Author(s): SCHAFFARCZYK, M., RÜCKEMANN, L., SPERLICH, B., GRONWALD, T., Institution: UNIVERSITY OF WÜRZBURG, Country: GERMANY, Abstract-ID: 517

INTRODUCTION:
Cardiac autonomic regulation, commonly indexed by heart rate variability (HRV), is frequently hypothesized to fluctuate systematically across the menstrual cycle. Empirical findings, however, remain inconsistent, likely due to sparse phase-based sampling and limited consideration of interindividual variability. The aim was to examine cyclic HRV trajectories employing dense longitudinal monitoring and phase-aligned cycle time scaling (PACTS) to distinguish population-level from individual-specific patterns and to evaluate the role of daily subjective well-being.
METHODS:
Daily HRV using smartphone photoplethysmography and well-being data were collected across and up to four menstrual cycles in 21 naturally cycling women. Cycle time was aligned using PACTS, with ovulation set to 0 and follicular and luteal phases rescaled to a common −1 to +1 interval. Short-term HRV was indexed using vagally-mediated log-transformed RMSSD, and coefficient of variation (CV) of ln(RMSSD), computed from a centered 3-day rolling window. Non-linear cyclic trajectories were analyzed using cyclic generalized additive mixed models (GAMMs), separating population-level smooths from participant-specific cyclic deviations. Associations between daily well-being (sleep, stress, fatigue, pain) and HRV were examined using linear mixed-effects models distinguishing within-person from between-person effects.
RESULTS:
For ln(RMSSD) (n=1213), the population-level cyclic smooth over PACTS time was flat and non-significant (s(t_pacts): edf≈0, p=0.483), whereas participant-specific cyclic deviations were significant (ti(t_pacts, ID): edf=20.47, p=0.013), indicating pronounced interindividual heterogeneity despite a stable average trajectory (adjusted R²=0.443). Comparable results were observed for CV ln(RMSSD) (n=1110), with no significant population-level cyclic pattern (edf=0.007, p=0.067) but strong individual-specific cyclic variability (edf=54.52, p<0.001; adjusted R²=0.349). Within-person analyses showed that worse daily sleep, higher stress, and greater fatigue were associated with lower ln(RMSSD) (all p≤0.003), whereas between-person differences were not related to mean ln(RMSSD). Further, within-person increases in stress and pain were associated with modest increases in CV ln(RMSSD) (both p=0.030).
CONCLUSION:
When menstrual cycle timing is precisely aligned to ovulation, vagally-mediated HRV does not exhibit a consistent population-level rhythm. Instead, HRV dynamics are highly individual and strongly modulated by day-to-day fluctuations in subjective well-being. Previously reported inconsistencies in menstrual cycle effects on HRV may therefore reflect unmodeled interindividual variability and short-term contextual influences rather than weak or absent physiological effects.