Context

Multilevel church-based interventions may help address racial/ethnic disparities in obesity in the United States since churches are often trusted institutions in vulnerable communities. These types of interventions affect at least two levels of socio-ecological influence which could mean an intervention that targets individual congregants as well as the congregation as a whole. However, the extent to which such interventions are developed using a collaborative partnership approach and are effective with diverse racial/ethnic populations is unclear, and these crucial features of well-designed community-based interventions.

Objective

The present systematic literature review of church-based interventions was conducted to assess their efficacy for addressing obesity across different racial/ethnic groups (eg, African Americans, Latinos).

Data sources and extraction

In total, 43 relevant articles were identified using systematic review methods developed by the Center for Disease Control and Prevention (CDC)'s Task Force on Community Preventive Services. The extent to which each intervention was developed using community-based participatory research principles, was tailored to the particular community in question, and involved the church in the study development and implementation were also assessed.

Data analysis

Although 81% of the studies reported significant results for between- or within-group differences according to the study design, effect sizes were reported or could only be calculated in 56% of cases, and most were small. There was also a lack of diversity among samples (eg, few studies involved Latinos, men, young adults, or children), which limits knowledge about the ability of church-based interventions to reduce the burden of obesity more broadly among vulnerable communities of color. Further, few interventions were multilevel in nature, or incorporated strategies at the church or community level.

Conclusions

Church-based interventions to address obesity will have greater impact if they consider the diversity among populations burdened by this condition and develop programs that are tailored to these different populations (eg, men of color, Latinos). Programs could also benefit from employing multilevel approaches to move the field away from behavioral modifications at the individual level and into a more systems-based framework. However, effect sizes will likely remain small, especially since individuals only spend a limited amount of time in this particular setting.