Mario A. Cristancho, MD

[Editor's Note: Originally presented as an independent educational activity under the direction of CME LLC, this article was published by 

). The ability to receive CME credits has expired. The article is presented here for informational purposes.]

The existence of different subtypes of depressive episodes (ie, endogenous and nonendogenous) was initially postulated at least 80 years ago.

 In the early formulations, the term “endogenous depression” was used to describe a more severe biologically mediated illness, whereas “nonendogenous depression” or “exogenous depression” referred to a less severe and environmentally mediated condition characterized by mood reactivity.

 It was not until the introduction of the first monoamine oxidase inhibitor (MAOI)-iproniazid-that the term “atypical depression” began to emerge to describe a particular variant of nonendogenous depression.

Originally, endogenous, or melancholic, depression was thought to be the prototypical form of depression.

 Endogenous depression manifested with neurovegetative symptoms and nonreactive mood and regularly responded to the tricyclic antidepressant (TCA) imipramine and/or 

 (ECT). A different subgroup of patients was found to be responsive to iproniazid (the first commercially available MAOI, but currently off the market because of significant toxicity) and nonresponsive to well-established treatments for depression (ie, imipramine and ECT).

 Furthermore, those patients presented an unusual constellation of symptoms characterized by the absence of endogenous-type neurovegetative symptoms and the presence of emotional reactivity.

On the basis of these early observations, the existence of a unique subtype of nonendogenous depressive episodes characterized by mood reactivity with 

 neurovegetative symptoms (ie, hypersomnia and hyperphagia) and a predictable response to certain antidepressants was proposed and termed “atypical depression.” The hypothesis that such depressions were relatively nonresponsive to TCAs and more responsive to MAOIs was further supported by studies in the 1970s and 1980s.

 Atypical depression was formally recognized in 1994, when it was included as an “episode specifier” in DSM-IV.

” can be used to characterize the current or most recent depressive episode in patients with either unipolar or bipolar type mood disorder and in patients with dysthymic disorder.

, the DSM-IV specifier requires the presence of mood reactivity (criterion A) and at least 2 of 4 criterion B features (significant weight gain or hyperphagia, hypersomnia, leaden paralysis, and interpersonal rejection sensitivity resulting in social or occupational impairment). If the patient meets criteria for either melancholic or catatonic features during the same major depressive episode, a diagnosis of atypical depression cannot be made (criterion C).

To avoid overdiagnosis or underdiagnosis, bear in mind definitional aspects of the clinical features that constitute the criteria for atypical depression. Mood reactivity means that clinically depressed patients have the capacity to feel at least 50% better and even become transiently euthymic when exposed to positive events (eg, an invitation for a date, a compliment).

 Although never studied in a rigorous prospective manner, it has been reported that patients with atypical depression can remain euthymic for extended periods if the external circumstances remain positive.

 With respect to hyperphagia, a clear and sustained increase in appetite or a 5-lb weight gain during the depressive episode would satisfy criterion B1 (

Table. DSM-IV criteria for atypical depression features specifier

Hypersomnia refers to either a total of at least 10 hours of sleep per day, including nighttime sleep and daytime naps, or at least 2 hours more than when not depressed.

 Leaden paralysis is defined as a sensation of heaviness in the arms or legs as if they were made of lead; it is generally present for at least an hour a day but can last for many hours at a time.

Interpersonal rejection sensitivity in the context of atypical depression implies a lifelong trait (during both periods of depression and periods of euthymia) that is typically exacerbated during depressive episodes. It is characterized by an excessive or pathological sensitivity to rejection and/or criticism leading to functional impairment (eg, stormy relationships, inability to sustain long-term relationships, problems at work, avoidance of relationships because of fear of rejection, avoidance and fear of embarrassment) or maladaptive behavioral responses such as substance abuse.

 Rejection sensitivity seems to be the most common clinical feature in atypical depression, as demonstrated by a study of 332 patients, of whom 71% reported rejection sensitivity, 47% hyperphagia, 47% leaden paralysis, and 35% oversleeping.

Although the term “atypical” implies an unusual or uncommon condition, depression with atypical features is in fact a common clinical presentation and is one of the most common forms of depression in outpatient settings.

