Diagnosis, Classification, and Differential Diagnosis of the Mood Disorders

The diagnosis, classification, and differential diagnosis of the mood disorders. 

Topics covered:

  • Introduction
  • Definitions
    • Depression
    • Mania/hypomania ‘
    • Mixed states'
    • Bipolar/unipolar depression specifics
    • Bipolar categories
  • Depressive disorders: one, two, or three principal types?
  • Classification of affective mood disorders
    • Formal classification
    • Clinical classification—based on a hierarchical or tiered model
    • Step 1: Is a depressive disorder present?
    • Step 2: If a depressive disorder is present, what subtype?
    • Step 3: If a non-melancholic depressive disorder, what clinical subgroup?
    • Step 4: Distinguishing a melancholic subtype
    • Step 5: Distinguishing psychotic depression
  • Differential diagnosis and ascertainment difficulties
    • Depression
    • Mania/hypomania
  • Conclusion
  • References


Varying expressions and subtypes of depressed and elated mood states make for difficulties in definition, diagnosis, and classification. Some definitional and boundary issues will first be detailed, to assist later consideration of subtyping and differential diagnostic issues.



The term ‘depression' can variably define an affect, mood state, a disorder or syndrome, or a specific entity. A depressed ‘affect' usually occurs in response to a specific situation and is best defined as a relatively transient state of feeling ‘depressed', ‘sad', or ‘blue'.

A depressed ‘mood' is more pervasive, more likely to be experienced as unusual or atypical, associated with negative ideas (e.g. hopelessness, helplessness, pessimism about the future, and feeling like giving up), and may influence behaviour. The quintessential construct defining a depressed mood is lowering of the individual's intrinsic self-esteem and, as a corollary, self-criticism. The extent to which self-esteem is lowered usually defines the severity of the mood disturbance. Depressed mood states are experienced by most people, although in non-clinical subjects they generally last only minutes to days.

A depressive ‘condition' (be it a disorder, syndrome, or specific disease entity) is generally distinguished by a longer duration, by more or a greater number of clinical features, and by distinct social impairment. A duration criterion ensures that a transient depressed mood (even if severe) does not alone establish psychiatric ‘case' status, with diagnostic criteria sets generally requiring a minimum duration of 2 weeks for conditions other than the so-called ‘adjustment disorders'. The addition of several clinical features (detailed shortly) informs us about severity (e.g. ‘major' and ‘minor' depressive disorders) and subtyping, while the social impairment criterion assists cleavage between ‘normal' mood states and psychiatric ‘case' status.

At times, depressive conditions are described as ‘primary' or ‘secondary'. Any such distinction is necessarily imprecise. Thus, we concede ‘secondary depression' when depression emerges during the course of a substantive psychiatric condition (e.g. schizophrenia) or medical condition, or following certain aetiologically defined triggers (e.g. organic states; substance abuse, including alcoholism). As depression is commonly associated with certain other psychiatric disorders (e.g. severe anxiety states, personality disorders), it might be logical to also call these ‘secondary' depressive disorders. While some authors (1) use the term in this way, formal diagnostic systems operate to the restricted definition.


As described in the preceding article, such conditions are the converse of depression and fundamentally represent hedonistic states. Here self-esteem is almost invariably increased, the mood is generally infectious, the individual feels energized or ‘wired' (ideas and thoughts race from topic to topic), creativity and religiosity are often enhanced, there is a tendency to spend (or wish to spend) more money, while any psychotic features are generally in synchrony with the mood state.

There is no formal distinction between ‘hypomania' and ‘mania', and the lack of an agreed definition is regrettable. To some theorists, the presence of psychotic features determines manic (as against hypomanic) status. Goodwin and Jamison (2) suggest that hypomania and mania differ little in mood components, but that cognition, perception, and behaviour differ in severity and manifestation. Both DSM-IV ( 3) and ICD-10 (4) disallow a diagnosis of hypomania if psychotic features are present but, conversely, do not require psychotic features for a diagnosis of mania. ICD-10 views hypomania as ‘an intermediate state without delusions, hallucinations, or complete disruption of normal activities'. DSM-IV lists essentially similar clinical criteria for hypomanic and manic episodes, but distinguishes the two by the presence of marked impairment in social functioning, requirement for hospitalization, or the presence of psychotic features in a manic episode.

‘Mixed states'

Here the individual with the affective disorder may show depressive features during a manic episode or manic features during a depressive episode. While the term is sometimes used to describe the transition from one polar mood disturbance to another, the more common use is for a genuine coterminous presence of manic and depressive features.

Bipolar/unipolar depression specifics

Turning from cross-sectional to longitudinal definition, the course specifier ‘bipolar' has long been applied to describe patients who have had at least one manic or hypomanic episode, whether preceded by any depressive episode or not. Originally, Leonhard (5) introduced the concept of ‘monopolar' (or ‘unipolar') depression to distinguish those who had had episodes of the melancholic subtype of depression, but no manic episode. Regrettably, the term ‘unipolar' depression is now used to define a residual (i.e. non-bipolar) category, so heterogeneous as to be of limited utility for aetiological and treatment studies.

