Abstract
BACKGROUND AND OBJECTIVES: Proximity to health care, healthy foods, and recreation is linked to improved health in older adults while deterioration of the built environment is a risk factor for poor health. Yet, it remains unclear whether individuals prone to good health self-select into favorable built environments and how long-term exposure to deteriorated environments impacts health. This study uses a longitudinal framework to address these questions. RESEARCH DESIGN AND METHODS: The study analyzes 3,240 Americans aged 45 or older from the Panel Study of Income Dynamics with good self-reported health at baseline, and follows them from 1999 to 2013. At each biennial survey wave, individual data are combined with data on services in the neighborhood of residence (defined as the zip code) from the Economic Census. The analysis overcomes the problem of residential self-selection by employing marginal structural models and inverse probability of treatment weights. RESULTS: Logistic regression estimates indicate that long-term exposure to neighborhood built environments that lack health-supportive services (e.g., physicians, pharmacies, grocery stores, senior centers, and recreational facilities) and are commercially declined (i.e., have a high density of liquor stores, pawn shops, and fast food outlets) increases the risk of fair/poor self-rated health compared to more average neighborhoods. Short-term exposure to the same environments as compared to average neighborhoods has no bearing on self-rated health after adjusting for self-selection. DISCUSSION AND IMPLICATIONS: Results highlight the importance of expanding individuals' access to health-supportive services prior to their reaching old age, and expanding access for people unlikely to attain residence in service-dense neighborhoods.