Urban environmental determinants and their effects on mental health, physical function, and quality of life in older adults: a multidimensional study in Shiraz, Iran.

The SVM machine learning model demonstrated superior predictive performance compared to Decision Tree models for frailty outcomes.

• SVM model achieved R²=0.75 for frailty prediction
• Decision Tree model achieved R²=0.71 for frailty prediction
• Both models were applied to a cross-sectional sample of 3,000 older persons aged 60 years and above
• Statistical analyses included both regression models and machine learning approaches

Each 1 m² increase in green space per capita was associated with a 0.8-point reduction in depression scores among older adults.

• The reduction in depression scores had a 95% CI of -1.2 to -0.4
• Each 1 m² increase in green space per capita also predicted a 0.3-point lower frailty index
• Depression was assessed using the validated GDS-4 tool
• Environmental data on green space per capita were extracted from municipal records across 11 municipal zones of Shiraz

Waste production exceeding 250 kg/capita was associated with a 35% greater fall risk in older adults.

• OR = 1.35 (95% CI: 1.12–1.63) for fall risk when waste production exceeded 250 kg/capita
• Waste production data were extracted from municipal records
• The study sample included 3,000 older persons aged 60 and above across 11 municipal zones
• Stratified random sampling was used to recruit participants

Population density showed nonlinear associations with health outcomes in SVM models.

• Nonlinear associations were identified between population density and health outcomes
• Thresholds for these associations varied by health indicator
• Population density data were extracted from municipal records
• SVM models were used to detect these nonlinear relationships, which may not have been captured by standard regression models

A cross-sectional study of 3,000 older adults across 11 municipal zones of Shiraz assessed multiple environmental and health variables.

• Study was conducted from December 2024 to January 2025
• Participants were aged 60 years and above
• Stratified random sampling was used across 11 municipal zones
• Health outcomes assessed included BMI, frailty, depression, anxiety, and life satisfaction using validated tools (GDS-4, GAI-5, LSI-Z)
• Environmental data on green space per capita, population density, and waste production were extracted from municipal records