Bipolar categories

In recent years, bipolar disorder has been subcategorized into bipolar I and bipolar II expressions. The DSM-IV definition effectively requires an initial or previousmanic episode for the former, while bipolar II disorder requires hypomanic episodes and one or more previous episode of major depression. ICD-10 catalogues a bipolar II disorder (without definition) but does not list or describe a bipolar I disorder.

Depressive disorders: one, two, or three principal types?

The extended debate as to whether the depressive disorders are best conceptualized as comprising one or more distinct disorders warrants overview. The ‘unitarian' view posits one depressive disorder, varying essentially by severity. The strict ‘binarian' view argues for two separate types.

Debate did not commence this century. Altschule (6) suggested that St Paul distinguished between two types of depression, one ‘from God' and the other ‘of the world' (Corinthians 7:10), and that several distinguished Christian leaders from the fifth to the eighth century recognized a distinction between ‘rational' and ‘irrational' depression. Jackson (7) has extensively detailed historic descriptions of melancholic depression from early Roman and Greek views through to the twentieth century, charting the key defining clinical features, with an early emphasis on observable signs—particularly of psychomotor disturbance. Similarly, Berrios (8) informs us that, in classical antiquity ‘melancholia was defined in terms of overt behavioural features ... [and] that symptoms ... were not part of the concept'.

Rich descriptions of melancholia as a depressive condition marked by psychomotor retardation or agitation emerged in the late nineteenth century (e.g. Maudsley (9)), while Kraepelin (10) brought mania and melancholia together as manic–depressive insanity. Kraepelin defined a ‘melancholia simplex' condition characterized by a key feature of psychic inhibition, causing paralysis of thought, memory difficulties, and a sense of weariness and enervation, but without psychotic features. He also defined a second melancholic group with delusions and hallucinations in addition to psychomotor retardation or stupor, as well as a third group of ‘agitated melancholia' marked by distinct motor agitation. In 1917, Freud (11) published his influential paper on melancholia, and Jackson argued that this changed the descriptive focus of melancholia in the twentieth century, with theorists and analysts then focusing on cognitive and intrapsychic features. Subsequently, and continuing still, clinical description of depression weighted symptoms, and minimized or ignored behavioural signs.

In the 1920s, Mapother (12) argued that it was pointless to distinguish between ‘psychotic' and ‘neurotic' forms of depression, as both lay along a continuum. This provocative paper, presenting the unitarian position, challenged diagnostic practice. The ‘binarians' responded, with Gillespie (13) arguing for two types of depression: first, an ‘autonomous' type (previously and subsequently termed ‘endogenous depression'), which, once precipitated, tended to run an independent course and be unresponsive to the environment; second, a ‘reactive' type, in which the depression was responsive to the environment (although earlier, the concept of ‘reactivity' had been tied to onset induced by a psychogenic factor).

The debate was strongly influenced by Lewis's study of 61 patients. Lewis (14) concluded that he could find no clear demarcation between depressive types, examined both cross-sectionally and longitudinally, thus delivering support to the unitarian view. This view prevailed until the introduction of multivariate statistical approaches and computers led to the debate being reactivated in the 1960s, with the so-called Newcastle School arguing strongly that their analyses supported a binary view. In a representative paper, Kiloh and Garside (15) used a factor-analytic strategy to argue for separate ‘endogenous' and ‘neurotic' depressive conditions. Factor analysis is not ideal for developing a typology, in that it produces dimensions (here symptoms) rather than groupings of patients. Subsequently, more appropriate strategies have been used, such as cluster analysis (16) and latent class analysis, (17) and with those studies providing support for separate classes. Critics suggest, however, that such classes or subgroups could still be determined by severity or, even if subtypes can be identified, question whether subclassification is of any moment.

This latter challenge is fundamental, taking us to the heart of any consideration of the diagnosis and classification of the depressive disorders. To the unitarians (despite evidence of quite varying aetiological factors), depression essentially varies only by degree, allowing treatment decisions (e.g. electroconvulsive therapy (ECT), antidepressant drugs, psychotherapy, or cognitive–behavioural therapy) to be decided on the basis of severity or frequency of episode. The opposing argument—for conceding subtypes—has been well put by Kendell, (18) who drew on important historical analogies: distinguishing between cardiac and renal forms of ‘dropsy' allowed prediction of those who would respond to digitalis, and it was only when ‘the pox' had been identified as comprising smallpox and chickenpox that it was possible to predict who was likely to live or die.

Thus, if there are valid depressive subtypes, the contribution of putative psychosocial and biological risk factors may vary across each, so that the subtypes may have distinctly differing neurobiological determinants and may have differential response to the broad treatment modalities. If this is true, then to force homogeneity by creating dimensionally based categories such as ‘major depression' is to ensure muddied results. This is not merely a theoretical objection, when, as noted by Hickie,(19) large numbers of studies of patients with DSM-defined ‘major depression' have failed to demonstrate any coherent pattern of neurobiological changes, replicate key biological correlates, and demonstrate any specific pattern of treatment response outside inpatient treatment settings.

