The Association of the Planetary Health Diet With Type 2 Diabetes Incidence: Findings From the EPIC-Norfolk Study
Current developments in nutrition 2025 ; 9: .
Forouhi NG, Forouhi N, Wareham NJ, Wareham N, Sowah SA, Imamura F, Ibsen DB
DOI : 10.1016/j.cdnut.2025.106490
URL : https://doi.org/10.1016/j.cdnut.2025.106490
Abstract
Objectives: The EAT-Lancet reference diet was proposed as a planetary health diet (PHD) in 2019 to help improve dietary impacts on both population health and the environment globally. However, evidence is limited on the association of adherence to the PHD with T2D incidence in different world regions. We aimed to assess the association in a long-term observational study in the UK.
Methods: We assessed adherence to the PHD by scoring individuals’ diets (score range: 0 – 140) using a published algorithm and dietary data collected with food frequency questionnaires up to three times, from the baseline (1993-1997) to the second follow-up by 2011 in the EPIC-Norfolk study in the UK (n adults=23,722). Participants were followed-up until March 2020, and Cox regression models, accounting for time-varying covariates where available, were used to analyze the association of the PHD adherence with T2D incidence. We modelled the population attributable fraction (PAF) for T2D incidence that could have been prevented if all the participants had achieved a score equal to the 80th percentile, and causally experienced lower T2D incidence. Finally, we explored the association between the PHD and four causal risk factors for T2D: BMI, waist circumference, plasma glucose, and hemoglobin A1c.
Results: Over a mean of 19.4 (SD 6.8) years of follow-up, 3,496 cases of incident T2D were recorded. Greater adherence to the PHD was associated with lower T2D incidence; comparing the highest PHD quintile (85.7–117.8 points) to the lowest (33.9–68.4 points), the HR (95% CI) was 0.68 (0.61, 0.76) in the most adjusted model including socio-demographic, behavioural factors, energy intake, adiposity, and prevalent cardiovascular disease or cancer. We estimated that 12.3% (95% CI: 9.2, 15.3) of the population burden of incident T2D could be attributable to lower adherence to the PHD, defined as a PHD score below the 80th percentile (85.7 points). Adherence to the PHD showed negative associations with adiposity markers and hyperglycemia.
Conclusions: The EAT-Lancet PHD was associated inversely with T2D incidence, adiposity and glycemia. These findings suggest that an environmentally sustainable approach to reducing the risk of T2D may be achieved through increased adherence to the PHD diet.