Evidence-Based Reviews

Many of the 28 million people who suffer from migraine headaches each year¹ need psychiatric care in addition to headache relief. Migraine headaches often coexist with depression,² anxiety/panic disorders,^2,3 bipolar disorder,⁴ and phobias,⁵ as well as with stroke⁶ and epilepsy.⁷ A study of 995 young adults found that anxiety disorders, phobias, major depression, panic disorder, and obsessive-compulsive disorder were two to five times more prevalent among migraine sufferers than among a control group (Table 1 ).²

Migraine sufferers know that at any time an attack could hamper their ability to work, care for their families, or engage in social activities. A nationwide study of migraineurs found that attacks often impaired their relationships with family and friends.⁸

Psychiatrists should screen patients for a history of migraine or other headaches and carefully consider the relationship between migraines and psychiatric disorders when prescribing treatment. In this article, we outline acute and preventive headache treatments and present two cases to help you treat these patients appropriately.

Table 1

PSYCHIATRIC COMORBIDITIES IN PATIENTS WITH VS. WITHOUT MIGRAINES*

Table 2

THREE TYPES OF PRIMARY HEADACHE: DIAGNOSTIC CRITERIA

Headache definitions and diagnosis

Primary or secondary. Under the International Headache Society’s (IHS) 1988 headache classification and diagnostic criteria,⁹ headaches are primary or secondary:

• Primary headaches are benign recurrent headaches that commonly present in practice.
• Secondary headaches occur much less frequently and are caused by underlying pathology.

The possibility of secondary headache should be ruled out before a primary headache can be diagnosed. The following headache features should cause concern:

• Severe headache with abrupt onset
• Subacute or progressive headache over days or months
• Headache, nausea, vomiting, and fever not explained by systemic illness
• New-onset headache late in life
• Headache with neurologic signs or symptoms such as confusion, decreased level of consciousness, meningismus, or papilledema
• Headache following head trauma
• Patient history of sickle cell disease, malignancy, or HIV.

Headache types. The three major types of primary headache are migraine, tension-type, and cluster (Table 2 ). Tension-type is the most common, is often mild, and is either self-treated with over-the-counter medications or ignored. Migraine is the most troublesome headache in everyday practice. Cluster is the most severe and fortunately is rare.

Migraine with aura and migraine without aura are separate diagnoses. IHS criteria for diagnosing migraines without aura are listed in Table 3. According to the IHS, migraine with aura (or “classic migraine”) fulfills all the criteria for migraine without aura, with fully reversible neurologic symptoms indicating focal cerebral cortical and/or brain stem dysfunction.

Auras. About 15% of migraineurs experience auras. Symptoms develop within 5 to 20 minutes, usually last less than 1 hour, and fade before the headache’s onset. Gradual onset and history of previous attacks helps to distinguish aura from transient ischemic attacks. Auras may manifest as visual, sensory, motor, or brain-stem symptoms, or as combinations of these:

• Visual auras are most common, presenting as localized visual loss (scotoma), with flashing lights (scintillation) at margins or jagged edges (fortification).
• Sensory auras present as facial or limb paresthesias.
• Motor auras manifest as weakness or lack of coordination.
• Brain stem auras manifest as vertigo or double vision.

Migraine aura is considered part of the headache’s prodrome, which may occur days or hours before the headache’s onset. The aura may bring about:

• an altered mental state (e.g., depression, hyperactivity, euphoria, difficulty concentrating, dysphasia)
• neurologic symptoms (e.g., photophobia, phonophobia, hyperosmia, yawning)
• general bodily discomforts (e.g., anorexia, food craving, diarrhea, thirst, urination, fluid retention, cold feeling).

Despite their sometimes severe effects, migraines often remain undiagnosed.¹⁰ Migraine should be suspected in patients with recurrent moderate to severe disabling headaches (Box).^11-15

Case 1: “Bad, sick headaches”

Ms. A, 23, a single parent with a 2-year-old child, has had trouble staying employed because of repeated illnesses. She made 17 visits to her primary care physician within 26 months. While her main complaint was headache, she also complained of other aches and pains, a lack of energy, and insomnia. Numerous examinations revealed no physical abnormalities.

She reported having “bad, sick headaches” that sometimes lasted 2 to 3 days. Bed rest helped but this was not always possible. The headache was throbbing and usually one-sided. She had no aura, and ibuprofen gave partial relief. She noted that her mother gets similar headaches.

Table 3

DIAGNOSTIC REQUIREMENTS FOR MIGRAINE WITHOUT AURA

Ms. A was diagnosed with migraine without aura, and she was treated with sumatriptan, 100 mg (1 or 2 doses, as needed). Her headaches responded well to this treatment, but the frequency of attacks remained unchanged. She requested a change in her medicine.

Box

Treating migraines

Acute treatment. Migraineurs whose attacks are infrequent and mild may find OTC analgesics or NSAIDs adequate. Most patients, however, require specific migraine treatment, usually with triptans. Acute oral treatment options include sumatriptan, 50 to 100 mg; rizatriptan, 10 mg; zolmitriptan, 2.5 to 5 mg; and eletriptan, 40 mg.

In case of vomiting or nausea, options include sumatriptan, 20 mg nasal spray or 6 mg SC; rizatriptan, 10 mg on a dissolving wafer; or dihydroergotamine, 2 mg nasal spray or 1 mg IM or SC. For severe nausea or vomiting, an anti-nauseant (e.g., prochlorperazine suppositories, 25 mg) may be of value.

Preventive treatment. Preventive treatment may be warranted, depending on attack frequency, severity, and the extent of disability caused. One prolonged, severe attack per month that responds poorly to acute treatment may indicate the need for preventive treatment. A range of preventative treatments is available (Table 4).

In Ms. A’s case, oral sumatriptan lessened the severity of the migraine attacks, and the addition of nortriptyline, 50 mg/d, reduced frequency by about 50%. She felt more energetic overall and was sleeping better.

Treating the psychiatric comorbidity

Behavioral therapy is used as an adjunct to pharmacologic headache treatment. This approach is usually considered after a poor or adverse response to treatment, or when pharmacologic treatment is contraindicated (e.g., during pregnancy).

Relaxation training, biofeedback, and cognitive-behavioral stress management are the most commonly used forms of behavioral therapy. Thirty-five to 55% improvement in migraine has been reported following such treatments.¹⁶

Cognitive-behavioral intervention has been shown to be effective in depression¹⁷ and anxiety disorders.¹⁸ When either psychiatric problem is comorbid with migraine, cognitive therapy can improve both the migraine and the psychiatric comorbidity.

Pharmacologic therapy. Depression is commonly associated with migraine and may be caused by living with chronic disabling headaches over time. In such cases, the depression will improve as the migraine responds to treatment. However, in cases where comorbid depression or anxiety trigger or exacerbate acute migraine attacks, neither the migraine nor the psychiatric problem responds until the underlying psychopathology is treated. In such cases, simultaneous psychiatric and migraine pharmacologic treatment is required.

We recommend that you treat the psychiatric comorbidity as it would be treated without a co-existing migraine. Be advised, however, that monoamine oxidase inhibitors are contraindicated in depression during the 2 weeks before treating the comorbid migraine with a triptan. If the patient does not respond or if there is concern regarding possible underlying pathology, consult with a clinician who specializes in headache treatment.

Precipitating and aggravating factors

Headache triggers. Helping patients to recognize headache triggers and aggravating factors is an important element in treating and preventing migraines. Identifying these factors in the patient history can help you establish a diagnosis and implement steps to avoid or reduce attack severity.

Table 4

TREATMENT OPTIONS FOR PREVENTING MIGRAINE ATTACKS

Common migraine headache triggers include menstruation, stress, relaxation after stress, fatigue, too much or too little sleep, skipping meals, weather changes, high humidity, glare and flickering lights, loud or high-pitched noises, smoke or dust, strong perfumes or cooking aromas. Food triggers cause 10% of migraine cases. Chocolate, strong cheeses, red wine, beer, citrus fruits, and foods with monosodium glutamate and nitrate preservatives are common food triggers.

Tension-type headaches are triggered by stress or the end of a stress-filled day. Triggers for cluster-type headaches include alcohol, smoking during the cluster phase, and lying down during an attack.

Case 2: Flying the unfriendly skies

Ms. B, 38, is a mother of three who works as a flight attendant. She is separated from her husband and had filed for divorce because of repeated spousal abuse. She has visited her primary care physician multiple times for migraine, sinus problems, backache, and coccygodynia. Orthopedic and rectal examinations revealed no abnormalities.

Her headaches met the IHS diagnostic criteria for migraine with aura, and these responded well to zolmitriptan, 5 mg. The headaches usually occurred during days off from work, but her sinus problems also led to headaches and nasal stuffiness when she flew. Her headaches eventually occurred almost daily.

Her supervisor was unsympathetic. An otolaryngologist had prescribed decongestants and a course of desensitization, both of which brought only transient relief.

A counselor at work recommended that Ms. B go on sick leave and accept a transfer to a non-flying job. The patient was tearful and felt overwhelmed by her problems. She felt that life was no longer worth living. She agreed to see a psychiatrist, who diagnosed depression and anxiety disorder. The psychiatrist prescribed citalopram, 20 mg/d, and agreed to see her regularly to monitor progress.

Discussion. As a migraineur, Ms. B was at increased risk for depression and anxiety disorders.¹⁹ Migraine with aura is associated with an increased lifetime prevalence of suicidal ideation and suicide attempts.²⁰

The exact mechanisms by which migraine and depression are related are unknown. Each disorder increases the risk for developing the other. The specificity of this relationship is strengthened by the fact that depression is not associated with a greater risk of severe nonmigrainous headache, even though a severe nonmigrainous headache may cause depression.²¹

The patient in case 1 responded well when an antidepressant was added to her treatment. In her case, the diagnosis of a depressive disorder remained an open question. Migraine attacks are known to be associated with mood change, lethargy, and cognitive changes. The picture may be further confounded because migraine without depression responds well to prophylaxis with antidepressants.

The patient in case 2, however, presented with a complex of interrelated headache and psychiatric problems of potentially dangerous proportions. Psychiatric problems in migraineurs may be deep-seated, and these patients may require urgent, specialized attention to avoid further serious disability and a possible tragic outcome.

Related resources

• Silberstein SD, Lipton RB, Goadsby PJ, Smith, R, eds. Headache in primary care. Oxford, UK: Isis Medical Media, 1999.
• Silberstein SD, Lipton RB, Dalessio DJ. Wolff’s headache and other head pain (7th ed). New York: Oxford University Press, 2001.
• Davidoff RA. Migraine. Manifestations, pathogenesis and management (2nd ed). New York: Oxford University Press, 2002.
• International Headache Society. http://www.i-h-s.org

Drug brand names

• Citalopram • Celexa
• Dihydroergotamine • Migranal
• Eletriptan • Relpax
• Rizatriptan • Maxalt
• Sumatriptan • Imitrex
• Valproate sodium • Depakote
• Zolmitriptan • Zomig

Disclosure

Dr. Smith reports that he serves as a consultant to and is on the speakers’ bureau of AstraZeneca Pharmaceuticals.

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

Many of the 28 million people who suffer from migraine headaches each year¹ need psychiatric care in addition to headache relief. Migraine headaches often coexist with depression,² anxiety/panic disorders,^2,3 bipolar disorder,⁴ and phobias,⁵ as well as with stroke⁶ and epilepsy.⁷ A study of 995 young adults found that anxiety disorders, phobias, major depression, panic disorder, and obsessive-compulsive disorder were two to five times more prevalent among migraine sufferers than among a control group (Table 1 ).²

Migraine sufferers know that at any time an attack could hamper their ability to work, care for their families, or engage in social activities. A nationwide study of migraineurs found that attacks often impaired their relationships with family and friends.⁸

Psychiatrists should screen patients for a history of migraine or other headaches and carefully consider the relationship between migraines and psychiatric disorders when prescribing treatment. In this article, we outline acute and preventive headache treatments and present two cases to help you treat these patients appropriately.

Table 1

PSYCHIATRIC COMORBIDITIES IN PATIENTS WITH VS. WITHOUT MIGRAINES*

Table 2

THREE TYPES OF PRIMARY HEADACHE: DIAGNOSTIC CRITERIA

Headache definitions and diagnosis

Primary or secondary. Under the International Headache Society’s (IHS) 1988 headache classification and diagnostic criteria,⁹ headaches are primary or secondary:

• Primary headaches are benign recurrent headaches that commonly present in practice.
• Secondary headaches occur much less frequently and are caused by underlying pathology.

The possibility of secondary headache should be ruled out before a primary headache can be diagnosed. The following headache features should cause concern:

• Severe headache with abrupt onset
• Subacute or progressive headache over days or months
• Headache, nausea, vomiting, and fever not explained by systemic illness
• New-onset headache late in life
• Headache with neurologic signs or symptoms such as confusion, decreased level of consciousness, meningismus, or papilledema
• Headache following head trauma
• Patient history of sickle cell disease, malignancy, or HIV.

Headache types. The three major types of primary headache are migraine, tension-type, and cluster (Table 2 ). Tension-type is the most common, is often mild, and is either self-treated with over-the-counter medications or ignored. Migraine is the most troublesome headache in everyday practice. Cluster is the most severe and fortunately is rare.

Migraine with aura and migraine without aura are separate diagnoses. IHS criteria for diagnosing migraines without aura are listed in Table 3. According to the IHS, migraine with aura (or “classic migraine”) fulfills all the criteria for migraine without aura, with fully reversible neurologic symptoms indicating focal cerebral cortical and/or brain stem dysfunction.

