Nicole K. Y. Tang, Mark Fiecas, Esther F. Afolalu, Dieter Wolke

Sleep is a plausible target for public health promotion. We examined the association of changes in sleep with subsequent health and well-being in the general population. We analyzed data from the UK Household Longitudinal Survey, involving 30594 people (aged > 16) who provided data on sleep and health and well-being at both Wave 1 (2009-2011) and Wave 4 (2012-2014) assessments. Predicting variables were changes in sleep quantity, sleep quality, and sleep medication use over the 4-year period. Outcome variables were the General Health Questionnaire (GHQ-12) and the 12-Item Short-Form Health Survey (SF-12) mental (MCS) and physical (PCS) component scores at Wave 4. Linear regression on each outcome was fully adjusted for potential confounders and baseline values of the relevant predicting and outcome variables. Better outcomes were associated with an increase in sleep duration (GHQ: β = 1.031 [95% confidence interval {CI}: -1.328, -0.734]; MCS: 1.531 [1.006, 2.055]; PCS: -0.071 [-0.419, 0.56]), sleep quality (GHQ: β = -2.031 [95% CI: -2.218, -1.844]; MCS: 3.027 [2.692, 3.361]; PCS: 0.924 [0.604, 1.245]), and a reduction in sleep medication use (GHQ: β = -1.929 [95% CI: -2.400, -1.459]; MCS: 3.106 [2.279, 3.933]; PCS: 2.633 [1.860, 3.406]). Poorer outcomes were on the other hand associated with a reduction in sleep duration, a decrease in sleep quality, and an increase in sleep medication use. Changes in sleep quality yielded the largest effects on the health and well-being outcomes. Changes in sleep were temporally associated with subsequent health and well-being. Initiatives that aim to protect a critical amount of sleep, promote sleep quality, and reduce sleep medication use may have public health values.