clozapine > aripiprazole.⁸ Even though risperidone ranks highest in the hierarchy to cause hyperprolactinemia, it is accepted as the first-line antipsychotic in children and adolescents. This is probably because risperidone:

• has been in clinical use longer than other atypical antipsychotics except clozapine
• has received FDA approval for 3 pediatric indications.

Table 2

Factors that regulate prolactin secretion

Table 3

Sample age- and sex-specific reference ranges for serum prolactin (ng/mL)*

Prolactin and the HPG axis

Elevated serum prolactin inhibits the hypothalamus’ pulsatile release of gonadotrophin-releasing hormone (GnRH), which in turn decreases the pituitary’s secretion of follicle-stimulating hormone (FSH) and luteinizing hormone (LH). In women, prolactin also blocks the feedback effect of estradiol on LH secretion (Figure). The prolactin level that triggers gonadal hypofunction appears to vary substantially among individuals.⁹

Symptoms of elevated prolactin can occur as a direct result of prolactin’s physiologic effect on breast tissue or indirectly through hypogonadism related to decreased FSH and LH. Symptoms of hyperprolactinemia—which can be seen more readily in sexually mature adolescents than in children—include:

• amenorrhea or oligorrhea
• breast enlargement or engorgement in females and males
• galactorrhea (females > males)
• decreased libido
• erectile dysfunction.

Although evidence is inconclusive, other problems may be associated with increased prolactin in children and adolescents. These include failure to enter or progress through puberty,⁸ increased risk of benign breast tumors,²² and reduced bone density.¹⁰

Bone changes. Decreased estrogen related to hyperprolactinemia may inhibit bone mineralization, causing osteopenia, osteoporosis, and increased fracture risk.¹⁰ The mechanism of bone density loss may be estrogen’s osteoclast activating and osteoblast inhibiting action. The level and duration of prolactin elevation that can hamper bone growth has not been defined, although evidence suggests a pervasive effect:

• 65% of a group of 38 premenstrual patients developed osteoporosis or osteopenia when taking risperidone or typical antipsychotics for schizophrenia for a mean of 8 years.¹¹
• Bone loss has persisted 2 years after prolactin normalized in adolescents with prolactinomas.¹²

Prolactin secretion is controlled by stimulatory and inhibitory influences (A). Antipsychotic blockade of dopamine’s inhibitory influence (B) increases serum prolactin and its effect on mammary tissue (C). In the brain, hyperprolactinemia inhibits the release of gonadotropin-releasing hormone (GnRH) by the hypothalamus, which results in decreased follicle-stimulating hormone (FSH) and luteinizing hormone (LH) secretion by the pituitary (D). FSH and LH are important determinants of male and female gonadal maturation by their direct action on testes and ovaries within the hypothalamic-pituitary-gonadal axis.
Source: Developed by Manpreet Khemka, MBBS, and Jeffrey Ali, MD, MSc. Current Psychiatry Illustration by Rob Flewell

Hyperprolactinemia in children and adolescents

We suggest that children and adolescents receiving prolonged risperidone treatment can present with symptoms similar to those associated with hyperprolactinemia secondary to other causes, including:

• prolactinomas (the most common cause)¹³
• thyrotropin-releasing hormone stimulation in primary hypothyroidism
• hypoglycemia
• inherited endocrine syndromes
• physiologic stress
• medications.

The most common presenting symptoms of prolactinomas are headache, amenorrhea, and galactorrhea. A few patients have delayed puberty.¹³ In a review of hyperprolactinemia in children, Massart and Saggese¹⁴ proposed a correlation between elevated serum prolactin and underlying pathology:

• >100 ng/mL usually suggests organic pathology and requires MRI or CT confirmation
• <100 ng/mL usually indicates functional pathology.

What the evidence says. Using the key words “risperidone and hyperprolactinemia in children and adolescents” in a PubMed search, we identified 7 prospective, cross-sectional, and retrospective studies.^14-20 We then analyzed these studies in terms of subjects’ age, sex, primary psychiatric disorder, dosage of risperidone used, prolactin elevation pattern, reported clinical consequences, and interventions used to ameliorate asymptomatic or symptomatic hyperprolactinemia.

Prolactin and antipsychotics. In a cross-sectional study, Staller¹⁵ compared serum prolactin at baseline and after 6 months in 50 children treated with atypical antipsychotics. Patients taking risperidone showed greater increases in prolactin than those taking quetiapine or olanzapine. Saito et al¹⁶ reached a similar conclusion in a prospective study of 40 children treated with atypical antipsychotics for 4 to 15 weeks.

Ups and downs. Prolactin levels increase sharply in the first weeks of risperidone treatment, peak at around 6 to 8 weeks, and then trend downward toward normal.¹⁷ In a post hoc analysis of pooled data from 5 clinical trials totaling 700 patients age 5 to 15, Findling et al¹⁷ reported that mean serum prolactin:

• peaked in the first 1 to 2 months of patients’ starting risperidone, 0.02 to 0.06 mg/kg/d
• returned to within or close to normal range by 3 to 5 months.

No correlation was seen between prolactin elevation and side effects that could be attributed to prolactin.

In a 2-part study, Anderson et al¹⁸ examined the short- and long-term effects of risperidone treatment on prolactin in children age 5 to 17 with autism. In the initial double-blind, placebo-controlled trial, 101 children were randomly assigned to risperidone 1.8 mg/d, or placebo. After 8 weeks, 63 children continued with open-label risperidone, mean dose 1.96 mg/d, for up to a total 22 months. Serum prolactin was measured at baseline (9.3±7.5 ng/ mL) and then at 2, 6, and 22 months.

Serum prolactin increased sharply in children treated with risperidone in the placebo-controlled trial. After 8 weeks, prolactin levels were 4 times higher with risperidone treatment than with placebo (39.0±19.2 ng/mL vs 10.1±8.8 ng/mL [P< 0.0001]). In the open-label risperidone continuation trial, prolactin levels remained significantly higher than at baseline but decreased over time to:

• 32.4±17.8 ng/mL in 43 children who remained in the study at 6 months (P< 0.0001)
• 25.3±15.6 ng/mL in 30 children who remained at 22 months (P< 0.0001).

In this study, a sharp rise in serum prolactin in the first 2 months trended down to the upper limit of normal at 22 months. None of the children showed known clinical manifestations of elevated prolactin, including gynecomastia, galactorrhea, or menstrual disturbance.

A double-blind, placebo-controlled trial by Hellings et al¹⁸ examined risperidone’s effect on aggression and self-injury in children, adolescents, and adults with mental retardation and pervasive developmental disorders. In a subset of 10 children and adolescents whose serum prolactin was measured during the trial, prolactin remained elevated during at least 26 weeks of risperidone treatment. Mean prolactin levels were:

• 13.2±8.6 ng/mL at baseline
• 31.0±11.6 ng/mL during acute risperidone therapy
• 37.9±10.4 ng/mL during maintenance therapy.

Clinical features. Higher risperidone dosages—rather than longer duration of use—appear more likely to cause symptomatic elevated serum prolactin. A case series of 3 adolescents with symptomatic prolactin elevation showed:

• gynecomastia and galactorrhea in 2 adolescent males age 17 and 18, receiving risperidone, 4 mg/d and 5 mg/d, respectively
• amenorrhea within 2 to 6 weeks of starting treatment in a female patient age 15 receiving risperidone, 6 mg/d.²⁰

Holzer et al²¹ described 5 adolescents who showed symptoms of elevated prolactin after 3 to 15 months while taking risperidone, 2 to 6 mg/d, as treatment for psychosis.

Long-term health risks?

Some evidence suggests an association between elevated serum prolactin and carcinogenesis and infertility in adults. No studies have examined these long-term risks in children and adolescents who develop hyperprolactinemia from risperidone treatment. A link may be possible, however, if prolactin elevation affects postpubertal and HPG axis development.

Breast cancer. Halbriech et al²² reviewed mammograms and charts of 275 female psychiatric hospital patients age >40 and 928 women of similar age at a hospital radiology clinic. The incidence of breast cancer among psychiatric patients was:

• >3.5 times higher than among radiology clinic patients
• 9.5 times higher than in the general population.

The authors speculated that the observed increased breast cancer incidence in psychiatric patients could be associated with medications, although high rates of cigarette smoking and alcohol consumption also might have played a role.

