PTSD

Major depressive disorder (MDD) and bipolar spectrum disorders are associated with some symptoms of—and fully defined—posttraumatic stress disorder (PTSD). Many traumatic experiences can lead to this comorbidity, the most common being exposure to or witnessing combat for men and rape and sexual molestation for women.1

Trauma has major prognostic and treatment implications for affectively ill patients, including those whose symptoms do not meet PTSD’s full diagnostic criteria. This article aims to help clinicians by:

• presenting evidence characterizing the overlap between affective disorders and PTSD
• reviewing evidence that the bipolar spectrum may be broader than generally thought, an insight that affects PTSD treatment
• making a case for routine PTSD screening for all patients with affective illnesses
• recommending PTSD treatments tailored to the patient’s comorbid affective disorder.

Overlap of trauma and affective illness

PTSD is remarkably comorbid with mood disorders. Americans with MDD and bipolar disorder (BPD) are 7 and 9.4 times, respectively, more likely to meet criteria for PTSD than persons in the general population, according to odds ratios Kessler et al² calculated from the National Comorbidity Survey database.

I have never seen a patient with PTSD who did not also meet criteria for an affective disorder. The concurrence of PTSD and MDD is not the product of overlapping diagnostic criteria. Rather, evidence indicates these are distinct diagnostic entities.³ A review of diagnostic criteria for PTSD and hypomania/mania leads to the same conclusion.

Bipolar spectrum disorders

DSM-IV-TR assumes that mood disorders fall neatly into boxes. Other data (Table 1)^4–8 indicate that these disorders fall along a continuum or—more conservatively—that the scope of bipolarity is much wider than DSM-IV-TR recognizes. This is a controversial topic, and the individual clinician’s position could impact how one manages PTSD patients.

Table 1

Evidence of bipolar spectrum features in major depressive episodes

In this article, I include bipolar I disorder, bipolar II disorder, and mixed depression within the “bipolar spectrum disorders.” If one accepts this—and I do—it follows that 50% to 70% of all major depressive episodes (MDEs) are bipolar in nature.^4–9 Depending on your practice setting, you may see a higher or lower base rate of bipolar spectrum disorders.

Mixed depression is not recognized in DSM-IV-TR, and the purpose of this article is not to defend its inclusion as a bipolar spectrum phenomenon. A proposed definition of mixed depression⁹ requires the presence of an MDE contaminated by ≥3 features of hypomania or mania, without euphoria or inflated self-esteem/grandiosity (Table 2).¹⁰

Some experts believe episodes of hypomania and mania frequently occur in the illness course of persons with mixed depression; indeed, mixed depression is a predictor of a bipolar course. It is observed in outpatient⁹ and inpatient settings.¹¹ Common forms of mixed depression feature combinations of irritability, psychomotor agitation (mild to severe), increased talkativeness (which may fall short of frank pressured speech), racing or “crowded” thoughts (or “mental overactivity”), and distractibility. Other than increased self-esteem/grandiosity, any symptoms within DSM-IV-TR criterion B for a hypomanic or manic episode may be seen in mixed depression. Psychosis is an exclusion criterion for mixed depression.

Mixed depression responds poorly to antidepressant monotherapy. Validation studies suggest that mixed depression is a bipolar variant, as determined by its capacity to predict a bipolar course and its association with a family history of bipolar disorder and age of onset.⁹

Table 2

Diagnostic characteristics of a hypomanic episode, DSM-IV-TR criteria A and B

PTSD risk in affective illness

An adolescent sample. A preliminary cross-sectional study conducted by our group indicates that adolescents with affective disorders may have a much higher risk of developing PTSD than psychiatric comparison subjects.¹² We used modules from the Structured Clinical Interview for DSM-IV (SCID) to screen for intra-episode psychopathology (as opposed to lifetime prevalence of disorders) in 79 adolescents with MDD, 34 with BPD as defined in the DSM-IV-TR, and 26 with neither affective disorder (psychiatric controls). We found:

• 38.2% of subjects with BPD met criteria for PTSD, compared with 13.9% of those with MDD (OR 4.9; P =.001)
• 3.8% of adolescents without a mood disorder met criteria for PTSD.

We also found that comorbid PTSD was associated with a 4.5-fold higher risk of a suicide attempt, even after we controlled for BPD diagnosis. When we controlled for the presence of other concurrent anxiety disorders, the likelihood of an adolescent with PTSD having attempted suicide remained significant (OR 3.4; P=.023). This finding suggests that PTSD is an independent risk factor for a suicide attempt.

An adult sample. We then focused on adults meeting criteria for MDD or BPD. In a study of 187 consecutively presenting affectively ill patients, we used the SCID to screen for multiple anxiety disorders including PTSD.¹³ Lifetime—as opposed to intra-episode—PTSD prevalence was 23.8% among the 118 patients with MDD and 62.3% among the 69 patients with BPD. A patient with BPD was 5 times more likely to have PTSD than a patient with MDD (OR 5.3; P < .0001). The most common cause of trauma leading to PTSD was sexual molestation or rape as a child or adolescent in this predominantly female Latino population.

Populations at risk for PTSD

The prevalence of PTSD in clinical samples varies, depending on the population studied. For instance, women are at much higher risk for developing PTSD than men, even in comparisons where men are exposed to a greater number of traumatic events and analyses control for differences in the prevalence of sexual abuse. The gender difference is greater if the trauma occurs during childhood.¹⁴ Essentially all patients in our adolescent and adult studies developed PTSD in response to childhood or adolescent sexual trauma.^12,13

A population exposed to a high rate of violent crime would be expected to show a higher PTSD prevalence than one exposed to substantially less violence. The base rate of PTSD also is much higher in affectively ill patients than in the general population.

An analysis by Otto et al¹⁵ found a 16% lifetime prevalence of concomitant PTSD in 1,214 persons with BPD (not the manifold forms within the bipolar spectrum). Oquendo et al¹⁶ reported a 25.7% lifetime prevalence of PTSD in 230 patients with a history of MDD. Other epidemiologic² and clinical studies^12,13 suggest a considerably higher base rate of PTSD among persons with bipolar disorders than those with MDD.

