Alex Bierman1, Yeonjung Lee2, Scott Schieman3. 1. Department of Sociology, University of Calgary, Alberta, Canada. 2. Faculty of Social Work, University of Calgary, Alberta, Canada. 3. Department of Sociology, University of Toronto, Ontario, Canada.
Abstract
BACKGROUND AND OBJECTIVES: Contextual contributors to sleep problems are important to examine among older adults because sleep problems are associated with a number of adverse outcomes in late life. We examine whether disordered neighborhoods are a key contextual determinant of sleep problems in late life, as well as how subjective social power-a sense of personal control and subjective social status-mediates and moderates this association. Central to this contribution is the use of econometric techniques that holistically control for time-stable factors that may bias estimated associations. RESEARCH DESIGN AND METHODS: Three waves (2006, 2010, 2014) of the psychosocial subsample of the Health and Retirement Study (N = 7,130) are analyzed with random-effects models that adjust for repeated observations, as well as fixed-effects models that additionally control for all time-stable confounders. RESULTS: Neighborhood disorder is associated with greater sleep problems in random-effects models, but this association is substantially weakened in a fixed-effects model. Personal control mediates this association, but does not moderate it. Subjective social status does not mediate the association, but does moderate it. DISCUSSION AND IMPLICATIONS: Although neighborhood disorder is associated with sleep problems in older adults, this association is likely to be overestimated in analyses that do not compressively control for time-stable confounders. Rather than acting as dual mediators and moderators, perceived control and subjective social status play distinct roles in this association, with seniors at lower levels of subjective social status especially at risk for sleep problems due to neighborhood disorder.
BACKGROUND AND OBJECTIVES: Contextual contributors to sleep problems are important to examine among older adults because sleep problems are associated with a number of adverse outcomes in late life. We examine whether disordered neighborhoods are a key contextual determinant of sleep problems in late life, as well as how subjective social power-a sense of personal control and subjective social status-mediates and moderates this association. Central to this contribution is the use of econometric techniques that holistically control for time-stable factors that may bias estimated associations. RESEARCH DESIGN AND METHODS: Three waves (2006, 2010, 2014) of the psychosocial subsample of the Health and Retirement Study (N = 7,130) are analyzed with random-effects models that adjust for repeated observations, as well as fixed-effects models that additionally control for all time-stable confounders. RESULTS: Neighborhood disorder is associated with greater sleep problems in random-effects models, but this association is substantially weakened in a fixed-effects model. Personal control mediates this association, but does not moderate it. Subjective social status does not mediate the association, but does moderate it. DISCUSSION AND IMPLICATIONS: Although neighborhood disorder is associated with sleep problems in older adults, this association is likely to be overestimated in analyses that do not compressively control for time-stable confounders. Rather than acting as dual mediators and moderators, perceived control and subjective social status play distinct roles in this association, with seniors at lower levels of subjective social status especially at risk for sleep problems due to neighborhood disorder.
Authors: Safiyyah M Okoye; Adam P Spira; Nancy A Perrin; Jennifer A Schrack; Hae-Ra Han; Sarah Wanigatunga; Casandra Nyhuis; Sarah L Szanton Journal: Sleep Health Date: 2021-10-08