You are here:
Health effects associated with remotely sensed cyanobacteria blooms in a drinking water reservoir in Massachusetts
Citation:
Wu, J., E. Hilborn, B. Schaeffer, E. Urquhart, M. Coffer, C. Lin, AND A. Egorov. Health effects associated with remotely sensed cyanobacteria blooms in a drinking water reservoir in Massachusetts. American Geophysical Union, San Francisco, California, December 09 - 13, 2019.
Impact/Purpose:
Cyanobacteria harmful algal blooms are increasingly relevant to US surface water quality. Although blooms occur frequently in source water, associated health effects are infrequently reported. This is an Agency-relevant issue as cyanobacteria toxins are listed on the drinking water Contaminant Candidate Lists. We conducted an analysis of health effects associated with the remotely-sensed presence of cyanobacteria in a surface drinking water source.
Description:
Fresh water cyanobacterial blooms represent a potential human health risk. Exposure to harmful cyanobacteria is associated with several nonspecific health effects including respiratory, dermal, and gastrointestinal (GI) illnesses. We investigated the association between concentrations of cyanobacteria in a single public drinking water source and counts of emergency department (ED) visits for GI, respiratory, and dermal illnesses among Massachusetts residents in the drinking water service area. Cyanobacteria cell counts were estimated using satellite imagery (MERIS Cyanobacteria Index) from years 2008 – 2011, using pixels within the water body that were closest to the drinking water intake. Satellite images were available at approximately 3-day intervals in the spring-summer-fall period, and less regularly in the winter due to snow/ice/cloud. Hospital ED data were obtained from the Massachusetts Center for Health Information Analysis. Time-series data analysis was conducted in two stages. First, predictive autoregressive generalized additive models for Poisson-distributed outcome were fitted to daily ED visit counts using a spline function of time, lagged meteorological parameters, air quality data and indicator variables for weekdays and holidays as predictors. Second, residuals from the 1st stage models were then regressed against categorized cyanobacteria count data with a two-day lag corresponding to the average water residence time in the distribution system. We found that cyanobacteria counts above the 75th percentile were associated with additional 4.3 (95% confidence limits 0.6, 8.0) lagged cases of respiratory illness (p=0.02) compared to cyanobacteria counts below the detection limit of MERIS. In the population of approximately 1.4 million people this effect estimate corresponds to about 0.3 excess daily ED visits for respiratory illness per 100,000 individuals. There were no significant associations between estimated cyanobacterial counts and lagged data on ED visits for GI or dermal illnesses. Our study is the first to demonstrate an association between cyanobacteria in source water as measured using remote sensing and lagged respiratory illness suggesting adverse impacts of inhalation exposure to contaminated tap water. This abstract does not reflect EPA policy.