Project Summary/Abstract
The Flint water crisis and the resulting lead exposure to its population have brought national attention to
the issue of safe drinking water. Concerns about safe drinking water are not confined to lead nor to Flint. In
2015, 25 percent of the US population was served by a water district with at least one violation of the Safe
Drinking Water Act, the main policy regulating drinking water in the United States. This project focuses on two
contaminants, arsenic and nitrates, that are of a particular concern for human health. In determining the safe
levels of contaminants, the Environmental Protection Agency uses the estimated health effects of the
contaminants as a guide. However, policy discussions (e.g., the National Research Council 1999) point out that
the evidence on the health effects of these contaminants is incomplete, making it challenging to determine the
appropriate levels for regulation.
This project combines unique drinking water contaminant data tracking arsenic and nitrate levels
(California Division of Drinking Water Water Quality Analyses Database) with detailed infant health data
(California Birth Cohort data) and uses a quasi-experimental design to further understanding of the effects of
drinking water contaminants on infant health (Specific Aim 1). Cross-sectional comparisons of mothers with
varying exposure to contaminants may be fraught with biases. In particular, contrasts across residential locations
will likely combine the effects of contaminants and the effects of economic conditions. In contrast, this project
will primarily utilize sibling comparisons. As many of the factors affecting infant health are similar for siblings
(e.g., family background, family income), such a research design likely provides more accurate estimates of the
effects of drinking water contaminants by netting out the influence of these shared familial factors. Typical
datasets on siblings have limited sample sizes, and thus give imprecisely-measured effects. However, for this
project, a large dataset on siblings will be formed from the California Birth Cohort data by matching siblings
based on their mother’s name and date of birth. Using maternal residential location, the infant health data will be
merged with a unique dataset that tracks arsenic and nitrates for the 8,000 public water districts in California.
The sibling research design is complemented with a novel analysis of the impact of large and discontinuous
shifts in contaminant levels within water districts, such as those caused by California’s 2008 adoption of a
national revision to the maximum contaminant level for arsenic in public water systems.
This research will be informative for regulation and policy discussions relating to drinking water. In
addition to evaluating the causal health impact of arsenic and nitrate drinking water contamination (Specific Aim
1), this project will assess whether current regulatory thresholds for these contaminants sufficiently protect
infants from adverse outcomes (Specific Aim 2). To target policy, the project will offer an understanding of how
exposure to contaminants varies across racial, ethnic, age, and economic groups (Specific Aim 3).