BACKGROUND: While commonly accepted depressive subtypes reflect phenotypic differences, there has been minimal progress in identifying discrete pathophysiological pathways, biomarkers or differential therapeutic approaches which effectively guide clinical management.
AIMS: To test the biological validity and clinical utility of a circadian subtype of depression on the basis of clinical course, differential medication response (self-reported) and genetic risk profile.
METHODS: Cross-sectional data were drawn from the nationwide, genetically-informative Australian Genetics of Depression Study. Participants were classified as having a 'circadian' versus 'non-circadian' subtype of depression on the basis of meeting criteria for at least three binary circadian features: social jetlag, seasonality, delayed sleep midpoint, evening chronotype, sleep inertia, and hypersomnia. Clinical course characteristics were compared. Associations with response to commonly prescribed antidepressants and polygenic risk scores (PGS) for mental disorders and sleep, circadian, metabolic and inflammatory traits, were investigated using logistic regression models.
RESULTS: 2,604 participants (23%; 80% females; mean age=37.87, sd=13.62) had a circadian subtype. These cases reported an earlier age of onset (p<0.001), more severe clinical features including hypo/manic-like and psychotic-like experiences, suicidality, psychological distress and somatic complaints (p's<0.001), weight gain during depressive episodes (p<0.001), poorer response to SSRIs (OR=0.88 [0.82, 0.94]) and SNRIs (OR=0.89 [0.83, 0.97]) and more side-effects, compared to those with a non-circadian subtype. Having a circadian subtype was associated with higher PGS for attention-deficit/hyperactivity disorder (OR=1.11 [1.06, 1.17]), major depression (OR=1.11 [1.06, 1.16]), bipolar disorder (OR=1.09 [1.04, 1.14]), body mass index (OR=1.09 [1.05, 1.14]), triglycerides (OR=1.10 [1.06, 1.16]), interleukin-6 (OR=1.08 [1.03, 1.13]), higher insulin resistance (OR=1.08 [1.04, 1.13]), later sleep midpoint (OR=1.15 [1.10, 1.21]), insomnia (OR=1.08 [1.03, 1.13]), and later chronotype (OR=0.68 [0.65, 0.71]).
CONCLUSION: These findings support the face validity and potential clinical utility of circadian subtype of depression as a clinical profile. Pending independent replication, investigation of its biology and predictive utility are warranted.
