Addressing Hispanic Obesity Disparities Using a Community Health Worker Model Grounded in Motivational Interviewing

Abstract

Purpose: To determine whether Hispanic residents receiving the Healthy Fit intervention enhanced with Motivational Interviewing (MI) experienced greater improvements in body composition, relative to participants receiving the initial intervention.

Design: Quasi-experimental evaluation.

Setting: El Paso, Texas.

Sample: Among 656 baseline participants, 374 (54%) completed the 12-month assessment.

Intervention: In Healthy Fit, community health workers (CHWs) promote nutrition and exercise. To strengthen intrinsic motivation and help participants overcome barriers to change, we incorporated a 30-minute motivational interview into the baseline assessment. Follow-up phone calls at 1, 3, and 6 months were identical across conditions.

Measures: CHWs assessed body mass index (BMI) and body fat percentage (BFP) using a bioelectrical impedance scale.

Analysis: Regression models estimated differences between intervention conditions on change in BMI and BFP from baseline to the 12-month assessment.

Results: Participants receiving MI had 2.13 times higher odds of losing weight (OR = 2.14, 95% CI [1.30, 3.53], P = .003) and 2.59 times higher odds of reduced BFP (OR = 2.59, 95% CI [1.51, 4.41], P < .001), relative to initial intervention participants. MI participants lost an average of 1.23 kg (2.71 lbs.) and their BFP declined 2% over 12 months.

Conclusion: Findings suggest CHW use of MI is a promising approach for promoting incremental changes in diet and exercise, which Healthy Fit integrates into a low-cost intervention.

Keywords: community health workers; health equity; hispanic health disparities; motivational interviewing; obesity.

Figure 1.

Consort flow diagram of participant selection and allocation.

Figure 2.

Body mass index and body fat percentage from baseline to 12-month follow-up among Healthy Fit participants who received a motivational interview for diet and exercise, as compared to participants receiving the standard intervention, adjusted for covariates (n = 374).