The method of ascertaining the presence of this disorder may be another variable affecting the reported PTSD prevalence. Persistent avoidance—including “efforts to avoid thoughts, feelings, or conversations associated with the trauma”—is a diagnostic feature of PTSD.¹⁰ Researchers and clinicians who do not intentionally screen patients for PTSD are not likely to detect it. Determining the true prevalence of PTSD requires empathic inquiry about exposure to traumatic events.

PTSD screening

Humans are remarkably resilient, and most persons exposed to major trauma are thought not to develop PTSD. However, in my experience, because PTSD appears to be common among persons with affective illness, determining whether such patients have been traumatized is important for prognosis and treatment selection.

To get started, you could create a 1-page form to record traumatic events and identify features of PTSD according to DSM-IV-TR criteria (Checklist).¹⁰ PTSD screening without a form can become second nature with practice; an experienced clinician can screen a traumatized patient for the disorder within 3 to 5 minutes.

When screening for a history of trauma, ask patients in a straightforward manner if they have:

• been victims of violent crimes
• witnessed violent crimes
• been exposed to events in which people could have suffered grave injury
• experienced emotional, physical, or sexual abuse.

A person who has experienced emotional abuse but not physical or sexual abuse cannot meet DSM-IV-TR criterion A and therefore does not meet full criteria for PTSD. Many emotionally abused persons meet criteria B through F, however, and they are most reasonably managed similarly to persons who also meet criterion A. When formulating a treatment plan, I recommend using clinical judgment rather than rigid adherence to DSM-IV-TR.

Checklist

DSM-IV-TR diagnostic criteria for posttraumatic stress disorder

Treating PTSD in depression

Pharmacotherapy and psychotherapeutic interventions are important to PTSD patients’ recovery. Limited resources often prevent these patients from receiving expert psychotherapeutic intervention, however, leaving pharmacotherapy as the mainstay of treatment. This is unfortunate, because psychological interventions may be sufficient and preferred in some instances (Box).^17–20

Pharmacotherapy for comorbid MDD. Selective serotonin reuptake inhibitors (SSRIs) and venlafaxine are first-line interventions for PTSD in depressed patients who do not meet criteria for a bipolar spectrum disorder. Placebo-controlled studies suggest that sertraline,^21,22 fluoxetine,²³ and paroxetine,²⁴ are effective; doses higher than those used to treat depression may be required. Extended-release venlafaxine²⁵ in dosages similar to those needed for depressive disorders also can be effective. Bupropion does not appear to be beneficial in treating PTSD.

The monoamine oxidase inhibitor phenelzine was long used successfully in treating PTSD but for the most part has been replaced by SSRIs. Because of its associated dietary restrictions, risk of hypertensive crises, and other side effects, phenelzine probably is best reserved for patients who do not respond to treatment with SSRIs or venlafaxine.

Pharmacotherapy for comorbid bipolar spectrum. If one accepts that most patients meeting criteria for MDE have a bipolar spectrum disorder, then most affectively ill patients with PTSD would need to be treated as if they have bipolar disorder. Oddly enough, this creates difficulties for the use of not only antidepressants and benzodiazepines, but also mood stabilizers:

• Patients with BPD and comorbid anxiety disorders, including PTSD, may be resistant to mood stabilizers.^26,27
• Antidepressants can precipitate hypomanic or manic switches or onset of mixed hypomania, a mixed state, or rapid cycling in patients with a bipolar spectrum disorder.^28–30
• Benzodiazepines do not appear to relieve acute or chronic PTSD-related distress, and discontinuation could cause rebound symptoms.³¹

Because no outcome studies have addressed PTSD management in patients with bipolar spectrum disorders, clinicians must rely on their judgment when formulating treatment plans. One strategy is to treat patients with mood stabilizers, then leave well enough alone if both the mood and anxiety symptoms remit (which is possible but unlikely in my experience). I often start treatment for the bipolar spectrum disorder and co-existing PTSD using mood stabilizers (including atypical antipsychotics) and prazosin, an α-1antagonist originally used for treating hypertension.

Prazosin can help diminish nightmares, dreams, and other painful recollections of trauma.^32,33 The drug does not affect time to sleep onset. It also has been reported to reduce avoidance behavior and hyperarousal, such as irritability and anger.³⁴ This has been my experience.

Box

Prazosin to treat PTSD-related symptoms in children or adolescents has not been studied, but it can be useful in adults over a wide range of doses. As little as 1 mg at bedtime may confer benefit, although the mean prazosin dose in an 8-week, placebo-controlled study of 40 combat veterans was 13.3 mg in the evening.³³

I often initiate prazosin treatment as follows:

• 1 mg on the first night of treatment
• 2 mg on the second night
• 3 mg on the third night
• then, if tolerated, 1 mg upon waking, 1 mg 8 hours later, and 3 mg at bedtime. I then slowly adjust the dose schedule based on the patient’s needs, such as minimizing painful re-experiencing of the trauma. Reducing avoidance and hyperarousal also are reasonable targets. For example, when using prazosin to treat extremely angry men with PTSD stemming from exposure to violent crimes, I have observed that even “murderous” rage ceases with prazosin treatment, only to reappear when prazosin is discontinued.

In treating approximately 100 patients with prazosin, I have not exceeded 16 mg/d. Dosages used for treating hypertension usually are 5 to 20 mg/d. When using prazosin, I always:

• warn patients that faintness or fainting is a side effect and record this caveat in their chart
• obtain sitting and standing blood pressure and pulse before starting treatment and subsequently
• ask patients if they feel dizzy when changing posture before and after initiating treatment.

Most of my PTSD patients are suffering so much that they are willing to accept the risk of fainting associated with prazosin use. For PTSD comorbid with severe panic disorder,^12,13 I find that a benzodiazepine with antipanic properties such as alprazolam or clonazepam often works well in conjunction with prazosin.

Some patients with bipolar spectrum disorders might benefit from the addition of an SSRI after mood stabilizer treatment proves effective. However, I have never managed a patient in this manner, and like my own treatment strategy, this has not been subjected to rigorous empiric inquiry. In my view, psychological treatment is much preferred to antidepressant therapy.

