The community-based prevention of diabetes (ComPoD) study: a randomised, waiting list controlled trial of a voluntary sector-led diabetes prevention programme

Objective This two-site randomised trial compared the effectiveness of a voluntary sector-led, community-based diabetes prevention programme to a waiting-list control group at 6 months, and included an observational follow-up of the intervention arm to 12 months. Methods Adults aged 18-75 years at increased risk of developing type 2 diabetes due to elevated blood glucose and being overweight were recruited from primary care practices at two UK sites, with data collected in participants' homes or community venues. Participants were randomised using an online central allocation service. The intervention, comprising the prototype "Living Well, Taking Control" (LWTC) programme, involved four weekly two-hour group sessions held in local community venues to promote changes in diet and physical activity, plus planned follow-up contacts at two, three, six, nine and 12 months alongside 5 hours of additional activities/classes. Waiting list controls received usual care for 6 months before accessing the programme. The primary outcome was weight loss at 6 months. Secondary outcomes included glycated haemoglobin (HbA1c), blood pressure, physical activity, diet, health status and well-being. Only researchers conducting analyses were blinded. Results The target sample of 314 participants (157 each arm) was largely representative of local populations, including 44% men, 26% from ethnic minorities and 33% living in deprived areas. Primary outcome data were available for 285 (91%) participants (141 intervention, 144 control). Between baseline and 6 months, intervention participants on average lost more weight than controls (- 1.7 kg, 95% CI - 2.59 to - 0.85). Higher attendance was associated with greater weight loss (- 3.0 kg, 95% CI - 4.5 to - 1.5). The prototype LWTC programme more than doubled the proportion of participants losing > 5% of their body weight (21% intervention vs. 8% control, OR 2.83, 95% CI 1.36 to 5.90) and improved self-reported dietary behaviour and health status. There were no impacts on HbA1c, blood pressure, physical activity and well-being at 6 months and, amongst intervention participants, few further changes from six to 12-months (e.g. average weight re-gain 0.36 kg, 95% CI - 0.20 to 0.91). There were no serious adverse events but four exercise-related injuries were reported in the intervention arm. Conclusions This voluntary sector-led diabetes prevention programme reached a broad spectrum of the population and had modest effects on weight-related outcomes, but limited impacts on other diabetes risk factors.