Academic journal article Journal of Health Population and Nutrition

Reduction in Urinary Arsenic with Bottled-Water Intervention

Academic journal article Journal of Health Population and Nutrition

Reduction in Urinary Arsenic with Bottled-Water Intervention

Article excerpt


The study was conducted to measure the effectiveness of providing bottled water in reducing arsenic exposure. Urine, tap-water and toenail samples were collected from non-smoking adults residing in Ajo (n=40) and Tucson (n=33), Arizona, USA. The Ajo subjects were provided bottled water for 12 months prior to re-sampling. The mean total arsenic (µg/L) in tap-water was 20.3±3.7 in Ajo and 4.0±2.3 in Tucson. Baseline urinary total inorganic arsenic (µg/L) was significantly higher among the Ajo subjects (n=40, 29.1±20.4) than among the Tucson subjects (n=32, 11.0±12.0, p<0.001), as was creatinine-adjusted urinary total inorganic arsenic (µg/g) (35.5±25.2 vs 13.2±9.3, p<0.001). Baseline concentrations of arsenic (µg/g) in toenails were also higher among the Ajo subjects (0.51±0.72) than among the Tucson subjects (0.17±0.21) (p<0.001). After the intervention, the mean urinary total inorganic arsenic in Ajo (n=36) dropped by 21%, from 29.4±21.1 to 23.2±23.2 (p=0.026). The creatinine-adjusted urinary total inorganic arsenic and toenail arsenic levels did not differ significantly with the intervention. Provision of arsenic-free bottled water resulted in a modest reduction in urinary total inorganic arsenic.

Key words: Arsenic; Drinking-water; Bottled water; United States


Exposure of humans to inorganic arsenic is associated with an increased risk of lung, bladder, skin and other cancers (1,2). Although inorganic arsenic is found throughout the environment, drinking-water constitutes the most significant source of exposure for most populations. Moschandreas et al. have estimated the contribution of drinking-water to the overall daily arsenic exposure to be 35%. Exposure of inorganic arsenic through ingestion (food and water) also depends on certain demographic characteristics, such as age, race/ethnicity, and poverty level (3).

On 22 January 2001, the U.S. Environmental Protection Agency lowered the maximum contaminant level (MCL) allowable for arsenic in drinking-water in the public water systems, from 50 µg/L to 10 µg/L, beginning 23 January 2006 (4). Although this will reduce arsenic exposure of drinking-water from the public water systems that currently do not meet these standards, it will require significant expenditure for modifications to these systems. In addition, exposure at the new standard of the U.S. Environmental Protection Agency may still confer health risks. The costs involved in reduction of arsenic may not be a reasonable option for small municipal water systems or households using personal well-water. Hence, arsenicfree bottled water may be a safe alternative in these settings.

In this study, we compared baseline urinary and toenail arsenic in Ajo and Tucson, Arizona, USA and evaluated the impact of providing arsenic-free bottled water, for one year, on biomarkers of arsenic exposure in Ajo. The primary hypothesis of the study was that provision of bottled-water supplies would reduce arsenic exposure as measured by total inorganic arsenic species in urine and concentrations of arsenic in toenails.


Study area and selection of households

Census blocks within census tracts were initially selected at random. Based on a probability proportional to size (PPS) sampling protocol (5), our goal was to recruit a maximum of five households per census block. Our inclusion criteria required at least three years of continuous residence in Ajo at the time of recruitment, age over 18 years, exclusive use of tap-water for drinking and food preparation, and no current smoking. Given our restrictive inclusion criteria, we had difficulty in recruiting the sufficient number of subjects using the PPS protocol in Ajo. Subsequently, we resorted to a census of the entire community. In Tucson, the five census tracts that most closely resembled the Ajo population in age distribution (median age= 52 years) and percentage of Hispanic residents (38%) in the 2000 census were selected. …

Author Advanced search


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.