Auras. About 15% of migraineurs experience auras. Symptoms develop within 5 to 20 minutes, usually last less than 1 hour, and fade before the headache’s onset. Gradual onset and history of previous attacks helps to distinguish aura from transient ischemic attacks. Auras may manifest as visual, sensory, motor, or brain-stem symptoms, or as combinations of these:

• Visual auras are most common, presenting as localized visual loss (scotoma), with flashing lights (scintillation) at margins or jagged edges (fortification).
• Sensory auras present as facial or limb paresthesias.
• Motor auras manifest as weakness or lack of coordination.
• Brain stem auras manifest as vertigo or double vision.

Migraine aura is considered part of the headache’s prodrome, which may occur days or hours before the headache’s onset. The aura may bring about:

• an altered mental state (e.g., depression, hyperactivity, euphoria, difficulty concentrating, dysphasia)
• neurologic symptoms (e.g., photophobia, phonophobia, hyperosmia, yawning)
• general bodily discomforts (e.g., anorexia, food craving, diarrhea, thirst, urination, fluid retention, cold feeling).

Despite their sometimes severe effects, migraines often remain undiagnosed.¹⁰ Migraine should be suspected in patients with recurrent moderate to severe disabling headaches (Box).^11-15

Case 1: “Bad, sick headaches”

Ms. A, 23, a single parent with a 2-year-old child, has had trouble staying employed because of repeated illnesses. She made 17 visits to her primary care physician within 26 months. While her main complaint was headache, she also complained of other aches and pains, a lack of energy, and insomnia. Numerous examinations revealed no physical abnormalities.

She reported having “bad, sick headaches” that sometimes lasted 2 to 3 days. Bed rest helped but this was not always possible. The headache was throbbing and usually one-sided. She had no aura, and ibuprofen gave partial relief. She noted that her mother gets similar headaches.

Table 3

DIAGNOSTIC REQUIREMENTS FOR MIGRAINE WITHOUT AURA

Ms. A was diagnosed with migraine without aura, and she was treated with sumatriptan, 100 mg (1 or 2 doses, as needed). Her headaches responded well to this treatment, but the frequency of attacks remained unchanged. She requested a change in her medicine.

Box

Treating migraines

Acute treatment. Migraineurs whose attacks are infrequent and mild may find OTC analgesics or NSAIDs adequate. Most patients, however, require specific migraine treatment, usually with triptans. Acute oral treatment options include sumatriptan, 50 to 100 mg; rizatriptan, 10 mg; zolmitriptan, 2.5 to 5 mg; and eletriptan, 40 mg.

In case of vomiting or nausea, options include sumatriptan, 20 mg nasal spray or 6 mg SC; rizatriptan, 10 mg on a dissolving wafer; or dihydroergotamine, 2 mg nasal spray or 1 mg IM or SC. For severe nausea or vomiting, an anti-nauseant (e.g., prochlorperazine suppositories, 25 mg) may be of value.

Preventive treatment. Preventive treatment may be warranted, depending on attack frequency, severity, and the extent of disability caused. One prolonged, severe attack per month that responds poorly to acute treatment may indicate the need for preventive treatment. A range of preventative treatments is available (Table 4).

In Ms. A’s case, oral sumatriptan lessened the severity of the migraine attacks, and the addition of nortriptyline, 50 mg/d, reduced frequency by about 50%. She felt more energetic overall and was sleeping better.

Treating the psychiatric comorbidity

Behavioral therapy is used as an adjunct to pharmacologic headache treatment. This approach is usually considered after a poor or adverse response to treatment, or when pharmacologic treatment is contraindicated (e.g., during pregnancy).

Relaxation training, biofeedback, and cognitive-behavioral stress management are the most commonly used forms of behavioral therapy. Thirty-five to 55% improvement in migraine has been reported following such treatments.¹⁶

Cognitive-behavioral intervention has been shown to be effective in depression¹⁷ and anxiety disorders.¹⁸ When either psychiatric problem is comorbid with migraine, cognitive therapy can improve both the migraine and the psychiatric comorbidity.

Pharmacologic therapy. Depression is commonly associated with migraine and may be caused by living with chronic disabling headaches over time. In such cases, the depression will improve as the migraine responds to treatment. However, in cases where comorbid depression or anxiety trigger or exacerbate acute migraine attacks, neither the migraine nor the psychiatric problem responds until the underlying psychopathology is treated. In such cases, simultaneous psychiatric and migraine pharmacologic treatment is required.

We recommend that you treat the psychiatric comorbidity as it would be treated without a co-existing migraine. Be advised, however, that monoamine oxidase inhibitors are contraindicated in depression during the 2 weeks before treating the comorbid migraine with a triptan. If the patient does not respond or if there is concern regarding possible underlying pathology, consult with a clinician who specializes in headache treatment.

Precipitating and aggravating factors

Headache triggers. Helping patients to recognize headache triggers and aggravating factors is an important element in treating and preventing migraines. Identifying these factors in the patient history can help you establish a diagnosis and implement steps to avoid or reduce attack severity.

Table 4

TREATMENT OPTIONS FOR PREVENTING MIGRAINE ATTACKS

Common migraine headache triggers include menstruation, stress, relaxation after stress, fatigue, too much or too little sleep, skipping meals, weather changes, high humidity, glare and flickering lights, loud or high-pitched noises, smoke or dust, strong perfumes or cooking aromas. Food triggers cause 10% of migraine cases. Chocolate, strong cheeses, red wine, beer, citrus fruits, and foods with monosodium glutamate and nitrate preservatives are common food triggers.

Tension-type headaches are triggered by stress or the end of a stress-filled day. Triggers for cluster-type headaches include alcohol, smoking during the cluster phase, and lying down during an attack.

Case 2: Flying the unfriendly skies

Ms. B, 38, is a mother of three who works as a flight attendant. She is separated from her husband and had filed for divorce because of repeated spousal abuse. She has visited her primary care physician multiple times for migraine, sinus problems, backache, and coccygodynia. Orthopedic and rectal examinations revealed no abnormalities.

Her headaches met the IHS diagnostic criteria for migraine with aura, and these responded well to zolmitriptan, 5 mg. The headaches usually occurred during days off from work, but her sinus problems also led to headaches and nasal stuffiness when she flew. Her headaches eventually occurred almost daily.

Her supervisor was unsympathetic. An otolaryngologist had prescribed decongestants and a course of desensitization, both of which brought only transient relief.

A counselor at work recommended that Ms. B go on sick leave and accept a transfer to a non-flying job. The patient was tearful and felt overwhelmed by her problems. She felt that life was no longer worth living. She agreed to see a psychiatrist, who diagnosed depression and anxiety disorder. The psychiatrist prescribed citalopram, 20 mg/d, and agreed to see her regularly to monitor progress.

Discussion. As a migraineur, Ms. B was at increased risk for depression and anxiety disorders.¹⁹ Migraine with aura is associated with an increased lifetime prevalence of suicidal ideation and suicide attempts.²⁰

The exact mechanisms by which migraine and depression are related are unknown. Each disorder increases the risk for developing the other. The specificity of this relationship is strengthened by the fact that depression is not associated with a greater risk of severe nonmigrainous headache, even though a severe nonmigrainous headache may cause depression.²¹

The patient in case 1 responded well when an antidepressant was added to her treatment. In her case, the diagnosis of a depressive disorder remained an open question. Migraine attacks are known to be associated with mood change, lethargy, and cognitive changes. The picture may be further confounded because migraine without depression responds well to prophylaxis with antidepressants.

The patient in case 2, however, presented with a complex of interrelated headache and psychiatric problems of potentially dangerous proportions. Psychiatric problems in migraineurs may be deep-seated, and these patients may require urgent, specialized attention to avoid further serious disability and a possible tragic outcome.

Related resources

• Silberstein SD, Lipton RB, Goadsby PJ, Smith, R, eds. Headache in primary care. Oxford, UK: Isis Medical Media, 1999.
• Silberstein SD, Lipton RB, Dalessio DJ. Wolff’s headache and other head pain (7th ed). New York: Oxford University Press, 2001.
• Davidoff RA. Migraine. Manifestations, pathogenesis and management (2nd ed). New York: Oxford University Press, 2002.
• International Headache Society. http://www.i-h-s.org

Drug brand names

• Citalopram • Celexa
• Dihydroergotamine • Migranal
• Eletriptan • Relpax
• Rizatriptan • Maxalt
• Sumatriptan • Imitrex
• Valproate sodium • Depakote
• Zolmitriptan • Zomig

Disclosure

Dr. Smith reports that he serves as a consultant to and is on the speakers’ bureau of AstraZeneca Pharmaceuticals.

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

Many of the 28 million people who suffer from migraine headaches each year¹ need psychiatric care in addition to headache relief. Migraine headaches often coexist with depression,² anxiety/panic disorders,^2,3 bipolar disorder,⁴ and phobias,⁵ as well as with stroke⁶ and epilepsy.⁷ A study of 995 young adults found that anxiety disorders, phobias, major depression, panic disorder, and obsessive-compulsive disorder were two to five times more prevalent among migraine sufferers than among a control group (Table 1 ).²

Migraine sufferers know that at any time an attack could hamper their ability to work, care for their families, or engage in social activities. A nationwide study of migraineurs found that attacks often impaired their relationships with family and friends.⁸

Psychiatrists should screen patients for a history of migraine or other headaches and carefully consider the relationship between migraines and psychiatric disorders when prescribing treatment. In this article, we outline acute and preventive headache treatments and present two cases to help you treat these patients appropriately.

Table 1

PSYCHIATRIC COMORBIDITIES IN PATIENTS WITH VS. WITHOUT MIGRAINES*

Table 2

THREE TYPES OF PRIMARY HEADACHE: DIAGNOSTIC CRITERIA

Headache definitions and diagnosis

Primary or secondary. Under the International Headache Society’s (IHS) 1988 headache classification and diagnostic criteria,⁹ headaches are primary or secondary:

• Primary headaches are benign recurrent headaches that commonly present in practice.
• Secondary headaches occur much less frequently and are caused by underlying pathology.

The possibility of secondary headache should be ruled out before a primary headache can be diagnosed. The following headache features should cause concern:

• Severe headache with abrupt onset
• Subacute or progressive headache over days or months
• Headache, nausea, vomiting, and fever not explained by systemic illness
• New-onset headache late in life
• Headache with neurologic signs or symptoms such as confusion, decreased level of consciousness, meningismus, or papilledema
• Headache following head trauma
• Patient history of sickle cell disease, malignancy, or HIV.

Headache types. The three major types of primary headache are migraine, tension-type, and cluster (Table 2 ). Tension-type is the most common, is often mild, and is either self-treated with over-the-counter medications or ignored. Migraine is the most troublesome headache in everyday practice. Cluster is the most severe and fortunately is rare.

Migraine with aura and migraine without aura are separate diagnoses. IHS criteria for diagnosing migraines without aura are listed in Table 3. According to the IHS, migraine with aura (or “classic migraine”) fulfills all the criteria for migraine without aura, with fully reversible neurologic symptoms indicating focal cerebral cortical and/or brain stem dysfunction.

Auras. About 15% of migraineurs experience auras. Symptoms develop within 5 to 20 minutes, usually last less than 1 hour, and fade before the headache’s onset. Gradual onset and history of previous attacks helps to distinguish aura from transient ischemic attacks. Auras may manifest as visual, sensory, motor, or brain-stem symptoms, or as combinations of these:

• Visual auras are most common, presenting as localized visual loss (scotoma), with flashing lights (scintillation) at margins or jagged edges (fortification).
• Sensory auras present as facial or limb paresthesias.
• Motor auras manifest as weakness or lack of coordination.
• Brain stem auras manifest as vertigo or double vision.

Migraine aura is considered part of the headache’s prodrome, which may occur days or hours before the headache’s onset. The aura may bring about:

• an altered mental state (e.g., depression, hyperactivity, euphoria, difficulty concentrating, dysphasia)
• neurologic symptoms (e.g., photophobia, phonophobia, hyperosmia, yawning)
• general bodily discomforts (e.g., anorexia, food craving, diarrhea, thirst, urination, fluid retention, cold feeling).

Despite their sometimes severe effects, migraines often remain undiagnosed.¹⁰ Migraine should be suspected in patients with recurrent moderate to severe disabling headaches (Box).^11-15

Case 1: “Bad, sick headaches”

Ms. A, 23, a single parent with a 2-year-old child, has had trouble staying employed because of repeated illnesses. She made 17 visits to her primary care physician within 26 months. While her main complaint was headache, she also complained of other aches and pains, a lack of energy, and insomnia. Numerous examinations revealed no physical abnormalities.

She reported having “bad, sick headaches” that sometimes lasted 2 to 3 days. Bed rest helped but this was not always possible. The headache was throbbing and usually one-sided. She had no aura, and ibuprofen gave partial relief. She noted that her mother gets similar headaches.

Table 3

DIAGNOSTIC REQUIREMENTS FOR MIGRAINE WITHOUT AURA

Ms. A was diagnosed with migraine without aura, and she was treated with sumatriptan, 100 mg (1 or 2 doses, as needed). Her headaches responded well to this treatment, but the frequency of attacks remained unchanged. She requested a change in her medicine.

Box

Treating migraines

Acute treatment. Migraineurs whose attacks are infrequent and mild may find OTC analgesics or NSAIDs adequate. Most patients, however, require specific migraine treatment, usually with triptans. Acute oral treatment options include sumatriptan, 50 to 100 mg; rizatriptan, 10 mg; zolmitriptan, 2.5 to 5 mg; and eletriptan, 40 mg.

In case of vomiting or nausea, options include sumatriptan, 20 mg nasal spray or 6 mg SC; rizatriptan, 10 mg on a dissolving wafer; or dihydroergotamine, 2 mg nasal spray or 1 mg IM or SC. For severe nausea or vomiting, an anti-nauseant (e.g., prochlorperazine suppositories, 25 mg) may be of value.