Related resource

• Benazzi F. Bipolar disorder—focus on bipolar II disorder and mixed depression. Lancet. 2007;369:935-945.

Drug brand names

• Alprazolam • Xanax
• Bupropion • Wellbutrin
• Clonazepam • Klonopin
• D-cycloserine • Seromycin
• Fluoxetine • Prozac
• Paroxetine • Paxil
• Phenelzine • Nardil
• Prazosin • Minipress
• Sertraline • Zoloft
• Venlafaxine • Effexor

Disclosure

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

Major depressive disorder (MDD) and bipolar spectrum disorders are associated with some symptoms of—and fully defined—posttraumatic stress disorder (PTSD). Many traumatic experiences can lead to this comorbidity, the most common being exposure to or witnessing combat for men and rape and sexual molestation for women.1

Trauma has major prognostic and treatment implications for affectively ill patients, including those whose symptoms do not meet PTSD’s full diagnostic criteria. This article aims to help clinicians by:

• presenting evidence characterizing the overlap between affective disorders and PTSD
• reviewing evidence that the bipolar spectrum may be broader than generally thought, an insight that affects PTSD treatment
• making a case for routine PTSD screening for all patients with affective illnesses
• recommending PTSD treatments tailored to the patient’s comorbid affective disorder.

Overlap of trauma and affective illness

PTSD is remarkably comorbid with mood disorders. Americans with MDD and bipolar disorder (BPD) are 7 and 9.4 times, respectively, more likely to meet criteria for PTSD than persons in the general population, according to odds ratios Kessler et al² calculated from the National Comorbidity Survey database.

I have never seen a patient with PTSD who did not also meet criteria for an affective disorder. The concurrence of PTSD and MDD is not the product of overlapping diagnostic criteria. Rather, evidence indicates these are distinct diagnostic entities.³ A review of diagnostic criteria for PTSD and hypomania/mania leads to the same conclusion.

Bipolar spectrum disorders

DSM-IV-TR assumes that mood disorders fall neatly into boxes. Other data (Table 1)^4–8 indicate that these disorders fall along a continuum or—more conservatively—that the scope of bipolarity is much wider than DSM-IV-TR recognizes. This is a controversial topic, and the individual clinician’s position could impact how one manages PTSD patients.

Table 1

Evidence of bipolar spectrum features in major depressive episodes

In this article, I include bipolar I disorder, bipolar II disorder, and mixed depression within the “bipolar spectrum disorders.” If one accepts this—and I do—it follows that 50% to 70% of all major depressive episodes (MDEs) are bipolar in nature.^4–9 Depending on your practice setting, you may see a higher or lower base rate of bipolar spectrum disorders.

Mixed depression is not recognized in DSM-IV-TR, and the purpose of this article is not to defend its inclusion as a bipolar spectrum phenomenon. A proposed definition of mixed depression⁹ requires the presence of an MDE contaminated by ≥3 features of hypomania or mania, without euphoria or inflated self-esteem/grandiosity (Table 2).¹⁰

Some experts believe episodes of hypomania and mania frequently occur in the illness course of persons with mixed depression; indeed, mixed depression is a predictor of a bipolar course. It is observed in outpatient⁹ and inpatient settings.¹¹ Common forms of mixed depression feature combinations of irritability, psychomotor agitation (mild to severe), increased talkativeness (which may fall short of frank pressured speech), racing or “crowded” thoughts (or “mental overactivity”), and distractibility. Other than increased self-esteem/grandiosity, any symptoms within DSM-IV-TR criterion B for a hypomanic or manic episode may be seen in mixed depression. Psychosis is an exclusion criterion for mixed depression.

Mixed depression responds poorly to antidepressant monotherapy. Validation studies suggest that mixed depression is a bipolar variant, as determined by its capacity to predict a bipolar course and its association with a family history of bipolar disorder and age of onset.⁹

Table 2

Diagnostic characteristics of a hypomanic episode, DSM-IV-TR criteria A and B

PTSD risk in affective illness

An adolescent sample. A preliminary cross-sectional study conducted by our group indicates that adolescents with affective disorders may have a much higher risk of developing PTSD than psychiatric comparison subjects.¹² We used modules from the Structured Clinical Interview for DSM-IV (SCID) to screen for intra-episode psychopathology (as opposed to lifetime prevalence of disorders) in 79 adolescents with MDD, 34 with BPD as defined in the DSM-IV-TR, and 26 with neither affective disorder (psychiatric controls). We found:

• 38.2% of subjects with BPD met criteria for PTSD, compared with 13.9% of those with MDD (OR 4.9; P =.001)
• 3.8% of adolescents without a mood disorder met criteria for PTSD.

We also found that comorbid PTSD was associated with a 4.5-fold higher risk of a suicide attempt, even after we controlled for BPD diagnosis. When we controlled for the presence of other concurrent anxiety disorders, the likelihood of an adolescent with PTSD having attempted suicide remained significant (OR 3.4; P=.023). This finding suggests that PTSD is an independent risk factor for a suicide attempt.

An adult sample. We then focused on adults meeting criteria for MDD or BPD. In a study of 187 consecutively presenting affectively ill patients, we used the SCID to screen for multiple anxiety disorders including PTSD.¹³ Lifetime—as opposed to intra-episode—PTSD prevalence was 23.8% among the 118 patients with MDD and 62.3% among the 69 patients with BPD. A patient with BPD was 5 times more likely to have PTSD than a patient with MDD (OR 5.3; P < .0001). The most common cause of trauma leading to PTSD was sexual molestation or rape as a child or adolescent in this predominantly female Latino population.

Populations at risk for PTSD

The prevalence of PTSD in clinical samples varies, depending on the population studied. For instance, women are at much higher risk for developing PTSD than men, even in comparisons where men are exposed to a greater number of traumatic events and analyses control for differences in the prevalence of sexual abuse. The gender difference is greater if the trauma occurs during childhood.¹⁴ Essentially all patients in our adolescent and adult studies developed PTSD in response to childhood or adolescent sexual trauma.^12,13

A population exposed to a high rate of violent crime would be expected to show a higher PTSD prevalence than one exposed to substantially less violence. The base rate of PTSD also is much higher in affectively ill patients than in the general population.

