Abstract Introduction Controlled laboratory protocols and naturalistic studies reveal a 24-hour rhythm in positive affect (PA), which reaches its lowest point in the middle of the circadian night and peak in the afternoon. “Evening-types”, who experience more depression, report PA rhythms that are delayed, blunted, and decreased. Chronotherapeutic interventions to advance circadian phase can improve mood, but the mechanisms are unclear. In a sample of adolescents with later sleep timing, we examined whether a 2-week chronotherapeutic manipulation was associated with changes in PA rhythms (i.e., shifting the phase earlier, and increasing both the amplitude and rhythm-adjusted mean). Methods 79 high schoolers (60% female, mean age = 17.5 years) who reported a weekend bedtime of 1 AM or later underwent a one-week baseline and a two-week experimental period (random assignment to manipulation or control). Participants were asked to complete the Positive and Negative Affect Schedule (PANAS-SF) 5-6 times a day. DLMO was assessed once prior to randomization and once at the end of the protocol. Participants in the manipulation were asked to maintain a consistent rise time, advance their bedtime, reduce evening light (via blue light blocking glasses), and administer morning bright light (via ReTimer glasses) for two weeks. We used a cosinor variant of multilevel modeling with three-way interactions between cosine/sine terms, study condition, and study week, while controlling for age and sex. Results Model estimates indicated the overall presence of rhythmicity in PA (p.001), significant differences in PA rhythms during both experimental weeks compared to baseline (p=.02, p.01), and significant differences in the weekly changes of PA rhythms depending on study condition (ps.01). Bootstrapped contrasts revealed that the change in acrophase from Week 1 to Week 3 was significantly different between the manipulation and control (□=0.469 or 1.79 hours, p.001). No other contrasts of change were significant. Conclusion By the final week of the manipulation, participants’ PA rhythms were significantly advanced relative to the control. Important next steps include examining how changes in PA rhythms associate with changes in sleep and circadian metrics (e.g., DLMO) and assessing whether changes in PA rhythms predict meaningful clinical outcomes (e.g., depression). Support (if any) NIDA R01DA044143, T32 HL082160
Mirchandaney et al. (Fri,) studied this question.