Arsenic in Drinking Water and Human Biomonitoring in Eastern Croatia: An Exploratory Environmental Health Study
Abstract
Eastern Croatia includes settlements with historically elevated arsenic (As) concentrations in drinking water. Of 488 recruited adults, nine were excluded solely because no questionnaire was available; the exclusion was independent of measured arsenic concentrations, and the harmonised analytical cohort therefore comprised 479 adults from eight settlements. Total As concentrations in drinking water, hair, serum, and urine were measured in samples collected in June 2015 by inductively coupled plasma mass spectrometry. Kidney cancer data (ICD-10 C64) for 2000–2018 were obtained from the Croatian National Cancer Registry and analysed ecologically. MDCK1 cells were exposed to As2O3 at 10–200 µg/L for 48, 72, and 96 h, and acetylcholinesterase (AChE) activity was measured in human erythrocytes. Čepin had the highest median As concentrations in drinking water (191.75 µg/L) and urine (68.14 µg/L). Across all eight settlements, drinking-water As was strongly associated with urinary As (slope = 0.537; 95% CI, 0.352–0.722; R2 = 0.894; p < 0.001). However, in a sensitivity analysis excluding Čepin and using the seven non-Čepin settlement means reported, the association weakened (r = 0.740; R2 = 0.548; slope = 1.246; 95% CI, −0.054 to 2.547; p = 0.057), indicating substantial influence of the Čepin hotspot. Associations with hair As (R2 = 0.294; p = 0.165) and serum As (R2 = 0.239; p = 0.219) were not statistically significant. In a multivariable Gamma generalized linear model, residence in Čepin was associated with higher expected urinary As than residence in Vladislavci (adjusted mean ratio, 2.90; 95% CI, 1.90–4.42; p < 0.001), while female sex was associated with a lower expected concentration (adjusted mean ratio, 0.80; 95% CI, 0.66–0.97; p = 0.024). No statistically significant effect of As2O3 on MDCK1 viability or erythrocyte AChE activity was observed. The findings support urinary total As as a biomarker of recent environmental exposure, but do not establish causal associations with kidney toxicity or kidney cancer.