Modeling circadian misalignment and neuro-metabolic function using a synthetic human chronobiology cohort
Dinesh Reddy SagamObjectives
Circadian misalignment has been linked to cognitive and metabolic disturbances, but controlled large-scale human chronobiology studies remain difficult because of environmental heterogeneity, adherence issues, and the cost of laboratory protocols. This study aimed to develop a reproducible synthetic human chronobiology cohort for hypothesis generation and model testing.
Material and Methods
A synthetic cohort of 5000 simulated individuals was generated using a Python-based simulation framework grounded in published chronobiology literature. Each simulated individual was assigned age, sex, body mass index, chronotype, preferred mid-sleep timing, actual mid-sleep timing, circadian misalignment, sleep duration, night-time light exposure, clock-gene amplitude, cognitive score, mood score, and metabolic risk. Variable distributions, correlation structure, and noise terms were specified a priori to generate biologically plausible but non-empirical profiles.
Results
Within the prespecified model, increasing circadian misalignment was associated with lower modeled cognitive score. Lower clock-gene amplitude was associated with higher modeled metabolic risk. Evening chronotypes showed a right-shifted metabolic risk distribution compared with morning and intermediate chronotypes. These outputs reflect internal model behavior under predefined assumptions rather than observed associations in human participants.
Conclusion
This synthetic chronobiology framework offers a transparent and reproducible platform for hypothesis generation, sensitivity testing, and future validation against real-world chronobiology datasets.