Abstract SY02-03: Mediators of the effect of overweight and obesity on cardiovascular disease and cancer: Evidence from pooling of prospective studies

Abstract

Abstract Cardiovascular disease (CVD) and cancer share lifestyle risk factors such as smoking, alcohol use, physical inactivity and overweight/obesity. The massive increase in prevalence of obesity worldwide was once predicted to reverse the gains in life expectancy in the developed world. However, successful control of major mediators of the effect of obesity on CVD, including hypertension and dyslipidemia, in the developed world has led to substantial and continuous declines in CVD incidence and mortality. Obesity also increases the risk of several important cancers, including colorectal, kidney and breast cancer. Therefore, it is essential to investigate the potential mediators of the effect of obesity on cancer and ways to control these mediators to reduce the harmful effects of obesity on cancer. We have previously published using data from 97 prospective cohort studies on the direct and mediated effects of overweight/obesity on CVD. Our results indicated that about half of the excess risk of overweight/obesity on coronary heart disease and about two thirds of its effect on stroke is mediated by blood pressure, serum cholesterol and blood glucose collectively. Blood pressure was the most important mediator for the effect of obesity on CVD. Here, we apply similar methods of mediation analysis to evaluate the role of type-2 diabetes as a potential mediator of the effect of overweight/obesity on cancer using data from two cohorts in the US. Methods and Results We used data from Framingham Offspring Study and Women's Health Initiative - Clinical Trial in the US. The two cohorts combined included 517 cancer events. Body mass index was measured at baseline in all cohorts and diabetes was defined as either a self-reported diagnosis of diabetes or having a high blood glucose measurement in the same or a subsequent visit. We used methods developed recently in the causal inference literature to estimate what proportion of the effect of obesity on cancer was mediated by type-2 diabetes. We estimated the direct and indirect hazard ratios using two separate regression models in each cohort: we used a logistic regression model to estimate the association between obesity and diabetes and used a Cox proportional hazards model to quantify the impact of obesity on cancer. We pooled these hazard ratios using random-effect models in a meta-analyses. We also investigated the potential interaction between obesity and diabetes in causing cancer by adding product terms to the outcome regression model. Our results indicated that 30% of the effect of overweight and 10% of the effect of obesity on the selected cancers were mediated by diabetes. We found no interaction between overweight/obesity and diabetes (P values ranged from 0.4 to 0.8). Conclusion Controlling diabetes may reduce the harmful effect of obesity on cancer. Recent advances in causal inference and mediation analysis can be readily applied to estimate the direct and mediated effects of lifestyle and metabolic risk factors on cancer. Citation Format: Goodarz Danaei. Mediators of the effect of overweight and obesity on cardiovascular disease and cancer: Evidence from pooling of prospective studies. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr SY02-03. doi:10.1158/1538-7445.AM2015-SY02-03