Residents of vulnerable neighborhoods show lower levels of, and faster longitudinal declines in cognition across most, but not all studies. Mixed results may exist, in part, because residents’ self-reported (individual-level) perceptions of their neighborhood do not always reflect Census-level measurements. We investigated both types of neighborhood characteristics to reexamine previously reported null associations between neighborhood vulnerability and cognition within an interactive socio-environmental framework. Self-identified Latinos (N = 224; x¯age = 70.4) without dementia completed a modified Perception of Neighborhood Environment Scale (mPNES; higher scores = more positive perceptions of total and domain-specific community cohesiveness, health opportunities, and ambience), annual cognitive assessments, and had baseline addresses geocoded to a Census-derived Social Vulnerability Index (SVI; higher scores = higher vulnerability). Separate mixed effects regression models adjusted for relevant confounders tested relationships between mPNES and cognitive composite z-scores, and modifying effects of mPNES on the SVI-cognition associations. Higher total mPNES scores were associated with higher baseline global cognition and working memory (driven by community cohesiveness). Higher domain-specific health opportunities associated with slower rates of global cognitive decline. mPNES modified relationships between SVI and cognition, particularly baseline global cognition and episodic memory. Results demonstrate the importance of incorporating individual-level neighborhood perceptions when examining relationships between Census-level measures of neighborhood health and cognition.
Lamar et al. (Wed,) studied this question.