Abstract
Objectives/Goals: Arsenic (As) in drinking water and rice is a major public health issue. In areas where rice is a staple food, it is unclear whether As in water, raw rice, or cooked rice best represents total exposure. Here, we disentangle the relative contributions of dietary rice and drinking water to total As exposure to help prioritize As measurement and intervention. Methods/Study Population: We used samples from the Health Effects of Arsenic Longitudinal Study (HEALS) cohort based in Araihazar, Bangladesh, an area where there is substantial As contamination in well water and where the diet is rice-rich. We measured total As in matched water, raw rice, cooked rice, and urine samples collected in 2016 from 396 households in HEALS. We also measured As species (organic vs. inorganic As) from a 194-household subsample. We compared the levels of total, organic, and inorganic As in rice before and after cooking across levels of As in cooking water. Using urinary As as a biomarker of internal As dose, we calculated which matrix (raw rice, cooked rice, or cooking water) or combination of matrices explained the most variance in true As exposure using linear models adjusted for age and sex. Results/Anticipated Results: Inorganic As (iAs) is the most toxic As form. We found a nonlinear relationship between water and rice iAs wherein rice iAs decreased if cooked with low-As water but increased with high-As water. Accordingly, cooked rice was overall a better indicator of urinary iAs than raw rice, since it captured variation in both rice and water. However, when As levels in rice and water were below safety limits, neither matrix explained meaningful variance in urinary As. This suggests that in populations where rice and water As are low (e.g., the USA), other exposure sources predominate. Organic As in rice moderately decreased after cooking, but starting abundance of organic As in rice varied greatly and correlated with organic As in urine. Therefore, rice may be a previously under-appreciated exposure source of organic As. Discussion/Significance of Impact: Cooked rice As was the most informative single exposure measurement, but neither rice nor water As predict internal As dose when at low levels. These results highlight that As exposure measurements should be tailored to the study population to maximize research resources and that multiple As exposure routes should be considered simultaneously.