An analysis by Otto et al¹⁵ found a 16% lifetime prevalence of concomitant PTSD in 1,214 persons with BPD (not the manifold forms within the bipolar spectrum). Oquendo et al¹⁶ reported a 25.7% lifetime prevalence of PTSD in 230 patients with a history of MDD. Other epidemiologic² and clinical studies^12,13 suggest a considerably higher base rate of PTSD among persons with bipolar disorders than those with MDD.

The method of ascertaining the presence of this disorder may be another variable affecting the reported PTSD prevalence. Persistent avoidance—including “efforts to avoid thoughts, feelings, or conversations associated with the trauma”—is a diagnostic feature of PTSD.¹⁰ Researchers and clinicians who do not intentionally screen patients for PTSD are not likely to detect it. Determining the true prevalence of PTSD requires empathic inquiry about exposure to traumatic events.

PTSD screening

Humans are remarkably resilient, and most persons exposed to major trauma are thought not to develop PTSD. However, in my experience, because PTSD appears to be common among persons with affective illness, determining whether such patients have been traumatized is important for prognosis and treatment selection.

To get started, you could create a 1-page form to record traumatic events and identify features of PTSD according to DSM-IV-TR criteria (Checklist).¹⁰ PTSD screening without a form can become second nature with practice; an experienced clinician can screen a traumatized patient for the disorder within 3 to 5 minutes.

When screening for a history of trauma, ask patients in a straightforward manner if they have:

• been victims of violent crimes
• witnessed violent crimes
• been exposed to events in which people could have suffered grave injury
• experienced emotional, physical, or sexual abuse.

A person who has experienced emotional abuse but not physical or sexual abuse cannot meet DSM-IV-TR criterion A and therefore does not meet full criteria for PTSD. Many emotionally abused persons meet criteria B through F, however, and they are most reasonably managed similarly to persons who also meet criterion A. When formulating a treatment plan, I recommend using clinical judgment rather than rigid adherence to DSM-IV-TR.

Checklist

DSM-IV-TR diagnostic criteria for posttraumatic stress disorder

Treating PTSD in depression

Pharmacotherapy and psychotherapeutic interventions are important to PTSD patients’ recovery. Limited resources often prevent these patients from receiving expert psychotherapeutic intervention, however, leaving pharmacotherapy as the mainstay of treatment. This is unfortunate, because psychological interventions may be sufficient and preferred in some instances (Box).^17–20

Pharmacotherapy for comorbid MDD. Selective serotonin reuptake inhibitors (SSRIs) and venlafaxine are first-line interventions for PTSD in depressed patients who do not meet criteria for a bipolar spectrum disorder. Placebo-controlled studies suggest that sertraline,^21,22 fluoxetine,²³ and paroxetine,²⁴ are effective; doses higher than those used to treat depression may be required. Extended-release venlafaxine²⁵ in dosages similar to those needed for depressive disorders also can be effective. Bupropion does not appear to be beneficial in treating PTSD.

The monoamine oxidase inhibitor phenelzine was long used successfully in treating PTSD but for the most part has been replaced by SSRIs. Because of its associated dietary restrictions, risk of hypertensive crises, and other side effects, phenelzine probably is best reserved for patients who do not respond to treatment with SSRIs or venlafaxine.

Pharmacotherapy for comorbid bipolar spectrum. If one accepts that most patients meeting criteria for MDE have a bipolar spectrum disorder, then most affectively ill patients with PTSD would need to be treated as if they have bipolar disorder. Oddly enough, this creates difficulties for the use of not only antidepressants and benzodiazepines, but also mood stabilizers:

• Patients with BPD and comorbid anxiety disorders, including PTSD, may be resistant to mood stabilizers.^26,27
• Antidepressants can precipitate hypomanic or manic switches or onset of mixed hypomania, a mixed state, or rapid cycling in patients with a bipolar spectrum disorder.^28–30
• Benzodiazepines do not appear to relieve acute or chronic PTSD-related distress, and discontinuation could cause rebound symptoms.³¹

Because no outcome studies have addressed PTSD management in patients with bipolar spectrum disorders, clinicians must rely on their judgment when formulating treatment plans. One strategy is to treat patients with mood stabilizers, then leave well enough alone if both the mood and anxiety symptoms remit (which is possible but unlikely in my experience). I often start treatment for the bipolar spectrum disorder and co-existing PTSD using mood stabilizers (including atypical antipsychotics) and prazosin, an α-1antagonist originally used for treating hypertension.

Prazosin can help diminish nightmares, dreams, and other painful recollections of trauma.^32,33 The drug does not affect time to sleep onset. It also has been reported to reduce avoidance behavior and hyperarousal, such as irritability and anger.³⁴ This has been my experience.

Box

Prazosin to treat PTSD-related symptoms in children or adolescents has not been studied, but it can be useful in adults over a wide range of doses. As little as 1 mg at bedtime may confer benefit, although the mean prazosin dose in an 8-week, placebo-controlled study of 40 combat veterans was 13.3 mg in the evening.³³

I often initiate prazosin treatment as follows:

• 1 mg on the first night of treatment
• 2 mg on the second night
• 3 mg on the third night
• then, if tolerated, 1 mg upon waking, 1 mg 8 hours later, and 3 mg at bedtime. I then slowly adjust the dose schedule based on the patient’s needs, such as minimizing painful re-experiencing of the trauma. Reducing avoidance and hyperarousal also are reasonable targets. For example, when using prazosin to treat extremely angry men with PTSD stemming from exposure to violent crimes, I have observed that even “murderous” rage ceases with prazosin treatment, only to reappear when prazosin is discontinued.

In treating approximately 100 patients with prazosin, I have not exceeded 16 mg/d. Dosages used for treating hypertension usually are 5 to 20 mg/d. When using prazosin, I always:

• warn patients that faintness or fainting is a side effect and record this caveat in their chart
• obtain sitting and standing blood pressure and pulse before starting treatment and subsequently
• ask patients if they feel dizzy when changing posture before and after initiating treatment.

Most of my PTSD patients are suffering so much that they are willing to accept the risk of fainting associated with prazosin use. For PTSD comorbid with severe panic disorder,^12,13 I find that a benzodiazepine with antipanic properties such as alprazolam or clonazepam often works well in conjunction with prazosin.