A case-control study by Hankinson et al²³ of blood samples collected from women in the Nurses’ Health Study found a statistically significant association between hyperprolactinemia and breast cancer. This analysis included 306 postmenopausal women diagnosed with breast cancer and 448 controls matched for age, postmenopausal hormone use, and time of day and month when blood samples were drawn.

Several putative mechanisms have been proposed to explain a possible role of prolactin in breast carcinoma. Breast tissue—whether normal or cancerous—expresses the prolactin receptor, but the density of prolactin receptors is higher in tumor tissue. In several mouse models, prolactin induces tumor formation and increases tumor growth rates.^23,24

Infertility. As noted, hyperprolactinemia can cause HPG axis dysfunction.⁹ A retrospective review by Sigman and Jarow²⁵ linked endocrine disorders with infertility in 10% of 1,035 consecutive men attending 2 infertility centers. Hyperprolactinemia accounted for infertility in 0.4% of that population. No studies have associated hyperprolactinemia with female infertility.

Targeting hyperprolactinemia

When prescribing risperidone, consider obtaining a baseline serum prolactin level, especially in sexually mature patients. Repeat after 2 months, and ask the patient about menstruation, nipple discharge, sexual functioning, and pubertal development. Sexual side effects may be difficult to ascertain in patients receiving antipsychotics because of psychiatric comorbidities in this population.

Elevation without symptoms. If serum prolactin is elevated after 2 months but the patient has no clinical symptoms, repeat evaluation after another 2 months without altering the risperidone dosage. As discussed, serum prolactin tends to decline and may normalize with continued antipsychotic therapy in adults and children. A reasonable approach may be to wait 6 to 12 months for symptoms to resolve and hyperprolactinemia to diminish in patients who benefit from risperidone and have no or mild prolactin-related symptoms.⁸

Elevation with symptoms. Intervene if serum prolactin is elevated and your patient has clinical symptoms of hyperprolactinemia. Consider gradually tapering risperidone over 2 weeks and switching to a prolactin-sparing antipsychotic such as aripiprazole. If serum prolactin is >200 ng/mL or is persistently elevated despite switching to a prolactin-sparing antipsychotic, obtain an MRI of the sella turcica to look for a pituitary adenoma or parasellar tumor.

Use of dopamine agonists. Few studies have evaluated the safety and efficacy of using dopamine agonists such as cabergoline or amantadine to resolve the effects of hyperprolactinemia.^26,27 Further research is warranted before this approach can be recommended.

Related resources

• Masi G, Cosenza A. Prolactin levels in young children with pervasive developmental disorders during risperidone treatment. J Child Adolesc Psychopharmacol 2001;11:389-94.
• Cheng-Shannon J, McGough JJ, Pataki C, McCracken JT. Second-generation antipsychotic medications in children and adolescents. J Child Adolesc Psychopharmacol 2004;14:372-94.

Drug brand names

• Amantadine • Symmetrel
• Aripiprazole • Abilify
• Cabergoline • Dostinex
• Clozapine • Clozaril
• Haloperidol • Haldol
• Olanzapine • Zyprexa
• Quetiapine • Seroquel
• Risperidone • Risperdal
• Ziprasidone • Geodon

Disclosure

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

Patients with bipolar disorder show an unpredictable range of responses to stimulants, from virtually no ill effects to emerging manic-like symptoms.¹ Thus, although stimulants may be beneficial to some bipolar patients, there is a great deal of concern about using stimulants in this population. Even so, stimulants may be a rational adjunct for treating certain aspects of bipolar illness, particularly resistant depression, iatrogenic sedation, and comorbid attention-deficit/hyperactivity disorder (ADHD).

To help you decide if and when your patient might be a candidate for stimulant therapy, this article:

• reviews the evidence on stimulants’ safety and tolerability for patients with bipolar disorder
  
• weighs potential benefits and risks of using stimulants in this population
  
• addresses stimulants’ possible adverse effects on illness course and from interactions with other psychotropics
  
• discusses treatment options based on the limited evidence and our clinical experience.
  

Limited support

We are aware that using stimulants to treat patients with bipolar disorder is not an uncommon clinical practice, but supportive evidence is limited (Table 1). In searching the literature, we found only 2 randomized controlled studies—Frye et al² and Scheffer et al³—that addressed this practice. (One author of this review [TS] participated as a coinvestigator with Frye et al.²) Other evidence that suggests a role for stimulants in bipolar disorder comes from case reports, retrospective case series, and open-label studies.^4-11

• “traditional” stimulants (including amphetamine-based compounds such as dextroamphetamine, methylphenidate, dexmethylphenidate, and lisdexamfetamine) thought to affect the dopamine transporter, resulting in increased dopamine in nerve terminals
  
• the “novel” psychostimulant modafinil, thought to affect multiple neurotransmitter systems (dopamine, GABA, serotonin, histamine, and glutamate), although its mechanism of action is unclear.
  

Table 1

Clinical studies of stimulant use in patients with bipolar disorder

Depression and iatrogenic sedation

Small, uncontrolled trials have reported some benefit and tolerability in bipolar disorder patients when stimulants are used to treat residual depressive symptoms or iatrogenic sedation associated with mood stabilizers.

Traditional stimulants. A retrospective chart review of 16 patients treated with adjunctive methylphenidate noted improved functioning, as measured by the Global Assessment of Functioning scale. Some patients’ depressive symptoms and concentration also appeared to improve, but how these parameters were assessed is not clear. Some patients tolerated stimulants well, whereas others experienced irritability, agitation, and sleep disturbances.¹²

Another retrospective chart review described 8 patients with iatrogenic sedation or depression who received adjunctive methylphenidate, mean 20 to 40 mg/d, or a racemic mixture of amphetamine salts, mean 20 to 40 mg/d. Overall bipolar symptoms decreased in severity, as measured by Clinical Global Impression (CGI) scores, but the authors did not directly measure sedation or depression. The stimulants were well-tolerated, with no evidence of stimulant-induced mania.¹³

In a 12-week open-label trial of methylphenidate in 14 patients with bipolar disorder, depressive symptoms improved as measured by the Hamilton Depression Rating Scale (HAM-D). Mean doses were 10 mg/d for the 3 patients who discontinued because of anxiety, agitation, or hypomania and 16.6 mg/d for those who completed the trial.¹⁴

Modafinil may have some efficacy in treating bipolar depression. In a case series of 7 depressed patients (4 unipolar and 3 bipolar), 5 patients showed a 50% decrease in HAM-D scores with adjunctive modafinil. Dosages ranged from 100 to 200 mg/d, although most patients took 200 mg/d. In this series, modafinil was added to a variety of treatments, including bupropion, nefazodone, paroxetine, venlafaxine, an unspecified tricyclic antidepressant (TCA), divalproex sodium, lamotrigine, lithium, electroconvulsive therapy, olanzapine, and gabapentin.¹⁵

The only randomized, double-blind, placebo-controlled trial of adjunctive modafinil for bipolar depression enrolled 85 patients with moderate or more severe depression. In this 6-week trial by Frye et al,² 41 patients received modafinil, 100 to 200 mg/d (mean dose 174.2 mg/d), and 44 received placebo.

Bipolar disorder plus ADHD

An estimated 10% to 21% of bipolar patients meet criteria for ADHD,^16-19 although at times the line differentiating these 2 disorders is unclear. Co-occurring ADHD worsens the course of bipolar illness,^20-22 and data from the Systematic Treatment Enhancement Program for Bipolar Disorder (STEP-BD) trial suggest that only 2% of dual-diagnosis patients are receiving treatment specifically for ADHD symptoms.²³

Theoretically, overlapping symptoms such as talkativeness, distractibility, and physical activity remain relatively constant in ADHD but wax and wane with bipolar disorder’s manic and depressive phases. Recent evidence suggests, however, that many bipolar patients experience prodromal symptoms that may resemble ADHD, including cognitive impairment, distractibility, and increased psychomotor activity.²⁴ In addition, medications used to treat bipolar disorder may impair cognitive function, making ADHD diagnosis difficult in this population.

We are not aware of any clinical trials that examined stimulants’ safety and efficacy in adult bipolar patients with co-occurring ADHD. One of the only studies to examine stimulant treatment of ADHD symptoms in a bipolar population was a retrospective chart review of 34 adolescents hospitalized with bipolar mania. An earlier age of bipolar illness onset was reported in adolescents who had been exposed to stimulants, whether or not they also had ADHD.²⁵

Possible adverse events

Some bipolar disorder patients tolerate stimulants well, whereas others experience serious side effects, toxicities, and illness destabilization (Table 2). Because mood-stabilizer treatment may attenuate stimulants’ undesirable effects in bipolar disorder patients,^26,27 be sure to use adequate dosing of a mood stabilizer if you determine a stimulant trial is warranted in your patient.

