The EPIC-Norfolk Study
Consumption of fatty foods and incident type 2 diabetes in populations from eight European countries.
European Journal of Clinical Nutrition 2014 ; 69: 455-61.
Buijsse B, Boeing H, Drogan D, Schulze MB, Feskens EJ, Amiano P, Barricarte A, Clavel-Chapelon F, de Lauzon-Guillain B, Fagherazzi G, Fonseca-Nunes A, Franks PW, Huerta JM, Jakobsen MU, Kaaks R, Key TJ, Khaw KT, Masala G, Moskal A, Nilsson PM, Overvad K, Pala V, Panico S, Redondo ML, Ricceri F, Rolandsson O, Sánchez MJ, Sluijs I, Spijkerman AM, Tjonneland A, Tumino R, van der A DL, van der Schouw YT, Langenberg C, Sharp SJ, Forouhi NG, Riboli E, Wareham NJ, and InterAct Consortium
DOI : 10.1038/ejcn.2014.249
PubMed ID : 25424603
PMCID :
URL : https://www.nature.com/articles/ejcn2014249
Abstract
Diets high in saturated and trans fat and low in unsaturated fat may increase type 2 diabetes (T2D) risk, but studies on foods high in fat per unit weight are sparse. We assessed whether the intake of vegetable oil, butter, margarine, nuts and seeds and cakes and cookies is related to incident T2D.
A case-cohort study was conducted, nested within eight countries of the European Prospective Investigation into Cancer (EPIC), with 12,403 incident T2D cases and a subcohort of 16,835 people, identified from a cohort of 340,234 people. Diet was assessed at baseline (1991-1999) by country-specific questionnaires. Country-specific hazard ratios (HRs) across four categories of fatty foods (nonconsumers and tertiles among consumers) were combined with random-effects meta-analysis.
After adjustment not including body mass index (BMI), nonconsumers of butter, nuts and seeds and cakes and cookies were at higher T2D risk compared with the middle tertile of consumption. Among consumers, cakes and cookies were inversely related to T2D (HRs across increasing tertiles 1.14, 1.00 and 0.92, respectively; P-trend <0.0001). All these associations attenuated upon adjustment for BMI, except the higher risk of nonconsumers of cakes and cookies (HR 1.57). Higher consumption of margarine became positively associated after BMI adjustment (HRs across increasing consumption tertiles: 0.93, 1.00 and 1.12; P-trend 0.03). Within consumers, vegetable oil, butter and nuts and seeds were unrelated to T2D.
Fatty foods were generally not associated with T2D, apart from weak positive association for margarine. The higher risk among nonconsumers of cakes and cookies needs further explanation.