Both insufficient and excessive sleep duration have been associated with lower-quality diets in adult populations. However, investigations in Latin America, where different sleep norms may exist (e.g., daily napping), are scarce. Therefore, we examined whether weekday sleep duration and inconsistencies between weekday and weekend sleep duration were related to adherence to the Mediterranean diet among Costa Rican adults. The study population included 2169 controls (74% men) from a population-based case-control study. Usual sleep duration (weekday versus weekend) was self-reported and defined as short, recommended, and long (<7 h, 7-9 h, >9 h, respectively). Inconsistent weekday-weekend sleep duration was defined as >1-h difference. Diet was assessed with a food frequency questionnaire, and adherence to the Mediterranean diet was calculated with the Alternative Mediterranean Diet Score (AMED). Sex-stratified linear regression models were conducted with AMED score as a continuous outcome and sleep variables as categorical or dichotomous exposures (in separate models). Models were adjusted for age, area of residence, education, napping, caffeine intake, smoking status, type 2 diabetes mellitus, hypertension, and physical activity. Average (SD) hours of sleep per night reported were 7.0 (1.4) on weekdays and 7.3 (1.6) on weekends for men, and 7.0 (1.5) on weekdays and 7.2 (1.6) on weekends for women. Among women, sleep duration <7-h per night was associated with a lower AMED score compared to those with recommended sleep duration (β: -0.35, CI: -0.63 to -0.07). Unstratified models showed a suggestive association between inconsistent weekday-weekend sleep and lower AMED scores that did not vary by sex (β: -0.08, CI: -0.16, 0.006; P, interaction with sex = 0.93). Lastly, the napping frequency was not associated with AMED scores in any model. In conclusion, short and inconsistent sleep duration may affect the dietary patterns of Costa Ricans.
Keywords: Fruits; Mediterranean diet; Sleep duration; Vegetables.
Copyright © 2021 Elsevier Ltd. All rights reserved.