Destabilization. Stimulants can have a direct negative effect on mood; they can cause restlessness, irritability, anxiety, and mood lability. Some bipolar patients may be more sensitive to these adverse effects than others. Particularly concerning is the possibility of switching to mania or worsening of manic symptoms.^28,29 Other potential destabilizing effects include:

• changing cycling patterns, such as inducing rapid cycling
  
• sleep disturbance because stimulants promote wakefulness.
  

If you are considering stimulant treatment for a bipolar disorder patient in whom substance abuse is a concern, modafinil or lisdexamfetamine may have a lower abuse potential compared with immediate-release psychostimulants. Lisdexamfetamine is metabolized in the GI tract and does not produce high d-amphetamine blood levels or cause reinforcing effects if injected or snorted.³⁴

Table 2

Possible stimulant side effects, signs of toxicity, and contraindications

Drug-drug interactions

Polypharmacy is the rule in treating bipolar disorder, and stimulants can interact with many other psychotropics (Table 3).

Antidepressants. Never use traditional stimulants with monoamine oxidase inhibitors, as this combination may precipitate a hypertensive crisis. Coadministered stimulants also may decrease the metabolism of serotonergic agents—such as selective serotonin reuptake inhibitors (SSRIs) or serotonin-norepinephrine reuptake inhibitors (SNRIs)—and cause side effects associated with increased serotonin neurotransmission, including serotonin syndrome.

Combining traditional stimulants with TCAs may increase TCA concentrations. When coadministered with bupropion, stimulants can increase the risk of seizures.

Modafinil is both an inducer and inhibitor of cytochrome P450 isoenzymes. Because it induces CYP3A4 and inhibits CYP2C19 and CYP2C9, modafinil interacts with many other psychopharmacologic agents:

• Its induction of CYP3A4 may increase the metabolism of commonly used medications such as carbamazepine, aripiprazole, and triazolam.
  
• Its inhibition of CYP2C19 may decrease the metabolism of many SSRIs, TCAs, diazepam, and clozapine, increasing these drugs’ effects and adverse events.
  

Possible stimulant interactions with other psychotropics

Treatment considerations

Without evidence to support stimulants’ safety and efficacy in patients with bipolar disorder, we cannot make specific recommendations. We would, however, like to offer some general recommendations if you decide to use stimulants when treating patients with bipolar disorder (Table 4).

Carefully assess and—in many cases—reassess the patient’s symptoms to clarify the diagnosis. As mentioned, ADHD and bipolar disorder share many symptoms, particularly in the manic phase of bipolar illness. Overlapping symptoms include decreased ability to concentrate and focus, distractibility, hyperactivity and psychomotor agitation, racing thoughts, and impulsivity.

Substance abuse can negatively impact bipolar illness and present as clinical scenarios in which stimulants are used (such as treatment-resistant depression, impulsivity, somnolence, or fatigue).

Treat medical conditions such as thyroid disease, diabetes, and sleep apnea, which may worsen depression, cause somnolence and sedation, and present with symptoms similar to those of ADHD.

When possible, use lifestyle techniques to help patients manage the course of bipolar illness. Encourage good sleep hygiene, exercise, stable social rhythms, and limited use of alcohol and caffeine (both of which can impair sleep quality, which affects illness stability).

The next step. When you have explored all medication options and ruled out all other causes for the patient’s symptoms, stimulant treatment may be an appropriate next step. In these cases:

Encourage patients to participate in treatment, particularly in monitoring mood changes (as with life charts), symptoms associated with mood episodes, and emergence of side effects. When possible, involve family members in monitoring for adverse events.

Administration. Start stimulants only when bipolar illness is well-stabilized, especially regarding manic symptoms. We highly caution against using stimulants in patients with manic or hypomanic symptoms, including mixed states. We recommend not using stimulants in patients with:

• clinically significant insomnia or sleep fragmentation
  
• active suicidal ideation or psychotic symptoms, particularly if associated with manic symptoms.
  

Schedule frequent office visits when prescribing stimulants. At least initially, see patients every other week to assess for the emergence of adverse events.

Table 4

6 recommendations when using stimulants in bipolar disorder

• The Texas Medication Algorithm Project. Texas Department of State Health Services. www.dshs.state.tx.us/mhprograms/tmapover.shtm.
  
• The Cochrane Collaboration. www.cochrane.org.
  

• Amphetamine and dextroamphetamine • Adderall
  
• Aripiprazole • Abilify
  
• Bupropion • Wellbutrin
  
• Carbamazepine • Tegretol
  
• Citalopram • Celexa
  
• Clozapine • Clozaril
  
• Dexmethylphenidate • Focalin
  
• Dextroamphetamine • Dexedrine, DextroStat
  
• Diazepam • Valium
  
• Divalproex sodium • Depakote
  
• Escitalopram • Lexapro
  
• Fluoxetine • Prozac
  
• Fluvoxamine • Luvox
  
• Gabapentin • Neurontin
  
• Lamotrigine • Lamictal
  
• Lisdexamfetamine • Vyvanse
  
• Lithium • various
  
• Methylphenidate • Ritalin, Concerta, others
  
• Modafinil • Provigil
  
• Nefazodone • Serzone
  
• Olanzapine • Zyprexa
  
• Paroxetine • Paxil
  
• Quetiapine • Seroquel
  
• Sertraline • Zoloft
  
• Triazolam • Halcion
  
• Valproic acid • Depakene
  
• Venlafaxine • Effexor
  

Dr. Gonzalez reports no financial relationship with any company whose products are mentioned in the article or with manufacturers of competing products. He is a recipient of a T32 Ruth L. Kirschstein National Research Service Awards training fellowship sponsored by the National Institutes of Health.

Dr. Suppes receives grants/research support from Abbott Laboratories, AstraZeneca, GlaxoSmithKline, JDS Pharmaceuticals, Janssen Pharmaceutica, National Institute of Mental Health, Novartis, Pfizer Inc., and the Stanley Medical Research Institute.

Patients with bipolar disorder show an unpredictable range of responses to stimulants, from virtually no ill effects to emerging manic-like symptoms.¹ Thus, although stimulants may be beneficial to some bipolar patients, there is a great deal of concern about using stimulants in this population. Even so, stimulants may be a rational adjunct for treating certain aspects of bipolar illness, particularly resistant depression, iatrogenic sedation, and comorbid attention-deficit/hyperactivity disorder (ADHD).

To help you decide if and when your patient might be a candidate for stimulant therapy, this article:

• reviews the evidence on stimulants’ safety and tolerability for patients with bipolar disorder
  
• weighs potential benefits and risks of using stimulants in this population
  
• addresses stimulants’ possible adverse effects on illness course and from interactions with other psychotropics
  
• discusses treatment options based on the limited evidence and our clinical experience.
  

Limited support

We are aware that using stimulants to treat patients with bipolar disorder is not an uncommon clinical practice, but supportive evidence is limited (Table 1). In searching the literature, we found only 2 randomized controlled studies—Frye et al² and Scheffer et al³—that addressed this practice. (One author of this review [TS] participated as a coinvestigator with Frye et al.²) Other evidence that suggests a role for stimulants in bipolar disorder comes from case reports, retrospective case series, and open-label studies.^4-11

• “traditional” stimulants (including amphetamine-based compounds such as dextroamphetamine, methylphenidate, dexmethylphenidate, and lisdexamfetamine) thought to affect the dopamine transporter, resulting in increased dopamine in nerve terminals
  
• the “novel” psychostimulant modafinil, thought to affect multiple neurotransmitter systems (dopamine, GABA, serotonin, histamine, and glutamate), although its mechanism of action is unclear.
  

Table 1

Clinical studies of stimulant use in patients with bipolar disorder

Depression and iatrogenic sedation

Small, uncontrolled trials have reported some benefit and tolerability in bipolar disorder patients when stimulants are used to treat residual depressive symptoms or iatrogenic sedation associated with mood stabilizers.

