Residential modifications associated with less physical decline in two years among community-dwelling elderly: Results of a survey-weighted propensity model

Residential modifications and personal assistive devices may prevent disability by reducing task demand. While use of personal assistive devices has been associated with lower self-reports of disability, the effect of residential modifications alone has not been well-studied. Using the nationally representative Second Longitudinal Study on Aging (1994-1998), we examined the effect of residential modifications on reducing risk of physical functional decline. Participants (n=9,447) were non-institutionalized civilians, age 70 and older at the time of baseline interview who were re-interviewed two years later for self-reported changes in difficulties with activities of daily living (ADL) . Respondents who died or who reported more ADL loss at follow-up compared to baseline were considered to experience functional decline. To adjust for confounding, we developed a propensity score modeling the presence of residential modifications at baseline as a function of 32 covariates (c-statistic=0.70). Using both indirect and direct post-stratification adjustment techniques to take survey weights into consideration, we estimated a risk difference representative of the US population. Approximately 38% of the participants had residential modifications including ramps, railings, bathroom modifications, alerting devices, and elevators or lifts. Over the 2 year follow-up, 44% of the study sample experienced a decline in physical function. Respondents with baseline residential modifications experienced less decline than those without any baseline modifications (survey-weighted effect size of 3.1%; 95% Confidence Interval: 2.9% to 3.4%). Since physical decline is associated with higher medical costs and hospitalization rates, understanding modifiable factors that reduce functional decline can help formulate policy and public health interventions.