Data-Driven Child-Friendly Street Renewal for Health Equity in Older Urban Districts: Latent Activity–Health Profiles in Xi’an, China

Data-driven urban governance increasingly seeks to incorporate the needs of different population groups, yet child-sensitive evidence for public street-space renewal remains limited in older urban districts. Most studies still evaluate environmental conditions through population averages, with insufficient attention to heterogeneous child groups that may require differentiated planning responses. Based on an analytic sample of 314 children retained from 343 usable questionnaire responses collected from children aged 6–12 in the older urban districts of Xi’an, China, this study integrates street-activity characteristics and age- and sex-standardized body mass index (zBMI) using an established person-centered analytical approach. Latent Class Analysis (LCA) was used to identify children’s activity–health profiles, and multinomial logistic regression was used to examine associations between individual, family, and perceived street-environment factors and profile membership. Three profiles were identified: high-activity–healthy, high-intensity active, and low-activity–high-risk. The model-estimated low-activity–high-risk profile represented 35.7% of the analytic sample, and children assigned to this profile reported the lowest perceived safety and convenience. The findings suggest that profile-based analysis may inform child-sensitive street-renewal prioritization. Perceived safety and convenience showed the strongest and most consistent associations with membership in either the high-activity–healthy or high-intensity active profile relative to the low-activity–high-risk profile. These findings are associative and do not establish the effects of street interventions. The study therefore represents a context-specific extension and planning application of established analytical methods.