Heterogeneity in the predictors of long and short sleepers in the ELSA-Brasil study.

OBJECTIVES

The potential reasons why long sleepers are at increased risk, including cardiovascular events and mortality, are intriguing. This study aimed to explore independent variables related to long (primary aim) and short sleep duration (SDUR) (secondary aim) using objective measurements in a large sample of adults.

METHODS

Participants from the ELSA-Brasil study performed a standard clinical evaluation, sleep questionnaires, a home sleep monitoring to determine obstructive sleep apnea (OSA; apnea-hypopnea index ≥15 events/hour), and a 7-day actigraphy to measure SDUR. A multinomial regression analysis was used to identify the predictors of long (≥8 hours) and short (<6 hours) SDUR relative to the reference group (6 to <8 hours).

RESULTS

A total of 2062 participants were included in the analysis (age: 49 ± 8 years; 57.3% women). A third of them had OSA. The mean SDUR was 6.55 ± 0.49 hours. Long and short SDUR were observed in 6.5% and 26.9%, respectively. In the multinomial logistic regression analysis, older individuals (OR = 1.36; 95% CI 1.07-1.75) and depression disorder (OR = 2.59; 95% CI 1.26-5.3) were independently associated with long SDUR. In contrast, the independent variables associated with short SDUR were male sex (OR = 1.9; 95% CI 1.51-2.4), Black participants (OR = 1.75; 95% CI 1.29-2.36), excessive daytime sleepiness (OR = 1.49; 95% CI 1.20-1.86), and OSA (OR = 1.48; 95% CI 1.16-1.88).

CONCLUSIONS

In a large multiethnic cohort, we observed significant heterogeneity in the long and short SDUR predictors. Important characteristics such as age and depressive disorder might provide tips for the curious higher cardiovascular risk associated with long SDUR.