How then have the official classificatory systems addressed such a substantive issue? In developing the DSM-III system, (20) the working group was required to make a decision on the competing unitarian or binarian models. While the binarians were at the door, they had, until then, failed to prove their case and the DSM-III committee chose a compromise. Thus, classification was predicated on an initial dimensional component (i.e. major versus minor disorders). If criteria for a major disorder were met, second-order and more categorical decisions about the presence of melancholia or psychotic depression were specified. This model proved unsatisfactory for melancholia. For example, Zimmerman and colleagues (21) noted that the DSM-III melancholia criteria set, unlike the definition provided in the predecessor (DSM-II), ‘did not predict treatment response'. Thus, in DSM-IIIR (22) the criteria set for melancholia was revised to include complete recovery after previous episodes, previous good response to somatic treatments, and no significant personality disturbance, to overcome the lack of predictive validity by building into the definition some of the ‘givens' held by many clinicians about melancholia. However, the criteria set for melancholia developed for DSM-IV returned essentially to the DSM-III set, with limitations which are considered below. The contrasting system, ICD-10, is essentially based on a dimensional or unitarian view of the depressive disorders.

Thus, there has been an extended debate as to whether a categorical and more ‘biological' type of depression exists. There have been many ascriptions to this condition, variably termed ‘endogenous', ‘endogenomorphic', ‘autonomous', and ‘melancholic' depression. Definitions over time include a distinctive pattern of symptoms and signs, the greater relevance of genetic and other biological determinants, as contrasted with psychosocial precipitants (so leading to the term ‘endogenous depression'), minimal response to placebo, and a selective response to antidepressant medication and ECT. There is evidence to support each of these propositions, (23) although the term ‘endogenous' has proved unsatisfactory, as those with melancholia may commonly report antecedent life events. However, only the first issue falls within this article's purview, and will be addressed shortly.

Whether psychotic (or delusional) depression is a ‘severe' form of melancholia or a separate entity also remains problematic. DSM-III had a category ‘major depression with psychotic features' for use when delusions or hallucinations were present or when there was ‘depressive stupor (the individual is mute and unresponsive)', thus viewing ‘psychotic depression' as a subtype of the generic ‘major depression' category rather than a subtype of melancholia. The practical advantage to that definition was in recognizing that psychomotor disturbance can be so severe that some patients will not volunteer or admit to psychotic features—which may only be confirmed by the patient after improvement. While ‘depressive stupor' may then be a useful marker or proxy for the condition, this criterion was not retained in DSM-IIIR or DSM-IV, but is included in ICD-10. Two points argue for psychotic depression as a distinct entity: the presence of psychotic features, and its poor response to antidepressant medication alone or neuroleptic medication alone.(24)

A strict interpretation of the ‘binary' view would place the non-psychotic and non-melancholic depressive conditions in a pure second class. Variably termed ‘neurotic' or ‘reactive' depression over time, this class is best regarded as a heterogeneous residue category (i.e. non-melancholic depression), with its heterogeneity expressed widely—across aetiological factors, clinical expression, and natural and treated history.

It is suggested here that there are three relatively separate depressive classes: psychotic, melancholic, and the non-melancholic disorders. Clinical differentiation of each will be described shortly.

Classification of affective mood disorders

Formal classification

For the depressive disorders, both ICD-10 and DSM-IV have multiple conditions and specifiers. The ICD-10 system allows mild and moderate depressive episodes (with or without a ‘somatic syndrome' conceptualized as reflecting ‘melancholic' features), and severe depressive episode (with or without psychotic symptoms). Thereare separate codes for a similar set of ‘recurrent' disorders, while several ‘persistent' mood disorders (including cyclothymia and dysthymia) and residual conditions are listed. DSM-IV has two principal ‘stem' disorders (major depressive episode and dysthymia), with the first having a number of optional specifiers including ‘with' melancholic, catatonic, psychotic, or atypical features, as well as including disorders showing longitudinal patterns of rapid cycling or a seasonal pattern. Both systems have categories for affective disorders secondary to organic disease, while DSM-IV includes mood disorders due to a general medical condition or substance use, or occurring in the post-partum period. Both classificatory systems include adjustment disorders with depression.

Both ICD-10 and DSM-IV have course specifiers for bipolar disorder containing 10 and four subgroups, respectively. In addition to the number of subgroups, differences include a greater emphasis on distinguishing bipolar I and II in DSM-IV, and cyclothymia being listed as a ‘bipolar disorder' in DSM-IV as against being a ‘persistent' mood disorder overlapping with a personality style in ICD-10.

For episode disorders, both systems list mania and hypomania, with ICD-10 favouring descriptive diagnostic criteria and DSM-IV specifying duration and required criterion numbers. ICD-10 allows manic subtypes with and without psychotic features, while DSM-IV puts little emphasis on psychotic features in its criterion set, although both favour a duration of at least a week. For hypomania, both DSM-IV and ICD-10 exclude those with psychotic features, and impose a minimum duration (4 days for DSM-IV and ‘several days' for ICD-10). DSM-IV requires the presence of at least three specific features, while ICD-10 requires mood and behavioural features to be intermediate between cyclothymia and mania.

Formal classifications are therefore built principally on severity, features of current episode, patterns of disorder expression over time, as well as persistence and recurrence. Few diagnoses are consistent across the ICD-10 and DSM-IV systems and, while each provides definitions that allow a ‘shared language' to be used by clinicians and researchers, the extent to which their severity-weighted groupings capture ‘meaningful' depressive subtypes remains problematic.