Traditional stimulants. A retrospective chart review of 16 patients treated with adjunctive methylphenidate noted improved functioning, as measured by the Global Assessment of Functioning scale. Some patients’ depressive symptoms and concentration also appeared to improve, but how these parameters were assessed is not clear. Some patients tolerated stimulants well, whereas others experienced irritability, agitation, and sleep disturbances.¹²

Another retrospective chart review described 8 patients with iatrogenic sedation or depression who received adjunctive methylphenidate, mean 20 to 40 mg/d, or a racemic mixture of amphetamine salts, mean 20 to 40 mg/d. Overall bipolar symptoms decreased in severity, as measured by Clinical Global Impression (CGI) scores, but the authors did not directly measure sedation or depression. The stimulants were well-tolerated, with no evidence of stimulant-induced mania.¹³

In a 12-week open-label trial of methylphenidate in 14 patients with bipolar disorder, depressive symptoms improved as measured by the Hamilton Depression Rating Scale (HAM-D). Mean doses were 10 mg/d for the 3 patients who discontinued because of anxiety, agitation, or hypomania and 16.6 mg/d for those who completed the trial.¹⁴

Modafinil may have some efficacy in treating bipolar depression. In a case series of 7 depressed patients (4 unipolar and 3 bipolar), 5 patients showed a 50% decrease in HAM-D scores with adjunctive modafinil. Dosages ranged from 100 to 200 mg/d, although most patients took 200 mg/d. In this series, modafinil was added to a variety of treatments, including bupropion, nefazodone, paroxetine, venlafaxine, an unspecified tricyclic antidepressant (TCA), divalproex sodium, lamotrigine, lithium, electroconvulsive therapy, olanzapine, and gabapentin.¹⁵

The only randomized, double-blind, placebo-controlled trial of adjunctive modafinil for bipolar depression enrolled 85 patients with moderate or more severe depression. In this 6-week trial by Frye et al,² 41 patients received modafinil, 100 to 200 mg/d (mean dose 174.2 mg/d), and 44 received placebo.

Bipolar disorder plus ADHD

An estimated 10% to 21% of bipolar patients meet criteria for ADHD,^16-19 although at times the line differentiating these 2 disorders is unclear. Co-occurring ADHD worsens the course of bipolar illness,^20-22 and data from the Systematic Treatment Enhancement Program for Bipolar Disorder (STEP-BD) trial suggest that only 2% of dual-diagnosis patients are receiving treatment specifically for ADHD symptoms.²³

Theoretically, overlapping symptoms such as talkativeness, distractibility, and physical activity remain relatively constant in ADHD but wax and wane with bipolar disorder’s manic and depressive phases. Recent evidence suggests, however, that many bipolar patients experience prodromal symptoms that may resemble ADHD, including cognitive impairment, distractibility, and increased psychomotor activity.²⁴ In addition, medications used to treat bipolar disorder may impair cognitive function, making ADHD diagnosis difficult in this population.

We are not aware of any clinical trials that examined stimulants’ safety and efficacy in adult bipolar patients with co-occurring ADHD. One of the only studies to examine stimulant treatment of ADHD symptoms in a bipolar population was a retrospective chart review of 34 adolescents hospitalized with bipolar mania. An earlier age of bipolar illness onset was reported in adolescents who had been exposed to stimulants, whether or not they also had ADHD.²⁵

Possible adverse events

Some bipolar disorder patients tolerate stimulants well, whereas others experience serious side effects, toxicities, and illness destabilization (Table 2). Because mood-stabilizer treatment may attenuate stimulants’ undesirable effects in bipolar disorder patients,^26,27 be sure to use adequate dosing of a mood stabilizer if you determine a stimulant trial is warranted in your patient.

Destabilization. Stimulants can have a direct negative effect on mood; they can cause restlessness, irritability, anxiety, and mood lability. Some bipolar patients may be more sensitive to these adverse effects than others. Particularly concerning is the possibility of switching to mania or worsening of manic symptoms.^28,29 Other potential destabilizing effects include:

• changing cycling patterns, such as inducing rapid cycling
  
• sleep disturbance because stimulants promote wakefulness.
  

If you are considering stimulant treatment for a bipolar disorder patient in whom substance abuse is a concern, modafinil or lisdexamfetamine may have a lower abuse potential compared with immediate-release psychostimulants. Lisdexamfetamine is metabolized in the GI tract and does not produce high d-amphetamine blood levels or cause reinforcing effects if injected or snorted.³⁴

Table 2

Possible stimulant side effects, signs of toxicity, and contraindications

Drug-drug interactions

Polypharmacy is the rule in treating bipolar disorder, and stimulants can interact with many other psychotropics (Table 3).

Antidepressants. Never use traditional stimulants with monoamine oxidase inhibitors, as this combination may precipitate a hypertensive crisis. Coadministered stimulants also may decrease the metabolism of serotonergic agents—such as selective serotonin reuptake inhibitors (SSRIs) or serotonin-norepinephrine reuptake inhibitors (SNRIs)—and cause side effects associated with increased serotonin neurotransmission, including serotonin syndrome.

Combining traditional stimulants with TCAs may increase TCA concentrations. When coadministered with bupropion, stimulants can increase the risk of seizures.

Modafinil is both an inducer and inhibitor of cytochrome P450 isoenzymes. Because it induces CYP3A4 and inhibits CYP2C19 and CYP2C9, modafinil interacts with many other psychopharmacologic agents:

• Its induction of CYP3A4 may increase the metabolism of commonly used medications such as carbamazepine, aripiprazole, and triazolam.
  
• Its inhibition of CYP2C19 may decrease the metabolism of many SSRIs, TCAs, diazepam, and clozapine, increasing these drugs’ effects and adverse events.
  

Possible stimulant interactions with other psychotropics

Treatment considerations

Without evidence to support stimulants’ safety and efficacy in patients with bipolar disorder, we cannot make specific recommendations. We would, however, like to offer some general recommendations if you decide to use stimulants when treating patients with bipolar disorder (Table 4).

Carefully assess and—in many cases—reassess the patient’s symptoms to clarify the diagnosis. As mentioned, ADHD and bipolar disorder share many symptoms, particularly in the manic phase of bipolar illness. Overlapping symptoms include decreased ability to concentrate and focus, distractibility, hyperactivity and psychomotor agitation, racing thoughts, and impulsivity.

Substance abuse can negatively impact bipolar illness and present as clinical scenarios in which stimulants are used (such as treatment-resistant depression, impulsivity, somnolence, or fatigue).

Treat medical conditions such as thyroid disease, diabetes, and sleep apnea, which may worsen depression, cause somnolence and sedation, and present with symptoms similar to those of ADHD.

When possible, use lifestyle techniques to help patients manage the course of bipolar illness. Encourage good sleep hygiene, exercise, stable social rhythms, and limited use of alcohol and caffeine (both of which can impair sleep quality, which affects illness stability).

The next step. When you have explored all medication options and ruled out all other causes for the patient’s symptoms, stimulant treatment may be an appropriate next step. In these cases:

Encourage patients to participate in treatment, particularly in monitoring mood changes (as with life charts), symptoms associated with mood episodes, and emergence of side effects. When possible, involve family members in monitoring for adverse events.

Administration. Start stimulants only when bipolar illness is well-stabilized, especially regarding manic symptoms. We highly caution against using stimulants in patients with manic or hypomanic symptoms, including mixed states. We recommend not using stimulants in patients with:

• clinically significant insomnia or sleep fragmentation
  
• active suicidal ideation or psychotic symptoms, particularly if associated with manic symptoms.
  

Schedule frequent office visits when prescribing stimulants. At least initially, see patients every other week to assess for the emergence of adverse events.

Table 4

6 recommendations when using stimulants in bipolar disorder

• The Texas Medication Algorithm Project. Texas Department of State Health Services. www.dshs.state.tx.us/mhprograms/tmapover.shtm.
  
• The Cochrane Collaboration. www.cochrane.org.
  

