A programme tested in routine care
The trial involved adults aged 18 to 75 receiving care at two New York healthcare facilities. Their HbA1c, an indicator of blood glucose over time, was between 5.7 and 6.4%. None had a history of diabetes. Use of diabetes medication was an exclusion criterion, except for metformin.
Primary care teams were randomly assigned to the programme or usual care. For 12 months, community health workers offered health action plans, motivational interviewing, health education and referrals to local resources. The primary outcome was the development of diabetes, with follow-up extending up to six months after the intervention.
No demonstrated metabolic benefit
Diabetes developed in 5.9% of participants in the support group, compared with 4.3% in the control group. This difference was not statistically significant: it demonstrates neither a benefit from the programme nor an increase in risk linked to it.
No significant differences were observed in HbA1c, weight, blood pressure or blood lipids either. The lack of demonstrated benefit therefore applies to several indicators, not just the diagnosis of diabetes.
More healthcare contact, with no observed metabolic improvement
The support group had greater use of primary care, with reported proportions of 77.6% versus 62.5%. Emergency department visits were also more frequent, at 35.0% versus 29.1%.
These additional contacts do not, on their own, constitute evidence of improved health. The abstract does not establish why emergency department visits were more frequent.
What this trial tells us
The result applies to this specific programme, in these facilities and over this period. It does not support the conclusion that all community-based support is ineffective. The authors propose testing more intensive or more standardised models, but their superiority remains to be demonstrated.
Limitations to keep in mind
Only the PubMed abstract was consulted. It does not specify the sample size, which limits assessment of the precision of the results. Follow-up after the intervention extended up to six months and provides no information about any potential later benefit. Funding and conflicts of interest were not checked.
What we can take away
In this trial, more contact with healthcare services did not lead to a demonstrated reduction in diabetes or improvement in the metabolic indicators measured.
Sources and provenance
- Community health worker intervention to prevent incident type 2 diabetes: A cluster randomized trial.
Source date: 2026-10-30.
Journal Article, Randomized Controlled Trial, Pragmatic Clinical Trial · Résumé PubMed uniquement, texte intégral non consulté