Abstract P314: Physical Activity and Sedentary Behavior Phenotypes in a Majority-Minority Community-Based Sample: Houston TRAIN Study

Abstract

Background: Inadequate physical activity and prolonged sedentary behavior is related to premature mortality and suboptimal cardiovascular health. Yet, the health impacts of these behaviors are typically considered in isolation using a single method assessment approach. Use of accelerometer and reported data from a diverse sample provides the novel opportunity to identify and characterize distinct phenotypes to improve exposure ascertainment. Methods: Data are from 366 participants (61.4% female; 70.3% non-white; aged 50.7 years) enrolled in the Houston Travel Related Activity in Neighborhoods (TRAIN) Study; a natural experiment designed to determine the effect of a light rail train expansion on transit use and physical activity. Latent class analysis using baseline accelerometer (ActiGraph wGT3X-BT) and reported (self-administered, 7-day Modifiable Activity Questionnaire) data were used to derive the phenotypes and multinomial regression models were used to determine correlates of membership. Results: Four phenotypes emerged with the majority classified as Phenotype 4 or insufficient activity and high sedentary ( Table 1 ). When compared to Phenotype 2 (reference group), women were less likely than men to be classified as Phenotype 1 or 3. Every 1 kg/m 2 increase in body mass index increases the odds of membership in Phenotype 3 or 4 when compared to Phenotype 2. Black participants were more likely, and participants identifying as mixed race were less likely, than white participants to be classified as Phenotype 3 compared to Phenotype 2 (all p <0.05). Age was not a significant correlate of phenotype membership. Conclusions: A phenotype characterized by insufficient physical activity and high sedentary behavior represented the majority of study participants. This supports continued development of physical activity interventions targeting minority groups. Study findings also support use of methodological approaches to combine accelerometer and reported data to reduce potential exposure misclassification.