Clinical classification—based on a hierarchical or tiered model

This model (see Fig. 1) argues for a shared mood disorder component across three principal subtypes (i.e. from the low-order heterogeneous ‘non-melancholic' to the higher order ‘melancholic' and ‘psychotic' depressive subtypes), and a rationale for the model will be detailed. The first procedural step for the clinician, however, is to determine whether the disorder meets ‘caseness' criteria.

Fig. 1 Hierarchical model for the three principal depressive subtypes comprising an obligatory mood disorder component (albeit varying in severity) and two subtype-specific components of psychomotor disturbance and psychotic features.

Step 1: Is a depressive disorder present?

For all the depressive disorders, the first building block requires evidence of a depressed mood. Useful questions include the following: ‘Do you feel depressed?', ‘Has there been any change in your self-esteem/the extent to which you generally value yourself?', and ‘Are you being more self-critical or harder on yourself than usual?'

The next clinical priority is to determine if the depression is sufficiently severe as to warrant ‘case' status, and here the DSM-IV criteria for a major depressive episode have common acceptance. That criteria set lists the following:

  • four mood items (depressed mood, loss of interest or pleasure, feelings of worthlessness or inappropriate guilt, recurrent thoughts of death and suicidal ideation)
  • weight change
  • sleep disturbance
  • fatigue
  • impaired concentration
  • psychomotor disturbance

A positive diagnosis requires five or more of the nine, evidence of functional impairment, and a minimum duration of 2 weeks.

Step 2: If a depressive disorder is present, what subtype?

If ‘caseness' criteria are met, the next decision should be to determine the diagnostic subtype. A hierarchical (three-tiered) model is described. As additions to the shared low-order mood construct, evidence of higher-order clinical features (i.e. observable psychomotor disturbance and psychotic features) should be sought to assist definition of melancholic and psychotic depressive subtypes.

Step 3: If a non-melancholic depressive disorder, what clinical subgroup?

There is no generally accepted subtyping system for this essentially residual group (if so defined after excluding psychotic and melancholic depression), and where symptoms reflecting the lower-order mood construct dominate the clinical picture. Historically, terms such as ‘neurotic depression' and ‘reactive depression' were used, with the former emphasizing a premorbid style of neuroticism and high anxiety, and the latter defining depression developing largely in response to a severe life-event stressor.

Such terminology has not persisted in DSM-IV or ICD-10 where, as described, subtyping proceeds largely on a severity dimension (e.g. ‘major' and ‘minor') but also on patterns of recurrence and persistence.

A broader clinical approach, albeit proceeding beyond symptoms alone, may have utility here, as various multivariate analytic strategies have identified two non-melancholic subtypes with quite striking consistency. Thus, an early factor-analytic study (25) suggested both a ‘hostile' type (evidenced by irritability as well as anxiety) and an ‘anxious–tense' type. Blashfield and Morey (26) reviewed 11 cluster analytic studies suggesting separate ‘hostile' and ‘anxious' depressive subgroups. In an extensive review of the then published studies, Roth and Barnes (27) suggested three principal subgroups, with depression associated with a personality disorder, in addition to ‘hostile' and ‘anxious' depression.

While such ‘hostile' and ‘anxious' subgroups have been identified for a lengthy period, clear and consistent descriptions are lacking. Grinker et al.(28) described those with ‘hostile' depression as unappreciative, actively angry, provocative, and making excessive demands of and complaints about their therapists. Rosenthal and Gudeman (29) described such group members as having a self-pitying constellation, together with external blame, hypochondriasis, anxiety, demanding and complaining behaviours, irritability, and hostility. They, like Paykel, (16) noted that such patients tended to be younger. The second (‘anxiety') subgroup is variably interpreted as defining either those with an anxious personality or temperament, or the presence of significant coterminous anxiety symptoms when primarily depressed.

It is probably most useful to view these two suggested subgroups as ‘spectrum disorders', with the term ‘spectrum' capturing a continuum between temperament/personality style and symptom states,(30) a model with the temperament component indirectly supported by a recent factor analysis of DSM personality disorders identifying two basic dimensions (i.e. ‘hostile' and ‘anxious') interpersonal attitudes. (31) The spectrum model allows the possibility that certain biological factorsmay influence both temperament and personality style, as well as shape surface marker symptoms (here ‘hostility' and ‘anxiety').

A recent research report (32) suggests that those who develop an ‘irritable/hostile' depression tend to be more likely to have a cluster B personality and to report ‘acting out' behaviours when stressed (e.g. breaking things, storming around and yelling, being reckless, self-injuring, or otherwise demonstrating a ‘short fuse' response to stress). This subgroup was young, so that their ‘hostility' may have been a reflection of age. A percentage of the subgroup also had significant anxiety and it may be that, for these, ‘irritability' is an externalization of significant anxiety.

By contrast, those with an ‘anxious depressive' spectrum disorder appeared more likely (32) to internalize anxiety. They tended to have shown shyness and behavioural inhibition in childhood, to have high rates of lifetime anxiety disorders, score high on trait anxiety, view themselves as ‘worriers', ‘nervy', or ‘tense', and to rate as having a cluster C personality style. They developed their first episode of depression at a young age, had frequent depressive episodes and their depression tended to be highly persistent. They also had a seemingly high rate of becoming dependent on anxiolytic drugs and alcohol, as well as being highly likely to self-injure and attempt suicide. When stressed, they were somewhat more likely to ‘act in' by becoming quiet, retiring to their room, crying, and ‘stewing'. The clinical picture of their depression was noteworthy for the high frequency of anxiety symptoms. Thus, anxiety was evident both in the temperament pattern and in the prominent symptom profile when depressed.

