Bidirectional longitudinal dynamics of self-reported total sleep time and perceived stress: Establishing potential causal relationships.

Objective: Most studies in developing countries suggest that less total sleep time (TST) increases subsequent perceived stress (PS) more consistently than the inverse, but have used statistical models that are not optimal when ratings are measured in close proximity.

Methods: We used multilevel dynamic structural equation modeling with Bayesian estimation, ideal for assessing longitudinal daily dynamic interplay between self-reported TST and PS (minimum of 30 days) in 92 Brazilian pre-university students.

Results: TST showed lower inertia than PS (autoregressive effect: TST perturbations influenced subsequent TST ratings less so than corresponding PS measures) and exerted negative prospective effects on PS (cross-lagged effect).

Conclusion: In our developing nation sample, PS deviations from mean ratings took longer to return to baseline values than TST and PS was more sensitive to changes in prior TST than vice-versa, confirming previous findings. Future studies should confirm these findings with objective TST and stress measures.