Summary: In 56 adults with obesity and elevated HbA1c (59% identifying as Māori) in Tairāwhiti, New Zealand, a 10-week whole-food plant-based community programme produced significantly greater weight loss than a waitlist control at 10 weeks (3.3 kg; 95% CI 0.8 to 5.7; p<0.001). The randomised between-group differences in HbA1c (p=0.08), cholesterol (p=0.69), waist circumference (p=0.16), and activity (p=0.97) were not statistically significant.
PICO Summary
| Element | Detail |
|---|---|
| Population | n=56 adults aged 30 to 72 years with obesity (BMI ≥30) and HbA1c ≥40 mmol/mol; 59% identified as Māori, 5% as Pasifika; recruited from a Very Low-Cost Access general practice in Gisborne, Tairāwhiti, New Zealand. Randomised waitlist-controlled trial. |
| Intervention | 10-week whole-food plant-based (WFPB) community programme: two 2-hour sessions per week (40 hours total) of skills-based learning and health education emphasising unprocessed plant foods. |
| Comparison | Waitlist control group receiving no dietary intervention during the 10-week randomised phase (later crossed over to receive the programme). |
| Outcome | Between-group (ITT, 10 weeks): weight loss 3.3 kg greater with WFPB (95% CI 0.8 to 5.7; p<0.001). Non-significant between-group differences for HbA1c (3.2 mmol/mol; 95% CI −0.4 to 6.7; p=0.08), cholesterol (p=0.69), waist circumference (p=0.16), and activity (p=0.97). Uncontrolled within-group analyses after all participants received the programme showed weight loss of about 5 kg post-treatment, sustained at 3 kg at 36 months (95% CI 1.2 to 4.7; p<0.001), with waist circumference down 6 cm; HbA1c (3.3 mmol/mol) and cholesterol (0.4 mmol/L) fell post-treatment but were not sustained. No ARR/NNT reported. |
Plant-Based Diet, Predominantly Māori Community
RCT · obesity + high HbA1c · 10 weeks
Only weight loss separated from control at 10 weeks; the HbA1c reduction was a non-significant trend. Read as a modest weight signal, not a glucose-lowering therapy.
Expert Commentary
This pragmatic, community-embedded randomised trial is best read as a modest positive signal for weight rather than as evidence of glycaemic benefit. Within the randomised, waitlist-controlled comparison, only weight loss reached statistical significance at 10 weeks; the apparent HbA1c reduction was a non-significant trend, and lipids, waist circumference, and activity did not separate from control. The frequently quoted figures of roughly 5 kg loss and sustained 3 kg at 36 months are drawn from the uncontrolled phase after every participant received the programme, so they cannot be attributed to the intervention with the same confidence and are vulnerable to regression to the mean and secular change. The headline claim that the programme reduced HbA1c and cholesterol versus control is not supported by the randomised data. A central limitation is the small sample (n=56), which leaves the trial underpowered for glycaemic and lipid endpoints, and the authors themselves flag inadequate HbA1c data capture. The design is unblinded and open-label, as is unavoidable for a dietary education programme, and outcome ascertainment is therefore susceptible to behaviour-change effects. Can I use this with my patients? Cautiously yes for motivated adults seeking a culturally grounded, group-based weight intervention, but not as a glucose-lowering therapy on this evidence. Larger, adequately powered trials with sustained follow-up of metabolic endpoints are needed before stronger claims are made.
References
Smith M, Wright N, McHugh P, Duncan B, Chwyl C. Randomised waitlist-controlled trial of a 10-week community programme using a plant-based diet in a predominantly Māori population in Tairāwhiti (Gisborne). BMJ Open. 2025;15(6):e080946. doi:10.1136/bmjopen-2023-080946