Abstract
Previous studies have emphasized that tightly knit networks influence health behaviors. However, effective network structures for behavioral adoption may vary by diffusion stage. This study examines how the association between personal network structures and health behaviors varies across behaviors with different prevalence degrees. We used data from the third-wave Japanese Study on Stratification, Health, Income, and Neighborhood (J-SHINE) conducted in 2017, targeting residents aged 32-58 years in Japanese metropolitan areas. Peer characteristics, behaviors, and interconnections were collected using the name generator method. Data from 1,705 respondents (egos) and 6,820 peers were analyzed. Structural holes, as the network structural characteristic, were evaluated using the reciprocal of the dyad constraint index of each ego-peer pair and categorized into tertiles. Logistic regression analyses examined the associations of structural holes with ego's exercise and preventive dental care use (intermediate prevalence stage) and non-smoking behavior (later prevalence stage), adjusting for covariates. Results showed that, compared to peers with middle-level structural holes, those with many structural holes were positively associated with ego's exercise habits (odds ratio [OR], 1.35; 95% confidence interval [CI], 1.19-1.52) and preventive dental care use (OR, 1.20; 95% CI, 1.07-1.35), while peers with few structural holes were negatively associated with ego's non-smoking behavior (OR, 0.81; 95% CI, 0.70-0.94). The findings suggest that the association between structural holes and health behaviors varies according to the diffusion stage. Considering social connections with different levels of structural holes by diffusion stage of the target behavior may be effective for public health interventions.