Association between actigraphy-based sleep duration variability and cardiovascular risk factors - Results of a population-based study.

Background: Although a few studies suggest an adverse effect of sleep duration variability on cardiovascular risk factor, others did not and this association remains controversial. Moreover, most studies were non-representative of the general population, used different sleep duration variability measures, and relied on self-reported sleep duration. We aimed to assess the association between different, actigraphy-based sleep duration variability measures and cardiovascular risk factors in a population-based sample.

Methods: In a middle-aged population-based cohort, 2598 subjects had data on sleep duration variability measured by actigraph over 14 days. Multivariable logistic regressions were performed to assess the relationship between different sleep duration variability measures [ie, night-to-night variability (NNV), range between shortest and longest sleep duration (RSL), range between average weekday and weekend sleep duration (RWW)] and cardiovascular risk factors including obesity, diabetes and hypertension.

Results: Subjects with highest sleep duration variability - measured as NNV, RSL and RWW, were more likely to be obese. These associations robust in most but not all sensitivity analyses, and no associations between sleep duration variability measures and diabetes or hypertension were found.

Conclusion: There is a possible association between high sleep duration variability and obesity, although results were not robust in all sensitivity analyses. Further, no associations between sleep duration variability and other cardiovascular risk factor such as diabetes or hypertension were found.