SummaryThis project investigates whether real-world, multimodal 24-hour wearable data, as well as 10-second ECG data, can be used to identify transdiagnostic autonomic nervous system (ANS) profiles that reflect shared regulatory mechanisms across mental and somatic disease burden. Using recordings from multiple waves of the NAKO cohort, physiological markers derived from ECG, respiration, and actigraphy are integrated to characterize ANS function. Data-driven ANS profiles are identified via latent profile and mixture modeling, with evaluation of model fit, reproducibility, and longitudinal stability. These profiles are then linked to psychosocial, socioeconomic, and lifestyle factors using covariate-adjusted regression models. Associations with a transdiagnostic disease burden index—integrating somatic multimorbidity, mental symptom burden, and functional health indicators—are assessed while accounting for classification uncertainty. Robustness is evaluated through sensitivity and subgroup analyses. The resulting ANS profile metrics are further used for longitudinal and prognostic analyses, including mortality as well as for mental and somatic health outcomes. The data requested here will also form the basis of a BMBFTR proposal.

Keywords ANS-Profile ECG HRV Somnowatch

InstitutionsUniversitätsmedizin Greifswald