Spontaneous pregnancy loss in humans and exposure to arsenic in drinking water

Abstract

Maternal exposure to high concentrations of inorganic arsenic (iAs) in naturally contaminated drinking groundwater sources has been associated with an increased risk for the spontaneous loss of clinically recognized pregnancies in several epidemiologic studies. Whereas a large worldwide population depends on drinking groundwater sources with high levels of iAs contamination, in quantities exceeding 10 parts per billion (ppb), an even larger population is likely to be exposed to mild-moderate drinking groundwater iAs contamination, in quantities <10 ppb. Only a single epidemiologic study to date has considered spontaneous pregnancy loss in association with consumption of drinking water with mild-moderate iAs contamination; the vast majority of published studies of spontaneous loss addressed populations with substantial exposure. The aim of this review is to evaluate the published literature to assess the plausibility for a causal association between exposure to iAs-contaminated drinking water and the spontaneous loss of clinically recognized pregnancy. In spite of numerous methodologic limitations resulting from circumstance or design, a consistent pattern of increased risk for loss is suggested by the epidemiologic literature. Moreover, these study results are corroborated by a large number of experimental studies, albeit usually conducted at concentrations exceeding that to which humans are exposed via contaminated drinking water. In this review, we discuss sources of human iAs exposure, highlight several experimental studies pertinent to a possible causal link between iAs and spontaneous pregnancy loss in humans, and provide a critical review of published epidemiologic studies of pregnancy loss and drinking water iAs exposures, and their limitations. Based on a review of the published literature, we recommend the future conduct of a two-stage comprehensive prospective study of low-moderate iAs drinking water exposure and spontaneous pregnancy loss.

Introduction

Arsenic (As), is a highly toxic and extensively distributed metalloid which is associated with a variety of adverse human health endpoints including spontaneous pregnancy loss (ATSDR, 2007). Widespread dietary exposure to the toxic, inorganic species of arsenic (iAs) occurs primarily through the consumption of geogenically contaminated drinking groundwater relied on by many human populations (WHO, 2004). Whereas reports frequently address the well-established carcinogenicity of iAs (WHO, 2004), a spectrum of non-carcinogenic chronic health effects, including adverse reproductive endpoints are also reported in association with the consumption of iAs contaminated drinking water (Smith and Steinmaus, 2009, Tchounwou et al., 2003). Several studies indicate an increase in the rate of spontaneous abortion, defined as loss of a clinically recognized pregnancy prior to 20 weeks completed gestation (Zegers-Hochschild et al., 2009), and the rate of stillbirth, defined as pregnancy losses following 20 weeks completed gestation (Golub et al., 1998, Vahter, 2009). Associations are also reported between consumption of iAs contaminated drinking water and rates of preterm birth, defined as live delivery prior to 37 weeks completed gestation (Ahmad et al., 2001), reduced birth weight (Huyck et al., 2007, Yang et al., 2003), congenital malformations (Kwok et al., 2006, Zierler et al., 1988), and neonatal and infant deaths, defined as deaths within the first 28 days following delivery and within the first postnatal year, respectively (Hopenhayn-Rich et al., 2000). Some of these adverse reproductive endpoints may share a common pathophysiology with regard to iAs exposure (He et al., 2007).

Whereas substantial experimental and limited human evidence underscore the likelihood of iAs associated spontaneous loss of clinically recognized pregnancies at high doses (Golub et al., 1998, Shalat et al., 1996, Vahter, 2009), little data is available to assess this likelihood at the low-moderate iAs concentrations to which large populations are exposed. The aim of this literature review is to evaluate and summarize the existing English language literature describing studies of drinking water iAs exposure and spontaneous pregnancy loss in humans, and to assess the plausibility for a causal association. Moreover, a secondary aim is the critical assessment of the methodologic limitations of previously published epidemiologic studies, and the provision of specific methodologic recommendations for the design of a future confirmatory study to elucidate the tenability of associations between low-moderate level exposure and spontaneous loss.