To compare and analyze the distribution characteristics of fluoride and arsenic in drinking water in coal-burning borne endemic fluorosis and arsenic poisoning areas and non-endemic areas of Guizhou Province, and to assess the health risk.

From January 2018 to December 2019, drinking water samples were collected from all endemic and non-endemic townships in 37 coal-burning borne endemic fluorosis counties (cities and districts) and 4 coal-burning borne endemic arsenic poisoning counties (cities and districts) in Guizhou Province. Fluoride and arsenic levels in drinking water were determined in all the affected and non-affected townships. Among them, 3 800 water samples were collected from fluorosis areas, 1 445 water samples from non-endemic areas, and 117 water samples from arsenic poisoning areas and non-endemic areas, respectively. Based on Monte Carlo simulation, a health risk assessment model recommended by the US Environmental Protection Agency (EPA) was applied to quantitatively evaluate the health risks caused by fluoride and arsenic in drinking water in different regions through different pathways (drinking water intake or skin contact) to different populations (adult males, adult females and children).

According to the "Hygienic Standards for Drinking Water" (GB 5749-2022), the qualified rates of fluoride and arsenic levels in drinking water in the study area were 100%. The results of the health risk evaluation showed that in the same evaluation area (endemic area or non-endemic area), there were statistically significant differences in the average annual health risks caused by fluoride and arsenic exposure through drinking water among different populations through two different routes (P < 0.001), with children's health risks > adult males > adult females (P < 0.05). The average annual health risks of fluoride and arsenic in water samples from different water periods and source types were compared via the two exposure routes, and the differences were statistically significant (P < 0.05). At high (P₉₅) and moderate (P₅₀) exposure levels, the average annual total health risks of fluoride in drinking water via the two exposure routes (drinking water intake and skin contact) for adult males, adult females, children and the general population in fluorosis areas were a^-1: 2.50 × 10^-9 and 8.64 × 10^-10, 2.20 × 10^-9 and 7.61 × 10^-10, 3.00 × 10^-9 and 1.04 × 10^-9, 2.36 × 10^-9 and 8.15 × 10^-10, respectively, and a^-1: 2.44 × 10^-9 and 6.92 × 10^-10, 2.15 × 10^-9 and 6.09 × 10^-10, 2.93 × 10^-9 and 8.32 × 10^-10, 2.30 × 10^-9 and 6.53 × 10^-10 in non-endemic areas, respectively; the average annual total health risks of adult males, adult females, children, and the general population caused by arsenic ingestion through drinking water and skin contact in arsenic poisoning areas were a^-1: 9.10 × 10^-7 and 3.67 × 10^-7, 8.11 × 10^-7 and 3.22 × 10^-7, 1.13 × 10^-6 and 4.45 × 10^-7, 8.95 × 10^-7 and 3.47 × 10^-7, respectively, and a^-1: 2.42 × 10^-6 and 5.61 × 10^-7, 2.11 × 10^-6 and 4.80 × 10^-7, 2.88 × 10^-6 and 6.55 × 10^-7, 2.21 × 10^-6 and 5.25 × 10^-7 in non-endemic areas, respectively; average annual total health risks of adult males, adult females, children and the general population were all in the safe range [fluoride in drinking water: negligible levels recommended by the Royal Society and the Dutch Ministry of Construction and Environment were 1.0 × 10^-7 a^-1 and 1.0 × 10^-8 a^-1, respectively; arsenic in drinking water: the maximum acceptable level recommended by the International Commission on Radiological Protection (ICRP) was 5.0 × 10^-5 a^-1].

The fluoride and arsenic levels in the drinking water of coal-burning borne endemic fluorosis and arsenic poisoning areas and non-endemic areas in Guizhou Province meet the requirements of hygienic standards for drinking water; the overall health risks of fluoride and arsenic to people of different ages and genders ingested through drinking water and through skin contact are within an acceptable range.

Drinking water; Fluoride; Arsenic; Health risk assessment; Monte Carlo simulation