, 1995 and Howard et al., 1999). Diabetes and albuminuria, as dominant independent risk factors for myocardial infarction and CHD overall in these populations, were particularly

strong risk factors for these diseases in women and in Arizona (Howard et al., 1995 and Howard Alectinib ic50 et al., 1999). A higher prior rate of diabetes in Arizona participants (70%) than in participants in Oklahoma and the Dakotas (>40%) (Howard et al., 1999) indicates that correction for risk factors such as LDL cholesterol and smoking but not diabetes and albuminuria could result in the appearance of higher CVD risk in Arizona relative to the other regions. Additional research using prospective cohort designs and mechanistic studies is needed to examine the relationship between low-level iAs exposure and diabetes and kidney function. Further evaluation of whether dietary or population specific factors underlie the association between DMA and diabetes, albuminuria, and CVD risk is also needed. A POD for a significantly increased risk of CVD mortality around 100 μg/L (about 9 μg/kg-day

for this population) based on the study by Chen et al. (2011) is broadly supported by other studies meeting our initial inclusion criteria (particularly those with larger sample size and narrower exposure ranges for groups in the low-exposure region) (Table 1). Nevertheless, statistically non-significant, positive associations that increase with exposure may appear to support a dose–response relationship below 100 μg/L.

Non-differential exposure misclassification, although often cited as an explanation for Selleckchem Z VAD FMK the lack of statistical significance at low doses (e.g., Cantor and Lubin, 2007), is actually more likely to cause an appearance of a monotonically increasing dose–response among exposure groups (particularly for broad exposure groupings), even when the underlying association is threshold in nature. Moreover, the HRs less than 1.0 (i.e., decreased risk with more exposure) for the middle dose group (25–114 μg/L arsenic water concentration) among never smokers in Chen et al. (2011) (see Fig. 2 in this study) suggest that low power, causing regression to the null, (-)-p-Bromotetramisole Oxalate is not the cause of a lack of a statistically significant positive association of arsenic with CVD below 100 μg/L. The association of arsenic exposure in well water with CVD mortality reported by Chen et al. (2011) may be confounded by well water manganese in the Araihazar region, where both constituents in well water have been reported to be correlated (r = 0.13; P < 0.03) ( Wasserman et al., 2011). Some evidence, primarily from occupational exposures, suggests that manganese may have cardiovascular effects, although the effects may differ from those associated with high arsenic exposure (e.g., opposite effect on heart rhythm; hypotension; Jiang and Zheng, 2005). Manganese was not included as a covariate in the CVD studies ( Chen et al., 2006b, Chen et al., 2011, Chen et al., 2013a and Chen et al.