 Estimates in both community and clinical settings suggest that 15.7% to 36.6% of patients with depression present with atypical features.

 This estimate is in harmony with the 18% prevalence of atypical depression detected in the Sequenced Treatment Alternatives to Relieve Depression (STAR*D) study series and with the 42% prevalence in a sample of 396 patients with depression.

 The presence of atypical features is even higher (up to 50%) in patients with bipolar II disorder and dysthymic disorder.

Studies have suggested that patients with atypical depression tend to have an earlier onset of symptoms and a more chronic course than their melancholic counterparts.

 is more common in younger women. Also, rates of comorbid conditions, such as anxiety, cluster B and C personality disorders, substance abuse, and somatization disorder, seem to be higher in patients with atypical depression than in those with other forms of depression.

Validity of atypical depression and its DSM-IV criteria 

Although atypical depression appears to be clinically and diagnostically well characterized, the DSM-IV criteria for diagnosing atypical depression and its validity as a subtype of depression have been questioned recently.

 Pertinent sources of controversy include the following

• The inconsistency between reactive mood and the presence or absence of criterion B features
• Both the sex bias and the definition of rejection sensitivity
• The exclusion of state-dependent anxiety

The validity of mood reactivity as a mandatory feature for diagnosing atypical depression has been challenged. Findings from 4 studies showed that mood reactivity does not significantly correlate with the presence of criterion B features, which suggests that mood reactivity should not be considered an obligatory feature for the diagnosis of atypical depression.

 Furthermore, regarding melancholia (which requires the loss of mood reactivity) as exclusionary of the diagnosis of atypical depression subtype makes the presence of reactive mood largely redundant. The inclusion of mood reactivity as an essential feature also neglects the fact that some depressive episodes, when quite severe, manifest with a nonreactive mood, even in the presence of reversed neurovegetative symptoms. The term “anergic depression” is sometimes used to describe depressive episodes that take this form.

One established characteristic of atypical depression is its differential response to MAOIs. The correlation between the presence or absence of reactive mood and a differential response to either 

s or MAOIs has been challenged by a number of pharmacological studies.

 Findings from those studies suggest that the effectiveness of MAOIs in depression is not necessarily associated with mood reactivity, implying that the presence of this specific feature may not be essential for diagnosing this syndrome.

The hypothesis that reactive mood as a mandatory criterion is not indispensable for the diagnosis of atypical depression was supported by the community study by Angst and colleagues.

 Although mood reactivity was the most common symptom reported by their sample of patients with atypical depression (89% to 90%), other symptoms (ie, rejection sensitivity, leaden paralysis, and hypersomnia) were also quite commonly present (78% to 89%). This suggests that atypical depression could also be effectively diagnosed when mood reactivity is not considered a mandatory criterion.

 In a more recent analysis, Angst and colleagues

 could be made with equal validity if 3 of 5 criteria (including mood reactivity) or 2 of 4 criteria (excluding mood reactivity) were used.

Clearly, the inclusion of mood reactivity as a mandatory or hierarchical criterion for the diagnosis of atypical depression should be reassessed. This could be done using the available literature or, ideally, through more specific studies (ie, compare subjects with 2 or more criterion B symptoms and reactive mood with subjects with 2 or more criterion B symptoms without reactive mood).

Biological features of atypical depression 

Evidence for the validity of atypical depression as a distinct subtype of depression includes distinct 

• Abnormalities of hypothalamic-pituitary-adrenal (HPA) axis activity
• Polysomnographic findings
• Asymmetry of hemispheric processing on psychophysiological testing
• Distinct regional cerebral blood flow patterns on single photon emission CT (SPECT)

The characteristic biological profile of atypical depression includes either normal or abnormally 

; a relatively normal polysomnographic sleep profile; increased frontal, temporal, and parietal perfusion; and abnormally increased right hemispheric processing when performing psychophysiological tasks.

 Thus, the neurobiological profile of patients with atypical depression is both distinct from melancholia and, on select measures, abnormal compared with that of healthy controls. Interestingly, the most abnormal profiles appear to be evident in patients whose atypical depression has an early onset and a chronic course.

A SPECT brain perfusion study in patients with atypical depression showed increased frontal, temporal, and parietal perfusion coupled with decreased occipital perfusion relative to that seen in patients with melancholic and undifferentiated depression (ie, neither atypical nor melancholic). A SPECT study also showed increased right frontal perfusion relative to that seen in individuals in the control group.

 These specific biological characteristics imply that atypical depression is in fact a distinct entity and, as such, requires a distinct treatment approach.

 response to standard treatments is historically the strongest determinant of atypical depression as a distinct subtype; patients with atypical depression have a preferential response to MAOIs compared with TCAs and less robust therapeutic responses to ECT.

 Although some studies have failed to demonstrate the relationship between atypical features and preferential response to certain medications, the therapeutic benefit of MAOIs in atypical depression is widely accepted in practice guidelines and has been confirmed by meta-analyses.

The first reports of preferential response to MAOIs were based on uncontrolled case series; early attempts to replicate these findings in larger-scale randomized controlled trials (RCTs) met with only limited success (see Thase and colleagues

 for detailed review). Subsequent prospective trials using specific diagnostic criteria for atypical depression have generally supported the superiority of MAOIs-particularly phenelzine-over TCAs. For example, 

 randomized 119 patients with atypical depression to phenelzine, imipramine, or placebo. Improvement in both active groups was significantly superior to that in the placebo group; response rates were 71%, 50%, and 28% for phenelzine, imipramine, and placebo, respectively. Although the 21% difference in response rate between the phenelzine and imipramine groups was not statistically significant, the overall trend favored the MAOI.

Similar findings were obtained by Quitkin and colleagues

 in an RCT of 90 outpatients with atypical depression. They observed response rates of 83%, 50%, and 19% for phenelzine, imipramine, and placebo, respectively. In this study, the difference in response rate between phenelzine and imipramine was both clinically and statistically significant (

The utility of phenelzine was also supported in a 2-stage, double-blind study of patients with atypical depression. During the first stage, response rates were significantly higher for phenelzine at both 6- and 12-week assessments (63% at 6 weeks and 51% at 12 weeks) than for imipramine (35% at 6 weeks and 24% at 12 weeks).

 During the second phase, nonresponders to imipramine were switched to phenelzine; a response rate of 67% was observed. Only 41% of the phenelzine nonresponders responded to imipramine.

Interestingly, another placebo-controlled RCT conducted by the same group of investigators compared phenelzine and imipramine in outpatients with a depressive episode characterized by reactive mood but none of the other criteria for atypical depression.

 In this trial, response rates to the MAOI and to the TCA were essentially identical (75% for phenelzine and 74% for imipramine); both drugs were superior to placebo. When taken together, these findings suggest that mood reactivity per se is not related to drug response, but rather it is the presence of criterion B features that predicts response to MAOIs.

It is also worth noting that results of a re-analysis suggest that the overall advantage of the MAOI over the TCA in the aforementioned studies may well be the result of the relative inferiority of imipramine in patients with early onset of illness (ie, 

 years) and chronic atypical depression (ie, duration of at least 2 years with well periods of no longer than 2 months).

 In fact, among the patients with a later illness onset and a less chronic course, the imipramine response rate was about 65% (approximating that with phenelzine). Although 4 different MAOIs are licensed by the FDA for the treatment of major depression (ie, tranylcypromine, phenelzine, isocarboxazid, and selegiline transdermal delivery system), only the label for phenelzine includes a clear indication for atypical features.

Although MAOIs are clearly effective in patients with atypical depression, they are not considered first-line choices because of the required dietary restrictions and potential adverse effects.

 In light of this, researchers have switched their focus to the potential role of non-MAOI/non-TCA medications in the treatment of atypical depression.

In the first study to directly compare an SSRI with an MAOI, Pande and colleagues

 randomized 40 patients with atypical depression to a 6-week course of either fluoxetine (20 to 60 mg/d) or phenelzine (15 to 90 mg/d). Response rates (defined as a decrease of 50% or more in 

 for Depression-17 [HAMD-17] or a Clinical Global Impressions Improvement [CGI-I] score of 1 or 2) for fluoxetine and phenelzine were reported as 80% to 85% and 85%, respectively.