Some patients with bipolar spectrum disorders might benefit from the addition of an SSRI after mood stabilizer treatment proves effective. However, I have never managed a patient in this manner, and like my own treatment strategy, this has not been subjected to rigorous empiric inquiry. In my view, psychological treatment is much preferred to antidepressant therapy.

Related resource

• Benazzi F. Bipolar disorder—focus on bipolar II disorder and mixed depression. Lancet. 2007;369:935-945.

Drug brand names

• Alprazolam • Xanax
• Bupropion • Wellbutrin
• Clonazepam • Klonopin
• D-cycloserine • Seromycin
• Fluoxetine • Prozac
• Paroxetine • Paxil
• Phenelzine • Nardil
• Prazosin • Minipress
• Sertraline • Zoloft
• Venlafaxine • Effexor

Disclosure

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

Major depressive disorder (MDD) and bipolar spectrum disorders are associated with some symptoms of—and fully defined—posttraumatic stress disorder (PTSD). Many traumatic experiences can lead to this comorbidity, the most common being exposure to or witnessing combat for men and rape and sexual molestation for women.1

Trauma has major prognostic and treatment implications for affectively ill patients, including those whose symptoms do not meet PTSD’s full diagnostic criteria. This article aims to help clinicians by:

• presenting evidence characterizing the overlap between affective disorders and PTSD
• reviewing evidence that the bipolar spectrum may be broader than generally thought, an insight that affects PTSD treatment
• making a case for routine PTSD screening for all patients with affective illnesses
• recommending PTSD treatments tailored to the patient’s comorbid affective disorder.

Overlap of trauma and affective illness

PTSD is remarkably comorbid with mood disorders. Americans with MDD and bipolar disorder (BPD) are 7 and 9.4 times, respectively, more likely to meet criteria for PTSD than persons in the general population, according to odds ratios Kessler et al² calculated from the National Comorbidity Survey database.

I have never seen a patient with PTSD who did not also meet criteria for an affective disorder. The concurrence of PTSD and MDD is not the product of overlapping diagnostic criteria. Rather, evidence indicates these are distinct diagnostic entities.³ A review of diagnostic criteria for PTSD and hypomania/mania leads to the same conclusion.

Bipolar spectrum disorders

DSM-IV-TR assumes that mood disorders fall neatly into boxes. Other data (Table 1)^4–8 indicate that these disorders fall along a continuum or—more conservatively—that the scope of bipolarity is much wider than DSM-IV-TR recognizes. This is a controversial topic, and the individual clinician’s position could impact how one manages PTSD patients.

Table 1

Evidence of bipolar spectrum features in major depressive episodes

In this article, I include bipolar I disorder, bipolar II disorder, and mixed depression within the “bipolar spectrum disorders.” If one accepts this—and I do—it follows that 50% to 70% of all major depressive episodes (MDEs) are bipolar in nature.^4–9 Depending on your practice setting, you may see a higher or lower base rate of bipolar spectrum disorders.

Mixed depression is not recognized in DSM-IV-TR, and the purpose of this article is not to defend its inclusion as a bipolar spectrum phenomenon. A proposed definition of mixed depression⁹ requires the presence of an MDE contaminated by ≥3 features of hypomania or mania, without euphoria or inflated self-esteem/grandiosity (Table 2).¹⁰

Some experts believe episodes of hypomania and mania frequently occur in the illness course of persons with mixed depression; indeed, mixed depression is a predictor of a bipolar course. It is observed in outpatient⁹ and inpatient settings.¹¹ Common forms of mixed depression feature combinations of irritability, psychomotor agitation (mild to severe), increased talkativeness (which may fall short of frank pressured speech), racing or “crowded” thoughts (or “mental overactivity”), and distractibility. Other than increased self-esteem/grandiosity, any symptoms within DSM-IV-TR criterion B for a hypomanic or manic episode may be seen in mixed depression. Psychosis is an exclusion criterion for mixed depression.

Mixed depression responds poorly to antidepressant monotherapy. Validation studies suggest that mixed depression is a bipolar variant, as determined by its capacity to predict a bipolar course and its association with a family history of bipolar disorder and age of onset.⁹

Table 2

Diagnostic characteristics of a hypomanic episode, DSM-IV-TR criteria A and B

PTSD risk in affective illness

An adolescent sample. A preliminary cross-sectional study conducted by our group indicates that adolescents with affective disorders may have a much higher risk of developing PTSD than psychiatric comparison subjects.¹² We used modules from the Structured Clinical Interview for DSM-IV (SCID) to screen for intra-episode psychopathology (as opposed to lifetime prevalence of disorders) in 79 adolescents with MDD, 34 with BPD as defined in the DSM-IV-TR, and 26 with neither affective disorder (psychiatric controls). We found:

• 38.2% of subjects with BPD met criteria for PTSD, compared with 13.9% of those with MDD (OR 4.9; P =.001)
• 3.8% of adolescents without a mood disorder met criteria for PTSD.

We also found that comorbid PTSD was associated with a 4.5-fold higher risk of a suicide attempt, even after we controlled for BPD diagnosis. When we controlled for the presence of other concurrent anxiety disorders, the likelihood of an adolescent with PTSD having attempted suicide remained significant (OR 3.4; P=.023). This finding suggests that PTSD is an independent risk factor for a suicide attempt.

An adult sample. We then focused on adults meeting criteria for MDD or BPD. In a study of 187 consecutively presenting affectively ill patients, we used the SCID to screen for multiple anxiety disorders including PTSD.¹³ Lifetime—as opposed to intra-episode—PTSD prevalence was 23.8% among the 118 patients with MDD and 62.3% among the 69 patients with BPD. A patient with BPD was 5 times more likely to have PTSD than a patient with MDD (OR 5.3; P < .0001). The most common cause of trauma leading to PTSD was sexual molestation or rape as a child or adolescent in this predominantly female Latino population.