Preventive treatment. Preventive treatment may be warranted, depending on attack frequency, severity, and the extent of disability caused. One prolonged, severe attack per month that responds poorly to acute treatment may indicate the need for preventive treatment. A range of preventative treatments is available (Table 4).

In Ms. A’s case, oral sumatriptan lessened the severity of the migraine attacks, and the addition of nortriptyline, 50 mg/d, reduced frequency by about 50%. She felt more energetic overall and was sleeping better.

Treating the psychiatric comorbidity

Behavioral therapy is used as an adjunct to pharmacologic headache treatment. This approach is usually considered after a poor or adverse response to treatment, or when pharmacologic treatment is contraindicated (e.g., during pregnancy).

Relaxation training, biofeedback, and cognitive-behavioral stress management are the most commonly used forms of behavioral therapy. Thirty-five to 55% improvement in migraine has been reported following such treatments.¹⁶

Cognitive-behavioral intervention has been shown to be effective in depression¹⁷ and anxiety disorders.¹⁸ When either psychiatric problem is comorbid with migraine, cognitive therapy can improve both the migraine and the psychiatric comorbidity.

Pharmacologic therapy. Depression is commonly associated with migraine and may be caused by living with chronic disabling headaches over time. In such cases, the depression will improve as the migraine responds to treatment. However, in cases where comorbid depression or anxiety trigger or exacerbate acute migraine attacks, neither the migraine nor the psychiatric problem responds until the underlying psychopathology is treated. In such cases, simultaneous psychiatric and migraine pharmacologic treatment is required.

We recommend that you treat the psychiatric comorbidity as it would be treated without a co-existing migraine. Be advised, however, that monoamine oxidase inhibitors are contraindicated in depression during the 2 weeks before treating the comorbid migraine with a triptan. If the patient does not respond or if there is concern regarding possible underlying pathology, consult with a clinician who specializes in headache treatment.

Precipitating and aggravating factors

Headache triggers. Helping patients to recognize headache triggers and aggravating factors is an important element in treating and preventing migraines. Identifying these factors in the patient history can help you establish a diagnosis and implement steps to avoid or reduce attack severity.

Table 4

TREATMENT OPTIONS FOR PREVENTING MIGRAINE ATTACKS

Common migraine headache triggers include menstruation, stress, relaxation after stress, fatigue, too much or too little sleep, skipping meals, weather changes, high humidity, glare and flickering lights, loud or high-pitched noises, smoke or dust, strong perfumes or cooking aromas. Food triggers cause 10% of migraine cases. Chocolate, strong cheeses, red wine, beer, citrus fruits, and foods with monosodium glutamate and nitrate preservatives are common food triggers.

Tension-type headaches are triggered by stress or the end of a stress-filled day. Triggers for cluster-type headaches include alcohol, smoking during the cluster phase, and lying down during an attack.

Case 2: Flying the unfriendly skies

Ms. B, 38, is a mother of three who works as a flight attendant. She is separated from her husband and had filed for divorce because of repeated spousal abuse. She has visited her primary care physician multiple times for migraine, sinus problems, backache, and coccygodynia. Orthopedic and rectal examinations revealed no abnormalities.

Her headaches met the IHS diagnostic criteria for migraine with aura, and these responded well to zolmitriptan, 5 mg. The headaches usually occurred during days off from work, but her sinus problems also led to headaches and nasal stuffiness when she flew. Her headaches eventually occurred almost daily.

Her supervisor was unsympathetic. An otolaryngologist had prescribed decongestants and a course of desensitization, both of which brought only transient relief.

A counselor at work recommended that Ms. B go on sick leave and accept a transfer to a non-flying job. The patient was tearful and felt overwhelmed by her problems. She felt that life was no longer worth living. She agreed to see a psychiatrist, who diagnosed depression and anxiety disorder. The psychiatrist prescribed citalopram, 20 mg/d, and agreed to see her regularly to monitor progress.

Discussion. As a migraineur, Ms. B was at increased risk for depression and anxiety disorders.¹⁹ Migraine with aura is associated with an increased lifetime prevalence of suicidal ideation and suicide attempts.²⁰

The exact mechanisms by which migraine and depression are related are unknown. Each disorder increases the risk for developing the other. The specificity of this relationship is strengthened by the fact that depression is not associated with a greater risk of severe nonmigrainous headache, even though a severe nonmigrainous headache may cause depression.²¹

The patient in case 1 responded well when an antidepressant was added to her treatment. In her case, the diagnosis of a depressive disorder remained an open question. Migraine attacks are known to be associated with mood change, lethargy, and cognitive changes. The picture may be further confounded because migraine without depression responds well to prophylaxis with antidepressants.

The patient in case 2, however, presented with a complex of interrelated headache and psychiatric problems of potentially dangerous proportions. Psychiatric problems in migraineurs may be deep-seated, and these patients may require urgent, specialized attention to avoid further serious disability and a possible tragic outcome.

Related resources

• Silberstein SD, Lipton RB, Goadsby PJ, Smith, R, eds. Headache in primary care. Oxford, UK: Isis Medical Media, 1999.
• Silberstein SD, Lipton RB, Dalessio DJ. Wolff’s headache and other head pain (7th ed). New York: Oxford University Press, 2001.
• Davidoff RA. Migraine. Manifestations, pathogenesis and management (2nd ed). New York: Oxford University Press, 2002.
• International Headache Society. http://www.i-h-s.org

Drug brand names

• Citalopram • Celexa
• Dihydroergotamine • Migranal
• Eletriptan • Relpax
• Rizatriptan • Maxalt
• Sumatriptan • Imitrex
• Valproate sodium • Depakote
• Zolmitriptan • Zomig

Disclosure

Dr. Smith reports that he serves as a consultant to and is on the speakers’ bureau of AstraZeneca Pharmaceuticals.

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

Not everyone agrees that borderline personality disorder (BPD) should be a diagnostic category. BPD became “official” with DSM-III in 1980, although the term had been used for 40 years to describe various patient groups. Being listed in DSM-III legitimized BPD, which was thought to represent a specific—though not necessarily distinct—diagnostic category.

The history of the BPD diagnosis and opinions as to its usefulness can be viewed as a microcosm of psychiatric diagnosis in general. Before DSM-III, diagnoses were broadly defined and did not contain specific inclusion or exclusion criteria.¹ For the 5 to 10 years prior to DSM-III, however, two assumptions developed:

• distinct diagnostic categories did, in fact, exist
• by rigorously defining and studying those categories we could develop more specific and effective treatments for our patients.²

The specificity and exclusivity that we assumed we could achieve by categorical diagnoses, however, remain a distant wish. Comorbidity appears more common in psychiatry than was originally thought and confounds both treatment and outcome.³ Also, many patients appear treatment-resistant, despite fitting neatly into diagnostic categories.⁴

As psychiatry considers DSM-V, questions linger as to whether BPD (and personality disorder in general) should remain as a categorical diagnosis or if dimensional measures may be more appropriate. Dimensions imply that no one ever fits into a given box because no specific box exists. Rather, patients are described as being closer to or more distant from a prototypic model of the diagnosis. In personality disorders, the dimensions most often mentioned are cognition, impulsivity, emotional lability, environmental hyperreactivity, and anxiety. The case report (above) illustrates the interplay of these dimensions in a typical patient with presumed BPD.

What’s in a name?

The symptom complex or syndrome that bears the name borderline personality disorder has probably existed for as long as people have thought about patients in psychopathologic terms.⁵ Before 1980, the term “borderline” applied primarily to two separate but overlapping concepts:

• Patients thought to reside on the “border” with psychosis, such as the patient in our case example. They seemed to have an underlying psychotic disorder, but the psychosis—if it surfaced—appeared briefly, was not exceptionally deep or firmly held, and was not regularly evident or immediately accessible to the clinician.
• Patients who appeared to occupy the space between neurosis and psychosis. This concept evolved into the idea of a character or personality disorder distinguished primarily by unstable interpersonal relationships, a confused or inconsistent sense of identity, and emotional instability.

How DSM is changing. Comparing the disorders listed in DSM-IV (1994)⁶ versus DSM-II (1968)¹ suggests that psychiatry has become enamored of the naming process. For example, DSM-II lists anxiety neurosis (300.0), phobic neurosis (300.2), and obsessive-compulsive neurosis (300.3), whereas DSM-IV lists 11 different categories of anxiety disorders.

But beyond naming, subsequent DSMs have differed even more dramatically from DSM-II. We have seen a shift from describing a diagnostic category with a simple explanatory paragraph to lists of specific inclusion and exclusion criteria. These more-specific lists imply that they define categories closer to some reality or authenticity than did previous definitions.

Before DSM-III, the borderline concept was conceived in broad object relational and psychodynamic terms. In contrast, DSM-III produced a definitive set of criteria and required that a subset be met before the diagnosis could be made.⁷ An example of this criteria-based model is shown in Box 1, which lists the DSM-IV-TR criteria for BPD.

Some psychiatrists objected that BPD was solely a psychoanalytic construct and too theoretical for inclusion in DSM-III. Others argued that if BPD were not defined, it would be difficult to study the clinical usefulness of that definition or any other. Nonetheless, many have argued that BPD does not exist, though to what category BPD patients should belong has changed over the years:

• Is BPD nothing more than a milder or unusual presentation of an affective disorder⁸ or actually bipolar II disorder?⁹
• Is it a presentation of posttraumatic stress disorder (PTSD) called “complex PTSD,”^10-11 or an adult presentation of attention-deficit/ hyperactivity or other brain disorder?¹²
• Is it a stigmatizing diagnosis that we apply to patients whom we do not like?¹³

In truth, the diagnosis of BPD reflects a particular clinical presentation no more or less accurately than many of the well-accepted axis I disorders. Despite recent advances in the neurosciences, the dilemma we face as psychiatrists is that we make a diagnosis based upon what we see in the clinical setting (i.e., a phenotype). Yet in labeling what we believe is a specific psychiatric disorder, we make assumptions—for better or for worse, consciously or unconsciously—about pathophysiology and indirectly about genotype.

Box 1

Defining the borderline personality

Stern first used the term “borderline” in 1938 to describe patients who appeared to occupy the border between neurosis and psychosis.¹⁴ In 1942, Deutsch described the “as if” personality in patients who seemed chameleon-like. They could adapt or play the role demanded of them in specific situations, yet elsewhere—as in the analyst’s office—they had little sense of themselves and were thought to be internally disorganized and probably psychotic.¹⁵

Border to psychosis. The idea that borderline-type patients were psychotic continued in Hoch and Polatin’s description of the “pseudoneurotic schizophrenic,”¹⁶ a patient who appeared severely neurotic but was thought to employ many defenses and interpersonal styles to ward off a fundamental inner psychosis. Knight used the label “borderline states”¹⁷ to describe severely ill patients who were not frankly psychotic but fell within the realm of psychosis without qualifying for a diagnosis of schizophrenia. Knight was the first person to use the term “borderline” as a diagnostic entity, though simultaneously he argued against its use as a label because the term lacked precision.

Psychotic character. About the same time, Schmideberg characterized a group of patients whose emotional lability or affective reactivity seemed to be a consistent aspect of their clinical presentation. She believed this appearance of “stable instability”¹⁸ represented the patient’s characterologic adaptation to the world.

Frosch coined the term “psychotic character”¹⁹ that aptly captured both the characterologic and the border-to-psychosis aspects of these patients’ clinical picture. According to Frosch, these patients appeared to regress readily into psychotic thinking, yet they did not lose their ability to test reality.

Affective and emotional instability. Thus until the 1960s, the term borderline was applied primarily to patients who appeared to occupy the border between neurosis and psychosis but were thought to be closer to psychotic than neurotic. And this sitting close to the edge of psychosis appeared to be a stable condition.

Most of the attention up until this point had been paid to how these patients thought—with little attention to their affective lability or emotional instability, save for Schmideberg’s comments. In the 1960s, however, the term borderline was applied somewhat differently—not completely divorced from previous concepts but with greater emphasis on borderline as a stable but psychopathologic functioning of the personality that included affective and emotional instability and an impaired sense of self.

Box 2

Impaired personality organization. In 1967, Kernberg published a seminal article in the history of BPD diagnosis. Although he did not discuss the diagnosis of BPD, Kernberg did develop a concept concerning a specific organization of the personality based upon impaired object relations. This impaired organization could apply across several personality disorders. The construct, named borderline personality organization (BPO),²⁰ was defined by:

• an impaired sense of identity and lack of integration of one’s own identity
• use of primitive defenses, including splitting, rage, and regression
• ability to test reality.

Kernberg’s theory is too complex to summarize here, but he—along with Roy Grinker—is responsible for placing BPD on the diagnostic map. He was the first to describe BPO (and by extension BPD) in terms of a personality disorder.

Grinker’s ‘core’ group. Almost simultaneously (in 1968), Grinker published a careful study of 50 hospitalized patients. His work on the “borderline syndrome”²¹ revealed four subgroups to which the label of borderline had been applied:

• those occupying the border with psychosis
• those occupying the border with neurosis
• those similar to Deutsch’s “as if” group
• the “core” borderline group.

The core group—with its symptoms of anger and loneliness, a nonintegrated sense of self, and labile and oscillating interpersonal relationships— defined patients closest to our current definition.

Six criteria for BPD. In 1975, Gunderson and Singer published an article that greatly influenced our definition of BPD. They reviewed major descriptive accounts of BPD or BPD-like syndromes²² and proposed six diagnostic criteria (Box 2), though they did not identify a specific number or subset of the criteria as needing to be met for the diagnosis. (It is important to note that the term BPD did not become official for 5 more years.)

