Variability of body mass index and risks of prostate, lung, colon, and ovarian cancers

OBJECTIVE: We investigated the association between cancer incidence and body mass index (BMI) variability calculated from the recall of weight at decades of age by participants in the USA Prostate, Lung, Colorectal and Ovarian (PLCO) Cancer Screening Trial. METHODS: A total of 89,822 individuals' BMI were recorded as recalled the participant's aged 30, 40, 50, 60, 70 years, and baseline. BMI variability was assessed using four indices: SD, coefficient of variation (CV), variability independent of the mean (VIM), and average real variability (ARV). The multivariate Cox regression analysis was performed to calculate hazard ratios (HRs) of these measures for incident cancers and corresponding 95% CIs. RESULTS: During the median follow-up of 11.8 years, there were newly diagnosed 5,012 cases of prostate cancer, 792 cases of lung cancer, 994 cases of colon cancer, and 132 cases of ovarian cancer. Compared with the lowest quartile (Q1) group, the highest quartile (Q4) group of BMI variability indices was associated with increased lung cancer risk, including BMI_SD (HR, 1.58; 95% CI, 1.17–2.12), BMI_CV (HR, 1.46; 95% CI, 1.10–1.94), BMI_VIM (HR, 1.73; 95% CI, 1.33–2.25), and BMI_ARV (HR, 2.17; 95% CI, 1.62–2.91). Associations between BMI variability and prostate, colon, and ovarian cancer incidences were of limited significance. CONCLUSION: The findings imply that maintaining a stable weight across adulthood is associated with a decreased incidence of lung cancer.