Community-wide intervention and population-level physical activity: a 5-year cluster randomized trial

• Published in: International journal of epidemiology Vol. 47; no. 2; pp. 642 - 653
• Main Authors: Kamada, Masamitsu, Kitayuguchi, Jun, Abe, Takafumi, Taguri, Masataka, Inoue, Shigeru, Ishikawa, Yoshiki, Bauman, Adrian, Lee, I-Min, Miyachi, Motohiko, Kawachi, Ichiro
• Format: Journal Article
• Language: English
• Published: England Oxford University Press 01-04-2018
• Subjects: Adult; Aged; Exercise; Female; Health Plan Implementation - methods; Health Promotion - methods; Health Promotion - statistics & numerical data; Humans; Japan; Low Back Pain - rehabilitation; Male; Middle Aged; Miscellaneous; Pain Measurement; Program Evaluation; Regression Analysis; Residence Characteristics; Surveys and Questionnaires; Japan
• ISSN: 0300-5771; 1464-3685
• DOI: 10.1093/ije/dyx248

Evidence from a limited number of short-term trials indicates the difficulty in achieving population-level improvements in physical activity (PA) through community-wide interventions (CWIs). We sought to evaluate the effectiveness of a 5-year CWI for promoting PA in middle-aged and older adults using a cluster randomized design.
We randomized 12 communities in Unnan, Japan, to either intervention (9) or control (3). Additionally, intervention communities were randomly allocated to three subgroups by different PA types promoted. Randomly sampled residents aged 40-79 years responded to the baseline survey (n = 4414; 74%) and were followed at 1, 3 and 5 years (78-83% response rate). The intervention was a 5-year CWI using social marketing to promote PA. The primary outcome was a change in recommended levels of PA.
Compared with control communities, adults achieving recommended levels of PA increased in intervention communities [adjusted change difference = 4.6 percentage points (95% confidence interval: 0.4, 8.8)]. The intervention was effective for promoting all types of recommended PAs, i.e. aerobic (walking, 6.4%), flexibility (6.1%) and muscle-strengthening activities (5.7%). However, a bundled approach, which attempted to promote all forms of PAs above simultaneously, was not effective (1.3-3.4%, P ≥ 0.138). Linear dose-response relationships between the CWI awareness and changes in PA were observed (P ≤ 0.02). Pain intensity decreased in shoulder (intervention and control) and lower back (intervention only) but there was little change difference in all musculoskeletal pain outcomes between the groups.
The 5-year CWI using the focused social marketing strategy increased the population-level of PA.