Dietary arsenic exposure in bangladesh

Abstract

Background: Millions of people in Bangladesh are at risk of chronic arsenic toxicity from drinking contaminated groundwater, but little is known about diet as an additional source of As exposure.

Methods: We employed a duplicate diet survey to quantify daily As intake in 47 women residing in Pabna, Bangladesh. All samples were analyzed for total As, and a subset of 35 samples were measured for inorganic arsenic (iAs) using inductively coupled plasma mass spectrometry equipped with a dynamic reaction cell.

Results: Median daily total As intake was 48 microg As/day [interquartile range (IQR), 33-67) from food and 4 microg As/day (IQR, 2-152) from drinking water. On average, iAs comprised 82% of the total As detected in dietary samples. After adjusting for the estimated inorganic fraction, 34% [95% confidence interval (CI), 21-49%] of all participants exceeded the World Health Organization's provisional tolerable daily intake (PTDI) of 2.1 microg As/kg-day. Two of the 33 women who used a well with < 50 microg As/L exceeded this recommendation.

Conclusions: When drinking water concentrations exceeded the Bangladesh drinking water standard of 50 microg As/L, ingested water was the dominant source of exposure. However, as drinking water As concentrations decrease, the relative contribution of dietary As sources becomes more important to ingested dose. The combined intake from both diet and drinking water can cause some individuals to exceed the PTDI in spite of using a tube well that contains < 50 microg As/L.

Figure 1

Distribution of average daily As intake (μg/day) from both drinking water and dietary sources for all 47 participants sorted by tube well As concentration.

Figure 2

Average dietary As intake (μg/day) plotted against the logarithm of drinking water As (μg/L). The mean and median regressions were obtained from the model Y = (α + β) × (log₁₀ X + ɛ), where either the mean or median of ɛ is zero. For the median regression line, α = 40.5 and β = 11.6; for the mean regression line, α = 46.0 and β = 19.5. Shaded areas represent 1 SE around each regression line.