Is atypical depression simply a typical depression with unusual symptoms?

The term ‘atypical depression’ (AD) was first used by West and Dally in 1959 to describe depressed patients who failed to respond to tricyclic antidepressants (TCAs) but responded to the monoamine oxidase inhibitor (MAOI), iproniazid. Subsequently, Klein and Davis described a condition they labelled hysteroid dysphoria consisting of a subgroup of depressed female patients who demonstrated ‘an extreme response to admiration, approval and personal rejection’ and who also had symptoms of hypersomnia and hyperphagia (for a review of the history, see Davidson, 2007; Parker et al., 2002). Much later, it was recognised that these patients were further characterised by the symptoms of mood reactivity, weight gain, reverse diurnal mood variation and profound energy loss – captured in Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) as ‘feeling heavy, leaden, or weighted down, usually in the arms or legs; this is generally present for at least an hour a day but often lasts for many hours at a time’ and described as ‘leaden paralysis’. These clinical features have come to be known as ‘atypical’, contrasting them from the ‘typical’ features of appetite loss, weight loss, decreased sleep and morning worsening over-represented in melancholic depression. ‘Rejection sensitivity’ was subsequently included as a clinical feature of atypical depression. DSM-5 classifies atypical features as a specifier for depressive disorder, with the diagnostic criteria primarily of mood reactivity with two or more of the following: significant weight gain or increased appetite, hypersomnia, leaden paralysis and a long-standing pattern of interpersonal rejection sensitivity that results in significant social or occupational impairment. But these criteria have been questioned (Parker et al., 2002), with particular reference to the primacy given to mood reactivity rather than the characteristic personality style of interpersonal/rejection sensitivity. Mood reactivity was found to correlate poorly with hyperphagia, hypersomnia, leaden paralysis and long-standing interpersonal sensitivity, with the low internal consistency of features arguing against a clear ‘syndromal’ pattern. Furthermore, the symptom of ‘leaden paralysis’ is difficult to define clinically and is seldom differentiated from the loss of energy and lethargy observed with other forms of depression. In addition to such challenges regarding the symptoms that characterise AD, there is ongoing debate over whether AD should continue to be seen as a specifier or whether it qualifies as a distinct subtype of depression with a characteristic biological profile.

Alongside the DSM-5 diagnostic criteria for AD, research has identified a number of associated clinical characteristics (Pae et al., 2009). For example, women are affected two-to-three times more than men; onset of AD is early, usually in adolescence; and it typically follows a chronic, relapsing course. The response to first-line antidepressants is generally disappointing, and there is often an adverse developmental history characterised by childhood trauma and/or neglect (Davidson, 2007), with subsequent development of personality vulnerabilities and difficulties in relationship to self and others. In practice, the feature of mood reactivity may lead observers to a mistaken impression that these patients are not really depressed because they seem to readily ‘brighten up’.

Common comorbidities include anxiety disorders (panic disorder, generalised anxiety disorder, specific phobias), drug and alcohol abuse, dependent/avoidant personality, chronic fatigue syndrome, fibromyalgia and obesity. It is important to emphasise that any syndrome of atypical depression needs to be differentiated from the simple presence of ‘atypical features’ (essentially hypersomnia and hyperphagia) which are not uncommon in those with bipolar depression (especially younger patients) and those with seasonal affective disorder.

In patients presenting with depression, AD is common, with epidemiological studies reporting that 18–36% of patients with major depressive disorder have atypical depressive features and that almost all (87.9%) are women (Pae et al., 2009). Given the comparatively poorer response to antidepressants, it could be expected that the prevalence of AD in treatment-resistant populations would be higher especially since evidence from longitudinal studies (Rodgers et al., 2014) suggests that the features of atypical depression generally remain stable over time.

The proposition that AD is a distinct biological subtype of a depressive illness remains controversial and unproven – despite years of investigation. Representative studies across phenomenology, neuroendocrinology, genetics, psychology and treatment responsivity domains have attempted to characterise AD with variable results.

Clinically, as has been noted, it has been argued that a lack of correlation between the diagnostic criteria of atypical depression (Parker et al., 2002) suggests that it does not qualify as a syndrome and that the features of hypersomnia and hyperphagia are more compensatory mechanisms that are engaged by depressed patients who have a particular personality style rather than ‘symptoms’ and that the primary feature is a personality style of interpersonal sensitivity (to both praise and rejection). This frames atypical depression as a non-melancholic syndrome determined primarily by personality factors in keeping with earlier conceptualisations, such as ‘neurotic depression’. This psychologically weighted model is supported by emerging evidence that patients with atypical depression often have adverse developmental histories and relationship difficulties as well as other features of personality dysfunction, such as avoidance and dependence, and anxious-ambivalent attachment.

Examining neuroendocrine aspects, a hypoactive hypothalamic–pituitary–adrenal (HPA) axis in atypical depression, compared to melancholic depression, has been a relatively consistent finding, although it is unclear whether there is a difference between atypical depressed subjects and healthy controls. Compared to melancholic depression, AD subjects have lower serum cortisol and corticotropin-releasing hormone (CRH) levels, and compared to healthy controls, AD patients have lower CRH levels (Pae et al., 2009). These findings are in keeping with the clinical observation that melancholic patients tend to be hyper-aroused compared to AD patients who are usually sluggish and hypo-aroused. In other areas, while melancholic depression is associated with more HPA axis disturbance than atypical depression, metabolic and inflammatory dysfunction has been identified more so in atypical depression, although it is possible that these are simply a consequence of obesity (Gold and Chrousos, 2013).