 Remission rates (HAMD-17 less than 5 and a CGI-I score of 1 or 2) were 80% for fluoxetine and 70% for phenelzine. No statistically significant differences were found between the groups, but the frequency of adverse effects was lower with fluoxetine.

 This study suggests that fluoxetine may be as effective as phenelzine in the treatment of atypical depression, with the SSRI having a better tolerability profile.

Further attempts to establish the utility of SSRIs in the treatment of atypical depression include trials of SSRIs (fluoxetine and sertraline) versus the reversible MAOI drug moclobemide.

 In the first trial, moclobemide (an investigational drug that is not commercially available in the United States) at a median dose of 379 mg displayed superiority as evidenced by response rates ranging from 71% to 91% (HAMD and CGI-I) over the 60% to 65% for fluoxetine at a median dose of 27 mg in the sample of 40 patients with atypical depression.

 In the second trial, there was a trend toward superiority for sertraline at 50 to 100 mg/d over moclobemide at 300 to 450 mg/d, as reflected by remission rates (CGI-I = 1) of 77.5% and 67.5% for sertraline and moclobemide, respectively, after 12 weeks of treatment (

Interpretation of these studies is limited by the absence of a placebo arm and also by concerns about the limited clinical efficacy of moclobemide at the dosages studied (see, for example, Lotufo-Neto and colleagues

). This work, nevertheless, provides some further support for the clinical impression that the SSRIs are more useful for treatment of patients with atypical depression than the TCAs.

The efficacy of fluoxetine was also demonstrated by McGrath and colleagues

 in their 20-week, double-blind, placebo-controlled study of atypical depression (N = 154), which used imipramine and placebo control groups. A significant advantage for both active groups compared with the placebo group was reported; however, no significant difference was found between the SSRI and TCA groups. Both active groups achieved comparable response rates (ie, 51% for fluoxetine and 53% for imipramine); fluoxetine had a better tolerability profile.

 Although both response rates are lower than the ones that have been reported for studies conducted by this group with phenelzine, they are still clinically significant and support the usefulness of SSRIs in the treatment of atypical depression.

The irreversible MAOI selegiline, which is selective for monoamine oxidase B at low oral doses, and 

 (CBT) both appear to be promising treatment options for patients with atypical depression. Initial studies of selegiline in atypical depression date back to 1984 and suggested a beneficial role with response rates of 59% at doses greater than 20 mg/d and up to 40 mg/d.

 Later, a placebo-controlled trial of selegiline (N = 98) also showed positive results, as evidenced by a response rate of 50% in the selegiline arm compared with 28% in the placebo arm.

 It is noteworthy that the doses of selegiline used in these studies were high enough to inhibit monoamine oxidase A. However, to date, no studies of atypical depression have been completed using the transdermal formulation of selegiline.

A 10-week, double-blind, placebo-controlled study found efficacy for CBT in the acute-phase treatment of atypical depression.

 A total of 108 subjects were randomly assigned to twice-weekly CBT, phenelzine, or placebo. A significant reduction in HAMD scores was reported with both active interventions. Response rates were 58% for both active modalities compared with 28% for placebo. There was no statistically significant difference between the active groups.

 in the subset of patients with atypical depression in secondary analyses of the NIMH Treatment of Depression Collaborative Research Program, which also studied interpersonal psychotherapy and clinical management with double-blind placebo or imipramine.

Identification of atypical features is important in the treatment of depression for both treatment selection and prognosis, especially when initial measures prove ineffective. The concept of atypical depression has evolved over many years, and now it appears timely for a further revision. Our review of the current literature suggests a need for optimizing the precision and reliability of the current DSM criteria for atypical features. This could be achieved by the elimination of reactive mood as mandatory for diagnosis, by a more specific definition of criterion B features, and by the inclusion of chronicity and early age at onset as criteria.

