The Effectiveness of Text Message Support for Weight Loss to Reduce Diabetes Risk

Objective: To reach the estimated 84.1 million U.S. adults with prediabetes, lower-cost alternatives to the National Diabetes Prevention Program (NDPP) are needed, which is estimated to cost $500 per person. In a previous randomized controlled trial, we demonstrated efficacy of a 12-month text message support program (SMS4PreDM) in individuals with prediabetes. We now explore effectiveness in a pragmatic study in a healthcare system, in addition to calculating per-person costs of SMS4PreDM. Research Design and Methods: English- and Spanish-speaking patients with diabetes risks (e.g., A1C 5.7-6.4) were referred by their healthcare providers and offered NDPP classes, SMS4PreDM, or both. This analysis focused on comparing weight outcomes among SMS4PreDM-only participants to a usual care control group of patients with diabetes risks who were not referred. As a pragmatic study, weights for both groups were collected from electronic health records at baseline and 12 months. Rates of achieving ≥3% weight loss were compared using logistic regression, including a sub-analysis by language. Results: Among 183 SMS4PreDM-only participants, 51.4% (N=94) had documented pre- and post-intervention weights and 30.9% (N=29) achieved ≥3% weight loss, compared with 23.4% of 1,871 control patients (p=0.10). English-speakers receiving SMS4PreDM trended toward ≥3% weight loss at a higher rate than English-speaking controls (36.5% vs. 25.6%, p=.07). There was no significant difference among Spanish-speakers. In an intention-to-treat analysis that assumed no weight loss among participants with missing data, there was no significant difference in achieving ≥3% weight loss as compared to controls (20.8% vs. 20.2%, p=.49). Costs of SMS4PreDM were $100.92 per capita. Conclusions: SMS4PreDM trends toward greater achievement of ≥3% weight loss compared to usual care and at a lower cost than the NDPP. However, results may not be sufficiently robust or generalizable to support long-term implementation. Disclosure H. Fischer: None. S. Raghunath: None. J. Durfee: None. N. Ritchie: None.