Engagement, retention, and progression to type 2 diabetes: a retrospective analysis of the cluster-randomized "Let's Prevent Diabetes" trial

Background: Prevention of type 2 diabetes (T2DM) is a global priority. Let’s Prevent Diabetes is a group based diabetes prevention programme, it was evaluated in a cluster randomised trial, in which the primary analysis showed a non-significant 26% reduction in T2DM (HR 0.74, 95% CI 0.48, 1.14, p=0.18). We examined the effects of engagement and retention with the Let’s Prevent Diabetes prevention programme on T2DM incidence. Methods and findings: We used data from a completed cluster randomised controlled trial including 43 General Practices randomised to receive either standard care or a six-hour group structured education programme with an annual refresher course for two years. The primary outcome was progression to T2DM at three years. The characteristics of those who attended the initial education session (engagers) versus non-engagers and those who attended all sessions (retainers) versus non-retainers were compared. Risk reduction of progression to T2DM by level of attendance was compared to standard care. 880 participants were recruited, 447 to the intervention arm of which 346 (77.4%) were engagers and 130 (29.1%) were retainers. Retainers and engagers were more likely to be older, leaner and non-smokers than non-retainers/non-engagers. Engagers were also more likely to be male and be from less socio-economically deprived areas than non-engagers. Statistically significant intervention effects were seen in those who attended the initial session and at least one refresher (30 people of 248 versus 67 people of 433, HR 0.38 (95% CI 0.236, 0.62) and in retainers (7 people of 130 versus 67 people of 433, HR 0.12 (95% CI 0.05, 0.28)) compared to standard care. Retaining was also associated with improvements in glucose, HbA1c, weight, waist circumference, anxiety, quality of life and daily step count. Given the data used is from a clinical trial those taking part might reflect a more engaged sample than the population, which should be taken into account when interpreting the results. Conclusions: This study suggests that a relatively low resource, pragmatic diabetes prevention programme can lead to significant reductions in the progression to T2DM in people at high risk who engage/retain. Non-engagers and non-retainers share similar high risk traits. Service providers of programmes should focus on reaching these hard to reach groups.