Mortality attributable to modifiable lifestyle factors in the Spanish cohort of the European prospective investigation into cancer and nutrition (EPIC) study

Abstract

BACKGROUND: There is consistent evidence on the relationship of unhealthy habits with mortality risk. The population attributable fraction (PAF) is a useful quantifier of the number of new disease cases attributable to modifiable risk factors. Our objective was to obtain the hazard ratios (HRs) and the PAFs for the major individual lifestyle factors of mortality in the Spanish cohort of the European Prospective Investigation into Cancer and Nutrition (EPIC) study. METHODS: Prospective study which included 40 307 EPIC-Spain participants (62.5% women). A total of 7 262 cases of all-cause mortality (53.9% were men) occurred during an average follow-up of 25.1 years. We calculated PAFs separately for men and women, using adjusted HRs of mortality (and 95% CI) from multivariate Cox regression models for binary categories of smoking, high alcohol consumption, obesity markers, poor Mediterranean diet adherence, and low physical activity. RESULTS: Most lifestyle factors studied were significantly associated with mortality in the study cohort. In final multivariate analyses, smoking registered the largest PAFs in men (21.0%, 95%CI 19.1-22.7%), while general obesity (BMI) accounted for the largest PAF in women (10.8%, 7.5-13.8%). Attributable fractions were similar in both sexes for poor adherence to the Mediterranean diet, whereas the lowest PAFs were estimated for physical activity. High alcohol consumption and central obesity did not pose a risk in women. CONCLUSIONS: Shifting from any risk factor to a healthier lifestyle would reduce mortality in both sexes. However, in women, alcohol consumption did not influence mortality risk and only BMI was associated with mortality risk attributable body adiposity.