Key points are not available for this paper at this time.
Abstract Introduction Depression is a leading cause of disability in older adults, yet it remains underdiagnosed and undertreated. Establishing common multivariable sleep health profiles may eventually help to identify at-risk older adults and match them with appropriate treatments. However, profiles identified across cohorts often differ because of inconsistent measures and methods, hampering progress towards large-scale initiatives. We aimed to overcome this challenge using harmonized data across six cohorts. Methods We harmonized five self-report sleep health indicators across six epidemiologic cohorts of adults aged 60 from the United States and Netherlands (N=613 - 3,123). We performed latent class analysis in each cohort. Generalizability and comparability of findings were assessed using several indices, including cluster stability and novel distance metrics. Generalized linear mixed-effects modeling was used to relate sleep health profiles to the risk of increased depressive symptoms over time in each cohort. Results Two sleep health profiles were common across all cohorts: ‘Good Sleep’ (GS; average sleep duration, high quality and efficiency) and ‘Poor Sleep’ (PS; short sleep duration, low quality and efficiency, high daytime sleepiness). Three cohorts indicated an ‘Inefficient Sleep’ profile (IS) and three cohorts indicated a ‘Long Sleep’ profile (LS). PS had high and sustained depressive symptoms over 3-15 years of follow-up, especially relative to GS and IS (Risk Ratios RRs=1.47-3.44 for PS vs.GS; 1.75-2.32 for PS vs. IS; and 1.10-1.15 for PS vs. LS). Conclusion GS and PS profiles were generalizable across cohorts. Older adults with the PS profile had heightened depressive symptoms over multiple years of follow-up, especially relative to GS and IS profiles. Next steps involve quantifying the impact of screening using sleep health profiles and gathering evidence that interventions targeting sleep health profiles reduce onset or severity of depression. Support (if any) Wallace (RF1AG056331); Rush Alzheimer’s Disease Center (R01AG017917, R01AG022018); Osteoporotic Fractures in Men Study (U01AG027810, U01AG042124, U01AG042139, U01AG042140, U01AG042143, U01AG042145, U01AG042168, U01AR066160, UL1TR000128, R01HL071194, R01HL070848, R01HL070847, R01HL070842, R01HL070841, R01HL070837, R01HL070838, R01HL070839); Study of Osteoporotic Fractures (R01AG005407, R01AR35582, R01AR35583, R01AR35584, R01AG005394, R01AG027574, R01AG027576, and R01AG026720); Rotterdam Study (Erasmus Medical Center and Erasmus University, Netherlands Organization for the Health Research and Development, Ministries of Education and Health, European Commission, Rotterdam Municipality).
Wallace et al. (Sat,) studied this question.