Beyond the Walkability Composite: Destination Access, Perceived Street Conditions, and Transport Walking in Four Thai Secondary Cities

Walkability measures often combine destination access and street conditions, although the components may not relate to walking in the same way. We examined item-level, group-level, and composite associations with transport walking in four highly motorised Thai secondary cities. A cross-sectional household survey included 1659 adults (1627 complete cases; models adjusted for sociodemographics, household vehicle ownership, and city). Regular transport walking was reported by 63.8%. After mutual adjustment and false-discovery-rate correction, only shorter self-reported walking time to a fresh market (odds ratio (OR) 1.31 per band closer, 95% confidence interval (CI) 1.19–1.44) and to a school (OR 1.29, 95% CI 1.17–1.42) remained associated; model-adjusted predicted prevalence fell from 71.9% to 49.6% across the market bands. The nine street-condition items showed no joint association (p = 0.947); that item group had low internal consistency (α = 0.60) and was unvalidated, so this is not evidence that pedestrian infrastructure is unimportant. The composite association was concentrated in the two destination items: a 17-item composite (OR 1.20 per standard deviation) fell to 1.07 with them excluded (bootstrapped difference p < 0.001). Improving walking routes to existing everyday destinations is the implication these data support. These cross-sectional associations are not causal effects.