In a pragmatic cluster-randomized trial across Veterans Affairs and Bellevue primary care teams, adults aged 18-75 with HbA1c 5.7-6.4% received either a 12-month community health worker (CHW) program (action plans, motivational interviewing, education, referrals) or usual care. Incident diabetes was 5.9% versus 4.3% (risk difference 1.6%; 95% CI -3.68% to 4.10%). HbA1c, weight, blood pressure and lipids did not differ. Intervention patients had more primary care visits (77.6% vs 62.5%) and emergency visits (35.0% vs 29.1%).
This null result matters because the evidence for structured prevention is strong where intensity is high: the Diabetes Prevention Program's lifestyle arm cut progression by 58% using a 16-session core curriculum, a 7% weight-loss goal and 150 minutes of weekly activity. A loosely standardized, pragmatic CHW model appears to lack that dose and fidelity, and weight, the main driver of risk reduction, did not budge. Limitations temper any verdict: the follow-up was only up to 6 months after the intervention, event rates were low, and the wide confidence interval cannot exclude modest benefit or harm. The population, largely veterans and safety-net patients, may not generalize. Higher emergency use suggests CHWs surfaced unmet needs rather than improved metabolic health. For adults with prediabetes, the takeaway is practical: generic outreach is no substitute for an evidence-based program with a weight-loss target. Confirmatory in direction, this trial reinforces that program intensity, not mere engagement, determines results.