The most persuasive support for atypical depression being biologically different from other forms of depression emerged from the early clinical observations that atypical depression was seemingly more responsive to MAOIs than to TCAs. Indeed, the demonstration of the superiority of phenelzine over imipramine in several randomised controlled trials conducted at Columbia University was largely responsible for the inclusion in Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) of atypical depression as a separate entity. Four randomised placebo controlled trials comparing phenelzine with imipramine and placebo showed the superiority of phenelzine over imipramine. However, when subjected to meta-analysis, the effect size was shown to be small, suggesting that in clinical practice the superiority of phenelzine may not be as significant as suggested by the clinical trials (Pae et al., 2009). Moreover, clinical trials support the efficacy of selective serotonin reuptake inhibitors (SSRIs) in treating atypical depression, while randomised trials have failed to demonstrate any difference in efficacy between SSRIs, TCAs and MAOIs. The purported superior efficacy of MAOIs in treating atypical depression may therefore be purely of historical relevance and may not provide sufficient support for the classification of atypical depression as a separate disorder. However, it has to be noted that the quality of evidence is poor either way, and therefore, further studies to clarify this issue are warranted.

What then is the clinician to make of this confusing mix of historical observation, empirical data and expert opinion? Clearly, many questions remain to be answered and warrant careful research to further our understanding of atypical depressive features (see Box 1).

Box 1. Questions to guide future research

• 1.Should the DSM-5 diagnostic criteria be accepted or should the primacy it gives to mood reactivity be replaced by interpersonal sensitivity (or rejection sensitivity), which may be more informative of the personality vulnerabilities of patients and may correlate more closely to the secondary criteria?

• 2.Is the predominance of hypersomnia and hyperphagia atypical features among women, bipolar depression and seasonal affective disorder a reflection of an underlying homeostatic process to settle emotional dysregulation?

• 3.Is the term ‘leaden paralysis’ meaningful? Many clinicians would have difficulty in defining it clinically and differentiating it from the energy loss seen with most forms of depression, so there is an argument for deleting it from the diagnostic criteria.

• 4.Does the recognition of atypical features have any relevance to management? Specifically, if monoamine oxidase inhibitors are no more efficacious than SSRIs or CBT, so do they justify their use given the risks involved?

DSM-5: Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition; SSRIs: serotonin reuptake inhibitors; CBT: cognitive behavioural therapy.

Notwithstanding these remaining unknowns, most clinicians would recognise that the clustering of some atypical symptoms is a common feature of depressive presentations in clinical practice. Hyperphagia, weight gain and hypersomnia are the symptoms that are most likely to be immediately evident, and with careful review, other features of atypical depression may also be identified. The archetypal patient is a depressed young female, often overweight and affectively reactive, who may not even appear depressed until her underlying despair is explored. At this point, it becomes evident and there may also be a history of failure to respond to first-line antidepressants.

Recognising atypical depression is important as there are particular management issues that need to be considered. Careful and thorough assessment is required, including a comprehensive developmental history, examining for early trauma, abuse or neglect. Interpersonal sensitivity and its effect on self-esteem and forming and maintaining relationships should be explored. The patient’s ability to appear at times to not be depressed may lead to underestimating the degree of psychosocial and occupational impairment and the risk of suicide, so these issues need careful evaluation. Comorbid anxiety and substance abuse are common and need to be elicited. Bipolar depression is commonly associated with atypical features, so patients should be screened for a history of mania or hypomania. It is particularly important to carefully look for bipolar II disorder as this can easily be missed. Likewise, screening for seasonal affective disorder should be routine. Medical evaluation should include screening for hypothyroidism, polycystic ovarian syndrome and, if relevant, obesity-related metabolic disturbances, such as diabetes and hypertension.

The treatment of atypical depression, like typical depression, involves both psychological treatment and pharmacotherapy. But surprisingly little research has been done to evaluate the efficacy of these various treatments – especially psychotherapeutic interventions.

The limited available evidence concerning antidepressants suggests that they can be effective in a proportion of patients, although clinically, antidepressant non-response is a common outcome – suggesting that psychological determinants play a pivotal role and arguing for prioritising psychological therapies.

In sum, although there is evidence for AD to be considered as a distinct subtype of depression (Gold and Chrousos, 2013), at present it is insufficient to justify its classification as a separate disorder. Therefore, it should remain a specifier. However, the criteria need to be modified (see Box 2), and in essence, there are only three features that appear to provide some specificity. Among these, interpersonal sensitivity should replace rejection sensitivity. In contrast, leaden paralysis is rarely observed clinically and is poorly defined, and so we suggest that it no longer be considered as a defining feature of AD. However, ultimately, despite these many descriptive difficulties, the features of AD are sufficiently common for clinicians to be aware of this ‘type’ of atypical presentation that may represent psychological complexities and comorbidities that affect treatment. And perhaps, until more is known, AD should be regarded as typical depression that presents with unusual symptoms.

Box 2. Recommendations

We propose amending the DSM-5 criteria for the atypical features specifier to better reflect what is observed in clinical practice.

Set of symptoms that characterise atypical features:

• A.A long-standing pattern of interpersonal sensitivity (not limited to episodes of mood disturbance) that results in significant social or occupational impairment

• B.Mood reactivity

• C.Hyperphagia

• D.Hypersomnia

Symptom to be removed:

Leaden paralysis (because of limited clinical utility)

DSM-5: Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition