Fig 1 BMI according to frequency of fried food consumption and thirds of genetic risk score in pooled data of three cohorts. Data adjusted for age, source of genotyping data, physical activity, television watching, smoking, alcohol intake, intake of sugar sweetened beverages, alternative healthy eating index, and total energy intake in Nurses’ Health Study (NHS) and Health Professionals Follow-Up Study (HPFS); and age, physical activity, smoking, alcohol intake, intake of sugar sweetened beverages, alternative healthy eating index, and total energy intake In the Women’s Genome Health Study (WGHS). Data from three cohorts were pooled by means of fixed effects meta-analyses (if P≥0.05 for heterogeneity between studies) or random effects meta-analyses (if P<0.05 for heterogeneity between studies)