The suggested profile of these spectrum disorders is not only important for clinical consideration, but in facilitating research into possible neurobiological determinants and to consider any treatment specificity. There have been few studies that have examined the last issue. Blashfield and Morey (33) concluded, from their review, that ‘anxious depressives respond well to major and minor tranquillizers but not to tricyclics, while hostile depressives show little improvement with conventional drug therapies', while Fava et al.(34) reported that anxious depressives were more likely to be non-responders to a selective serotonin reuptake inhibitor ( SSRI) than other depressive subtypes (including ‘hostile depressives'). However, as there is currently no formal classification of these suggested groups, research examining for any such treatment specificity or differentiation is rare and limited. Certainly, there is a strong clinical impression that those with an anxious worrying temperament and a non-melancholic depression do well with an SSRI.

‘Reactive depression' (or as used in DSM-IV and ICD-10, ‘adjustment disorder') is surprisingly difficult to confirm empirically as an ‘entity'. This is not to deny that a significant percentage of depressed patients present with a history of having decompensated only as a consequence of a major and stressful life event. Various interpretations have been reviewed, (32) and range from antecedent life events producing, triggering, and maintaining depressive episodes through to life events being epiphenomenal. In an empirical study, (35) the authors were unable to establish clinical, family history, and even life-event stress differences between those with ‘situational' and ‘non-situational' major depression. In an accompanying editorial, Glass (36) concluded that ‘the presence or absence of a precipitant does not seem to be useful for subdividing groups of patients with major depression' and questioned the validity of the intuitively appealing concept of ‘situational depression'.

Conflicting evidence of ‘reactive depression' may reflect sampling issues. In the community, ‘depression' is a common transient response to life-event stress. Thus, most ‘normal' people presumably possess mechanisms promoting spontaneous remission of depression within a few days. In clinical samples, where a duration criterion (usually of 2 weeks) is imposed, the utility of the construct is clearly less apparent, and raises the question as to why most patients in clinical samples fail to have a spontaneous remission. A likely explanation is that a significant number have perpetuating factors (e.g. prominent premorbid anxiety) with such factors (here ‘worrying' about life-event stressors) tending to maintain the non-melancholic depression. Thus, the life events here act only as provoking factors, seemingly requiring interaction with predispositional or other vulnerability factors.

Thus, while diagnoses of ‘reactive depression' and ‘adjustment disorder' appear to have clinical validity, they should not be accepted at face value, and should more encourage clinical consideration of predisposing vulnerabilities. The clinician may still seek to assist the patient to resolve or to come to terms with the life-event stressors, but, equally importantly, should seek to identify any predisposing variables that might also benefit from clinical intervention. Common scenarios are the presence of significant anxiety and personality disorder which, if treated or modified, may not only efficiently help resolution of the depression but reduce ongoing vulnerability to further depressive episodes.

Step 4: Distinguishing a melancholic subtype

In DSM-IV, the ‘melancholic features specifier' requires, in addition to a base diagnosis of major depression, either one of two A criteria and three (or more) of six B criteria, with most of the latter comprising the so-called ‘endogeneity symptoms' (it is therefore another hierarchical model). However, the DSM criteria for melancholia have a number of limitations. Firstly, criterion A (loss of pleasure and/or lack of mood reactivity) is met by most clinically depressed patients, whether ‘melancholic' or ‘non-melancholic'. Secondly, some of the criterion B features are vague. One, ‘distinct quality' is defined as a mood different to that experienced after ‘the death of a loved one'. As such, it is a negative definition (akin to defining ‘psychiatry' as ‘not cardiology'). A second, ‘excessive or inappropriate guilt', is a concatenated descriptor, able to subsume the excessive expression of normal guilt, as well as guilt held at an overvalued or delusional level. Several others are non-specific (e.g. early morning wakening, significant anorexia or weight loss), in that they occur commonly in other psychiatric conditions (e.g. anxiety disorders) as well as in other expressions of depression.

Nelson and Charney (37) undertook a review of 33 studies using several multivariate statistical approaches to identify melancholic clinical features or ‘endogeneity symptoms'. They found no support for appetite/weight loss and insomnia, little support for early morning wakening and ‘distinct quality', but some support for a severely depressed and non-reactive mood, loss of interest in pleasurable activities (or anhedonia), and psychotic features. The most strongly associated feature was psychomotor change (with retardation more consistently associated than agitation). When Rush and Weissenburger (38) examined nine diagnostic systems for diagnosing melancholia or endogenous depression, the only common criterion in all nine was psychomotor retardation (with psychomotor agitation included in six). Our research(39) has established that the specificity of psychomotor disturbance to ‘melancholia' is dependent on measuring it as a sign. If assessed as a ‘symptom', most depressed patients, independent of depressive type, will admit to feeling ‘slowed down' (or, less commonly, agitated), so that its subtyping potential is diffused or lost. If assessed as a sign, its specificity (i.e. presence in melancholic and absence in non-melancholic depression) is impressive.

