Yijia Zhang
Anne E. Nigra
Melissa A. Lombard
Matthew O. Gribble
Liping Lu
Frederick W. Unverzagt
Leslie A. McClure
Suzanne E. Judd
Mary Cushman
John Brockman
Ka Kahe
Meghan Angley
2026
<div id="abssec0010"><h3 id="sectitle0015" class="u-h4 u-margin-m-top u-margin-xs-bottom">Background</h3><div id="abspara0010" class="u-margin-s-bottom">There are several pathways by which inorganic arsenic (iAs) exposure can affect cognition among adults. Few epidemiologic studies evaluate iAs in water and inter-individual differences in urinary arsenic toxicokinetics. We aimed to estimate the association between individual-level urinary arsenic biomarkers, county-level iAs in drinking water, and cognitive impairment in a cohort of Black and White Americans.</div></div><div id="abssec0015"><h3 id="sectitle0020" class="u-h4 u-margin-m-top u-margin-xs-bottom">Methods</h3><div id="abspara0015" class="u-margin-s-bottom">We evaluated the association between county-level iAs in water and urinary iAs metabolites with incident cognitive impairment in REasons for Geographic and Racial Differences in Stroke (REGARDS). Participants were linked to county-level concentrations of arsenic in public water (n = 15,516) and county-level probabilities of private well arsenic exceeding 10 μg/L (n = 20,448). In addition, urinary concentrations of iAs, monomethylarsonic acid (MMA) and dimethylarsinic acid (DMA) were measured in a subset of participants (n = 1,013). Cognitive impairment was determined by the Six Item Screener (SIS) and the Enhanced Cognitive Battery (ECB). We used multivariable logistic regression to estimate odds ratios.</div></div><div id="abssec0020"><h3 id="sectitle0025" class="u-h4 u-margin-m-top u-margin-xs-bottom">Results</h3><div id="abspara0020" class="u-margin-s-bottom">Approximately 9 % of participants in REGARDS developed incident cognitive impairment on the SIS over 10 years of follow-up. County-level public water iAs concentrations and private well iAs probabilities were not associated with increased odds of cognitive impairment. Higher concentrations of MMA (OR: 1.74 [95 % CI: 1.22, 2.49]) and iAs (OR: 1.58 [95 % CI: 1.12, 2.22]) in urine were associated with greater odds of cognitive impairment on the SIS. Results for the ECB were similar.</div></div><div id="abssec0025"><h3 id="sectitle0030" class="u-h4 u-margin-m-top u-margin-xs-bottom">Conclusions</h3><div id="abspara0025" class="u-margin-s-bottom">iAs exposure may be associated with cognitive impairment. Our findings highlight the importance of considering multiple measures of iAs exposure.</div></div>
application/pdf
10.1016/j.envres.2026.123768
en
Elsevier
Drinking water arsenic, urinary arsenic biomarkers, and cognitive impairment in the REGARDS study
article