Abstract Introduction Identifying biological risk-markers for bipolar disorder (BD) remains challenging. Meta-analyses suggest individuals at-risk for BD exhibit significant variability in daily routines and irregular sleep patterns, potentially indicating circadian rhythm dysregulation. We investigated whether biological circadian measures predicted symptoms of mania over time in young people at clinical high-risk for BD. Methods In an ongoing study, 92 participants aged 16-24 (M= 21.9, SD=2.11) were recruited across a spectrum of lifetime subthreshold mania symptoms (Mood Spectrum Measure-Lifetime; MOODS-SR-L). At baseline, salivary dim-light melatonin onset (DLMO) assessed circadian phase on weekday (Thursday) and weekend (Sunday) nights; measures of circadian phase (average of weekday and weekend DLMO), phase instability (weekend-weekday DLMO difference) were derived. The post-illumination pupil response (PIPR) measured melanopsin-driven light responsivity by comparing blue to red 200 ms light stimuli at 6sec (PIPR6) and 10–40sec (PIPR30) after stimulus onset. Mania (Altman Self-Rating Mania Scale; ASRM) and depressive (Patient Health Questionnaire-9; PHQ-9) symptoms were evaluated monthly over up to 3-years follow-up. As high and/or variable levels of manic symptoms over time are associated with BD risk, mixed effect models assessed average levels of mood symptoms over follow-up and generalized linear mixed effect models with a dispersion parameter examined mood symptom variation over follow-up. All models varied for age, sex, psychiatric medication use, family bipolar history, retinal irradiance, and photoperiod. Results Greater circadian phase instability was associated with higher average mania symptoms over follow-up (βinstability=0.50, p=0.022). Both greater circadian phase instability (βinstability=0.05, p 0.001) and higher melanopsin responsivity (βPIPR6=2.38, p 0.001) were associated with more variability in mania symptoms over time. Additionally, later circadian phase (βphase=0.05, p 0.001) and phase instability (βinstability=0.10, p 0.001) were associated with more variability in depression symptoms over time. Conclusion Our interim results indicate that elevated melanopsin responsivity may be a promising biological marker of mania risk, whereas later circadian phase specifically relates to worsening depression over time. Circadian phase instability predicted longitudinal mood lability more broadly. These data provide insight into potential mechanisms underpinning the mood stabilizing effects of social rhythms therapy and dark therapy for BD; future experimental or intervention studies should evaluate causal relationships. Support (if any)
Keller et al. (Fri,) studied this question.