|| Checking for direct PDF access through Ovid
To determine whether 2012 to 2015 (times 1-3) ridership changes correlated with body mass index (BMI) changes after transit line completion in Salt Lake City, Utah.We used Global Positioning System/accelerometry-measured transit ridership measures in 2012 to 2013 (times 1-2) to compare objective and self-reported ridership. Regression models related changes in objectively measured ridership (times 1-2) and self-reported ridership (times 1-2 and times 1-3) to BMI changes, adjusting for control variables.Objective and self-reported ridership measures were consistent. From time 1 to 2 (P = .021) or to 3 (P = .015), BMI increased among self-reported former riders and decreased among new riders (P = .09 for both times 1-2 and times 1-3), although the latter was nonsignificant. Time 3 attrition adjustment had no effect on results. Adjusting for baseline BMI, the nonsignificant effect for new riders remained nonsignificant, indicating no BMI change; the BMI increase after discontinuing transit remained significant.Observed BMI increases subsequent to stopping transit ridership persisted for more than 2 years (postintervention). These results suggest that transit ridership protects against BMI gains and support the need to provide convenient transit for public health.