DSM-IV’s definition of BPD retains four of Gunderson and Singer‘s criteria among the nine it lists (five being necessary for a diagnosis of BPD). Missing are:

• social adaptiveness—though DSM-IV does say that social adaptiveness may be superficial (as in the “as if” personality) and may hide a disturbed identity⁶
• and the criterion relating to psychological test performance (this omission reflects a movement since 1980 away from listing “psychological” or psychodynamic criteria in the DSM).

DSM-III. BPD was included in DSM-III⁷ following an important study that tried to determine whether the term “borderline” refers to patients at the border of psychosis or to a stable group with mood instability and affective lability as part of a personality disorder. Spitzer et al²³ asked 808 clinicians to describe patients they would label as borderline and to use 22 items gleaned from the literature to score two of their own patients:

• one patient who the clinician felt truly had borderline personality, borderline personality organization, or borderline schizophrenia
• and a control patient who was not diagnosed as psychotic and did not fall into any borderline category.

Box 3

Their responses showed that BPD and schizotypal personality disorder (SPD) were separate, independent (though not mutually exclusive) disorders. Spitzer et al preserved the “schizotypal” label in DSM-III to describe the personality disorder that closely matched the border-to-psychosis subset. The other criteria set, which they labeled the “unstable personality item set,” was renamed “borderline” in DSM-III to describe the personality disorder that closely matched the emotional lability subset.

Box 4

Not everyone agrees with renaming the unstable set “borderline” because the word:

• has always been ambiguous
• does not connote or denote any specific criteria or characteristic of patients who bear the label
• brands the patient as untreatable, defiant, or just “bad.”

Post-DSM-III: Where are we now?

From DSM-III evolved the hope that psychiatry could describe valid, well-defined diagnostic categories. Lost in the DSM-III enthusiasm was the fact that the categories were based upon theoretic constructs—theories no more or less valid than other theories that had preceded them. Because some of these categories were based upon empiric data— such as the Spitzer et al study—these diagnoses were perceived as more valid and more related to pathophysiology and perhaps genotype than prior constructs and definitions.

In the 1980s and early 1990s, a proliferation of studies attempted to examine the validity and reliability of DSM-III definitions, and BPD became the most studied of the personality disorders. The BPD concept took hold, even though several studies did not support it and despite refinements in subsequent DSM editions (Box 3).

One refinement in the BPD construct applied to transient psychotic or psychotic-like experiences, including dissociative phenomena. Yet questions remain about the duration of these transient episodes (Box 4).

Categorical versus dimensional

The categorical concept of BPD is facing new scrutiny,^26,27 as recent studies have implied that biological disturbances may be spread across a number of personality disorders.²⁸ If biological findings are found to be more closely allied with genotypic variations (alleles),²⁹ then perhaps a dimensional classification system is needed for BPD and personality disorders in general.

On the other hand, categories provide a well-defined population that we can study and try to delineate from other populations, whereas dimensions—while perhaps closer to the reality of clinical presentation—may allow too much variability for research to proceed without confounding restraints.

BPD will continue to evolve, as will all psychiatric diagnostic categories, but the need to modify its definition does not negate its usefulness and clinical applicability. Most of our patients do not read the DSM before coming to us. They present with symptom complexes and problems that demand that we listen to what they say and understand who they are while we also try to fit them—as best we can—into categories or dimensions³⁰ that help us choose the most appropriate interventions.

Related resources

• BPD Sanctuary (borderline personality disorder education, communities, support, books, and resources) http://www.mhsanctuary.com/borderline/ Borderline Central (resources for people who care about someone with borderline personality disorder) http://www.bpdcentral.com/
• Gunderson JG. Borderline personality disorder: a clinical guide. Washington, DC: American Psychiatric Publishing, Inc., 200l.
• Paris J (ed). Borderline personality disorder. Psych Clin N Am 2000;23:1 (entire volume devoted to BPD).
• Silk KR (ed). Biological and neurobehavioral studies of borderline personality disorder. Washington, DC: American Psychiatric Press, 1994.

Not everyone agrees that borderline personality disorder (BPD) should be a diagnostic category. BPD became “official” with DSM-III in 1980, although the term had been used for 40 years to describe various patient groups. Being listed in DSM-III legitimized BPD, which was thought to represent a specific—though not necessarily distinct—diagnostic category.

The history of the BPD diagnosis and opinions as to its usefulness can be viewed as a microcosm of psychiatric diagnosis in general. Before DSM-III, diagnoses were broadly defined and did not contain specific inclusion or exclusion criteria.¹ For the 5 to 10 years prior to DSM-III, however, two assumptions developed:

• distinct diagnostic categories did, in fact, exist
• by rigorously defining and studying those categories we could develop more specific and effective treatments for our patients.²

The specificity and exclusivity that we assumed we could achieve by categorical diagnoses, however, remain a distant wish. Comorbidity appears more common in psychiatry than was originally thought and confounds both treatment and outcome.³ Also, many patients appear treatment-resistant, despite fitting neatly into diagnostic categories.⁴

As psychiatry considers DSM-V, questions linger as to whether BPD (and personality disorder in general) should remain as a categorical diagnosis or if dimensional measures may be more appropriate. Dimensions imply that no one ever fits into a given box because no specific box exists. Rather, patients are described as being closer to or more distant from a prototypic model of the diagnosis. In personality disorders, the dimensions most often mentioned are cognition, impulsivity, emotional lability, environmental hyperreactivity, and anxiety. The case report (above) illustrates the interplay of these dimensions in a typical patient with presumed BPD.

What’s in a name?

The symptom complex or syndrome that bears the name borderline personality disorder has probably existed for as long as people have thought about patients in psychopathologic terms.⁵ Before 1980, the term “borderline” applied primarily to two separate but overlapping concepts:

• Patients thought to reside on the “border” with psychosis, such as the patient in our case example. They seemed to have an underlying psychotic disorder, but the psychosis—if it surfaced—appeared briefly, was not exceptionally deep or firmly held, and was not regularly evident or immediately accessible to the clinician.
• Patients who appeared to occupy the space between neurosis and psychosis. This concept evolved into the idea of a character or personality disorder distinguished primarily by unstable interpersonal relationships, a confused or inconsistent sense of identity, and emotional instability.

How DSM is changing. Comparing the disorders listed in DSM-IV (1994)⁶ versus DSM-II (1968)¹ suggests that psychiatry has become enamored of the naming process. For example, DSM-II lists anxiety neurosis (300.0), phobic neurosis (300.2), and obsessive-compulsive neurosis (300.3), whereas DSM-IV lists 11 different categories of anxiety disorders.

But beyond naming, subsequent DSMs have differed even more dramatically from DSM-II. We have seen a shift from describing a diagnostic category with a simple explanatory paragraph to lists of specific inclusion and exclusion criteria. These more-specific lists imply that they define categories closer to some reality or authenticity than did previous definitions.

Before DSM-III, the borderline concept was conceived in broad object relational and psychodynamic terms. In contrast, DSM-III produced a definitive set of criteria and required that a subset be met before the diagnosis could be made.⁷ An example of this criteria-based model is shown in Box 1, which lists the DSM-IV-TR criteria for BPD.

Some psychiatrists objected that BPD was solely a psychoanalytic construct and too theoretical for inclusion in DSM-III. Others argued that if BPD were not defined, it would be difficult to study the clinical usefulness of that definition or any other. Nonetheless, many have argued that BPD does not exist, though to what category BPD patients should belong has changed over the years:

• Is BPD nothing more than a milder or unusual presentation of an affective disorder⁸ or actually bipolar II disorder?⁹
• Is it a presentation of posttraumatic stress disorder (PTSD) called “complex PTSD,”^10-11 or an adult presentation of attention-deficit/ hyperactivity or other brain disorder?¹²
• Is it a stigmatizing diagnosis that we apply to patients whom we do not like?¹³

In truth, the diagnosis of BPD reflects a particular clinical presentation no more or less accurately than many of the well-accepted axis I disorders. Despite recent advances in the neurosciences, the dilemma we face as psychiatrists is that we make a diagnosis based upon what we see in the clinical setting (i.e., a phenotype). Yet in labeling what we believe is a specific psychiatric disorder, we make assumptions—for better or for worse, consciously or unconsciously—about pathophysiology and indirectly about genotype.

Box 1

Defining the borderline personality

Stern first used the term “borderline” in 1938 to describe patients who appeared to occupy the border between neurosis and psychosis.¹⁴ In 1942, Deutsch described the “as if” personality in patients who seemed chameleon-like. They could adapt or play the role demanded of them in specific situations, yet elsewhere—as in the analyst’s office—they had little sense of themselves and were thought to be internally disorganized and probably psychotic.¹⁵

Border to psychosis. The idea that borderline-type patients were psychotic continued in Hoch and Polatin’s description of the “pseudoneurotic schizophrenic,”¹⁶ a patient who appeared severely neurotic but was thought to employ many defenses and interpersonal styles to ward off a fundamental inner psychosis. Knight used the label “borderline states”¹⁷ to describe severely ill patients who were not frankly psychotic but fell within the realm of psychosis without qualifying for a diagnosis of schizophrenia. Knight was the first person to use the term “borderline” as a diagnostic entity, though simultaneously he argued against its use as a label because the term lacked precision.

Psychotic character. About the same time, Schmideberg characterized a group of patients whose emotional lability or affective reactivity seemed to be a consistent aspect of their clinical presentation. She believed this appearance of “stable instability”¹⁸ represented the patient’s characterologic adaptation to the world.

Frosch coined the term “psychotic character”¹⁹ that aptly captured both the characterologic and the border-to-psychosis aspects of these patients’ clinical picture. According to Frosch, these patients appeared to regress readily into psychotic thinking, yet they did not lose their ability to test reality.

Affective and emotional instability. Thus until the 1960s, the term borderline was applied primarily to patients who appeared to occupy the border between neurosis and psychosis but were thought to be closer to psychotic than neurotic. And this sitting close to the edge of psychosis appeared to be a stable condition.

Most of the attention up until this point had been paid to how these patients thought—with little attention to their affective lability or emotional instability, save for Schmideberg’s comments. In the 1960s, however, the term borderline was applied somewhat differently—not completely divorced from previous concepts but with greater emphasis on borderline as a stable but psychopathologic functioning of the personality that included affective and emotional instability and an impaired sense of self.

Box 2

Impaired personality organization. In 1967, Kernberg published a seminal article in the history of BPD diagnosis. Although he did not discuss the diagnosis of BPD, Kernberg did develop a concept concerning a specific organization of the personality based upon impaired object relations. This impaired organization could apply across several personality disorders. The construct, named borderline personality organization (BPO),²⁰ was defined by:

• an impaired sense of identity and lack of integration of one’s own identity
• use of primitive defenses, including splitting, rage, and regression
• ability to test reality.

Kernberg’s theory is too complex to summarize here, but he—along with Roy Grinker—is responsible for placing BPD on the diagnostic map. He was the first to describe BPO (and by extension BPD) in terms of a personality disorder.

Grinker’s ‘core’ group. Almost simultaneously (in 1968), Grinker published a careful study of 50 hospitalized patients. His work on the “borderline syndrome”²¹ revealed four subgroups to which the label of borderline had been applied:

• those occupying the border with psychosis
• those occupying the border with neurosis
• those similar to Deutsch’s “as if” group
• the “core” borderline group.

The core group—with its symptoms of anger and loneliness, a nonintegrated sense of self, and labile and oscillating interpersonal relationships— defined patients closest to our current definition.

Six criteria for BPD. In 1975, Gunderson and Singer published an article that greatly influenced our definition of BPD. They reviewed major descriptive accounts of BPD or BPD-like syndromes²² and proposed six diagnostic criteria (Box 2), though they did not identify a specific number or subset of the criteria as needing to be met for the diagnosis. (It is important to note that the term BPD did not become official for 5 more years.)

DSM-IV’s definition of BPD retains four of Gunderson and Singer‘s criteria among the nine it lists (five being necessary for a diagnosis of BPD). Missing are:

• social adaptiveness—though DSM-IV does say that social adaptiveness may be superficial (as in the “as if” personality) and may hide a disturbed identity⁶
• and the criterion relating to psychological test performance (this omission reflects a movement since 1980 away from listing “psychological” or psychodynamic criteria in the DSM).

DSM-III. BPD was included in DSM-III⁷ following an important study that tried to determine whether the term “borderline” refers to patients at the border of psychosis or to a stable group with mood instability and affective lability as part of a personality disorder. Spitzer et al²³ asked 808 clinicians to describe patients they would label as borderline and to use 22 items gleaned from the literature to score two of their own patients:

• one patient who the clinician felt truly had borderline personality, borderline personality organization, or borderline schizophrenia
• and a control patient who was not diagnosed as psychotic and did not fall into any borderline category.

Box 3

Their responses showed that BPD and schizotypal personality disorder (SPD) were separate, independent (though not mutually exclusive) disorders. Spitzer et al preserved the “schizotypal” label in DSM-III to describe the personality disorder that closely matched the border-to-psychosis subset. The other criteria set, which they labeled the “unstable personality item set,” was renamed “borderline” in DSM-III to describe the personality disorder that closely matched the emotional lability subset.

Box 4

Not everyone agrees with renaming the unstable set “borderline” because the word:

• has always been ambiguous
• does not connote or denote any specific criteria or characteristic of patients who bear the label
• brands the patient as untreatable, defiant, or just “bad.”

Post-DSM-III: Where are we now?