• Amphetamine and dextroamphetamine • Adderall
  
• Aripiprazole • Abilify
  
• Bupropion • Wellbutrin
  
• Carbamazepine • Tegretol
  
• Citalopram • Celexa
  
• Clozapine • Clozaril
  
• Dexmethylphenidate • Focalin
  
• Dextroamphetamine • Dexedrine, DextroStat
  
• Diazepam • Valium
  
• Divalproex sodium • Depakote
  
• Escitalopram • Lexapro
  
• Fluoxetine • Prozac
  
• Fluvoxamine • Luvox
  
• Gabapentin • Neurontin
  
• Lamotrigine • Lamictal
  
• Lisdexamfetamine • Vyvanse
  
• Lithium • various
  
• Methylphenidate • Ritalin, Concerta, others
  
• Modafinil • Provigil
  
• Nefazodone • Serzone
  
• Olanzapine • Zyprexa
  
• Paroxetine • Paxil
  
• Quetiapine • Seroquel
  
• Sertraline • Zoloft
  
• Triazolam • Halcion
  
• Valproic acid • Depakene
  
• Venlafaxine • Effexor
  

Dr. Gonzalez reports no financial relationship with any company whose products are mentioned in the article or with manufacturers of competing products. He is a recipient of a T32 Ruth L. Kirschstein National Research Service Awards training fellowship sponsored by the National Institutes of Health.

Dr. Suppes receives grants/research support from Abbott Laboratories, AstraZeneca, GlaxoSmithKline, JDS Pharmaceuticals, Janssen Pharmaceutica, National Institute of Mental Health, Novartis, Pfizer Inc., and the Stanley Medical Research Institute.

Patients with bipolar disorder show an unpredictable range of responses to stimulants, from virtually no ill effects to emerging manic-like symptoms.¹ Thus, although stimulants may be beneficial to some bipolar patients, there is a great deal of concern about using stimulants in this population. Even so, stimulants may be a rational adjunct for treating certain aspects of bipolar illness, particularly resistant depression, iatrogenic sedation, and comorbid attention-deficit/hyperactivity disorder (ADHD).

To help you decide if and when your patient might be a candidate for stimulant therapy, this article:

• reviews the evidence on stimulants’ safety and tolerability for patients with bipolar disorder
  
• weighs potential benefits and risks of using stimulants in this population
  
• addresses stimulants’ possible adverse effects on illness course and from interactions with other psychotropics
  
• discusses treatment options based on the limited evidence and our clinical experience.
  

Limited support

We are aware that using stimulants to treat patients with bipolar disorder is not an uncommon clinical practice, but supportive evidence is limited (Table 1). In searching the literature, we found only 2 randomized controlled studies—Frye et al² and Scheffer et al³—that addressed this practice. (One author of this review [TS] participated as a coinvestigator with Frye et al.²) Other evidence that suggests a role for stimulants in bipolar disorder comes from case reports, retrospective case series, and open-label studies.^4-11

• “traditional” stimulants (including amphetamine-based compounds such as dextroamphetamine, methylphenidate, dexmethylphenidate, and lisdexamfetamine) thought to affect the dopamine transporter, resulting in increased dopamine in nerve terminals
  
• the “novel” psychostimulant modafinil, thought to affect multiple neurotransmitter systems (dopamine, GABA, serotonin, histamine, and glutamate), although its mechanism of action is unclear.
  

Table 1

Clinical studies of stimulant use in patients with bipolar disorder

Depression and iatrogenic sedation

Small, uncontrolled trials have reported some benefit and tolerability in bipolar disorder patients when stimulants are used to treat residual depressive symptoms or iatrogenic sedation associated with mood stabilizers.

Traditional stimulants. A retrospective chart review of 16 patients treated with adjunctive methylphenidate noted improved functioning, as measured by the Global Assessment of Functioning scale. Some patients’ depressive symptoms and concentration also appeared to improve, but how these parameters were assessed is not clear. Some patients tolerated stimulants well, whereas others experienced irritability, agitation, and sleep disturbances.¹²

Another retrospective chart review described 8 patients with iatrogenic sedation or depression who received adjunctive methylphenidate, mean 20 to 40 mg/d, or a racemic mixture of amphetamine salts, mean 20 to 40 mg/d. Overall bipolar symptoms decreased in severity, as measured by Clinical Global Impression (CGI) scores, but the authors did not directly measure sedation or depression. The stimulants were well-tolerated, with no evidence of stimulant-induced mania.¹³

In a 12-week open-label trial of methylphenidate in 14 patients with bipolar disorder, depressive symptoms improved as measured by the Hamilton Depression Rating Scale (HAM-D). Mean doses were 10 mg/d for the 3 patients who discontinued because of anxiety, agitation, or hypomania and 16.6 mg/d for those who completed the trial.¹⁴

Modafinil may have some efficacy in treating bipolar depression. In a case series of 7 depressed patients (4 unipolar and 3 bipolar), 5 patients showed a 50% decrease in HAM-D scores with adjunctive modafinil. Dosages ranged from 100 to 200 mg/d, although most patients took 200 mg/d. In this series, modafinil was added to a variety of treatments, including bupropion, nefazodone, paroxetine, venlafaxine, an unspecified tricyclic antidepressant (TCA), divalproex sodium, lamotrigine, lithium, electroconvulsive therapy, olanzapine, and gabapentin.¹⁵

The only randomized, double-blind, placebo-controlled trial of adjunctive modafinil for bipolar depression enrolled 85 patients with moderate or more severe depression. In this 6-week trial by Frye et al,² 41 patients received modafinil, 100 to 200 mg/d (mean dose 174.2 mg/d), and 44 received placebo.

Bipolar disorder plus ADHD

An estimated 10% to 21% of bipolar patients meet criteria for ADHD,^16-19 although at times the line differentiating these 2 disorders is unclear. Co-occurring ADHD worsens the course of bipolar illness,^20-22 and data from the Systematic Treatment Enhancement Program for Bipolar Disorder (STEP-BD) trial suggest that only 2% of dual-diagnosis patients are receiving treatment specifically for ADHD symptoms.²³

Theoretically, overlapping symptoms such as talkativeness, distractibility, and physical activity remain relatively constant in ADHD but wax and wane with bipolar disorder’s manic and depressive phases. Recent evidence suggests, however, that many bipolar patients experience prodromal symptoms that may resemble ADHD, including cognitive impairment, distractibility, and increased psychomotor activity.²⁴ In addition, medications used to treat bipolar disorder may impair cognitive function, making ADHD diagnosis difficult in this population.

We are not aware of any clinical trials that examined stimulants’ safety and efficacy in adult bipolar patients with co-occurring ADHD. One of the only studies to examine stimulant treatment of ADHD symptoms in a bipolar population was a retrospective chart review of 34 adolescents hospitalized with bipolar mania. An earlier age of bipolar illness onset was reported in adolescents who had been exposed to stimulants, whether or not they also had ADHD.²⁵

Possible adverse events

Some bipolar disorder patients tolerate stimulants well, whereas others experience serious side effects, toxicities, and illness destabilization (Table 2). Because mood-stabilizer treatment may attenuate stimulants’ undesirable effects in bipolar disorder patients,^26,27 be sure to use adequate dosing of a mood stabilizer if you determine a stimulant trial is warranted in your patient.

Destabilization. Stimulants can have a direct negative effect on mood; they can cause restlessness, irritability, anxiety, and mood lability. Some bipolar patients may be more sensitive to these adverse effects than others. Particularly concerning is the possibility of switching to mania or worsening of manic symptoms.^28,29 Other potential destabilizing effects include:

• changing cycling patterns, such as inducing rapid cycling
  
• sleep disturbance because stimulants promote wakefulness.
  

If you are considering stimulant treatment for a bipolar disorder patient in whom substance abuse is a concern, modafinil or lisdexamfetamine may have a lower abuse potential compared with immediate-release psychostimulants. Lisdexamfetamine is metabolized in the GI tract and does not produce high d-amphetamine blood levels or cause reinforcing effects if injected or snorted.³⁴

Table 2

Possible stimulant side effects, signs of toxicity, and contraindications

Drug-drug interactions

Polypharmacy is the rule in treating bipolar disorder, and stimulants can interact with many other psychotropics (Table 3).

Antidepressants. Never use traditional stimulants with monoamine oxidase inhibitors, as this combination may precipitate a hypertensive crisis. Coadministered stimulants also may decrease the metabolism of serotonergic agents—such as selective serotonin reuptake inhibitors (SSRIs) or serotonin-norepinephrine reuptake inhibitors (SNRIs)—and cause side effects associated with increased serotonin neurotransmission, including serotonin syndrome.

Combining traditional stimulants with TCAs may increase TCA concentrations. When coadministered with bupropion, stimulants can increase the risk of seizures.

Modafinil is both an inducer and inhibitor of cytochrome P450 isoenzymes. Because it induces CYP3A4 and inhibits CYP2C19 and CYP2C9, modafinil interacts with many other psychopharmacologic agents:

• Its induction of CYP3A4 may increase the metabolism of commonly used medications such as carbamazepine, aripiprazole, and triazolam.
  