Populations at risk for PTSD

The prevalence of PTSD in clinical samples varies, depending on the population studied. For instance, women are at much higher risk for developing PTSD than men, even in comparisons where men are exposed to a greater number of traumatic events and analyses control for differences in the prevalence of sexual abuse. The gender difference is greater if the trauma occurs during childhood.¹⁴ Essentially all patients in our adolescent and adult studies developed PTSD in response to childhood or adolescent sexual trauma.^12,13

A population exposed to a high rate of violent crime would be expected to show a higher PTSD prevalence than one exposed to substantially less violence. The base rate of PTSD also is much higher in affectively ill patients than in the general population.

An analysis by Otto et al¹⁵ found a 16% lifetime prevalence of concomitant PTSD in 1,214 persons with BPD (not the manifold forms within the bipolar spectrum). Oquendo et al¹⁶ reported a 25.7% lifetime prevalence of PTSD in 230 patients with a history of MDD. Other epidemiologic² and clinical studies^12,13 suggest a considerably higher base rate of PTSD among persons with bipolar disorders than those with MDD.

The method of ascertaining the presence of this disorder may be another variable affecting the reported PTSD prevalence. Persistent avoidance—including “efforts to avoid thoughts, feelings, or conversations associated with the trauma”—is a diagnostic feature of PTSD.¹⁰ Researchers and clinicians who do not intentionally screen patients for PTSD are not likely to detect it. Determining the true prevalence of PTSD requires empathic inquiry about exposure to traumatic events.

PTSD screening

Humans are remarkably resilient, and most persons exposed to major trauma are thought not to develop PTSD. However, in my experience, because PTSD appears to be common among persons with affective illness, determining whether such patients have been traumatized is important for prognosis and treatment selection.

To get started, you could create a 1-page form to record traumatic events and identify features of PTSD according to DSM-IV-TR criteria (Checklist).¹⁰ PTSD screening without a form can become second nature with practice; an experienced clinician can screen a traumatized patient for the disorder within 3 to 5 minutes.

When screening for a history of trauma, ask patients in a straightforward manner if they have:

• been victims of violent crimes
• witnessed violent crimes
• been exposed to events in which people could have suffered grave injury
• experienced emotional, physical, or sexual abuse.

A person who has experienced emotional abuse but not physical or sexual abuse cannot meet DSM-IV-TR criterion A and therefore does not meet full criteria for PTSD. Many emotionally abused persons meet criteria B through F, however, and they are most reasonably managed similarly to persons who also meet criterion A. When formulating a treatment plan, I recommend using clinical judgment rather than rigid adherence to DSM-IV-TR.

Checklist

DSM-IV-TR diagnostic criteria for posttraumatic stress disorder

Treating PTSD in depression

Pharmacotherapy and psychotherapeutic interventions are important to PTSD patients’ recovery. Limited resources often prevent these patients from receiving expert psychotherapeutic intervention, however, leaving pharmacotherapy as the mainstay of treatment. This is unfortunate, because psychological interventions may be sufficient and preferred in some instances (Box).^17–20

Pharmacotherapy for comorbid MDD. Selective serotonin reuptake inhibitors (SSRIs) and venlafaxine are first-line interventions for PTSD in depressed patients who do not meet criteria for a bipolar spectrum disorder. Placebo-controlled studies suggest that sertraline,^21,22 fluoxetine,²³ and paroxetine,²⁴ are effective; doses higher than those used to treat depression may be required. Extended-release venlafaxine²⁵ in dosages similar to those needed for depressive disorders also can be effective. Bupropion does not appear to be beneficial in treating PTSD.

The monoamine oxidase inhibitor phenelzine was long used successfully in treating PTSD but for the most part has been replaced by SSRIs. Because of its associated dietary restrictions, risk of hypertensive crises, and other side effects, phenelzine probably is best reserved for patients who do not respond to treatment with SSRIs or venlafaxine.

Pharmacotherapy for comorbid bipolar spectrum. If one accepts that most patients meeting criteria for MDE have a bipolar spectrum disorder, then most affectively ill patients with PTSD would need to be treated as if they have bipolar disorder. Oddly enough, this creates difficulties for the use of not only antidepressants and benzodiazepines, but also mood stabilizers:

• Patients with BPD and comorbid anxiety disorders, including PTSD, may be resistant to mood stabilizers.^26,27
• Antidepressants can precipitate hypomanic or manic switches or onset of mixed hypomania, a mixed state, or rapid cycling in patients with a bipolar spectrum disorder.^28–30
• Benzodiazepines do not appear to relieve acute or chronic PTSD-related distress, and discontinuation could cause rebound symptoms.³¹

Because no outcome studies have addressed PTSD management in patients with bipolar spectrum disorders, clinicians must rely on their judgment when formulating treatment plans. One strategy is to treat patients with mood stabilizers, then leave well enough alone if both the mood and anxiety symptoms remit (which is possible but unlikely in my experience). I often start treatment for the bipolar spectrum disorder and co-existing PTSD using mood stabilizers (including atypical antipsychotics) and prazosin, an α-1antagonist originally used for treating hypertension.

Prazosin can help diminish nightmares, dreams, and other painful recollections of trauma.^32,33 The drug does not affect time to sleep onset. It also has been reported to reduce avoidance behavior and hyperarousal, such as irritability and anger.³⁴ This has been my experience.

Box

Prazosin to treat PTSD-related symptoms in children or adolescents has not been studied, but it can be useful in adults over a wide range of doses. As little as 1 mg at bedtime may confer benefit, although the mean prazosin dose in an 8-week, placebo-controlled study of 40 combat veterans was 13.3 mg in the evening.³³

I often initiate prazosin treatment as follows:

• 1 mg on the first night of treatment
• 2 mg on the second night
• 3 mg on the third night
• then, if tolerated, 1 mg upon waking, 1 mg 8 hours later, and 3 mg at bedtime. I then slowly adjust the dose schedule based on the patient’s needs, such as minimizing painful re-experiencing of the trauma. Reducing avoidance and hyperarousal also are reasonable targets. For example, when using prazosin to treat extremely angry men with PTSD stemming from exposure to violent crimes, I have observed that even “murderous” rage ceases with prazosin treatment, only to reappear when prazosin is discontinued.