From DSM-III evolved the hope that psychiatry could describe valid, well-defined diagnostic categories. Lost in the DSM-III enthusiasm was the fact that the categories were based upon theoretic constructs—theories no more or less valid than other theories that had preceded them. Because some of these categories were based upon empiric data— such as the Spitzer et al study—these diagnoses were perceived as more valid and more related to pathophysiology and perhaps genotype than prior constructs and definitions.

In the 1980s and early 1990s, a proliferation of studies attempted to examine the validity and reliability of DSM-III definitions, and BPD became the most studied of the personality disorders. The BPD concept took hold, even though several studies did not support it and despite refinements in subsequent DSM editions (Box 3).

One refinement in the BPD construct applied to transient psychotic or psychotic-like experiences, including dissociative phenomena. Yet questions remain about the duration of these transient episodes (Box 4).

Categorical versus dimensional

The categorical concept of BPD is facing new scrutiny,^26,27 as recent studies have implied that biological disturbances may be spread across a number of personality disorders.²⁸ If biological findings are found to be more closely allied with genotypic variations (alleles),²⁹ then perhaps a dimensional classification system is needed for BPD and personality disorders in general.

On the other hand, categories provide a well-defined population that we can study and try to delineate from other populations, whereas dimensions—while perhaps closer to the reality of clinical presentation—may allow too much variability for research to proceed without confounding restraints.

BPD will continue to evolve, as will all psychiatric diagnostic categories, but the need to modify its definition does not negate its usefulness and clinical applicability. Most of our patients do not read the DSM before coming to us. They present with symptom complexes and problems that demand that we listen to what they say and understand who they are while we also try to fit them—as best we can—into categories or dimensions³⁰ that help us choose the most appropriate interventions.

Related resources

• BPD Sanctuary (borderline personality disorder education, communities, support, books, and resources) http://www.mhsanctuary.com/borderline/ Borderline Central (resources for people who care about someone with borderline personality disorder) http://www.bpdcentral.com/
• Gunderson JG. Borderline personality disorder: a clinical guide. Washington, DC: American Psychiatric Publishing, Inc., 200l.
• Paris J (ed). Borderline personality disorder. Psych Clin N Am 2000;23:1 (entire volume devoted to BPD).
• Silk KR (ed). Biological and neurobehavioral studies of borderline personality disorder. Washington, DC: American Psychiatric Press, 1994.

Not everyone agrees that borderline personality disorder (BPD) should be a diagnostic category. BPD became “official” with DSM-III in 1980, although the term had been used for 40 years to describe various patient groups. Being listed in DSM-III legitimized BPD, which was thought to represent a specific—though not necessarily distinct—diagnostic category.

The history of the BPD diagnosis and opinions as to its usefulness can be viewed as a microcosm of psychiatric diagnosis in general. Before DSM-III, diagnoses were broadly defined and did not contain specific inclusion or exclusion criteria.¹ For the 5 to 10 years prior to DSM-III, however, two assumptions developed:

• distinct diagnostic categories did, in fact, exist
• by rigorously defining and studying those categories we could develop more specific and effective treatments for our patients.²

The specificity and exclusivity that we assumed we could achieve by categorical diagnoses, however, remain a distant wish. Comorbidity appears more common in psychiatry than was originally thought and confounds both treatment and outcome.³ Also, many patients appear treatment-resistant, despite fitting neatly into diagnostic categories.⁴

As psychiatry considers DSM-V, questions linger as to whether BPD (and personality disorder in general) should remain as a categorical diagnosis or if dimensional measures may be more appropriate. Dimensions imply that no one ever fits into a given box because no specific box exists. Rather, patients are described as being closer to or more distant from a prototypic model of the diagnosis. In personality disorders, the dimensions most often mentioned are cognition, impulsivity, emotional lability, environmental hyperreactivity, and anxiety. The case report (above) illustrates the interplay of these dimensions in a typical patient with presumed BPD.

What’s in a name?

The symptom complex or syndrome that bears the name borderline personality disorder has probably existed for as long as people have thought about patients in psychopathologic terms.⁵ Before 1980, the term “borderline” applied primarily to two separate but overlapping concepts:

• Patients thought to reside on the “border” with psychosis, such as the patient in our case example. They seemed to have an underlying psychotic disorder, but the psychosis—if it surfaced—appeared briefly, was not exceptionally deep or firmly held, and was not regularly evident or immediately accessible to the clinician.
• Patients who appeared to occupy the space between neurosis and psychosis. This concept evolved into the idea of a character or personality disorder distinguished primarily by unstable interpersonal relationships, a confused or inconsistent sense of identity, and emotional instability.

How DSM is changing. Comparing the disorders listed in DSM-IV (1994)⁶ versus DSM-II (1968)¹ suggests that psychiatry has become enamored of the naming process. For example, DSM-II lists anxiety neurosis (300.0), phobic neurosis (300.2), and obsessive-compulsive neurosis (300.3), whereas DSM-IV lists 11 different categories of anxiety disorders.

But beyond naming, subsequent DSMs have differed even more dramatically from DSM-II. We have seen a shift from describing a diagnostic category with a simple explanatory paragraph to lists of specific inclusion and exclusion criteria. These more-specific lists imply that they define categories closer to some reality or authenticity than did previous definitions.

Before DSM-III, the borderline concept was conceived in broad object relational and psychodynamic terms. In contrast, DSM-III produced a definitive set of criteria and required that a subset be met before the diagnosis could be made.⁷ An example of this criteria-based model is shown in Box 1, which lists the DSM-IV-TR criteria for BPD.

Some psychiatrists objected that BPD was solely a psychoanalytic construct and too theoretical for inclusion in DSM-III. Others argued that if BPD were not defined, it would be difficult to study the clinical usefulness of that definition or any other. Nonetheless, many have argued that BPD does not exist, though to what category BPD patients should belong has changed over the years:

• Is BPD nothing more than a milder or unusual presentation of an affective disorder⁸ or actually bipolar II disorder?⁹
• Is it a presentation of posttraumatic stress disorder (PTSD) called “complex PTSD,”^10-11 or an adult presentation of attention-deficit/ hyperactivity or other brain disorder?¹²
• Is it a stigmatizing diagnosis that we apply to patients whom we do not like?¹³

In truth, the diagnosis of BPD reflects a particular clinical presentation no more or less accurately than many of the well-accepted axis I disorders. Despite recent advances in the neurosciences, the dilemma we face as psychiatrists is that we make a diagnosis based upon what we see in the clinical setting (i.e., a phenotype). Yet in labeling what we believe is a specific psychiatric disorder, we make assumptions—for better or for worse, consciously or unconsciously—about pathophysiology and indirectly about genotype.

Box 1

Defining the borderline personality

Stern first used the term “borderline” in 1938 to describe patients who appeared to occupy the border between neurosis and psychosis.¹⁴ In 1942, Deutsch described the “as if” personality in patients who seemed chameleon-like. They could adapt or play the role demanded of them in specific situations, yet elsewhere—as in the analyst’s office—they had little sense of themselves and were thought to be internally disorganized and probably psychotic.¹⁵

Border to psychosis. The idea that borderline-type patients were psychotic continued in Hoch and Polatin’s description of the “pseudoneurotic schizophrenic,”¹⁶ a patient who appeared severely neurotic but was thought to employ many defenses and interpersonal styles to ward off a fundamental inner psychosis. Knight used the label “borderline states”¹⁷ to describe severely ill patients who were not frankly psychotic but fell within the realm of psychosis without qualifying for a diagnosis of schizophrenia. Knight was the first person to use the term “borderline” as a diagnostic entity, though simultaneously he argued against its use as a label because the term lacked precision.

Psychotic character. About the same time, Schmideberg characterized a group of patients whose emotional lability or affective reactivity seemed to be a consistent aspect of their clinical presentation. She believed this appearance of “stable instability”¹⁸ represented the patient’s characterologic adaptation to the world.

Frosch coined the term “psychotic character”¹⁹ that aptly captured both the characterologic and the border-to-psychosis aspects of these patients’ clinical picture. According to Frosch, these patients appeared to regress readily into psychotic thinking, yet they did not lose their ability to test reality.

Affective and emotional instability. Thus until the 1960s, the term borderline was applied primarily to patients who appeared to occupy the border between neurosis and psychosis but were thought to be closer to psychotic than neurotic. And this sitting close to the edge of psychosis appeared to be a stable condition.

Most of the attention up until this point had been paid to how these patients thought—with little attention to their affective lability or emotional instability, save for Schmideberg’s comments. In the 1960s, however, the term borderline was applied somewhat differently—not completely divorced from previous concepts but with greater emphasis on borderline as a stable but psychopathologic functioning of the personality that included affective and emotional instability and an impaired sense of self.

Box 2

Impaired personality organization. In 1967, Kernberg published a seminal article in the history of BPD diagnosis. Although he did not discuss the diagnosis of BPD, Kernberg did develop a concept concerning a specific organization of the personality based upon impaired object relations. This impaired organization could apply across several personality disorders. The construct, named borderline personality organization (BPO),²⁰ was defined by:

• an impaired sense of identity and lack of integration of one’s own identity
• use of primitive defenses, including splitting, rage, and regression
• ability to test reality.

Kernberg’s theory is too complex to summarize here, but he—along with Roy Grinker—is responsible for placing BPD on the diagnostic map. He was the first to describe BPO (and by extension BPD) in terms of a personality disorder.

Grinker’s ‘core’ group. Almost simultaneously (in 1968), Grinker published a careful study of 50 hospitalized patients. His work on the “borderline syndrome”²¹ revealed four subgroups to which the label of borderline had been applied:

• those occupying the border with psychosis
• those occupying the border with neurosis
• those similar to Deutsch’s “as if” group
• the “core” borderline group.

The core group—with its symptoms of anger and loneliness, a nonintegrated sense of self, and labile and oscillating interpersonal relationships— defined patients closest to our current definition.

Six criteria for BPD. In 1975, Gunderson and Singer published an article that greatly influenced our definition of BPD. They reviewed major descriptive accounts of BPD or BPD-like syndromes²² and proposed six diagnostic criteria (Box 2), though they did not identify a specific number or subset of the criteria as needing to be met for the diagnosis. (It is important to note that the term BPD did not become official for 5 more years.)

DSM-IV’s definition of BPD retains four of Gunderson and Singer‘s criteria among the nine it lists (five being necessary for a diagnosis of BPD). Missing are:

• social adaptiveness—though DSM-IV does say that social adaptiveness may be superficial (as in the “as if” personality) and may hide a disturbed identity⁶
• and the criterion relating to psychological test performance (this omission reflects a movement since 1980 away from listing “psychological” or psychodynamic criteria in the DSM).

DSM-III. BPD was included in DSM-III⁷ following an important study that tried to determine whether the term “borderline” refers to patients at the border of psychosis or to a stable group with mood instability and affective lability as part of a personality disorder. Spitzer et al²³ asked 808 clinicians to describe patients they would label as borderline and to use 22 items gleaned from the literature to score two of their own patients:

• one patient who the clinician felt truly had borderline personality, borderline personality organization, or borderline schizophrenia
• and a control patient who was not diagnosed as psychotic and did not fall into any borderline category.

Box 3

Their responses showed that BPD and schizotypal personality disorder (SPD) were separate, independent (though not mutually exclusive) disorders. Spitzer et al preserved the “schizotypal” label in DSM-III to describe the personality disorder that closely matched the border-to-psychosis subset. The other criteria set, which they labeled the “unstable personality item set,” was renamed “borderline” in DSM-III to describe the personality disorder that closely matched the emotional lability subset.

Box 4

Not everyone agrees with renaming the unstable set “borderline” because the word:

• has always been ambiguous
• does not connote or denote any specific criteria or characteristic of patients who bear the label
• brands the patient as untreatable, defiant, or just “bad.”

Post-DSM-III: Where are we now?

From DSM-III evolved the hope that psychiatry could describe valid, well-defined diagnostic categories. Lost in the DSM-III enthusiasm was the fact that the categories were based upon theoretic constructs—theories no more or less valid than other theories that had preceded them. Because some of these categories were based upon empiric data— such as the Spitzer et al study—these diagnoses were perceived as more valid and more related to pathophysiology and perhaps genotype than prior constructs and definitions.

In the 1980s and early 1990s, a proliferation of studies attempted to examine the validity and reliability of DSM-III definitions, and BPD became the most studied of the personality disorders. The BPD concept took hold, even though several studies did not support it and despite refinements in subsequent DSM editions (Box 3).

One refinement in the BPD construct applied to transient psychotic or psychotic-like experiences, including dissociative phenomena. Yet questions remain about the duration of these transient episodes (Box 4).

Categorical versus dimensional

The categorical concept of BPD is facing new scrutiny,^26,27 as recent studies have implied that biological disturbances may be spread across a number of personality disorders.²⁸ If biological findings are found to be more closely allied with genotypic variations (alleles),²⁹ then perhaps a dimensional classification system is needed for BPD and personality disorders in general.

On the other hand, categories provide a well-defined population that we can study and try to delineate from other populations, whereas dimensions—while perhaps closer to the reality of clinical presentation—may allow too much variability for research to proceed without confounding restraints.

BPD will continue to evolve, as will all psychiatric diagnostic categories, but the need to modify its definition does not negate its usefulness and clinical applicability. Most of our patients do not read the DSM before coming to us. They present with symptom complexes and problems that demand that we listen to what they say and understand who they are while we also try to fit them—as best we can—into categories or dimensions³⁰ that help us choose the most appropriate interventions.

Related resources

• BPD Sanctuary (borderline personality disorder education, communities, support, books, and resources) http://www.mhsanctuary.com/borderline/ Borderline Central (resources for people who care about someone with borderline personality disorder) http://www.bpdcentral.com/
• Gunderson JG. Borderline personality disorder: a clinical guide. Washington, DC: American Psychiatric Publishing, Inc., 200l.
• Paris J (ed). Borderline personality disorder. Psych Clin N Am 2000;23:1 (entire volume devoted to BPD).
• Silk KR (ed). Biological and neurobehavioral studies of borderline personality disorder. Washington, DC: American Psychiatric Press, 1994.

