Negative Spousal Interactions As a Mediator between Perceived Neighborhood Disorder and Depressive Symptoms Among Married Older Adults: A Structural Equation Model

Background and Purpose: This study examines a structural equation model that tests the mediating role of negative spousal interactions to understand how or why perceived neighborhood disorder is associated with depressive symptoms in a national sample of married older adults.  Addressing depressive symptoms is of particular importance for older adults in that depressive symptoms can affect physical decline, mortality and quality of life among this population.  In social disorganization theory and neighborhood effects literature, perceived neighborhood disorder (physical and social conditions/activities that residents perceive as signs of the breakdown of social order) has been shown to be a key predictor of psychological distress.  In stress process theory, positive and negative social interactions can be a mediator to explain how or why primary stressors lead to mental health problems.  For older married individuals, the spouse is usually a key social partner.  However, there is a gap in knowledge about the potential mediating role of spousal interactions on the association between perceived neighborhood disorder and depressive symptoms and a lack of attention to the older adult population.  This study hypothesizes that (a) higher perceived neighborhood disorder will be associated with higher depressive symptoms and that this relationship will be partially mediated by (b) higher perceived neighborhood disorder being associated with higher negative spousal interactions and (c) higher negative spousal interactions being associated with higher depressive symptoms.

Methods: This cross-sectional study analyzed secondary data from the Health and Retirement Study (2014), a nationally representative panel study of adults over the age of 50.  The study’s final sample after deletion of missing data is N= 3,764.  Hypotheses were tested in a recursive structural equation model with maximum likelihood estimation.  Perceived neighborhood disorder (8 indicators) was the exogenous latent construct, negative interactions with the spouse (4 indicators) was the latent mediator, and depressive symptoms (3 subscales from the CESD-8) was the latent endogenous outcome.

Results: The model produced an acceptable fit to the data, χ² (87) = 1,024.924, p < .001; GFI = .961, RMSEA = .054 (90% CI = .051 - .056), CFI = .964.  All parameter estimates were statistically significant at p < .001.  Higher perceived neighborhood disorder was positively associated with higher negative spousal interactions (.208) and with higher depressive symptoms (.173).  Higher negative spousal interactions were also positively associated with higher depressive symptoms (.320).  The bootstrapped standardized indirect effect of perceived neighborhood disorder on depressive symptoms was .067 (p < .01, bootstrapped 90% CI = .053 - .080).

Conclusions and Implications: Negative spousal interactions partially mediated the association between perceived neighborhood disorder and depressive symptoms among married older adults in this study.  Given the importance of depressive symptomatology and interactions with the spouse/partner for older adults, this study highlights a need for social work interventions to address older adults’ perceptions of neighborhood disorder and the link between these perceptions and depressive symptomatology through negative marital interactions.  Greater attention should be paid to older adults living in disadvantaged neighborhoods, as they may be particularly vulnerable to negative effects.