Table 1 New goals for enhanced diagnostic categories and worked examples for major depression

1. Focus primarily on enhancing clinical practice

1. Abandon use of the single term ‘major depression’ as on its own it does not predict response to specific psychological or physical treatments [62];

2. Only use in association with specifiers that predict likely response to specific treatments - for example, major depression with psychotic features; melancholia associated with psychomotor changes [63–65]; depression following manic episode [62];

3. Differentiate other risk or comorbidity factors from the diagnosis itself - notably risk of self-harm or suicide or misuse of alcohol and other substances [66–69].

2. Link directly to objective markers of pathophysiological processes

Require the cross-sectional and longitudinal recording of objective markers that may predict response to treatment or risk of recurrence:

• Neurohormonal - for example, presence of non-suppression to dexamethasone [70];

• Circadian or sleep - for example, actigraphic evidence of phase-delay [58, 60, 71];

• Psychomotor change - for example, observer or automated measures [63–65];

• Neuropsychological - for example, neuropsychological evidence of delayed reaction time [72–75];

• Brain imaging - for example, presence of subcortical white matter changes [59, 74, 76–82].

3. Incorporate known facts about developmental paths, environmental risk factors, course of illness or family history

1. Differentiate early-onset (<30 years) from late-onset (>50 years) forms [83–87];

2. Differentiate first major episode from recurrence, relapse or chronicity [21, 22, 52, 88];

3. Record clear environmental (for example, seasonal onset, exposure to traumatic events) or medical illness (for example, post-stroke) exposures that are concurrent with depression [89];

4. Record clear earlier (notably childhood) phenotypes such as childhood anxiety;

5. Record clear family history data related to presence of psychosis, mania or suicide in first-degree relatives [35, 37, 90–95];

6. Record clear history of exposure to social adversity or interpersonal stressors or ongoing evidence of major socio-economic, interpersonal or other relevant social circumstances [96–98].

4. Be consistent with data from family, twin or genetic studies

1. Restrict the diagnosis to those sub-categories with strong evidence of high heritability - for example, depression in those with previous mania; depression in those with psychotic features [35, 37, 38, 99–101];

2. Support the concepts of depressive disorders preceded by childhood anxiety or early- versus late-onset depressive disorders [66, 84–86, 102].

5. Capture key aspects of illness stage

Use a clinical staging format for depressive disorders that differentiates early stages that are strongly linked to other childhood and adolescent phenotypes (for example, anxiety, phase-delay in sleep and circadian systems, fatigue, hypomania, mood instability) from later early adult or mid-adult stages (which may also be associated with different phenotypes such as psychomotor change, phase-advance in sleep and circadian systems) [21, 22, 58–61, 88, 102].

6. Best predict future illness course or response to specific treatments.

Use known factors about response/non-response to specific treatments - for example, for acute episode: classify as selective serotonin reuptake inhibitor responder or non-responder; classify as responder or non-responder to cognitive behavioral therapy [103–105].