Abstract
BACKGROUND: Intervention trials that have demonstrated significant effects may not always replicate those effects when scaled up. This study aimed to test whether scaling-up a successful cluster randomized trial (the COMMUNICATE study, 9 intervention communities), which promoted population-level physical activity (PA), could promote PA in a broader citywide setting (29 communities) after two years, as a mid-term evaluation of the six-year scaled-up trial. METHODS: This is a single-arm, pre-post comparison of a multi-strategic community-wide intervention covering the entire Unnan City, Japan. The intervention for middle-aged and older people consisted of three components: information delivery, education, and support delivery. The intervention method followed the COMMUNICATE study but adapted and introduced new initiatives tailored to local resources. A baseline survey (n = 3,718) among randomly selected residents aged 40-79 years in 2016 and a follow-up survey with the same respondents two years later were conducted. The primary outcome was the change in the percentage of people who practiced the recommended levels of PA, analyzed with a generalized linear mixed model to account for clusters at individual and community levels. Additionally, we examined the dose-response relation of the effect based on the intervention doses in each community. The RE-AIM framework assessed intervention dissemination and implementation. RESULTS: The two-year intervention was implemented in all communities, reaching and involving various groups. The rate of people engaged in PA significantly increased in two years (adjusted change: + 8.0 percentage points [95% confidence interval: 6.1, 10.0]). Based on the type of PA, only muscle-strengthening activity showed a significant increase (+ 11.5% points [9.6, 13.5]), whereas walking (-1.8% points [-3.6, 0.1]) and flexibility activities (+ 0.3% points [-1.5, 2.0]) did not. The increase in PA in higher-dose areas was not significantly different but slightly larger than that in lower-dose areas (+ 8.4% points vs. + 7.6% points, adjusted difference in change: 0.8% points [-3.8, 5.5]). CONCLUSION: The scaled-up citywide intervention promoted PA, especially muscle-strengthening activity. Collaboration with diverse organizations in different settings is crucial for multi-faceted interventions and requires balancing uncertainty in its implementation quality and quantity owing to collaborative decision-making. TRIAL REGISTRATION: UMIN-CTR, UMIN000024682. Registered 02 November 2016, https://center6.umin.ac.jp/cgi-open-bin/ctr/ctr_view.cgi?recptno=R000028377.