Should you prescribe a stimulant to treat attention and hyperactivity problems in teenagers and adults with a history of substance abuse? Evidence suggests that using a stimulant to treat attention-deficit/hyperactivity disorder (ADHD) may place such patients at risk for stimulant abuse or for relapse into abuse of other substances. But a stimulant may be the only option for patients whose ADHD symptoms do not respond to alternate medications, such as antidepressants.

Growing numbers of adults are being treated for ADHD. Because substance abuse problems are common in adults with ADHD (Box 1 ),^1-3 prescribing an antidepressant instead of a stimulant in some cases may be prudent. Consider the following factors when choosing ADHD therapy for patients with a history of substance abuse.

Prevalence of stimulant use and abuse

In the United States, more than 95% of medications prescribed for children and adults with ADHD are stimulants—usually methylphenidate.⁴ Stimulant use has increased as more children and adults are diagnosed with ADHD. Methylphenidate prescriptions increased five-fold from 1990 to 1995.⁵ Visits to psychiatrists and physicians that included stimulant prescriptions grew from 570,000 to 2.86 million from 1985 to 1994, with most of that increase occurring during visits to primary care and other physicians.⁶

When used as prescribed, methylphenidate is safe and effective for treating most children and adults with ADHD. Methylphenidate’s pharmacologic properties, however, are similar to those of amphetamines and cocaine (Box 2, Figure 1),^7,8 which is why methylphenidate is a schedule-II controlled substance.

Published data. Fifteen reports of methylphenidate abuse were published in the medical literature between 1960 and 1999,⁷ but little is known about the prevalence of stimulant abuse among patients with ADHD. Banov and colleagues recently published what may be the only data available, when they reported that 3 of 37 (8%) patients abused the stimulants they were prescribed for ADHD.⁹ The three patients who abused stimulants had histories of drug and alcohol abuse at study entry. In all three cases, stimulant abuse did not develop immediately but became apparent within 6 months after the study began.

In a study of 651 students ages 11 to 18 in Wisconsin and Minnesota, more than one-third of those taking stimulants reported being asked to sell or trade their medications. More than one-half of those not taking ADHD medications said they knew someone who sold or gave away his or her medication.¹⁰

Stimulant theft, recreational use. Methylphenidate has been identified as the third most abused prescribed substance in the United States.¹¹ It was the 10th most frequently stolen controlled drug from pharmacies between 1990 and 1995, and 700,000 dosage units were reported stolen in 1996 and 1997.¹²

Box 1

The popular media have reported recreational use of methylphenidate—with street names such as “R-Ball” and “Vitamin R”—among teens and college students.¹³ Illegal stimulants are perceived to be easily accessible on college campuses, but no data have been reported.

The use of stimulant medication for ADHD patients with substance abuse problems remains controversial. For such patients, this author reserves stimulant medication for those:

• whose ADHD symptoms have not responded adequately to alternate treatments
• who have been reliable with prescription medications
• and whose functional level is seriously impaired by their ADHD.

Antidepressants vs. stimulants

Although few well-designed controlled studies have been published, four antidepressants appear to be reasonably equivalent in effectiveness for adults with ADHD and do not carry potential for stimulant abuse.¹⁴

Desipraime, bupropion, venlafaxine, and the experimental drug atomoxetine (Table 1) all increase norepinephrine at the synapse by inhibiting presynaptic reuptake. Though dopamine has traditionally been considered the neurotransmitter of choice for ADHD treatment, norepinephrine may be equally potent.

Impulse control center. Several lines of research have recently established a connection between the prefrontal cortex, norepinephrine, and ADHD.¹⁵ This evidence suggests that the prefrontal cortex plays a major role in inhibiting impulses and responses to distractions:

Box 2

HOW STIMULANTS—THERAPEUTIC OR ILLICIT—AFFECT THE BRAIN

Figure 1

PET scans of the brain using carbon 11 (¹¹C)-labeled cocaine and methylphenidate HCl show similar distributions in the striatium when the drugs are administered intravenously.

Source: Reproduced with permission from Volkow ND, Ding YS, Fowler JS, et al. Is methylphenidate like cocaine? Studies on their pharmacokinetics and distribution in the human brain. Arch Gen Psychiatry 1995;52:456-63.

Stimulants are classified as schedule-II drugs because they produce powerful reinforcing effects by increasing synaptic dopamine.⁷ Positron-emission tomography (PET) scans using carbon labeling have shown similar distributions of methylphenidate and cocaine in the brain (Figure 1).⁸ When administered intravenously, both drugs occupy the same receptors in the striatum and produce a “high” that parallels rapid neuronal uptake. Methylphenidate and cocaine similarly increase stimulation of the postsynaptic neuron by blocking the dopamine reuptake pump (dopamine transporter).

How a substance gets to the brain’s euphoric receptors greatly affects its addictive properties. Delivery systems with rapid onset—smoking, “snorting,” or IV injection—have much greater ability to produce a “high” than do oral or transdermal routes. The greater the “high,” the greater the potential for abuse.

Because methylphenidate is prescribed for oral use, the potential for abuse is minimal. However, we need to be extremely cautious when giving methylphenidate or similar stimulant medications to patients who have shown they are unable to control their abuse of other substances.

Stimulants can also re-ignite a dormant substance abuse problem. Though little has been written about this in the medical literature, Elizabeth Wurtzel, author of the controversial Prozac Nation, chronicles the resumption of her cocaine abuse in More, Now, Again: A Memoir of Addiction. She contends that after her doctor added methylphenidate to augment treatment of partially remitting depression, she began abusing it and eventually was using 40 tablets per day before slipping back into cocaine dependence.

• Patients with prefrontal cortex deficits can have problems with inattention and poor impulse control.
• Patients with ADHD have frontal lobe impairments, as neuropsychological testing and imaging studies have shown.
• Norepinephrine neurons, with cell bodies in the locus coeruleus, have projections that terminate in the prefrontal cortex.
• Agents with norepinephrine activity, but without mood-altering properties (e.g., clonidine), have been shown to improve ADHD symptoms.

Table 1

ADHD IN ADULTS: ANTIDEPRESSANT DOSAGES AND SIDE EFFECTS

Additional evidence suggests that the prefrontal cortex has projections back to the locus coeruleus, which may explain the relationship between the two areas. It may be that the brain’s higher-functioning areas, such as the prefrontal cortex, provide intelligent screening of impulses from the brain’s older areas, such as the locus coeruleus. Therefore, increased prefrontal cortex activity may modulate some impulses that ADHD patients cannot control otherwise.

No ‘high’ with antidepressants. Patients with ADHD who have experienced the powerful effects of street drugs such as cocaine, methamphetamine, or even alcohol may report that antidepressants do not provide the effect they desire. It is difficult to know if these patients are reporting a lack of benefit or simply the absence of a euphoric “high” they are used to experiencing with substances of abuse. The newer antidepressants do not activate the brain’s euphoric receptors to an appreciable degree.

Patients who take stimulants as prescribed also do not report a “high” but can detect the medication’s presence and absence. Most do not crave this feeling, but substance abusers tend to like it. A patient recently told me he didn’t think stimulants improved his ADHD, but said, “I just liked the way they made me feel.”

Desipramine

The tricyclic antidepressant desipramine is a potent norepinephrine reuptake inhibitor that is effective in treating ADHD in children and adults. In a double-blind, placebo-controlled study of adults with ADHD, subjects receiving desipramine showed robust improvement in symptom scores on the ADHD Rating Scale,¹⁶ compared with those receiving the placebo (Figure 2).¹⁷

During the 6-week trial, 41 adults with ADHD received desipramine, 200 mg/d, or a placebo. Those receiving desipramine showed significant improvement in 12 of 14 ADHD symptoms and less hyperactivity, impulsivity, and inattentiveness, whereas those receiving the placebo showed no improvement. According to the study criteria, 68% of those who received desipramine and none who received the placebo were considered positive responders.

Though no head-to-head studies have compared desipramine with methylphenidate, the same researchers conducted a similar placebo-controlled study with methylphenidate. The 6-week symptom score on the ADHD Rating Scale was 12.5 for methylphenidate, compared with a score of 12 for desipramine.¹⁸

Recommendation. Desipramine may be the most effective of the antidepressant treatments for patients with ADHD. Because of its side effects, however, it is not this author’s first choice and is usually reserved for patients whose symptoms fail to respond to other antidepressants. Desipramine can cause sedation, dry mouth, and constipation, which are related to blockade of adrenergic, histamine, and muscarinic cholinergic receptors. It also can be lethal in overdose.

Some substance abusers lose confidence in a medication that they cannot feel working. The side effects of desipramine, which can be intolerable for some patients, can reassure others with a history of substance abuse that they are being medicated.

Bupropion

Bupropion is a unique antidepressant that inhibits the presynaptic reuptake of dopamine and norepinephrine. Open-label studies demonstrate good responses to bupropion by adults with ADHD. One placebo-controlled, double-blind study found improved ADHD symptoms in 76% of patients receiving bupropion SR, compared with 37% of those receiving a placebo; the difference was statistically significant.¹⁹

Figure 2 IMPROVED ADHD SYMPTOMS WITH DESIPRAMINE

Adults with ADHD who received desipramine, 200 mg/d, in a double-blind trial showed significantly less hyperactivity, impulsivity, and inattentiveness after 6 weeks of therapy than a control group that received a placebo.

Source: Adapted with permission from Wilens TE, Biederman J, Prince J, et al. Six-week, double-blind, placebo-controlled study of desipramine for adult attention-deficit/hyperactivity disorder. Am J Psychiatry 1996;153:1147-53.Another study of adults with ADHD compared bupropion SR to methylphenidate and a placebo.²⁰ Using a primary outcome of the Clinical Global Impression (CGI) scale, response rates were 64% for bupropion SR, 50% for methylphenidate, and 27% for the placebo. The difference in response rates between the two agents was not statistically significant (p = 0.14).

Recommendation. The risk of seizures with bupropion is about 1 in 1,000. Therefore, bupropion should not be given to patients with a seizure disorder or to those with conditions that alter the seizure threshold (e.g., eating disorders, recent head trauma, or benzodiazepine withdrawal).²¹

This author uses bupropion as first-line treatment for appropriate patients with ADHD and a substance abuse history. Bupropion’s mild benefit with smoking cessation may provide some crossover effect for other substances of abuse. The low incidence of sexual side effects is another benefit. Drawbacks include twice-daily dosing and lack of a robust effect on attention and concentration.

Venlafaxine

Venlafaxine is a potent inhibitor of serotonin reuptake, a moderate inhibitor of norepinephrine, and a mild inhibitor of dopamine. Venlafaxine has displayed response rates similar to those of desipramine and bupropion in open-label studies in adults with ADHD,¹⁴ but no placebo-controlled studies exist.

As noted above, these antidepressants are believed to improve ADHD symptoms by making norepinephrine more available at the synapse. Hypothetically, then, one would need to administer venlafaxine at dosages that adequately inhibit the norepinephrine reuptake receptor. Venlafaxine XR, 150 mg/d, provides significant norepinephrine activity, according to several lines of evidence.

Recently, Upadhyaya et al reported the use of venlafaxine in one of the few treatment studies of patients with ADHD and comorbid alcohol/cocaine abuse. In an open-label trial, 10 subjects received venlafaxine, up to 300 mg/d, along with psychotherapy and attendance at Alcoholics Anonymous meetings. The nine who completed 4 weeks of treatment showed significantly improved ADHD symptoms and decreased alcohol craving.²²

Recommendation. Venlafaxine is this author’s second choice for patients with ADHD and substance abuse problems. Sexual side effects that some patients experience with venlafaxine can limit its use. Some clinicians are concerned about increases in blood pressure associated with venlafaxine, although significant changes do not seem to occur at dosages below 300 mg/d.²³

Atomoxetine

Atomoxetine is an investigational antidepressant in phase-III trials as a treatment for ADHD. Evidence shows atomoxetine to be a potent inhibitor of the presynaptic norepinephrine transporter, with minimal affinity for other neurotransmitter receptors. Initial studies suggest that atomoxetine is effective for adults and children with ADHD:

• In a small, double-blind, placebo-controlled, crossover trial, 11 of 20 adults showed improvement in ADHD symptoms within 3 weeks of starting atomoxetine.²⁴
• In 297 children and adolescents, atomoxetine at dosages averaging approximately 1.2 mg/kg/d was more effective than a placebo in reducing ADHD symptoms and improving social and family functioning. Treatment was well tolerated and without significant side effects.²⁵
• In a randomized open-label trial, 228 children received atomoxetine or methylphenidate. Both treatments significantly reduced inattention and hyperactive/impulsive symptoms.²⁶

Related resources

• Arnsten AF. Genetics of childhood disorders (XVIII). ADHD, Part 2: Norepinephrine has a critical modulatory influence on prefrontal cortical function. J Am Acad Child Adolesc Psychiatry. 2000;39:1201-3.
• Ward MF, Wender PH, Reimherr FW. The Wender Utah Rating Scale: An aid in the retrospective diagnosis of childhood attention-deficit/hyperactivity disorder. Am J Psychiatry. 1993;150:885-90.
• Research Report: Prescription Drugs—abuse and addiction. National Institute on Drug Abuse, Substance Abuse and Mental Health Services Administration.