In treating approximately 100 patients with prazosin, I have not exceeded 16 mg/d. Dosages used for treating hypertension usually are 5 to 20 mg/d. When using prazosin, I always:

• warn patients that faintness or fainting is a side effect and record this caveat in their chart
• obtain sitting and standing blood pressure and pulse before starting treatment and subsequently
• ask patients if they feel dizzy when changing posture before and after initiating treatment.

Most of my PTSD patients are suffering so much that they are willing to accept the risk of fainting associated with prazosin use. For PTSD comorbid with severe panic disorder,^12,13 I find that a benzodiazepine with antipanic properties such as alprazolam or clonazepam often works well in conjunction with prazosin.

Some patients with bipolar spectrum disorders might benefit from the addition of an SSRI after mood stabilizer treatment proves effective. However, I have never managed a patient in this manner, and like my own treatment strategy, this has not been subjected to rigorous empiric inquiry. In my view, psychological treatment is much preferred to antidepressant therapy.

Related resource

• Benazzi F. Bipolar disorder—focus on bipolar II disorder and mixed depression. Lancet. 2007;369:935-945.

Drug brand names

• Alprazolam • Xanax
• Bupropion • Wellbutrin
• Clonazepam • Klonopin
• D-cycloserine • Seromycin
• Fluoxetine • Prozac
• Paroxetine • Paxil
• Phenelzine • Nardil
• Prazosin • Minipress
• Sertraline • Zoloft
• Venlafaxine • Effexor

Disclosure

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

Posttraumatic stress disorder (PTSD) is a confusing diagnostic category because it includes victims of trauma as well as individuals exposed to trauma. Also, PTSD encompasses exposure to different types of trauma, which can have significant implications for symptom development and treatment.

Consider the treatment history of a male combat veteran who exhibits multiple PTSD symptoms, including nightmares, flashbacks, social isolation, anger, and guilt related to his war experiences. Several psychiatrists saw the patient, which resulted in multiple medication changes but little benefit. On further assessment, the practitioners noted that the veteran’s war experiences were minimally problematic; the prominent nightmares, ruminations, flashbacks, and guilt were related to his witnessing a civilian female being sexually assaulted. The veteran’s guilt about not intervening was the basis of his PTSD. This led to a change in treatment from pharmacotherapy to a focus on supportive therapy.

Conceptualizing subtypes of PTSD—similar to many DSM-IV-TR diagnoses such as phobias or delusional disorders—might help better define the diagnosis. Each sub-type, as conceptualized below, might have its own prognosis and treatment. Our hope is that this strategy will benefit the patient by improving research and evidence-based practice.

PTSD subtypes

Victim-related trauma. Related to witnessing a criminal act or being a victim of a criminal act such as rape or assault. The patient is in a passive role.

Natural disasters, such as a tornado, earthquake, or hurricane.

Survivor guilt. The patient is not a perpetrator and might have been exposed to trauma, but symptoms are related to surviving while others close to the patient did not.

Perpetrator guilt. It is debatable whether this should be a PTSD subtype but our experience suggests that this pattern severely complicates PTSD diagnosis and treatment. It often is not initially disclosed by patients but surfaces when treatment is not working despite a strong therapeutic alliance.

PTSD not otherwise specified. This subtype is typical in patients who were not directly involved in a traumatic event but experienced symptoms related to it. Examples include picking up dead bodies, cleaning up a tornado site, or observing siblings being beaten. This category also may reflect an unclear picture if no primary subtype accounts for the majority of symptoms.

Qualifiers

Individuals who previously have been exposed to trauma are more vulnerable to subsequent trauma. Experiencing ongoing multiple traumatic events—such as in military combat—can have a cumulative effect. Thus, identifying episodes of trauma also should be part of the PTSD assessment.

Single event. The patient is exposed to a single traumatic episode, such as being the victim of a crime.

Multiple events/single episode. The patient is exposed to repeated, related traumatic events. Examples include ongoing military combat or child abuse.

Multiple events. The patient is exposed to ≥2 separate traumatic events. A combination such as this might include a serious motor vehicle accident followed by a natural disaster.

As the diagnosis of PTSD evolves, utilizing subtypes and qualifiers might clarify treatment strategies because some subtypes might be more amenable to certain psychopharmacologic or psychotherapeutic treatment regimens.

Diagnostic confusion

Some researchers question whether traumatic stress causes PTSD syndrome,¹ whereas others recommend “tightening” the diagnostic criteria.² Concerns regarding PTSD diagnosis are multiple and include:

• the importance of ruling out malingering³
• the effects of different diagnostic criteria resulting in disparate prevalence rates
• emphasizing the importance of dysfunction as a criterion for PTSD.⁴

Conceptual inconsistencies in DSM-IV-TR diagnostic criteria also can lead to confusion. Although there is a category of arousal symptoms, Criterion B4 (intense psychological distress) and Criterion B5 (physiological reactivity) are listed as re-experiencing symptoms rather than arousal symptoms. Finally, the criteria presented do not follow a logical progression. Research suggests that re-experiencing symptoms do not lead to avoidance but result in arousal symptoms, which in turn trigger avoidance.⁵

Posttraumatic stress disorder (PTSD) is a confusing diagnostic category because it includes victims of trauma as well as individuals exposed to trauma. Also, PTSD encompasses exposure to different types of trauma, which can have significant implications for symptom development and treatment.

Consider the treatment history of a male combat veteran who exhibits multiple PTSD symptoms, including nightmares, flashbacks, social isolation, anger, and guilt related to his war experiences. Several psychiatrists saw the patient, which resulted in multiple medication changes but little benefit. On further assessment, the practitioners noted that the veteran’s war experiences were minimally problematic; the prominent nightmares, ruminations, flashbacks, and guilt were related to his witnessing a civilian female being sexually assaulted. The veteran’s guilt about not intervening was the basis of his PTSD. This led to a change in treatment from pharmacotherapy to a focus on supportive therapy.

Conceptualizing subtypes of PTSD—similar to many DSM-IV-TR diagnoses such as phobias or delusional disorders—might help better define the diagnosis. Each sub-type, as conceptualized below, might have its own prognosis and treatment. Our hope is that this strategy will benefit the patient by improving research and evidence-based practice.

