Objective:

To explore regional differences in the impact of surface water consumption on Parkinson disease (PD) risk in the United States (U.S.)

Background:

Although private well water and its quality has been a suspected risk factor for PD, public water, including from surface water sources, is also subject to contamination by a variety of environmental pollutants. We previously reported a greater PD risk in people living in residences served by public water, based on a large, nationwide, population-based sample. 

Design/Methods:

Using the same population-based sample (89,790 incident PD cases; 21,549,400 comparable controls) from Medicare data in 2009, we linked residential zip code to 1990 Census block data for residential water source and assigned individuals to one of 18 watersheds in the U.S. For each watershed, we constructed a logistic regression model to estimate the relative risk (RR) of PD when using public water vs. private well water. We adjusted for demographics, smoking, and healthcare utilization. We tested formally for interaction between PD, reliance on public water, and watershed using a likelihood ratio test, and mapped public water-PD RRs. 

Results:

We found the relationship between public water consumption and PD depends on which of the 18 watersheds in which each individual resides (interaction p-value<0.001). Residents of the Rio Grande watershed had the greatest risk of PD in relation to reliance on public water vs. private well water (RR=1.63, 95% CI 1.18-2.25). Overall, risk associated with public water tended to be greater in watersheds in the western vs. eastern U.S. (interaction p-value<0.001). 

Conclusions:

The possible increased risk of PD in relation to surface water consumption is due to associations in several watersheds. Further study will help determine the potential causal contaminants in each watershed.