SLEEP · 2018 · 75 citations · 35 references
MedicineEndocrinologyInternal MedicineObesity PreventionSleep MedicineWomen's HealthSleep AssessmentSleep DisturbanceSleep HealthSleep PhysiologySleep-related Breathing DisordersSleep-related FactorsCross-sectional EpidemiologyMenopausal StatusObjective Sleep AssessmentObjective Sleep MeasurementOxygen Desaturation TestingSleepSleep Disturbance SeveritySleep Quality Assessment
Study Objectives: The individual prevalence of sleep-disordered breathing (SDB), short sleep duration, and obesity is high and increasing. The study aimed to investigate potential associations between SDB, objective sleep duration, obesity, diabetes and hypertension across genders, and the effect of pre- or post-menopausal status. Methods: A cross-sectional study evaluated 7051 community participants with wrist actigraphy for a week, and nocturnal oximetry ≥ 2 nights. SDB was assessed by 3 per cent oxygen desaturation index (ODI) corrected for sleep duration obtained from wrist actigraphy. Moderate-to-severe SDB was defined as ODI3% levels ≥ 15 per hour. Results: Both logODI3% and body mass index showed independent negative associations with sleep duration (β = -0.16, p < 0.001 and β = -0.07, p < 0.001, respectively). Moderate-to-severe SDB (men/premenopausal women/postmenopausal women; 23.7/1.5/9.5%, respectively) was associated with a higher risk of diabetes in premenopausal women (OR 28.1; 95%CI 6.35-124.6; p < 0.001) and postmenopausal women (OR 3.25; 95%CI 1.94-5.46; p < 0.001), but not in men (OR 1.47; 95%CI 0.90-2.40; p = 0.119). Moderate-to-severe SDB was associated with a higher risk of hypertension in men (OR 3.11; 95%CI 2.23-4.33; p < 0.001), premenopausal women (OR 3.88; 95%CI 1.42-10.6; p = 0.008), and postmenopausal women (OR 1.96; 95%CI 1.46-2.63; p < 0.001). Short sleep duration was not associated with diabetes or hypertension. The associations of obesity with diabetes or hypertension were indirectly mediated by SDB (24.0% and 21.5%, respectively), with possible sex differences emerging (men/women; 15.3/27.8% and 27.0/16.9%, respectively). Conclusions: Notwithstanding the cross-sectional design, SDB and obesity, but not short sleep duration, were independently associated with diabetes and hypertension, with gender and menopausal status-related differences in risk emerging.
35
José M. Marı́n, Santiago Carrizo, Eugenio Vicente et al. · The Lancet · 2005 · 4.7K citations
A general approach to causal mediation analysis.
Kosuke Imai, Luke Keele, Dustin Tingley · Psychological Methods · 2010 · 3.9K citations · Full text
Shahrad Taheri, Ling Lin, Diane Austin et al. · PLoS Medicine · 2004 · 2.5K citations · Full text