PTSD subtypes

Victim-related trauma. Related to witnessing a criminal act or being a victim of a criminal act such as rape or assault. The patient is in a passive role.

Natural disasters, such as a tornado, earthquake, or hurricane.

Survivor guilt. The patient is not a perpetrator and might have been exposed to trauma, but symptoms are related to surviving while others close to the patient did not.

Perpetrator guilt. It is debatable whether this should be a PTSD subtype but our experience suggests that this pattern severely complicates PTSD diagnosis and treatment. It often is not initially disclosed by patients but surfaces when treatment is not working despite a strong therapeutic alliance.

PTSD not otherwise specified. This subtype is typical in patients who were not directly involved in a traumatic event but experienced symptoms related to it. Examples include picking up dead bodies, cleaning up a tornado site, or observing siblings being beaten. This category also may reflect an unclear picture if no primary subtype accounts for the majority of symptoms.

Qualifiers

Individuals who previously have been exposed to trauma are more vulnerable to subsequent trauma. Experiencing ongoing multiple traumatic events—such as in military combat—can have a cumulative effect. Thus, identifying episodes of trauma also should be part of the PTSD assessment.

Single event. The patient is exposed to a single traumatic episode, such as being the victim of a crime.

Multiple events/single episode. The patient is exposed to repeated, related traumatic events. Examples include ongoing military combat or child abuse.

Multiple events. The patient is exposed to ≥2 separate traumatic events. A combination such as this might include a serious motor vehicle accident followed by a natural disaster.

As the diagnosis of PTSD evolves, utilizing subtypes and qualifiers might clarify treatment strategies because some subtypes might be more amenable to certain psychopharmacologic or psychotherapeutic treatment regimens.

Diagnostic confusion

Some researchers question whether traumatic stress causes PTSD syndrome,¹ whereas others recommend “tightening” the diagnostic criteria.² Concerns regarding PTSD diagnosis are multiple and include:

• the importance of ruling out malingering³
• the effects of different diagnostic criteria resulting in disparate prevalence rates
• emphasizing the importance of dysfunction as a criterion for PTSD.⁴

Conceptual inconsistencies in DSM-IV-TR diagnostic criteria also can lead to confusion. Although there is a category of arousal symptoms, Criterion B4 (intense psychological distress) and Criterion B5 (physiological reactivity) are listed as re-experiencing symptoms rather than arousal symptoms. Finally, the criteria presented do not follow a logical progression. Research suggests that re-experiencing symptoms do not lead to avoidance but result in arousal symptoms, which in turn trigger avoidance.⁵

Posttraumatic stress disorder (PTSD) is a confusing diagnostic category because it includes victims of trauma as well as individuals exposed to trauma. Also, PTSD encompasses exposure to different types of trauma, which can have significant implications for symptom development and treatment.

Consider the treatment history of a male combat veteran who exhibits multiple PTSD symptoms, including nightmares, flashbacks, social isolation, anger, and guilt related to his war experiences. Several psychiatrists saw the patient, which resulted in multiple medication changes but little benefit. On further assessment, the practitioners noted that the veteran’s war experiences were minimally problematic; the prominent nightmares, ruminations, flashbacks, and guilt were related to his witnessing a civilian female being sexually assaulted. The veteran’s guilt about not intervening was the basis of his PTSD. This led to a change in treatment from pharmacotherapy to a focus on supportive therapy.

Conceptualizing subtypes of PTSD—similar to many DSM-IV-TR diagnoses such as phobias or delusional disorders—might help better define the diagnosis. Each sub-type, as conceptualized below, might have its own prognosis and treatment. Our hope is that this strategy will benefit the patient by improving research and evidence-based practice.

PTSD subtypes

Victim-related trauma. Related to witnessing a criminal act or being a victim of a criminal act such as rape or assault. The patient is in a passive role.

Natural disasters, such as a tornado, earthquake, or hurricane.

Survivor guilt. The patient is not a perpetrator and might have been exposed to trauma, but symptoms are related to surviving while others close to the patient did not.

Perpetrator guilt. It is debatable whether this should be a PTSD subtype but our experience suggests that this pattern severely complicates PTSD diagnosis and treatment. It often is not initially disclosed by patients but surfaces when treatment is not working despite a strong therapeutic alliance.

PTSD not otherwise specified. This subtype is typical in patients who were not directly involved in a traumatic event but experienced symptoms related to it. Examples include picking up dead bodies, cleaning up a tornado site, or observing siblings being beaten. This category also may reflect an unclear picture if no primary subtype accounts for the majority of symptoms.

Qualifiers

Individuals who previously have been exposed to trauma are more vulnerable to subsequent trauma. Experiencing ongoing multiple traumatic events—such as in military combat—can have a cumulative effect. Thus, identifying episodes of trauma also should be part of the PTSD assessment.

Single event. The patient is exposed to a single traumatic episode, such as being the victim of a crime.

Multiple events/single episode. The patient is exposed to repeated, related traumatic events. Examples include ongoing military combat or child abuse.

Multiple events. The patient is exposed to ≥2 separate traumatic events. A combination such as this might include a serious motor vehicle accident followed by a natural disaster.

As the diagnosis of PTSD evolves, utilizing subtypes and qualifiers might clarify treatment strategies because some subtypes might be more amenable to certain psychopharmacologic or psychotherapeutic treatment regimens.

Diagnostic confusion

Some researchers question whether traumatic stress causes PTSD syndrome,¹ whereas others recommend “tightening” the diagnostic criteria.² Concerns regarding PTSD diagnosis are multiple and include:

• the importance of ruling out malingering³
• the effects of different diagnostic criteria resulting in disparate prevalence rates
• emphasizing the importance of dysfunction as a criterion for PTSD.⁴

Conceptual inconsistencies in DSM-IV-TR diagnostic criteria also can lead to confusion. Although there is a category of arousal symptoms, Criterion B4 (intense psychological distress) and Criterion B5 (physiological reactivity) are listed as re-experiencing symptoms rather than arousal symptoms. Finally, the criteria presented do not follow a logical progression. Research suggests that re-experiencing symptoms do not lead to avoidance but result in arousal symptoms, which in turn trigger avoidance.⁵