Community social networks, individual social participation and dietary behavior among older Japanese adults: Examining mediation using nonlinear structural equation models for three-wave longitudinal data

• Published in: Preventive medicine 2021-08, Vol.149, p.106613-106613, Article 106613
• Main Authors: Nishio, Marisa, Takagi, Daisuke, Shinozaki, Tomohiro, Kondo, Naoki
• Format: Article
• Language: English
• Subjects: Cross-lagged panel model; Health disparities; Internal Medicine; Living arrangements; Social support
• ISSN: 0091-7435; 1096-0260
• DOI: 10.1016/j.ypmed.2021.106613

AbstractCommunity social networks positively affect older adults' dietary behavior, but the underlying mechanisms remain uncertain. This study investigated if the relationship between community social networks and dietary behavior is mediated by social participation, and whether the influence of community social networks differs across sociodemographic groups. We conducted a cross-lagged panel mediation analysis employing nonlinear structural equation modeling using panel data from the Japan Gerontological Evaluation Study (JAGES), a longitudinal study of Japanese older adults in 2010, 2013, and 2016, which included 11,347 men and 14,105 women aged over 65. Monodirectional relationships in community social networks, individual social participation, and fruit/vegetable intake were examined. Community social networks were associated with social participation regardless of sociodemographic conditions. Individual social participation was in turn, positively associated with fruit/vegetable intake, with this association being stronger among those living alone. Analyses further showed that individuals' social participation positively mediated the relationship between community social networks and fruit/vegetable intake. For both genders, the mediating effects were stronger among people living alone than in those living with someone. However, community networks could also directly and negatively affect fruit/vegetable intake among men who live alone unless they participated in community activities. Community social networks may promote social participation, thereby facilitating healthier dietary behavior regardless of sociodemographic status. Social participation in turn may positively contribute to the dietary health of people living alone. Community-based interventions to encourage people living alone to participate in social activities may help reduce inequality in dietary behavior related to cohabitation status.