Thus, and returning to a hierarchical model, differentiation between the non-melancholic and melancholic disorders (on the basis of clinical features) appears assisted principally by the specific feature of behaviourally evident psychomotor disturbance. As measured by the sign-based CORE system, (39) psychomotor disturbance is reflected along three dimensions, essentially assessing motor retardation and agitation, as well as cognitive processing difficulties, although components are not mutually exclusive. For example, those with significant agitation may have it present for much of the time or, and more commonly, have a base of retardation with intermittent epochs of agitation.

The mood disorder tends to be more severe in melancholic than in non-melancholic depression—either intrinsically or because some of the so-called ‘endogeneity symptoms' (e.g. non-reactive mood, anhedonia, diurnal variation in mood and energy) are more severe or more likely to be present in melancholia, but such a difference is dimensional and does not, of itself, allow clear distinction of melancholic and non-melancholic depression.

Step 5: Distinguishing psychotic depression

Proceeding to psychotic (or delusional) depression, here the central mood component generally appears to the clinical observer to be even more severe than in melancholic depression but such patients not uncommonly deny or minimize a depressed mood. A number of ‘endogeneity symptoms' are also frequently more severe (particularly non-reactive mood, anhedonia, insomnia, and constipation). One frequent feature in melancholic depression (i.e. diurnal variation of mood) is, however, rarely present at episode nadir in psychotic depression, as the patient is more likely to be consistently depressed across the days. Observable psychomotor disturbance is present and generally markedly more severe than in melancholic depression. In some, the combination of the cognitive processing problems and motor change (retardation in particular) can give the impression of a dementing process, effectively an example of ‘pseudodementia'.

The key specific feature, however, that distinguishes psychotic depression from the melancholic and non-melancholic disorders is the presence of psychotic features. Delusions are almost invariably present while hallucinations (auditory most commonly) are present in 10 to 20 per cent according to representative studies. DSM-IV subdivides delusions and hallucinations as ‘mood congruent' (where themes of guilt, disease, death, nihilism, and personal inadequacy dominate) and ‘mood-incongruent' (where psychotic features appear independent of the depressive theme and might include persecutory themes, delusions of control, as well as thought insertion and thought broadcasting) states. It is important to emphasize that mood-incongruent states are common and do not, by themselves, challenge a diagnosis of delusional depression.

A significant percentage (approximately one-third) of patients with psychotic depression develop constipation. In a significant percentage it appears to be a primary feature (not merely a consequence of psychotropic medication), and was so described in historical descriptions (7) of severe depression over the centuries. One term was ‘costiveness', with ‘costive' being defined as ‘slow in action or in expressing ideas, opinions' and derived from the Latin word constipatus. Constipation may serve as a nidus of delusional interpretation (e.g. the depressed patient who believes that their bowels have turned to concrete, or that they have a bowel cancer).

When psychomotor disturbance is extremely severe, it may not be possible for psychotic features to be elicited, particularly if the patient is mute. Many patients are diffident to reveal psychotic material, and here indirect questions can often be useful. In particular, pursuing the presence of ‘guilt' can assist, with guilt here defined as a sense of self-blame and not merely self-criticism, together with a sense of remorse for wrong acts or omissions that are independent of any concern about potential evaluation by others. While a high percentage of depressed patients experience guilt, for most it is explainable by circumstances (e.g. the patient concerned by their depression preventing them from optimally looking after their family). In psychotic depression, the guilt is more likely to be held at the level of an overvalued idea or at a formally delusional level. If direct inquiry about guilt does not elicit delusional material, then asking the question ‘Do you, in any way, feel any sense that you deserve to be punished?' can often be helpful in eliciting previously unexpressed psychotic material.

Differential diagnosis and ascertainment difficulties


The three key features of lowered self-esteem, increased self-criticism, and acknowledgment of a depressed mood distinguish depression phenomenologically from states such as grief or bereavement—where there is a distinct sense of ‘loss', but no primary ‘loss' of self-esteem. They also assist phenomenological differentiation from anxiety, where the individual is more likely to report a sense of insecurity, fear, apprehension, worry, panic, or of ‘going mad'. In practice, such features, together with less specific concomitants such as crying, non-reactive mood, and anhedonia, assist in making a diagnosis of depression in the majority of instances.

At the practical level, such a definitional approach may fail in certain groups. The first group comprises some patients with psychotic or melancholic depression, as a percentage appear more emotionally blunted and ‘flat' rather than depressed. They may deny depression, self-blame. and worthlessness, and instead note a lack of feeling (or vitality), sluggishness, enervation, or emphasize a physical state, termed ‘corporization' by Schneider, (40) with reports of pain or physical sensations in the head, chest, or stomach. As noted earlier, others may have such profound psychomotor disturbance that they do not respond to questioning and appear as if they have a dementia (here a so-called ‘pseudodementia'). In such instances, pursuit of proxy items (severe psychomotor disturbance, pathological guilt, overvalued ideas) may assist. If unsuccessful, the diagnosis may require corroborative reports and clinical observation, as well as certain investigations (e.g. CT scans, single-photon emission CT scans, EEG) to exclude a dementia. In a percentage of the elderly, however, a depressive episode and dementia may coexist, and reflect a shared aetiological process.