• Its inhibition of CYP2C19 may decrease the metabolism of many SSRIs, TCAs, diazepam, and clozapine, increasing these drugs’ effects and adverse events.
  

Possible stimulant interactions with other psychotropics

Treatment considerations

Without evidence to support stimulants’ safety and efficacy in patients with bipolar disorder, we cannot make specific recommendations. We would, however, like to offer some general recommendations if you decide to use stimulants when treating patients with bipolar disorder (Table 4).

Carefully assess and—in many cases—reassess the patient’s symptoms to clarify the diagnosis. As mentioned, ADHD and bipolar disorder share many symptoms, particularly in the manic phase of bipolar illness. Overlapping symptoms include decreased ability to concentrate and focus, distractibility, hyperactivity and psychomotor agitation, racing thoughts, and impulsivity.

Substance abuse can negatively impact bipolar illness and present as clinical scenarios in which stimulants are used (such as treatment-resistant depression, impulsivity, somnolence, or fatigue).

Treat medical conditions such as thyroid disease, diabetes, and sleep apnea, which may worsen depression, cause somnolence and sedation, and present with symptoms similar to those of ADHD.

When possible, use lifestyle techniques to help patients manage the course of bipolar illness. Encourage good sleep hygiene, exercise, stable social rhythms, and limited use of alcohol and caffeine (both of which can impair sleep quality, which affects illness stability).

The next step. When you have explored all medication options and ruled out all other causes for the patient’s symptoms, stimulant treatment may be an appropriate next step. In these cases:

Encourage patients to participate in treatment, particularly in monitoring mood changes (as with life charts), symptoms associated with mood episodes, and emergence of side effects. When possible, involve family members in monitoring for adverse events.

Administration. Start stimulants only when bipolar illness is well-stabilized, especially regarding manic symptoms. We highly caution against using stimulants in patients with manic or hypomanic symptoms, including mixed states. We recommend not using stimulants in patients with:

• clinically significant insomnia or sleep fragmentation
  
• active suicidal ideation or psychotic symptoms, particularly if associated with manic symptoms.
  

Schedule frequent office visits when prescribing stimulants. At least initially, see patients every other week to assess for the emergence of adverse events.

Table 4

6 recommendations when using stimulants in bipolar disorder

• The Texas Medication Algorithm Project. Texas Department of State Health Services. www.dshs.state.tx.us/mhprograms/tmapover.shtm.
  
• The Cochrane Collaboration. www.cochrane.org.
  

• Amphetamine and dextroamphetamine • Adderall
  
• Aripiprazole • Abilify
  
• Bupropion • Wellbutrin
  
• Carbamazepine • Tegretol
  
• Citalopram • Celexa
  
• Clozapine • Clozaril
  
• Dexmethylphenidate • Focalin
  
• Dextroamphetamine • Dexedrine, DextroStat
  
• Diazepam • Valium
  
• Divalproex sodium • Depakote
  
• Escitalopram • Lexapro
  
• Fluoxetine • Prozac
  
• Fluvoxamine • Luvox
  
• Gabapentin • Neurontin
  
• Lamotrigine • Lamictal
  
• Lisdexamfetamine • Vyvanse
  
• Lithium • various
  
• Methylphenidate • Ritalin, Concerta, others
  
• Modafinil • Provigil
  
• Nefazodone • Serzone
  
• Olanzapine • Zyprexa
  
• Paroxetine • Paxil
  
• Quetiapine • Seroquel
  
• Sertraline • Zoloft
  
• Triazolam • Halcion
  
• Valproic acid • Depakene
  
• Venlafaxine • Effexor
  

Dr. Gonzalez reports no financial relationship with any company whose products are mentioned in the article or with manufacturers of competing products. He is a recipient of a T32 Ruth L. Kirschstein National Research Service Awards training fellowship sponsored by the National Institutes of Health.

Dr. Suppes receives grants/research support from Abbott Laboratories, AstraZeneca, GlaxoSmithKline, JDS Pharmaceuticals, Janssen Pharmaceutica, National Institute of Mental Health, Novartis, Pfizer Inc., and the Stanley Medical Research Institute.

Ms. J, age 32, comes to our mental health clinic seeking treatment for depression and anxiety. She reports she has attempted suicide 3 times. Ms. J describes the first 2 attempts—both of which occurred when she was in her 20s after the end of a relationship—as “cries for attention” that were relatively innocuous. Her third suicide attempt, however, was an acetaminophen overdose approximately 1 year ago that resulted in hospitalization and irreversible liver damage.

Ms. J acknowledges that over the last several weeks she has been thinking about suicide almost constantly, especially as the anniversary of her fiancé’s death approaches. She says she has a nearly full bottle of zolpidem in her medicine cabinet and fantasizes about taking all of them and just “going to sleep.”

Many patients—especially those with depression—experience recurrent thoughts of death or a wish to die, but only about 10% attempt suicide.¹ To identify those who are at highest risk and warrant hospitalization, it is vital to assess how a history of suicidal behavior and other factors impact the risk for future suicide attempts. This article:

• examines research on patients who have attempted suicide and risk factors for repeat suicide attempts
  
• describes characteristics of patients with multiple attempts
  
• explores the link between a history of self-injurious behavior and suicide attempts.
  

A strong predictor

A previous suicide attempt is among the strongest predictors of future suicide attempts.^2-4 In a sample of clinically referred European adolescents, those who had attempted suicide were 3 times more likely to try again during the 1-year follow-up compared with those who had never attempted suicide.⁵ In addition, Harris et al⁶ found that patients with a previous suicide attempt were 38 times more likely to eventually commit suicide than those with no past attempts.

Other risk factors

Other factors might help predict which individuals will continue to engage in suicidal behavior after a first attempt (Table 1).^7,8 Spirito et al⁷ followed 58 adolescent suicide attempters for 3 months after their initial attempt. Seven (12%) made a subsequent attempt, and 26 (45%) reported continued suicidal ideation. Depressed mood was the strongest predictor of subsequent suicidal behavior, followed by poor family functioning, affect regulation difficulty, and hopelessness.

Hopelessness. Beck et al⁹ found that patients who scored ≥9 on the Beck Hopelessness Scale (BHS)—the most common self-report measure of hopelessness—were approximately 11 times more likely to commit suicide than patients who scored ≤8. A study of hospitalized suicide attempters found that BHS scores were unique predictors of future suicide attempts.¹⁰ Several studies have found that persons who remain consistently hopeless are more likely to kill themselves compared with those who have fluctuating hopelessness levels.^11,12

History of abuse—specifically sexual abuse—is associated with suicidal behavior. A study of depressed women age >50 found that among those who were sexually abused before age 18, 83% reported 1 suicide attempt and 67% made multiple attempts.¹³ Among women who had not experienced childhood sexual abuse, 58% reported a past suicide attempt and 27% made multiple attempts.¹³

In a separate study of psychiatric inpatients, those who had been physically or sexually abused were more likely to have made a suicide attempt than patients with no such history.¹⁴ This study did not find a difference in reported abuse between single and multiple suicide attempters.

Stressors. In many cases suicide attempts are precipitated by acute or chronic stressors, including:

• job stress
  
• chronic illness
  
• financial problems
  
• relationship discord
  
• retirement and declining physical health (especially for older men)
  
• death of a loved one.¹⁵
  

Risk is not necessarily cumulative—and not all risk factors are weighted equally. In general, however, the more risk factors a patient has, the greater the likelihood that he or she may attempt suicide.¹⁷

Table 1

Repeated suicidal behavior: Factors that increase risk

Red flag: Multiple attempts

When assessing a patient’s suicide history, ask about the number of attempts. A person who makes >1 suicide attempt—a multiple attempter—has a significantly higher chance of making subsequent attempts compared with those with 1 or no attempts.^18,19

Persons who make multiple attempts share certain characteristics (Table 2).^19-21 Rudd et al¹⁹ compared 68 multiple attempters with 128 single attempters and found that multiple attempters had higher levels of:

• suicide ideation
  
• depression
  
• hopelessness
  
• perceived stress.
  

• diagnosed with substance use disorders, psychotic disorder, or borderline personality disorder
  
• unemployed and have relationship difficulties, a history of emotional abuse, and a family history of psychiatric problems and suicide.
  

