Community Trial of a Faith-Based Lifestyle Intervention to Prevent Diabetes Among African-Americans

Abstract

About 75% of African-Americans (AAs) ages 20 or older are overweight and nearly 50% are obese, but community-based programs to reduce diabetes risk in AAs are rare. Our objective was to reduce weight and fasting plasma glucose (FPG) and increase physical activity (PA) from baseline to week-12 and to month-12 among overweight AA parishioners through a faith-based adaptation of the Diabetes Prevention Program called Fit Body and Soul (FBAS). We conducted a single-blinded, cluster randomized, community trial in 20 AA churches enrolling 604 AAs, aged 20-64 years with BMI ≥ 25 kg/m(2) and without diabetes. The church (and their parishioners) was randomized to FBAS or health education (HE). FBAS participants had a significant difference in adjusted weight loss compared with those in HE (2.62 vs. 0.50 kg, p = 0.001) at 12-weeks and (2.39 vs. -0.465 kg, p = 0.005) at 12-months and were more likely (13%) than HE participants (3%) to achieve a 7% weight loss (p < 0.001) at 12-weeks and a 7% weight loss (19 vs. 8%, p < 0.001) at 12-months. There were no significant differences in FPG and PA between arms. Of the 15.2% of participants with baseline pre-diabetes, those in FBAS had, however, a significant decline in FPG (10.93 mg/dl) at 12-weeks compared with the 4.22 mg/dl increase in HE (p = 0.017), and these differences became larger at 12-months (FBAS, 12.38 mg/dl decrease; HE, 4.44 mg/dl increase) (p = 0.021). Our faith-based adaptation of the DPP led to a significant reduction in weight overall and in FPG among pre-diabetes participants. CLINICALTRIALS.

Gov identifier: NCT01730196.

Keywords: African-American; Diabetes prevention; Obesity; Translation research.

Figures

Figure 1

Flow diagram of the randomization, consenting, screening, and allocation to the intervention process.

Figure 2. Weight, Fasting Plasma Glucose, and Physical Activity at Baseline, Week 12, and Week 52

Shown are estimated marginal means for those who participated in the Fit Body and Soul (FBAS) intervention arm or the Health Education comparison arm. Means were estimated with the use of general linear mixed models for continuous measures. For weight, there was a significant interaction of arm and time (p = 0.001). There was not a significant interaction of arm and time for either fasting plasma glucose (p = 0.221) or for physical activity (p = 0.667). MET denotes metabolic equivalents (minutes/week).

Figure 2. Weight, Fasting Plasma Glucose, and Physical Activity at Baseline, Week 12, and Week 52

Shown are estimated marginal means for those who participated in the Fit Body and Soul (FBAS) intervention arm or the Health Education comparison arm. Means were estimated with the use of general linear mixed models for continuous measures. For weight, there was a significant interaction of arm and time (p = 0.001). There was not a significant interaction of arm and time for either fasting plasma glucose (p = 0.221) or for physical activity (p = 0.667). MET denotes metabolic equivalents (minutes/week).