Secondly, there are some patients and certain cultures where psychological issues are either denied or expressed somatically, although careful and directed questioning about central depressive descriptors will usually clarify the possibility. If unsuccessful, again the diagnosis may require corroborative reports and clinical observation.

Thirdly, it is commonly difficult to define ‘clinical depression' in those with a medical illness. Here, general depression criteria sets risk false-positive diagnoses by including certain features which may be secondary to the medical illness (e.g. fatigue, insomnia, anorexia) rather than reflecting depression per se. Common corrective strategies include the ‘aetiological approach', where only symptoms viewed as independent of the medical condition are counted, the ‘exclusive approach', eliminating potentially confounded items such as anorexia, the ‘inclusive approach', where all symptoms are counted irrespective of their origin, and the ‘substitutive approach', excluding features that could be due to the medical illness and substituting features such as social withdrawal and crying.

In some cases where depression has been established, the salient clinical difficulty is in determining whether depression is or is not the primary disorder in those with concomitant major medical problems, excessive alcohol intake, organic central nervous system disease, and certain psychiatric conditions (e.g. anxiety states, depressive personality disorder). Clinical judgement is generally required with two alternate logical approaches: either weighting the disorders hierarchically or sequentially. The hierarchical approach assumes that the more severe disorder is the primary one, while the sequential approach weights the antecedent condition (e.g. organic disorder, schizophrenia, anxiety). Acceptance of one approach does not logically bind the clinician to any therapeutic consequence (such as necessarily treating only the more severe or primary condition).


The differential diagnoses for manic states essentially include other psychotic conditions (e.g. schizophrenia, drug-induced psychosis) and, rarely, a primary organic state. While cross-sectional dissection of the phenomenology can be helpful, there is wisdom in also weighting the longitudinal course. Thus, those with manic states are more likely to describe complete restoration of function between episodes (of mania and/or depression), while this is less likely for those with schizophrenia. Definitive distinction is not always possible, and a diagnosis of ‘schizoaffective' disorder may then be appropriate. In severe mania, an ‘organic' picture may be suggested, and require exclusion of a dementia or delirium.

The differential diagnosis of hypomanic states is often quite difficult. Questioned about having ‘highs', a number of depressed people will present a remission to a euthymic state as a ‘high'. Highly creative people may affirm many hypomanic descriptors when possessed by the muse (e.g. less need for sleep, feeling creative and overconfident, being enthused and energized), as may those with a distinctly extroverted or cyclothymic personality when stimulated. Some patients with a cluster B personality style (especially of the borderline type) may also describe mood states that approximate to hypomania. Regrettably, current criteria sets appear to result in a percentage of patients with a primary personality disorder or even a (healthy) cyclothymic personality style being diagnosed as having a hypomanic (or bipolar II) disorder, risking inappropriate treatment. (41) Clarification is probably best assisted by interview of a corroborative witness to determine if hypomanic behaviours are observable.


Current formal classificatory systems allow a large number of depressive subtypes, with criteria designed to assist diagnostic reliability. Most reflect attempts to create classes on the basis of severity-weighted dimensional models. Terms such as ‘major depression' and ‘unipolar depression' have achieved acceptance in recent years, for such diagnoses are easily made (and easily reified), but the limitations inherent to their heterogeneity should not be ignored. Until the dissonance between the formal classifications and clinician-derived models has been resolved, practitioners should proceed by recognizing the advantages and limitations to competing approaches.


1.Kessler, R.C., Nelson, C.B., McGonagle, K.A., Liu, J., Swartz, M., and Blazer, D.G. (1996). Co-morbidity of DSM-III-R major depressive disorder in the general population: results from the US National Comorbidity Survey. British Journal of Psychiatry, 168(Supplement 30), 17–30.

2.Goodwin, F.K. and Jamison, K.R. (1990). Manic-depressive illness. Oxford University Press, New York.

3.American Psychiatric Association (1994). Diagnostic and statistical manual of mental disorders(4th edn). American Psychiatric Association, Washington, DC.

4.World Health Organization (1992). The ICD-10 classification of mental and behavioural disorders—clinical descriptions and diagnostic guidelines. World Health Organization, Geneva.

5.Leonhard, K. (1957). The classification of endogenous psychoses(transl. R. Berman). Irvington, New York.

6.Altschule, M.D. (1967) The two kinds of depression according to St. Paul. British Journal of Psychiatry, 113, 779–80.

7.Jackson, S.W. (1986). Melancholia and depression: from Hippocratic times to modern times. Yale University Press, New Haven, CT.

8.Berrios, G.E. (1988). Melancholia and depression during the 19th century. British Journal of Psychiatry, 153, 298–304.

9.Maudsley, H. (1985). The pathology of the mind: a study of its distempers, deformities and disorders. Macmillan, London.

10.Kraepelin, E. (1921). Manic depressive insanity and paranoia. Livingstone, Edinburgh.

11.Freud, S. (1917). Mourning and melancholia. In Standard edition of the complete psychological works of Sigmund Freud, Vol. 14 (ed. J. Strachey), pp. 243–60. Hogarth Press, London, 1957.

12.Mapother, E. (1926). Manic depressive psychosis. British Medical Journal, 2, 872–7.

13.Gillespie, R.D. (1929). The clinical differentiation of types of depression. Guy's Hospital Report, 2, 306–44.

