Onset of impaired sleep as a predictor of change in health-related behaviours; analysing observational data as a series of non-randomized pseudo-trials

Background: Changes in health-related behaviour may be a key mechanism linking impaired sleep to poor health, but evidence on this is limited. In this study, we analysed observational data to determine whether onset of impaired sleep is followed by changes in health-related behaviours. Methods: We used data from 37 508 adults from the longitudinal Finnish Public Sector Study. In analysis of 59 152 person-observations on duration and quality of sleep and health-related behaviours (alcohol consumption, smoking, physical activity and weight control), data were treated as a series of non-randomized pseudo-trials with strict pre-defined criteria for data inclusion and temporality. Results: Smokers who experienced onset of short sleep were less likely to quit smoking than those with persistent normal sleep [odds ratio (OR) = 0.78, 95% confidence interval (CI): 0.64-0.97]. Onset of short sleep also predicted initiating high-risk alcohol consumption (OR = 1.17, 95% CI: 1.00-1.37). Onset of disturbed sleep was associated with changes in all assessed health-related behaviours: initiation of high-risk alcohol consumption (OR = 1.23, 95% CI: 1.05-1.45), quitting smoking (OR = 0.80, 95% CI: 0.63-1.00), becoming physically inactive (OR = 1.17, 95% CI: 1.06-1.30) and becoming overweight or obese (OR = 1.12, 95% CI: 1.01-1.23). Conclusions: Findings suggest that the onset of short or disturbed sleep are risk factors for adverse changes in health-related behaviours. These findings highlight potential pathways linking impaired sleep to the development of lifestyle-related morbidity and mortality.