Sleep clusters and modifiable risk behaviors for noncommunicable diseases in young adults: Data from a birth cohort in Brazil

Abstract

ObjectivesTo determine sleep clusters among young adults and describe the prevalence of modifiable risk behaviors for noncommunicable diseases (NCDs) according to these clusters. MethodsA cross-sectional analysis was carried out with the 1993 Pelotas Birth Cohort participants at the 22-year follow-up. Sleep onset, sleep offset, total sleep time (TST), TST variability, and sleep efficiency were evaluated by a triaxial accelerometer. We asked participants to wear the devices for 7 days on a nondominant wrist. Excessive daytime sleepiness was assessed by the Epworth sleep scale and sleep quality by the Pittsburgh Sleep Quality Index. Sleep clusters of 2738 individuals were determined using k-means cluster analysis. Crude and adjusted prevalence of modifiable risk behaviors for NCDs (smoking, harmful alcohol intake, leisure physical inactivity, overweight, screen time, and ultra-processed food consumption) were presented according to the sleep clusters. Adjustments included wealth index, skin color, years of schooling, current occupation, shift work, and having children under 2 years. All analyses were stratified according to sex. ResultsWe identified 3 sleep clusters for men (Healthy sleepers, Late and variant sleepers, and Shorter and poorer sleepers) and 3 for women (Healthy sleepers, Late and poor-quality sleepers, and Shorter, variant, and inefficient sleepers). Both males and females classified as Healthy sleepers presented a lower prevalence of modifiable risk behaviors for NCDs compared to individuals from other sleep clusters. ConclusionsPoor sleep health is associated with higher prevalence of modifiable risk behaviors for NCDs. Prevention strategies for NCDs should also focus on sleep health.