Drug brand names

• Bupropion • Wellbutrin
• Desipramine • Norpramin
• Methylphenidate • Ritalin, Concerta
• Venlafaxine • Effexor

Disclosure

Dr. Higgins reports that he is on the speakers’ bureaus for Wyeth-Ayerst Pharmaceuticals and Eli Lilly and Co.

Should you prescribe a stimulant to treat attention and hyperactivity problems in teenagers and adults with a history of substance abuse? Evidence suggests that using a stimulant to treat attention-deficit/hyperactivity disorder (ADHD) may place such patients at risk for stimulant abuse or for relapse into abuse of other substances. But a stimulant may be the only option for patients whose ADHD symptoms do not respond to alternate medications, such as antidepressants.

Growing numbers of adults are being treated for ADHD. Because substance abuse problems are common in adults with ADHD (Box 1 ),^1-3 prescribing an antidepressant instead of a stimulant in some cases may be prudent. Consider the following factors when choosing ADHD therapy for patients with a history of substance abuse.

Prevalence of stimulant use and abuse

In the United States, more than 95% of medications prescribed for children and adults with ADHD are stimulants—usually methylphenidate.⁴ Stimulant use has increased as more children and adults are diagnosed with ADHD. Methylphenidate prescriptions increased five-fold from 1990 to 1995.⁵ Visits to psychiatrists and physicians that included stimulant prescriptions grew from 570,000 to 2.86 million from 1985 to 1994, with most of that increase occurring during visits to primary care and other physicians.⁶

When used as prescribed, methylphenidate is safe and effective for treating most children and adults with ADHD. Methylphenidate’s pharmacologic properties, however, are similar to those of amphetamines and cocaine (Box 2, Figure 1),^7,8 which is why methylphenidate is a schedule-II controlled substance.

Published data. Fifteen reports of methylphenidate abuse were published in the medical literature between 1960 and 1999,⁷ but little is known about the prevalence of stimulant abuse among patients with ADHD. Banov and colleagues recently published what may be the only data available, when they reported that 3 of 37 (8%) patients abused the stimulants they were prescribed for ADHD.⁹ The three patients who abused stimulants had histories of drug and alcohol abuse at study entry. In all three cases, stimulant abuse did not develop immediately but became apparent within 6 months after the study began.

In a study of 651 students ages 11 to 18 in Wisconsin and Minnesota, more than one-third of those taking stimulants reported being asked to sell or trade their medications. More than one-half of those not taking ADHD medications said they knew someone who sold or gave away his or her medication.¹⁰

Stimulant theft, recreational use. Methylphenidate has been identified as the third most abused prescribed substance in the United States.¹¹ It was the 10th most frequently stolen controlled drug from pharmacies between 1990 and 1995, and 700,000 dosage units were reported stolen in 1996 and 1997.¹²

Box 1

The popular media have reported recreational use of methylphenidate—with street names such as “R-Ball” and “Vitamin R”—among teens and college students.¹³ Illegal stimulants are perceived to be easily accessible on college campuses, but no data have been reported.

The use of stimulant medication for ADHD patients with substance abuse problems remains controversial. For such patients, this author reserves stimulant medication for those:

• whose ADHD symptoms have not responded adequately to alternate treatments
• who have been reliable with prescription medications
• and whose functional level is seriously impaired by their ADHD.

Antidepressants vs. stimulants

Although few well-designed controlled studies have been published, four antidepressants appear to be reasonably equivalent in effectiveness for adults with ADHD and do not carry potential for stimulant abuse.¹⁴

Desipraime, bupropion, venlafaxine, and the experimental drug atomoxetine (Table 1) all increase norepinephrine at the synapse by inhibiting presynaptic reuptake. Though dopamine has traditionally been considered the neurotransmitter of choice for ADHD treatment, norepinephrine may be equally potent.

Impulse control center. Several lines of research have recently established a connection between the prefrontal cortex, norepinephrine, and ADHD.¹⁵ This evidence suggests that the prefrontal cortex plays a major role in inhibiting impulses and responses to distractions:

Box 2

HOW STIMULANTS—THERAPEUTIC OR ILLICIT—AFFECT THE BRAIN

Figure 1

PET scans of the brain using carbon 11 (¹¹C)-labeled cocaine and methylphenidate HCl show similar distributions in the striatium when the drugs are administered intravenously.

Source: Reproduced with permission from Volkow ND, Ding YS, Fowler JS, et al. Is methylphenidate like cocaine? Studies on their pharmacokinetics and distribution in the human brain. Arch Gen Psychiatry 1995;52:456-63.

Stimulants are classified as schedule-II drugs because they produce powerful reinforcing effects by increasing synaptic dopamine.⁷ Positron-emission tomography (PET) scans using carbon labeling have shown similar distributions of methylphenidate and cocaine in the brain (Figure 1).⁸ When administered intravenously, both drugs occupy the same receptors in the striatum and produce a “high” that parallels rapid neuronal uptake. Methylphenidate and cocaine similarly increase stimulation of the postsynaptic neuron by blocking the dopamine reuptake pump (dopamine transporter).

How a substance gets to the brain’s euphoric receptors greatly affects its addictive properties. Delivery systems with rapid onset—smoking, “snorting,” or IV injection—have much greater ability to produce a “high” than do oral or transdermal routes. The greater the “high,” the greater the potential for abuse.

Because methylphenidate is prescribed for oral use, the potential for abuse is minimal. However, we need to be extremely cautious when giving methylphenidate or similar stimulant medications to patients who have shown they are unable to control their abuse of other substances.

Stimulants can also re-ignite a dormant substance abuse problem. Though little has been written about this in the medical literature, Elizabeth Wurtzel, author of the controversial Prozac Nation, chronicles the resumption of her cocaine abuse in More, Now, Again: A Memoir of Addiction. She contends that after her doctor added methylphenidate to augment treatment of partially remitting depression, she began abusing it and eventually was using 40 tablets per day before slipping back into cocaine dependence.

• Patients with prefrontal cortex deficits can have problems with inattention and poor impulse control.
• Patients with ADHD have frontal lobe impairments, as neuropsychological testing and imaging studies have shown.
• Norepinephrine neurons, with cell bodies in the locus coeruleus, have projections that terminate in the prefrontal cortex.
• Agents with norepinephrine activity, but without mood-altering properties (e.g., clonidine), have been shown to improve ADHD symptoms.

Table 1

ADHD IN ADULTS: ANTIDEPRESSANT DOSAGES AND SIDE EFFECTS

Additional evidence suggests that the prefrontal cortex has projections back to the locus coeruleus, which may explain the relationship between the two areas. It may be that the brain’s higher-functioning areas, such as the prefrontal cortex, provide intelligent screening of impulses from the brain’s older areas, such as the locus coeruleus. Therefore, increased prefrontal cortex activity may modulate some impulses that ADHD patients cannot control otherwise.

No ‘high’ with antidepressants. Patients with ADHD who have experienced the powerful effects of street drugs such as cocaine, methamphetamine, or even alcohol may report that antidepressants do not provide the effect they desire. It is difficult to know if these patients are reporting a lack of benefit or simply the absence of a euphoric “high” they are used to experiencing with substances of abuse. The newer antidepressants do not activate the brain’s euphoric receptors to an appreciable degree.

Patients who take stimulants as prescribed also do not report a “high” but can detect the medication’s presence and absence. Most do not crave this feeling, but substance abusers tend to like it. A patient recently told me he didn’t think stimulants improved his ADHD, but said, “I just liked the way they made me feel.”

Desipramine

The tricyclic antidepressant desipramine is a potent norepinephrine reuptake inhibitor that is effective in treating ADHD in children and adults. In a double-blind, placebo-controlled study of adults with ADHD, subjects receiving desipramine showed robust improvement in symptom scores on the ADHD Rating Scale,¹⁶ compared with those receiving the placebo (Figure 2).¹⁷

During the 6-week trial, 41 adults with ADHD received desipramine, 200 mg/d, or a placebo. Those receiving desipramine showed significant improvement in 12 of 14 ADHD symptoms and less hyperactivity, impulsivity, and inattentiveness, whereas those receiving the placebo showed no improvement. According to the study criteria, 68% of those who received desipramine and none who received the placebo were considered positive responders.

Though no head-to-head studies have compared desipramine with methylphenidate, the same researchers conducted a similar placebo-controlled study with methylphenidate. The 6-week symptom score on the ADHD Rating Scale was 12.5 for methylphenidate, compared with a score of 12 for desipramine.¹⁸

Recommendation. Desipramine may be the most effective of the antidepressant treatments for patients with ADHD. Because of its side effects, however, it is not this author’s first choice and is usually reserved for patients whose symptoms fail to respond to other antidepressants. Desipramine can cause sedation, dry mouth, and constipation, which are related to blockade of adrenergic, histamine, and muscarinic cholinergic receptors. It also can be lethal in overdose.

Some substance abusers lose confidence in a medication that they cannot feel working. The side effects of desipramine, which can be intolerable for some patients, can reassure others with a history of substance abuse that they are being medicated.

Bupropion

Bupropion is a unique antidepressant that inhibits the presynaptic reuptake of dopamine and norepinephrine. Open-label studies demonstrate good responses to bupropion by adults with ADHD. One placebo-controlled, double-blind study found improved ADHD symptoms in 76% of patients receiving bupropion SR, compared with 37% of those receiving a placebo; the difference was statistically significant.¹⁹

Figure 2 IMPROVED ADHD SYMPTOMS WITH DESIPRAMINE

Adults with ADHD who received desipramine, 200 mg/d, in a double-blind trial showed significantly less hyperactivity, impulsivity, and inattentiveness after 6 weeks of therapy than a control group that received a placebo.

Source: Adapted with permission from Wilens TE, Biederman J, Prince J, et al. Six-week, double-blind, placebo-controlled study of desipramine for adult attention-deficit/hyperactivity disorder. Am J Psychiatry 1996;153:1147-53.Another study of adults with ADHD compared bupropion SR to methylphenidate and a placebo.²⁰ Using a primary outcome of the Clinical Global Impression (CGI) scale, response rates were 64% for bupropion SR, 50% for methylphenidate, and 27% for the placebo. The difference in response rates between the two agents was not statistically significant (p = 0.14).

Recommendation. The risk of seizures with bupropion is about 1 in 1,000. Therefore, bupropion should not be given to patients with a seizure disorder or to those with conditions that alter the seizure threshold (e.g., eating disorders, recent head trauma, or benzodiazepine withdrawal).²¹

This author uses bupropion as first-line treatment for appropriate patients with ADHD and a substance abuse history. Bupropion’s mild benefit with smoking cessation may provide some crossover effect for other substances of abuse. The low incidence of sexual side effects is another benefit. Drawbacks include twice-daily dosing and lack of a robust effect on attention and concentration.

Venlafaxine

Venlafaxine is a potent inhibitor of serotonin reuptake, a moderate inhibitor of norepinephrine, and a mild inhibitor of dopamine. Venlafaxine has displayed response rates similar to those of desipramine and bupropion in open-label studies in adults with ADHD,¹⁴ but no placebo-controlled studies exist.

As noted above, these antidepressants are believed to improve ADHD symptoms by making norepinephrine more available at the synapse. Hypothetically, then, one would need to administer venlafaxine at dosages that adequately inhibit the norepinephrine reuptake receptor. Venlafaxine XR, 150 mg/d, provides significant norepinephrine activity, according to several lines of evidence.

Recently, Upadhyaya et al reported the use of venlafaxine in one of the few treatment studies of patients with ADHD and comorbid alcohol/cocaine abuse. In an open-label trial, 10 subjects received venlafaxine, up to 300 mg/d, along with psychotherapy and attendance at Alcoholics Anonymous meetings. The nine who completed 4 weeks of treatment showed significantly improved ADHD symptoms and decreased alcohol craving.²²

Recommendation. Venlafaxine is this author’s second choice for patients with ADHD and substance abuse problems. Sexual side effects that some patients experience with venlafaxine can limit its use. Some clinicians are concerned about increases in blood pressure associated with venlafaxine, although significant changes do not seem to occur at dosages below 300 mg/d.²³

Atomoxetine

Atomoxetine is an investigational antidepressant in phase-III trials as a treatment for ADHD. Evidence shows atomoxetine to be a potent inhibitor of the presynaptic norepinephrine transporter, with minimal affinity for other neurotransmitter receptors. Initial studies suggest that atomoxetine is effective for adults and children with ADHD:

• In a small, double-blind, placebo-controlled, crossover trial, 11 of 20 adults showed improvement in ADHD symptoms within 3 weeks of starting atomoxetine.²⁴
• In 297 children and adolescents, atomoxetine at dosages averaging approximately 1.2 mg/kg/d was more effective than a placebo in reducing ADHD symptoms and improving social and family functioning. Treatment was well tolerated and without significant side effects.²⁵
• In a randomized open-label trial, 228 children received atomoxetine or methylphenidate. Both treatments significantly reduced inattention and hyperactive/impulsive symptoms.²⁶

Related resources

• Arnsten AF. Genetics of childhood disorders (XVIII). ADHD, Part 2: Norepinephrine has a critical modulatory influence on prefrontal cortical function. J Am Acad Child Adolesc Psychiatry. 2000;39:1201-3.
• Ward MF, Wender PH, Reimherr FW. The Wender Utah Rating Scale: An aid in the retrospective diagnosis of childhood attention-deficit/hyperactivity disorder. Am J Psychiatry. 1993;150:885-90.
• Research Report: Prescription Drugs—abuse and addiction. National Institute on Drug Abuse, Substance Abuse and Mental Health Services Administration.

Drug brand names

• Bupropion • Wellbutrin
• Desipramine • Norpramin
• Methylphenidate • Ritalin, Concerta
• Venlafaxine • Effexor

Disclosure

Dr. Higgins reports that he is on the speakers’ bureaus for Wyeth-Ayerst Pharmaceuticals and Eli Lilly and Co.