Findings from the literature on the treatment of atypical depression show that phenelzine, and by implication the MAOIs as an antidepressant class, is the most effective pharmacological agent for atypical depression. Imipramine, and by implication the TCAs, is not as effective but still represents a treatment option for patients with atypical depression, particularly those with later onset of illness and less chronic courses. Given that neither MAOIs nor TCAs are widely prescribed now, other important treatment options include the SSRIs, notably fluoxetine and sertraline. Unfortunately, other antidepressants that may be useful in atypical depression, including bupropion and venlafaxine, have not been systematically studied. One form of depression-specific psychotherapy, CBT, has also been found to be efficacious and should be included in the treatment plan for atypical depression.

There is still a clear need for medication trials to better characterize the initial steps (before consideration of MAOIs or TCAs) in the treatment algorithm for atypical depression. Although historically 

 has been thought to be ineffective for atypical depression, recent reports suggest that it, too, may be effective in well-selected cases.

 The role of newer neuromodulation techniques, such as transcranial magnetic stimulation and, for patients with more advanced degrees of treatment resistance, vagus nerve stimulation and deep brain stimulation, still needs to be explored.

 Gillespie RD. The clinical differentiation of types of depression. 

Kiloh LG, Garside RF. Depression: a multivariate study of Sir Aubrey Lewis’s data on melancholia. 

Thase ME. Atypical depression: useful concept, but it’s time to revise the DSM-IV criteria. 

 Thase ME. Recognition and diagnosis of atypical depression.

West ED, Dally PJ. Effect of iproniazid in depressive syndromes. 

Robinson DS, Nies A, Ravaris CL, Lamborn KR. The monoamine oxidase inhibitor, phenelzine, in the treatment of depressive-anxiety states. A controlled clinical trial. 

 Quitkin F, Rifkin A, Klein DF. Monoamine oxidase inhibitors. A review of antidepressant effectiveness. 

Davidson JR, Miller RD, Turnbull CD, Sullivan JL. Atypical depression. 

Quitkin FM, Stewart JW, McGrath PJ, et al. Phenelzine versus imipramine in the treatment of probable atypical depression: defining syndrome boundaries of selective MAOI responders. 

Diagnostic and Statistical Manual of Mental Disorders. 

4th ed. Washington, DC: American Psychiatric Association; 1994.


Nierenberg AA, Alpert JE, Pava J, et al. Course and treatment of atypical depression.

 Quitkin FM. Depression with atypical features: diagnostic validity, prevalence, and treatment. 

 McGrath PJ, Stewart JW, Harrison WM, et al. Predictive value of symptoms of atypical depression for differential drug treatment outcome. 

 Kendler KS, Eaves LJ, Walters EE, et al. The identification and validation of distinct depressive syndromes in a population-based sample of female twins. 

Sullivan PF, Kessler RC, Kendler KS. Latent class analysis of lifetime depressive symptoms in the national comorbidity survey. 

 Horwath E, Johnson J, Weissman MM, Hornig CD. The validity of major depression with atypical features based on a community study. 

 Levitan RD, Lesage A, Parikh SV, et al. Reversed neurovegetative symptoms of depression: a community study of Ontario.

Asnis GM, McGinn LK, Sanderson WC. Atypical depression: clinical aspects and noradrenergic function. 

Robertson HA, Lam RW, Stewart JN, et al. Atypical depressive symptoms and clusters in unipolar and bipolar depression. 

 Zisook S, Shuchter SR, Gallagher T, Sledge P. Atypical depression in an outpatient psychiatric population. 

 Angst J, Gamma A, Sellaro R, et al. Toward validation of atypical depression in the community: results of the Zurich cohort study.


 Novick JS, Stewart JW, Wisniewski SR, et al; STAR*D Investigators. Clinical and demographic features of atypical depression in outpatients with major depressive disorder: preliminary findings from STARD*D. 

. Benazzi F. Testing DSM-IV definition of atypical depression. 

Angst J, Gamma A, Benazzi F, et al. Atypical depressive syndromes in varying definitions.

Perugi G, Akiskal HS, Lattanzi L, et al. The high prevalence of “soft” bipolar (II) features in atypical depression. 

Articles

Identification of atypical features is important in the treatment of depression for both treatment selection and prognosis, especially when initial measures prove ineffective. The concept of atypical depression has evolved over many years, and now it appears timely for a further revision.

Mario A. Cristancho, MD

Articles

Atypical Depression in the 21st Century: Diagnostic and Treatment Issues