Association of predicted body compositions with the risk of colorectal cancer in the Chinese population: A 15-year prospective cohort.

Given that body mass index (BMI) fails to distinguish between fat and muscle, and that the relationship between body composition and colorectal cancer (CRC) risk in the Chinese population is insufficiently understood, we prospectively examined the associations between predicted body compositions and CRC risk in a large cohort. Cox proportional hazards regression models were used to evaluate hazard ratios (HRs) and 95% confidence intervals (CIs) for the relationships between predicted body compositions and CRC risk. Over a mean follow-up period of 14.88 years, we observed 825 incident cases of CRC. To address potential reverse causation, we additionally fitted lagged Cox models with delayed entry at 2 and 4 years after baseline. In multivariable-adjusted models, individuals in the fourth and fifth quintiles of predicted fat mass exhibited HRs of 1.31 (95% CI 1.04-1.64) and 1.38 (95% CI 1.10-1.73), respectively, compared with those in the first quintile. Predicted body fat percentage and BMI showed similar positive associations with CRC risk, whereas predicted lean body mass was not significantly associated; these associations were broadly similar in the 4-year lag analysis. Restricted cubic spline models revealed positive dose-response associations between these predicted fat mass, body fat percentage, and BMI, and CRC risk, with evidence of non-linearity for predicted fat mass. Elevated levels of predicted fat mass, predicted body fat percentage, and BMI were associated with an increased risk of CRC in this Chinese cohort.