Abstract Background: Body mass index (BMI), a measure of weight standardised for height is a commonly used indicator for excess adiposity where studies have identified at least 13 obesity-related (OBR) cancer types. The aims of this thesis were: first, identify at scale the associations between relatively novel adiposity overweight-, obese- and waist circumference-years metrics with cancer incidence and compare their performance characteristics with a single BMI measure. Second, identify effect modifiers of the adiposity-cancer link. Third, identify ages where excess adiposity is related to a change in association with cancer risk. Methods: The Atherosclerosis Risk in Communities (ARIC) study, a small but deeply rich cohort with prospectively repeated BMI measurements across adulthood, was used to calculate overweight- and obese-years metrics (sum of the exposure time and degree of BMI greater than or equal to 25 kg/m2 and greater than or equal to 30 kg/m2 respectively) (Ch. 3). The ABACus 2 Consortium of 1 European and 4 United States cohorts was assembled, and overweight- and obese-year metrics were calculated (Ch. 4). Using repeated WC measurements in the ARIC study, waist circumference-years (incorporating exposure time to WC measurements greater than or equal to 102 cm in men and greater than or equal to 88 cm in women respectively) were calculated (Ch. 5). Cox proportional hazards regression modelled associations of these metrics with cancer incidence and Harrells C-statistic was used to compare performance characteristics of metrics with a single BMI measure. Using the ABACus 2 Consortium, interactions by sex, race, smoking and hormone replacement therapy (HRT) of the adiposity-cancer link were explored (Ch. 6) and Sensitive Period Analysis across ages 30-65 identified periods associated with higher and lower risk of OBR cancers (Ch. 7). IPD meta-analyses combined results of cohorts in the ABACus 2 Consortium. Results III (Ch. 3): 13,463 ARIC participants had hazard ratios (HRs) per standard deviation (SD) overweight-years of 1.15 (95% CI:1.05,1.25) in men and 1.14 (95% CI:1.08,1.20) in women for OBR cancers. There were non-significant differences in the performance characteristics of overweight-years and single BMI. Results IV (Ch. 4): For 720,210 participants in the ABACus 2 consortium, HRs per SD overweight-years were 1.15 (95% CI:1.13,1.17) in men and 1.08 (95% CI:1.04,1.13) in women for OBR cancers. Marginal differences were found in performance characteristics of overweight-years and single BMI. Results V (Ch. 5): For 10,172 ARIC participants, HRs per SD waist circumference-years were 1.14 (95% CI:1.04,1.25) in men and 1.19 (95% CI:1.12,1.27) in women for OBR cancers. Minimal differences were found in performance characteristics between waist circumference-years, single BMI and WC. Results VI (Ch. 6): Significant interactions of the excess adiposity and cancer link were found by sex for colorectal cancer; by race for pancreatic cancer in men and women; by smoking for pancreatic cancer in men and by HRT for endometrial, ovarian, and postmenopausal breast cancers. Results IV (Ch. VII): 720,210 participants were included. No significant interactions were found by age per SD BMI for OBR cancers except at ages 30 and 40 in women and ages 35 to 65 for postmenopausal breast cancer in women. Conclusion: This thesis included an IPD meta-analyses of >720,000 participants from 10 countries which found that degree and duration of excess adiposity were associated with OBR cancer incidence, but performance characteristics were similar to that of single BMI. Sex, race, smoking and HRT interaction terms should be incorporated in research and prevention strategies and exposure to excess adiposity across ages 30 to 65 in women contributes to cancer risk.
- lifecourse
- BMI
- obesity
- overweight
- cancer
Adulthood BMI and Cancer Risk Using Adiposity Metrics (ABACus 2): An Individual Participant Data Meta-Analysis
Hawwash, N. (Author). 1 Jan 1824
Student thesis: Phd