Among 326 individuals in a military medical setting treated for suicidal behavior or severe suicidal ideation, multiple suicide attempters reported higher levels of ongoing distress that was unrelated to specific life stressors.²² This suggests these patients may not respond well to psychological interventions that focus on problem-solving.

Table 2

Common characteristics of multiple suicide attempters

Self-harm and suicidal behavior

Patients who engage in nonsuicidal self harm—also called self-injurious behavior (SIB)—may be mistaken for suicide attempters. Although differences exist between suicide attempters and those who engage in SIB, evidence suggests that a history of SIB increases risk for suicidal behavior.^23,24 In a retrospective study of 4,167 self-harmers, females who engaged in ≥4 acts of SIB were more likely to die from suicide than those who engaged in ≤3 acts.²⁵ A cross-sectional analysis of data from 3,069 students responding to a random Web-based survey found that an increased incidence of SIB significantly increased the odds of suicidal behavior.²⁶

Although the link between SIB and suicide attempts remains unclear, evidence suggests SIB is a risk factor for suicidal behavior and therefore should be assessed when evaluating a patient’s suicide risk.

CASE CONTINUED: At high risk

Ms. J has several risk factors for making another suicide attempt. She has 3 previous attempts, and because her last attempt caused liver damage we know she is capable of lethal behavior. In addition, the anniversary of the death of her fiancé is approaching. Ms. J also reports almost constant suicidal ideation, with a specific plan (to overdose). Her fantasies of taking pills could be interpreted as mental rehearsal and desensitization to the behavior.

Because we believe Ms. J is at high risk for a serious, if not lethal, suicide attempt we conduct a 4-question suicide inquiry. It is clear that Ms. J had suicidal thoughts and a plan. Her answer to “How likely is it that once you leave my office you will do something to hurt yourself?” is the key to determining whether or not she requires hospitalization. Ms. J states that she is “pretty certain she will hurt herself” once she leaves the office, so we hospitalize her.

To determine if a patient requires immediate hospitalization, perform a specific suicide inquiry. Although there is no surefire way to determine if a patient will kill himself or herself, asking specific questions can help you gauge risk. Based on evidence²⁸ and my clinical experience, I focus on patients’ answers to 4 questions (Table 3). Affirmative answers to these questions are a strong indication that a patient requires hospitalization.

Occasionally, patients are not truthful when asked about their suicidal intent. If you suspect a patient is lying, clinical judgment and the patient’s history guide the decision on hospitalization.

Table 3

Hospitalize? 4 questions to guide your decision

• Joiner TE. Why people die by suicide. Cambridge, MA: Harvard University Press; 2005:46-93,203-22.
  
• American Foundation for Suicide Prevention. www.afsp.org.
  
• SAVE: Suicide Awareness Voices of Education. www.save.org.
  

• Zolpidem • Ambien
  

The author reports no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.

Ms. J, age 32, comes to our mental health clinic seeking treatment for depression and anxiety. She reports she has attempted suicide 3 times. Ms. J describes the first 2 attempts—both of which occurred when she was in her 20s after the end of a relationship—as “cries for attention” that were relatively innocuous. Her third suicide attempt, however, was an acetaminophen overdose approximately 1 year ago that resulted in hospitalization and irreversible liver damage.

Ms. J acknowledges that over the last several weeks she has been thinking about suicide almost constantly, especially as the anniversary of her fiancé’s death approaches. She says she has a nearly full bottle of zolpidem in her medicine cabinet and fantasizes about taking all of them and just “going to sleep.”

Many patients—especially those with depression—experience recurrent thoughts of death or a wish to die, but only about 10% attempt suicide.¹ To identify those who are at highest risk and warrant hospitalization, it is vital to assess how a history of suicidal behavior and other factors impact the risk for future suicide attempts. This article:

• examines research on patients who have attempted suicide and risk factors for repeat suicide attempts
  
• describes characteristics of patients with multiple attempts
  
• explores the link between a history of self-injurious behavior and suicide attempts.
  

A strong predictor

A previous suicide attempt is among the strongest predictors of future suicide attempts.^2-4 In a sample of clinically referred European adolescents, those who had attempted suicide were 3 times more likely to try again during the 1-year follow-up compared with those who had never attempted suicide.⁵ In addition, Harris et al⁶ found that patients with a previous suicide attempt were 38 times more likely to eventually commit suicide than those with no past attempts.

Other risk factors

Other factors might help predict which individuals will continue to engage in suicidal behavior after a first attempt (Table 1).^7,8 Spirito et al⁷ followed 58 adolescent suicide attempters for 3 months after their initial attempt. Seven (12%) made a subsequent attempt, and 26 (45%) reported continued suicidal ideation. Depressed mood was the strongest predictor of subsequent suicidal behavior, followed by poor family functioning, affect regulation difficulty, and hopelessness.

Hopelessness. Beck et al⁹ found that patients who scored ≥9 on the Beck Hopelessness Scale (BHS)—the most common self-report measure of hopelessness—were approximately 11 times more likely to commit suicide than patients who scored ≤8. A study of hospitalized suicide attempters found that BHS scores were unique predictors of future suicide attempts.¹⁰ Several studies have found that persons who remain consistently hopeless are more likely to kill themselves compared with those who have fluctuating hopelessness levels.^11,12

History of abuse—specifically sexual abuse—is associated with suicidal behavior. A study of depressed women age >50 found that among those who were sexually abused before age 18, 83% reported 1 suicide attempt and 67% made multiple attempts.¹³ Among women who had not experienced childhood sexual abuse, 58% reported a past suicide attempt and 27% made multiple attempts.¹³

In a separate study of psychiatric inpatients, those who had been physically or sexually abused were more likely to have made a suicide attempt than patients with no such history.¹⁴ This study did not find a difference in reported abuse between single and multiple suicide attempters.

Stressors. In many cases suicide attempts are precipitated by acute or chronic stressors, including:

• job stress
  
• chronic illness
  
• financial problems
  
• relationship discord
  
• retirement and declining physical health (especially for older men)
  
• death of a loved one.¹⁵
  

Risk is not necessarily cumulative—and not all risk factors are weighted equally. In general, however, the more risk factors a patient has, the greater the likelihood that he or she may attempt suicide.¹⁷

Table 1

Repeated suicidal behavior: Factors that increase risk

Red flag: Multiple attempts

When assessing a patient’s suicide history, ask about the number of attempts. A person who makes >1 suicide attempt—a multiple attempter—has a significantly higher chance of making subsequent attempts compared with those with 1 or no attempts.^18,19

Persons who make multiple attempts share certain characteristics (Table 2).^19-21 Rudd et al¹⁹ compared 68 multiple attempters with 128 single attempters and found that multiple attempters had higher levels of:

• suicide ideation
  
• depression
  
• hopelessness
  
• perceived stress.
  

• diagnosed with substance use disorders, psychotic disorder, or borderline personality disorder
  
• unemployed and have relationship difficulties, a history of emotional abuse, and a family history of psychiatric problems and suicide.
  

Among 326 individuals in a military medical setting treated for suicidal behavior or severe suicidal ideation, multiple suicide attempters reported higher levels of ongoing distress that was unrelated to specific life stressors.²² This suggests these patients may not respond well to psychological interventions that focus on problem-solving.

Table 2

Common characteristics of multiple suicide attempters

Self-harm and suicidal behavior

Patients who engage in nonsuicidal self harm—also called self-injurious behavior (SIB)—may be mistaken for suicide attempters. Although differences exist between suicide attempters and those who engage in SIB, evidence suggests that a history of SIB increases risk for suicidal behavior.^23,24 In a retrospective study of 4,167 self-harmers, females who engaged in ≥4 acts of SIB were more likely to die from suicide than those who engaged in ≤3 acts.²⁵ A cross-sectional analysis of data from 3,069 students responding to a random Web-based survey found that an increased incidence of SIB significantly increased the odds of suicidal behavior.²⁶

Although the link between SIB and suicide attempts remains unclear, evidence suggests SIB is a risk factor for suicidal behavior and therefore should be assessed when evaluating a patient’s suicide risk.

CASE CONTINUED: At high risk

Ms. J has several risk factors for making another suicide attempt. She has 3 previous attempts, and because her last attempt caused liver damage we know she is capable of lethal behavior. In addition, the anniversary of the death of her fiancé is approaching. Ms. J also reports almost constant suicidal ideation, with a specific plan (to overdose). Her fantasies of taking pills could be interpreted as mental rehearsal and desensitization to the behavior.

