Factors Associated with Sleep Disruption among Community-Dwelling Older Adults in the Health and Retirement Study

The purpose of this study was to examine behavioral outcomes and sleep disruption in aging adults using data from the 2004 wave of the longitudinal Health and Retirement Study (HRS). It was hypothesized that sleep disruption is associated with behavioral outcomes and sociodemographic variables in a population-based sample. Methods. The 2004 HRS data (N=20,129) represents a cross-sectional analysis of community dwelling aging adults born in the US at or before 1923 through 1953. Data are stratified by date of birth to provide five cohorts of aging elderly. The HRS Psychosocial Leave-Behind Participant Lifestyle Questionnaire (PLBQ) was given to a random sample of participants (N=1,439; 52.5% male; 47.5% female; 91.5% White; 5.2% Black; Response Rate 76.8%) who completed the 2004 HRS wave. Portions of the HRS including age, gender, race, marital status, education, income, comorbidities, and sleep measures were matched to the subjects who completed the PLBQ. An index of sleep disruption items was computed and sleep sensitive behaviors (cynical hostility, optimism, pessimism and social participation) were computed according to instrument directions. Results. After controlling for sociodemographic variables, sleep disruption independently predicted social integration ( t =2.135, p <.0001), pessimism ( t =3.995, p <.0001), cynical hostility ( t =3.854, p <.0001), and negatively predicted optimism ( t =-4.876, p <.0001). Oldest-old adults had no greater sleep disruption than younger aging adults (? 2 =1.234, p =.872), although 59% of the oldest-old subset reported frequent waking during the night. Women reported more sleep disruption than men ( t =3.270, p <.001), but all subjects reported frequent waking during the night (Men=58%; Women=68%). Married participants reported more sleep disruption than divorced or widowed participants ( t =2.161, p =.03). Blacks had no greater sleep disruption than Whites ( t =.812, p =.417). Higher education was the most influential SES predictor for sleep disruption ( F =15.309, p <.0001), and income did not independently predict sleep disruption ( t =1.297, p =.195). Conclusion. Sleep disruption predicts negative behavioral outcomes, when controlling for sociodemographic variables. Sleep disruption was greater among women and married participants, but there were no racial differences. Education was also a strong influence on sleep disruption. Further research should examine the role of sleep disruption in other behavioral and attitudinal traits.

Personality Assessment

21020