A dose-response meta-analysis of chronic arsenic exposure and incident cardiovascular disease

Abstract

Background: Consistent evidence at high levels of water arsenic (≥100 µg/l), and growing evidence at low-moderate levels (<100 µg/l), support a link with cardiovascular disease (CVD). The shape of the dose-response across low-moderate and high levels of arsenic in drinking water is uncertain and critical for risk assessment.

Methods: We conducted a systematic review of general population epidemiological studies of arsenic and incident clinical CVD (all CVD, coronary heart disease (CHD) and stroke) with three or more exposure categories. In a dose-response meta-analysis, we estimated the pooled association between log-transformed water arsenic (log-linear) and restricted cubic splines of log-transformed water arsenic (non-linear) and the relative risk of each CVD endpoint.

Results: Twelve studies (pooled N = 408 945) conducted at high (N = 7) and low-moderate (N = 5) levels of water arsenic met inclusion criteria, and 11 studies were included in the meta-analysis. Compared with 10 µg/l, the estimated pooled relative risks [95% confidence interval (CI)] for 20 µg/l water arsenic, based on a log-linear model, were 1.09 (1.03, 1.14) (N = 2) for CVD incidence, 1.07 (1.01, 1.14) (N = 6) for CVD mortality, 1.11 (1.05, 1.17) (N = 4) for CHD incidence, 1.16 (1.07, 1.26) (N = 6) for CHD mortality, 1.08 (0.99, 1.17) (N = 2) for stroke incidence and 1.06 (0.93, 1.20) (N = 6) for stroke mortality. We found no evidence of non-linearity, although these tests had low statistical power.

Conclusions: Although limited by the small number of studies, this analysis supports quantitatively including CVD in inorganic arsenic risk assessment, and strengthens the evidence for an association between arsenic and CVD across low-moderate to high levels.

Keywords: Arsenic; cardiovascular disease; dose-response; meta-analysis.

Figure 1

Study exclusion criteria. ^a13 general population epidemiological studies (14 articles) were identified in Navas-Acien et al. 2005 (Tsai et al. 1999 and Wu et al. 1989 were considered as the same study population). ^bWang et al. (2005) was included in the meta-analysis only as a sensitivity analysis.

Figure 2

Pooled log-linear and non-linear relative risks of incident overall cardiovascular disease, coronary heart disease and stroke in relation to estimated water arsenic. Pooled linear (red) and non-linear (blue) relative risks of CVD endpoints (overall CVD, CHD and stroke, stratified by studies of incidence and mortality) were estimated for drinking water arsenic concentrations in reference to 10 µg/l. Dashed lines correspond to pooled relative risks, and shaded regions correspond to the 95% confidence intervals of the pooled relative risks. Log-linear associations were estimated from models with log-transformed estimated water arsenic concentrations. Non-linear associations were estimated from models with restricted cubic splines of log-transformed estimated water arsenic concentrations with knots at the 10th, 50th and 90th percentiles of log-transformed arsenic (exact knot locations vary by model; for CHD incidence, knots were placed at 5.1, 20.5 and 58.7 µg/l). A rug plot along the x-axis provides the median estimated water arsenic concentrations included in each model.

Figure 3

Individual study and pooled log-linear relative risks (95% confidence intervals) of incident overall cardiovascular disease, coronary heart disease and stroke, comparing 20 µg/l with 10 µg/l water arsenic. Forest plot of individual study pooled relative risks (95% confidence intervals) and overall pooled relative risks (95% confidence intervals) for CVD, CHD and stroke (stratified by incidence and mortality) comparing 20 µg/l with 10 µg/l (i.e. per doubling) of water arsenic. Individual study pooled linear relative risks were estimated in the first stage of the two-stage dose-response meta-analysis, in which a generalized least-squares linear model was fitted to each study’s categorical dose-response data. In the second stage, the individual study pooled linear relative risks were pooled together in a multivariate random-effects meta-analysis to estimate the overall pooled relative risks. Within each model, studies were ordered by average water arsenic levels, from highest to lowest. The sizes of the individual study relative risk squares were weighted by the inverse variance of the log relative risk within each model.