Spatial Variability of Geriatric Depression Risk in a High-Density City: A Data-Driven Socio-Environmental...

Previous studies found a relationship between geriatric depression and social deprivation. However, most studies did not include environmental factors in the statistical models, introducing a bias to estimate geriatric depression risk because the urban environment was found to have significant associations with mental health. We developed a cross-sectional study with a binomial logistic regression to examine the geriatric depression risk of a high-density city based on five social vulnerability factors and four environmental measures. We constructed a socio-environmental vulnerability index by including the significant variables to map the geriatric depression risk in Hong Kong, a high-density city characterized by compact urban environment and high-rise buildings. Crude and adjusted odds ratios (ORs) of the variables were significantly different, indicating that both social and environmental variables should be included as confounding factors. For the comprehensive model controlled by all confounding factors, older adults who were of lower education had the highest geriatric depression risks (OR: 1.60 (1.21, 2.12)). Higher percentage of residential area and greater variation in building height within the neighborhood also contributed to geriatric depression risk in Hong Kong, while average building height had negative association with geriatric depression risk. In addition, the socio-environmental vulnerability index showed that higher scores were associated with higher geriatric depression risk at neighborhood scale. The results of mapping and cross-section model suggested that geriatric depression risk was associated with a compact living environment with low socio-economic conditions in historical urban areas in Hong Kong. In conclusion, our study found a significant difference in geriatric depression risk between unadjusted and adjusted models, suggesting the importance of including environmental factors in estimating geriatric depression risk. We also developed a framework to map geriatric depression risk across a city, which can be used for identifying neighborhoods with higher risk for public health surveillance and sustainable urban planning.