RESUMO
BACKGROUND: Latina women are less likely to report engaging in leisure-time physical activity (PA) than non-Latina white women. This study evaluated the 24-month impact of a faith-based PA intervention targeting Latinas. METHODS: The study is a cluster randomized controlled trial of a PA intervention or cancer screening comparison condition, with churches as the randomization unit. A total of 436 Latinas (aged 18-65 years) from 16 churches who engaged in low levels of self-report and accelerometer-based PA were enrolled. The experimental condition was a 24-month PA intervention, with in-person classes, social support, and environmental changes, led by community health workers (i.e., promotoras). At baseline, 12-, and 24 months, we assessed changes in accelerometer-based and self-reported moderate to vigorous physical activity (MVPA; primary outcomes). Secondary outcomes were light intensity activity, sedentary time, body mass index (BMI), and waist circumference. RESULTS: After adjusting for sociodemographic factors, a mixed effects analysis found significant increases in self-reported leisure time MVPA (p < 0.005) and marginal increases in accelerometer-assessed MVPA (p < 0.08) 24 months post-baseline in the intervention compared to the attention-control condition. Data showed significant associations between PA class attendance and engaging in MVPA as assessed by self-report and accelerometry. No significant changes were found for light activity, sedentary time, BMI, or waist circumference. CONCLUSIONS: Participants who attended the PA classes at least once a month engaged in significantly higher MVPA compared to those who did not. Maximizing engagement and maintenance strategies to enhance PA maintenance could contribute to important long-term health benefits. TRIAL REGISTRATION: NCT01776632 , Registered March 18, 2011.