Association of social determinants of health (SDH), allostatic load (AL) and incident cancers in the multi-ethnic study of atherosclerosis (MESA).

10562 Background: AL may represent accumulated physiologic stress from adverse SDH. We assess the relationship of AL and incident cancer risk in MESA, a cohort free of cardiovascular disease (CVD) at baseline. Methods: MESA participants at Exam 1 (July 2000-August 2002) were included, except those with prevalent cancers other than non-melanoma skin, median follow-up, 16.5 years. We assigned the composite AL measure 1 point for each individual biomarker in the worst quartile for total cholesterol, high density lipoprotein cholesterol, C-reactive protein, and IL-6. One point each was assigned for BMI>25 or 1.2mg/dL (women) or >1.4mg/dL(men), systolic (≥140mmHg), diastolic (≥90mmHg) blood pressure; heart rate>100bpm; coronary artery calcium score>0 and history of diabetes. AL score (mean=3.09, SD=1.76) ranged 0-12. Other covariates included sociodemographic variables. Where missing, data for individual biomarkers were imputed using Multivariate Imputation by Chained Equations in R prior to calculating AL score. Cox regression models assessed AL in the whole population and stratified by race/ethnicity. Results: 6808 participants comprised the analytic sample; 1022 (15%) were diagnosed with new cancer during follow-up. Chinese (adjusted hazard ratio aHR, 95% confidence interval 95%CI, 0.46,0.36-0.60) and Hispanic/Latino (aHR,95%CI 0.60,0.50-0.72) participants were less likely to develop new cancers compared with White participants. Each AL point increase was associated with increased new cancer risk (aHR,95%CI 1.07,1.03-1.11). In models stratified by race/ethnicity, AL association with new cancer risk persisted only for Black (aHR,95%CI 1.07, 1.00-1.14, p=0.04) and Hispanic/Latino (aHR,95%CI 1.17, 1.07-1.28, p=0.001) participants. Conclusions: AL as a composite measure of physiologic stress and self-reported Chinese and Hispanic/Latino race/ethnicity independently correlated with incident cancer risk in a cohort free of CVD at baseline. Race/ethnicity may moderate the effects of AL. Table: see text