Should you prescribe a stimulant to treat attention and hyperactivity problems in teenagers and adults with a history of substance abuse? Evidence suggests that using a stimulant to treat attention-deficit/hyperactivity disorder (ADHD) may place such patients at risk for stimulant abuse or for relapse into abuse of other substances. But a stimulant may be the only option for patients whose ADHD symptoms do not respond to alternate medications, such as antidepressants.

Growing numbers of adults are being treated for ADHD. Because substance abuse problems are common in adults with ADHD (Box 1 ),^1-3 prescribing an antidepressant instead of a stimulant in some cases may be prudent. Consider the following factors when choosing ADHD therapy for patients with a history of substance abuse.

Prevalence of stimulant use and abuse

In the United States, more than 95% of medications prescribed for children and adults with ADHD are stimulants—usually methylphenidate.⁴ Stimulant use has increased as more children and adults are diagnosed with ADHD. Methylphenidate prescriptions increased five-fold from 1990 to 1995.⁵ Visits to psychiatrists and physicians that included stimulant prescriptions grew from 570,000 to 2.86 million from 1985 to 1994, with most of that increase occurring during visits to primary care and other physicians.⁶

When used as prescribed, methylphenidate is safe and effective for treating most children and adults with ADHD. Methylphenidate’s pharmacologic properties, however, are similar to those of amphetamines and cocaine (Box 2, Figure 1),^7,8 which is why methylphenidate is a schedule-II controlled substance.

Published data. Fifteen reports of methylphenidate abuse were published in the medical literature between 1960 and 1999,⁷ but little is known about the prevalence of stimulant abuse among patients with ADHD. Banov and colleagues recently published what may be the only data available, when they reported that 3 of 37 (8%) patients abused the stimulants they were prescribed for ADHD.⁹ The three patients who abused stimulants had histories of drug and alcohol abuse at study entry. In all three cases, stimulant abuse did not develop immediately but became apparent within 6 months after the study began.

In a study of 651 students ages 11 to 18 in Wisconsin and Minnesota, more than one-third of those taking stimulants reported being asked to sell or trade their medications. More than one-half of those not taking ADHD medications said they knew someone who sold or gave away his or her medication.¹⁰

Stimulant theft, recreational use. Methylphenidate has been identified as the third most abused prescribed substance in the United States.¹¹ It was the 10th most frequently stolen controlled drug from pharmacies between 1990 and 1995, and 700,000 dosage units were reported stolen in 1996 and 1997.¹²

Box 1

The popular media have reported recreational use of methylphenidate—with street names such as “R-Ball” and “Vitamin R”—among teens and college students.¹³ Illegal stimulants are perceived to be easily accessible on college campuses, but no data have been reported.

The use of stimulant medication for ADHD patients with substance abuse problems remains controversial. For such patients, this author reserves stimulant medication for those:

• whose ADHD symptoms have not responded adequately to alternate treatments
• who have been reliable with prescription medications
• and whose functional level is seriously impaired by their ADHD.

Antidepressants vs. stimulants

Although few well-designed controlled studies have been published, four antidepressants appear to be reasonably equivalent in effectiveness for adults with ADHD and do not carry potential for stimulant abuse.¹⁴

Desipraime, bupropion, venlafaxine, and the experimental drug atomoxetine (Table 1) all increase norepinephrine at the synapse by inhibiting presynaptic reuptake. Though dopamine has traditionally been considered the neurotransmitter of choice for ADHD treatment, norepinephrine may be equally potent.

Impulse control center. Several lines of research have recently established a connection between the prefrontal cortex, norepinephrine, and ADHD.¹⁵ This evidence suggests that the prefrontal cortex plays a major role in inhibiting impulses and responses to distractions:

Box 2

HOW STIMULANTS—THERAPEUTIC OR ILLICIT—AFFECT THE BRAIN

Figure 1

PET scans of the brain using carbon 11 (¹¹C)-labeled cocaine and methylphenidate HCl show similar distributions in the striatium when the drugs are administered intravenously.

Source: Reproduced with permission from Volkow ND, Ding YS, Fowler JS, et al. Is methylphenidate like cocaine? Studies on their pharmacokinetics and distribution in the human brain. Arch Gen Psychiatry 1995;52:456-63.

Stimulants are classified as schedule-II drugs because they produce powerful reinforcing effects by increasing synaptic dopamine.⁷ Positron-emission tomography (PET) scans using carbon labeling have shown similar distributions of methylphenidate and cocaine in the brain (Figure 1).⁸ When administered intravenously, both drugs occupy the same receptors in the striatum and produce a “high” that parallels rapid neuronal uptake. Methylphenidate and cocaine similarly increase stimulation of the postsynaptic neuron by blocking the dopamine reuptake pump (dopamine transporter).

How a substance gets to the brain’s euphoric receptors greatly affects its addictive properties. Delivery systems with rapid onset—smoking, “snorting,” or IV injection—have much greater ability to produce a “high” than do oral or transdermal routes. The greater the “high,” the greater the potential for abuse.

Because methylphenidate is prescribed for oral use, the potential for abuse is minimal. However, we need to be extremely cautious when giving methylphenidate or similar stimulant medications to patients who have shown they are unable to control their abuse of other substances.

Stimulants can also re-ignite a dormant substance abuse problem. Though little has been written about this in the medical literature, Elizabeth Wurtzel, author of the controversial Prozac Nation, chronicles the resumption of her cocaine abuse in More, Now, Again: A Memoir of Addiction. She contends that after her doctor added methylphenidate to augment treatment of partially remitting depression, she began abusing it and eventually was using 40 tablets per day before slipping back into cocaine dependence.

• Patients with prefrontal cortex deficits can have problems with inattention and poor impulse control.
• Patients with ADHD have frontal lobe impairments, as neuropsychological testing and imaging studies have shown.
• Norepinephrine neurons, with cell bodies in the locus coeruleus, have projections that terminate in the prefrontal cortex.
• Agents with norepinephrine activity, but without mood-altering properties (e.g., clonidine), have been shown to improve ADHD symptoms.

Table 1

ADHD IN ADULTS: ANTIDEPRESSANT DOSAGES AND SIDE EFFECTS

Additional evidence suggests that the prefrontal cortex has projections back to the locus coeruleus, which may explain the relationship between the two areas. It may be that the brain’s higher-functioning areas, such as the prefrontal cortex, provide intelligent screening of impulses from the brain’s older areas, such as the locus coeruleus. Therefore, increased prefrontal cortex activity may modulate some impulses that ADHD patients cannot control otherwise.

No ‘high’ with antidepressants. Patients with ADHD who have experienced the powerful effects of street drugs such as cocaine, methamphetamine, or even alcohol may report that antidepressants do not provide the effect they desire. It is difficult to know if these patients are reporting a lack of benefit or simply the absence of a euphoric “high” they are used to experiencing with substances of abuse. The newer antidepressants do not activate the brain’s euphoric receptors to an appreciable degree.

Patients who take stimulants as prescribed also do not report a “high” but can detect the medication’s presence and absence. Most do not crave this feeling, but substance abusers tend to like it. A patient recently told me he didn’t think stimulants improved his ADHD, but said, “I just liked the way they made me feel.”

Desipramine

The tricyclic antidepressant desipramine is a potent norepinephrine reuptake inhibitor that is effective in treating ADHD in children and adults. In a double-blind, placebo-controlled study of adults with ADHD, subjects receiving desipramine showed robust improvement in symptom scores on the ADHD Rating Scale,¹⁶ compared with those receiving the placebo (Figure 2).¹⁷

During the 6-week trial, 41 adults with ADHD received desipramine, 200 mg/d, or a placebo. Those receiving desipramine showed significant improvement in 12 of 14 ADHD symptoms and less hyperactivity, impulsivity, and inattentiveness, whereas those receiving the placebo showed no improvement. According to the study criteria, 68% of those who received desipramine and none who received the placebo were considered positive responders.

Though no head-to-head studies have compared desipramine with methylphenidate, the same researchers conducted a similar placebo-controlled study with methylphenidate. The 6-week symptom score on the ADHD Rating Scale was 12.5 for methylphenidate, compared with a score of 12 for desipramine.¹⁸

Recommendation. Desipramine may be the most effective of the antidepressant treatments for patients with ADHD. Because of its side effects, however, it is not this author’s first choice and is usually reserved for patients whose symptoms fail to respond to other antidepressants. Desipramine can cause sedation, dry mouth, and constipation, which are related to blockade of adrenergic, histamine, and muscarinic cholinergic receptors. It also can be lethal in overdose.

Some substance abusers lose confidence in a medication that they cannot feel working. The side effects of desipramine, which can be intolerable for some patients, can reassure others with a history of substance abuse that they are being medicated.

Bupropion

Bupropion is a unique antidepressant that inhibits the presynaptic reuptake of dopamine and norepinephrine. Open-label studies demonstrate good responses to bupropion by adults with ADHD. One placebo-controlled, double-blind study found improved ADHD symptoms in 76% of patients receiving bupropion SR, compared with 37% of those receiving a placebo; the difference was statistically significant.¹⁹

Figure 2 IMPROVED ADHD SYMPTOMS WITH DESIPRAMINE

Adults with ADHD who received desipramine, 200 mg/d, in a double-blind trial showed significantly less hyperactivity, impulsivity, and inattentiveness after 6 weeks of therapy than a control group that received a placebo.

Source: Adapted with permission from Wilens TE, Biederman J, Prince J, et al. Six-week, double-blind, placebo-controlled study of desipramine for adult attention-deficit/hyperactivity disorder. Am J Psychiatry 1996;153:1147-53.Another study of adults with ADHD compared bupropion SR to methylphenidate and a placebo.²⁰ Using a primary outcome of the Clinical Global Impression (CGI) scale, response rates were 64% for bupropion SR, 50% for methylphenidate, and 27% for the placebo. The difference in response rates between the two agents was not statistically significant (p = 0.14).

Recommendation. The risk of seizures with bupropion is about 1 in 1,000. Therefore, bupropion should not be given to patients with a seizure disorder or to those with conditions that alter the seizure threshold (e.g., eating disorders, recent head trauma, or benzodiazepine withdrawal).²¹

This author uses bupropion as first-line treatment for appropriate patients with ADHD and a substance abuse history. Bupropion’s mild benefit with smoking cessation may provide some crossover effect for other substances of abuse. The low incidence of sexual side effects is another benefit. Drawbacks include twice-daily dosing and lack of a robust effect on attention and concentration.

Venlafaxine

Venlafaxine is a potent inhibitor of serotonin reuptake, a moderate inhibitor of norepinephrine, and a mild inhibitor of dopamine. Venlafaxine has displayed response rates similar to those of desipramine and bupropion in open-label studies in adults with ADHD,¹⁴ but no placebo-controlled studies exist.

As noted above, these antidepressants are believed to improve ADHD symptoms by making norepinephrine more available at the synapse. Hypothetically, then, one would need to administer venlafaxine at dosages that adequately inhibit the norepinephrine reuptake receptor. Venlafaxine XR, 150 mg/d, provides significant norepinephrine activity, according to several lines of evidence.

Recently, Upadhyaya et al reported the use of venlafaxine in one of the few treatment studies of patients with ADHD and comorbid alcohol/cocaine abuse. In an open-label trial, 10 subjects received venlafaxine, up to 300 mg/d, along with psychotherapy and attendance at Alcoholics Anonymous meetings. The nine who completed 4 weeks of treatment showed significantly improved ADHD symptoms and decreased alcohol craving.²²

Recommendation. Venlafaxine is this author’s second choice for patients with ADHD and substance abuse problems. Sexual side effects that some patients experience with venlafaxine can limit its use. Some clinicians are concerned about increases in blood pressure associated with venlafaxine, although significant changes do not seem to occur at dosages below 300 mg/d.²³

Atomoxetine

Atomoxetine is an investigational antidepressant in phase-III trials as a treatment for ADHD. Evidence shows atomoxetine to be a potent inhibitor of the presynaptic norepinephrine transporter, with minimal affinity for other neurotransmitter receptors. Initial studies suggest that atomoxetine is effective for adults and children with ADHD:

• In a small, double-blind, placebo-controlled, crossover trial, 11 of 20 adults showed improvement in ADHD symptoms within 3 weeks of starting atomoxetine.²⁴
• In 297 children and adolescents, atomoxetine at dosages averaging approximately 1.2 mg/kg/d was more effective than a placebo in reducing ADHD symptoms and improving social and family functioning. Treatment was well tolerated and without significant side effects.²⁵
• In a randomized open-label trial, 228 children received atomoxetine or methylphenidate. Both treatments significantly reduced inattention and hyperactive/impulsive symptoms.²⁶

Related resources

• Arnsten AF. Genetics of childhood disorders (XVIII). ADHD, Part 2: Norepinephrine has a critical modulatory influence on prefrontal cortical function. J Am Acad Child Adolesc Psychiatry. 2000;39:1201-3.
• Ward MF, Wender PH, Reimherr FW. The Wender Utah Rating Scale: An aid in the retrospective diagnosis of childhood attention-deficit/hyperactivity disorder. Am J Psychiatry. 1993;150:885-90.
• Research Report: Prescription Drugs—abuse and addiction. National Institute on Drug Abuse, Substance Abuse and Mental Health Services Administration.

Drug brand names

• Bupropion • Wellbutrin
• Desipramine • Norpramin
• Methylphenidate • Ritalin, Concerta
• Venlafaxine • Effexor

Disclosure

Dr. Higgins reports that he is on the speakers’ bureaus for Wyeth-Ayerst Pharmaceuticals and Eli Lilly and Co.