Association of body composition with risk of overall and site-specific cancers: A population-based prospective cohort study.

Although excess adiposity has been linked with various cancers, association between body composition and some cancers remains unclear, like lung and prostate cancers. We investigated associations of body composition with risk of overall cancer and major site-specific cancers in a prospective cohort of 454 079 cancer-free participants from UK-Biobank. Body composition was measured with bioimpedance analysis. We evaluated hazard ratio (HR) and 95% confidence interval (CI) with multivariate Cox linear and nonlinear models in men and women separately. We identified 27 794 cancers over 7.6 years of follow-up. Multivariable adjusted models including fat-free mass (FFM) and fat mass (FM) showed that FFM was positively associated with overall cancer risk in men and women (HR 1.03, 95% CI 1.01-1.04 and 1.07, 1.04-1.10, respectively); while the association between FM and overall cancer disappeared after adjusting for FFM. FFM was associated with higher risks of obesity-related cancers combined, stomach (women only), malignant melanoma, postmenopausal breast, corpus uteri, prostate, kidney (men only), and blood cancers and lower risk of lung cancer. FM was associated with higher risks of obesity-related cancers combined, esophageal, colon, lung (men only), postmenopausal breast (at the lower end of FM range), and corpus uteri cancers and lower risks of rectal, malignant melanoma (women only), prostate and blood cancers. FFM and FM seemed to have different effects on cancer risk, and the effects varied substantially by cancer type, in both direction and size. Higher FM/FFM ratio was also associated with some cancers risk, and might be a useful predictor of cancer risk.