14.Lewis, A. (1934). Melancholia: a clinical survey of depressive states. Journal of Mental Science, 80, 277–378.

15.Kiloh, L.G. and Garside, R.F. (1963). The independence of neurotic depression and endogenous depression. British Journal of Psychiatry, 109, 451–63.

16.Paykel, E.S. (1971). Classification of depressed patients: a cluster analysis derived grouping. British Journal of Psychiatry, 118, 275–88.

17.Parker, G., Hadzi-Pavlovic, P., Mitchell, P., et al.(1994). Defining melancholia: properties of a refined sign-based measure. British Journal of Psychiatry, 164, 316–26.

18.Kendell, R.E. (1989). Clinical validity. In The validity of psychiatric diagnosis(ed. L.N. Robins and J.E. Barrett), pp. 305–23. Raven Press, New York.

19.Hickie, I. (1996). Issues in classification: III. Utilising behavioural constructs in melancholia research. In Melancholia: a disorder of movement and mood. A phenomenological and neurobiological review(ed. G. Parker and D. Hadzi-Pavlovic), pp. 38–56. Cambridge University Press, New York.

20.American Psychiatric Association (1980). Diagnostic and statistical manual of mental disorders(3rd edn). American Psychiatric Association, Washington, DC.

21.Zimmerman, M., Black, D.W., and Coryell, W. (1989). Diagnostic criteria for melancholia. The comparative validity of DSM-III and DSM-III-R. Archives of General Psychiatry, 46, 361–8.

22.American Psychiatric Association (1987). Diagnostic and statistical manual of mental disorders(3rd edn). American Psychiatric Association, Washington, DC.

23.Rush, A.J. and Weissenburger, J.E. (1995). Melancholic symptom features: a review and options for DSM-IV. In DSM-IV source book, Vol. 2. American Psychiatric Press, Washington, DC.

24.Spiker, D.G., Weiss J.C., Dealy, R.S., et al.(1985). The pharmacological treatment of delusional depression. American Journal of Psychiatry, 142, 430–5.

25.Overall, J.E., Hollister, C.E., Johnson, M., and Pennington, J. (1966). Nosology of depression and differential response to drugs. Journal of the American Medical Association, 195, 946–8.

26.Blashfield, R.K. and Morey, L.C. (1979). The classification of depression through cluster analysis. Comprehensive Psychiatry, 20, 516–27.

27.Roth, M. and Barnes, T.R.E. (1981). The classification of affective disorders: a synthesis of old and new concepts. Comprehensive Psychiatry, 22, 54–77.

28.Grinker, R.R., Miller, J., Sabshin, M., Nunn, R., and Nunnally, J.C. (1961). The phenomena of depressions. Harper and Row, New York.

29.Rosenthal, S.H. and Gudeman, J.E. (1967). The endogenous depressive pattern. An empirical investigation. Archives of General Psychiatry, 16, 241–9.

30.Cassano, G.B., Michelini, S., Shear, M.K., Coli, E., Maser, J.D., and Frank, E. (1997). The panic–agoraphobic spectrum: a descriptive approach to the assessment and treatment of subtle symptoms. American Journal of Psychiatry, 154(Supplement), 27–38.

31.Schotte, C.K.W., de Doncker, D., Vankerckhoven, C., Vertommen, H., and Cosyns, P. (1998). Self-report assessment of the DSM-IV personality disorders. Measurement of trait and distress characteristics: the ADP-IV. Psychological Medicine, 28, 1179–88.

32.Parker, G., Hadzi-Pavlovic, D., Roussos, J., et al.(1998). Non-melancholic depression: the contribution of personality, anxiety and life events to subclassification. Psychological Medicine, 28, 1209–19.

33.Blashfield, R.K. and Morey, L.C. (1979). The classification of depression through cluster analysis. Comprehensive Psychiatry, 20, 516–27.

34.Fava, M., Uebelacker, L.A., Alpert, J.E., Nierenberg, A.A., Pava, J.A., and Rosenbaum, J.F. (1997). Major depressive subtypes and treatment response. Biological Psychiatry, 42, 568–76.

35.Hirschfeld, R.M.A., Klerman, G.L., Andreasen, N.C., Clayton, P.J., and Keller, M.B. (1985). Situational major depressive disorder. Archives of General Psychiatry, 42, 1109–14.

36.Glass, R.M. (1985). Situational and neurotic-reactive depression. Archives of General Psychiatry, 42, 1126–7.

37.Nelson, J.C. and Charney, D.S. (1981). The symptoms of major depressive illness. American Journal of Psychiatry, 138, 1–13.

38.Rush, A.J. and Weissenburger, J.E. (1994). Melancholic symptom features and DSM-IV. American Journal of Psychiatry, 151, 489–98.

39.Parker, G. and Hadzi-Pavlovic, D. (1996). Melancholia: a disorder of movement and mood. A phenomenological and neurobiological review. Cambridge University Press, New York.

40.Schneider, K. (1920). The stratification of emotional life and the structure of the depressive states. Zentralblat für Neurologie, Psychiatrie 59, 281.

41.Bolton, S. and Gunderson, J.G. (1996). Distinguishing borderline personality disorder from bipolar disorder: differential diagnosis and implications. American Journal of Psychiatry 153, 1202–7.