Considerations when using longitudinal cohort studies to assess dietary exposure to inorganic arsenic and chronic health outcomes

Dietary arsenic exposure and chronic health outcomes are of interest, due in part to increased awareness and data available on inorganic arsenic levels in some foods. Recent concerns regarding levels of inorganic arsenic, the primary form of arsenic of human health concern, in foods are based on extrapolation from adverse health effects observed at high levels of inorganic arsenic exposure; the potential for the occurrence of these health effects from lower levels of dietary inorganic arsenic exposure has not been established. In this review, longitudinal cohort studies are evaluated for their utility in estimating dietary inorganic arsenic exposure and quantifying statistically reliable associations with health outcomes. The primary limiting factor in longitudinal studies is incomplete data on inorganic arsenic levels in foods combined with the aggregation of consumption of foods with varying arsenic levels into a single category, resulting in exposure misclassification. Longitudinal cohort studies could provide some evidence to evaluate associations of dietary patterns related to inorganic arsenic exposure with risk of arsenic-related diseases. However, currently available data from longitudinal cohort studies limit causal analyses regarding the association between inorganic arsenic exposure and health outcomes. Any conclusions should therefore be viewed with knowledge of the analytical and methodological limitations.