Measuring Livability and its Influence on Health

Introduction Livability (defined as the ability of urban systems to reliably meet basic human needs, provide opportunities for expression, and foster a sense of community) is a nuanced multidimensional construct that is growing in interest in public health and other fields in the social sciences. However, measurement of livability is skewed towards attracting new development instead of accurately representing the lived experiences of residents. Creation of strong measures of livability are needed to improve understanding of this construct, its operation in causal pathways of health, and facilitate comparisons of livability across contexts to advance knowledge sharing. Despite the promise of livability measurement to advance cross-sector collaboration for public health goals, there is little to no empirical investigation into livability and health and insufficient evidence about how livability affects health behaviors and outcomes. Methods We conducted a scoping review of livability measures and the assessment of their respective psychometric properties (Aim 1) to inform a secondary data analysis of the DataHaven Community Wellbeing Survey (DCWS). We employed factor analytic methods to create measures of livability in Connecticut, examined their psychometric and ecometric properties, and assessed convergent and divergent validity of these measures with other area-level measures (Aim 2). Multilevel generalized linear mixed models (GLMM) were fit to estimate associations between livability measures and four health outcomes: self-rated health, BMI, cardiovascular disease, and diabetes (Aim 3). Results The scoping review of livability measures included 24 studies, representing a wide geographic range and varying geographic scales. There was minimal consensus on the conceptualization of livability, but significant overlap in identified domains. Factor analysis of DCWS data identified three domains: safety, opportunity, and infrastructure. As a result, four livability measures were created, one overall measure and three subscale measures. All four measures had high reliability and demonstrated high convergent and divergent validity with other measures (e.g., place satisfaction). Adjusted multilevel models show that increases in overall livability, safety, and opportunity are associated with higher odds of excellent self-reported health and lower odds of cardiovascular disease. There was no statistically significant association between any livability measure and BMI or diabetes. Conclusion Further measurement research can support the creation of parsimonious, standardized measures of livability to facilitate shared goals across sectors for livability interventions. The largest disparities in livability were across income and geography, providing evidence for the long-lasting impact of racial residential segregation in the United States. Thus, livability measurement has the potential to describe and improve quality of life for marginalized populations experiencing persistent livability inequities.