Direct and Moderating Causal Effects of Network Support on Sleep Quality: Findings From the UC Berkeley Social Network Study

Background Sleep is an important, restorative behavior for health, yet many adults report troubled sleep. The existence of a support network may be beneficial for sleep quality, including as a buffer for stressful events, yet few studies have examined these relationships longitudinally. Purpose To examine the causal effect of changes in personal network support on sleep quality both directly and as a buffer of negative life events among young and older adults. Methods The UC Berkeley Social Network survey collected data from young (21-30 year old, n = 475) and late middle-age (50-70 year old, n = 637) adults across three waves between 2015 and 2018. Participants reported on personal network characteristics, negative life events, and number of nights with trouble falling and staying asleep. Fixed effects models are used to examine causal relationships among each age cohort. Results Direct effects of network support on sleep quality were observed among older adults. Insufficient practical support predicted higher rates of trouble falling asleep (incident rate ratio [IRR] = 1.40, p <.01), while a desire for more social companions predicted lower rates of trouble staying asleep (IRR = 0.81, p<.01). Buffering effects of network support on sleep quality were observed among young adults. Changes in partnership status buffer the negative effects of the death of a close tie on trouble falling asleep (IRR = 0.75, p<.01) and persistent difficulties paying bills on trouble staying asleep (IRR = 0.45, p < 0.001) among young adults. Conclusions This study provides evidence for the direct and buffering role of network support on sleep quality. Our results indicate that efforts to improve sleep quality should address personal networks and the support they provide, perhaps especially during times of stress for younger adults.