Characterizing Nonlinear Heterogeneous Sleep Trajectories across the Menstrual Cycle in a Transdiagnostic Psychiatric...

Sleep disturbance is closely linked to affective symptoms and suicidal ideation (SI). These symptoms fluctuate across the menstrual cycle; however, menstrual cycle-related sleep variability remains poorly characterized in psychiatric populations. This study examined how subjective and objective sleep fluctuate across the menstrual cycle in a psychiatric sample, considering individual differences and associations with negative affective and SI symptom trajectories.

A transdiagnostic psychiatric sample recruited for recent suicidality provided intensive daily subjective sleep and symptom ratings (N = 142,M = 26.59,years SD = 5.34) alongside wearable-derived (Oura) objective sleep measures (N = 39) across 1-3 menstrual cycles. Cycle timing was aligned to ovulation and menses using Phase-Aligned Cycle Time Scaling. Generalized additive mixed models quantified average cyclical effects and interindividual heterogeneity. Exploratory smooth mixture modeling (SMM) identified latent subgroups with shared temporal sleep trajectories which were compared to trait-level demographic and clinical characteristics as well as affective, SI, and pain cyclical symptom trajectory subgroups.

Most sleep outcomes showed substantial within-person cyclical variability. SMM identified distinct subgroups for Oura-derived sleep efficiency, and for subjective ratings of fatigue, insomnia, sleep quality, and hypersomnia, with patterns including perimenstrual or luteal worsening, as well as no cycle-related variability. Sleep subgroups, especially for subjective ratings, frequently significantly aligned with parallel affective, pain, and SI trajectories, suggesting temporally-coupled vulnerability.

Sleep trajectories across the menstrual cycle are heterogeneous and meaningfully coupled with psychiatric symptom patterns. These findings support cycle-aware sleep monitoring and suggest that cyclical sleep disruption may help identify windows of heightened psychiatric risk, motivating replication and personalized intervention approaches.