Examining the Associations and Cost-Effectiveness of Policy Level Changes to the Built and Transportation Environments Intended to Increase Physical Activity

Gregory Knell, The University of Texas School of Public Health


Physical inactivity is a significant public health problem. Less than one-half of United States (US) adults report engaging in sufficient physical activity for health benefits and almost one-third report no leisure-time physical activity at all. Efforts to increase population levels of physical activity should focus on systematic interventions targeting the physical, social, and economic environments. Collectively, the three articles included in this dissertation aimed to assess relations in the physical (transportation and built), and economic environments related to physical activity. ^ The Houston Travel Related Activity in Neighborhoods (TRAIN) Study served as the parent study for this dissertation and data collected from 2013 to 2016 were used to address the specific aims. Briefly, the specific aims include: 1) to determine the association between baseline physical activity and transit use (Article 1), 2) to assess the changes in physical activity associated with living near improved sidewalks (Article 2), and 3) to determine if improving sidewalks sufficiently increases physical activity to be considered cost-effective (Article 3). The TRAIN Study is a natural experiment aimed at prospectively evaluating the impact of a light rail transit (LRT) system on transit use and physical activity over a four year period (2013–2017), in Houston, Texas. To be eligible, participants must, 1) be at least 18 years of age, 2) reside within a 3 mile Euclidean buffer around the LRT, and 3) be the only TRAIN participant in their household. A further eligibility criteria for this dissertation work included complete data on self-reported physical activity at baseline (Article 1) and first follow-up (Articles 2 & 3). ^ Article 1 utilized multiple logistic regressions to model the odds of transit users achieving sufficient physical activity for health benefits versus transit non-users. Articles 2 and 3 included a follow-up assessment a median 14.3 months post baseline assessment. Multivariable regressions (two-step probit & OLS models in Article 2) were used to estimate the relation between living near sidewalk improvements and physical activity. Article 3 conducted a cost-effectiveness analysis of sidewalk improvements. Article 1 results indicated that transit users accumulated approximately 30 more weekly minutes of moderate intensity physical activity than transit non-users. Transit users had 2.46 (95% CI: 1.08–5.61) times the odds of health benefits from transit related physical activity than non-users. Article 2 found that reported leisure-time physical activity was 60% (expβ=0.46=1.58, p=0.009) higher among physically active participants living near two sidewalk improvements than not living near an improvement. Finally, Article 3 found that sidewalk improvements were cost-effective at a cost of $0.007–$0.07 per MET hour per year gained of self-reported physical activity. ^ These results highlight the importance of a multi-level or systems approach to increase physical activity. Future work should establish true baseline assessments of transit use and exposures to sidewalk improvements with longer follow-up periods to establish temporally sequenced and dose-response relations with physical activity.^

Subject Area


Recommended Citation

Knell, Gregory, "Examining the Associations and Cost-Effectiveness of Policy Level Changes to the Built and Transportation Environments Intended to Increase Physical Activity" (2017). Texas Medical Center Dissertations (via ProQuest). AAI10274894.