Because we believe Ms. J is at high risk for a serious, if not lethal, suicide attempt we conduct a 4-question suicide inquiry. It is clear that Ms. J had suicidal thoughts and a plan. Her answer to “How likely is it that once you leave my office you will do something to hurt yourself?” is the key to determining whether or not she requires hospitalization. Ms. J states that she is “pretty certain she will hurt herself” once she leaves the office, so we hospitalize her.

To determine if a patient requires immediate hospitalization, perform a specific suicide inquiry. Although there is no surefire way to determine if a patient will kill himself or herself, asking specific questions can help you gauge risk. Based on evidence²⁸ and my clinical experience, I focus on patients’ answers to 4 questions (Table 3). Affirmative answers to these questions are a strong indication that a patient requires hospitalization.

Occasionally, patients are not truthful when asked about their suicidal intent. If you suspect a patient is lying, clinical judgment and the patient’s history guide the decision on hospitalization.

Table 3

Hospitalize? 4 questions to guide your decision

• Joiner TE. Why people die by suicide. Cambridge, MA: Harvard University Press; 2005:46-93,203-22.
  
• American Foundation for Suicide Prevention. www.afsp.org.
  
• SAVE: Suicide Awareness Voices of Education. www.save.org.
  

• Zolpidem • Ambien
  

The author reports no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.

Ms. J, age 32, comes to our mental health clinic seeking treatment for depression and anxiety. She reports she has attempted suicide 3 times. Ms. J describes the first 2 attempts—both of which occurred when she was in her 20s after the end of a relationship—as “cries for attention” that were relatively innocuous. Her third suicide attempt, however, was an acetaminophen overdose approximately 1 year ago that resulted in hospitalization and irreversible liver damage.

Ms. J acknowledges that over the last several weeks she has been thinking about suicide almost constantly, especially as the anniversary of her fiancé’s death approaches. She says she has a nearly full bottle of zolpidem in her medicine cabinet and fantasizes about taking all of them and just “going to sleep.”

Many patients—especially those with depression—experience recurrent thoughts of death or a wish to die, but only about 10% attempt suicide.¹ To identify those who are at highest risk and warrant hospitalization, it is vital to assess how a history of suicidal behavior and other factors impact the risk for future suicide attempts. This article:

• examines research on patients who have attempted suicide and risk factors for repeat suicide attempts
  
• describes characteristics of patients with multiple attempts
  
• explores the link between a history of self-injurious behavior and suicide attempts.
  

A strong predictor

A previous suicide attempt is among the strongest predictors of future suicide attempts.^2-4 In a sample of clinically referred European adolescents, those who had attempted suicide were 3 times more likely to try again during the 1-year follow-up compared with those who had never attempted suicide.⁵ In addition, Harris et al⁶ found that patients with a previous suicide attempt were 38 times more likely to eventually commit suicide than those with no past attempts.

Other risk factors

Other factors might help predict which individuals will continue to engage in suicidal behavior after a first attempt (Table 1).^7,8 Spirito et al⁷ followed 58 adolescent suicide attempters for 3 months after their initial attempt. Seven (12%) made a subsequent attempt, and 26 (45%) reported continued suicidal ideation. Depressed mood was the strongest predictor of subsequent suicidal behavior, followed by poor family functioning, affect regulation difficulty, and hopelessness.

Hopelessness. Beck et al⁹ found that patients who scored ≥9 on the Beck Hopelessness Scale (BHS)—the most common self-report measure of hopelessness—were approximately 11 times more likely to commit suicide than patients who scored ≤8. A study of hospitalized suicide attempters found that BHS scores were unique predictors of future suicide attempts.¹⁰ Several studies have found that persons who remain consistently hopeless are more likely to kill themselves compared with those who have fluctuating hopelessness levels.^11,12

History of abuse—specifically sexual abuse—is associated with suicidal behavior. A study of depressed women age >50 found that among those who were sexually abused before age 18, 83% reported 1 suicide attempt and 67% made multiple attempts.¹³ Among women who had not experienced childhood sexual abuse, 58% reported a past suicide attempt and 27% made multiple attempts.¹³

In a separate study of psychiatric inpatients, those who had been physically or sexually abused were more likely to have made a suicide attempt than patients with no such history.¹⁴ This study did not find a difference in reported abuse between single and multiple suicide attempters.

Stressors. In many cases suicide attempts are precipitated by acute or chronic stressors, including:

• job stress
  
• chronic illness
  
• financial problems
  
• relationship discord
  
• retirement and declining physical health (especially for older men)
  
• death of a loved one.¹⁵
  

Risk is not necessarily cumulative—and not all risk factors are weighted equally. In general, however, the more risk factors a patient has, the greater the likelihood that he or she may attempt suicide.¹⁷

Table 1

Repeated suicidal behavior: Factors that increase risk

Red flag: Multiple attempts

When assessing a patient’s suicide history, ask about the number of attempts. A person who makes >1 suicide attempt—a multiple attempter—has a significantly higher chance of making subsequent attempts compared with those with 1 or no attempts.^18,19

Persons who make multiple attempts share certain characteristics (Table 2).^19-21 Rudd et al¹⁹ compared 68 multiple attempters with 128 single attempters and found that multiple attempters had higher levels of:

• suicide ideation
  
• depression
  
• hopelessness
  
• perceived stress.
  

• diagnosed with substance use disorders, psychotic disorder, or borderline personality disorder
  
• unemployed and have relationship difficulties, a history of emotional abuse, and a family history of psychiatric problems and suicide.
  

Among 326 individuals in a military medical setting treated for suicidal behavior or severe suicidal ideation, multiple suicide attempters reported higher levels of ongoing distress that was unrelated to specific life stressors.²² This suggests these patients may not respond well to psychological interventions that focus on problem-solving.

Table 2

Common characteristics of multiple suicide attempters

Self-harm and suicidal behavior

Patients who engage in nonsuicidal self harm—also called self-injurious behavior (SIB)—may be mistaken for suicide attempters. Although differences exist between suicide attempters and those who engage in SIB, evidence suggests that a history of SIB increases risk for suicidal behavior.^23,24 In a retrospective study of 4,167 self-harmers, females who engaged in ≥4 acts of SIB were more likely to die from suicide than those who engaged in ≤3 acts.²⁵ A cross-sectional analysis of data from 3,069 students responding to a random Web-based survey found that an increased incidence of SIB significantly increased the odds of suicidal behavior.²⁶

Although the link between SIB and suicide attempts remains unclear, evidence suggests SIB is a risk factor for suicidal behavior and therefore should be assessed when evaluating a patient’s suicide risk.

CASE CONTINUED: At high risk

Ms. J has several risk factors for making another suicide attempt. She has 3 previous attempts, and because her last attempt caused liver damage we know she is capable of lethal behavior. In addition, the anniversary of the death of her fiancé is approaching. Ms. J also reports almost constant suicidal ideation, with a specific plan (to overdose). Her fantasies of taking pills could be interpreted as mental rehearsal and desensitization to the behavior.

Because we believe Ms. J is at high risk for a serious, if not lethal, suicide attempt we conduct a 4-question suicide inquiry. It is clear that Ms. J had suicidal thoughts and a plan. Her answer to “How likely is it that once you leave my office you will do something to hurt yourself?” is the key to determining whether or not she requires hospitalization. Ms. J states that she is “pretty certain she will hurt herself” once she leaves the office, so we hospitalize her.

To determine if a patient requires immediate hospitalization, perform a specific suicide inquiry. Although there is no surefire way to determine if a patient will kill himself or herself, asking specific questions can help you gauge risk. Based on evidence²⁸ and my clinical experience, I focus on patients’ answers to 4 questions (Table 3). Affirmative answers to these questions are a strong indication that a patient requires hospitalization.

Occasionally, patients are not truthful when asked about their suicidal intent. If you suspect a patient is lying, clinical judgment and the patient’s history guide the decision on hospitalization.

Table 3

Hospitalize? 4 questions to guide your decision

• Joiner TE. Why people die by suicide. Cambridge, MA: Harvard University Press; 2005:46-93,203-22.
  
• American Foundation for Suicide Prevention. www.afsp.org.
  
• SAVE: Suicide Awareness Voices of Education. www.save.org.
  

• Zolpidem • Ambien
  

The author reports no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.