Low-level environmental metals and metalloids and incident pregnancy loss

- A preconception cohort had metal(loids) measured to assess risk of incident pregnancy loss.
- No significant associations were observed between metal(loids) concentrations and loss.
- The findings were robust across partner, biospecimen (blood or urine) and statistical model.

Environmental exposure to metals and metalloids is associated with pregnancy loss in some but not all studies. We assessed arsenic, cadmium, mercury, and lead concentrations in 501 couples upon trying for pregnancy and followed them throughout pregnancy to estimate the risk of incident pregnancy loss. Using Cox proportional hazard models, we estimated hazard ratios (HR) and 95% confidence intervals (CIs) for pregnancy loss after covariate adjustment for each partner modeled individually then we jointly modeled both partners' concentrations. Incidence of pregnancy loss was 28%. In individual partner models, the highest adjusted HRs were observed for female and male blood cadmium (HR = 1.08; CI 0.81, 1.44; HR = 1.09; 95% CI 0.84, 1.41, respectively). In couple based models, neither partner's blood cadmium concentrations were associated with loss (HR = 1.01; 95% CI 0.75, 1.37; HR = 0.92; CI 0.68, 1.25, respectively). We observed no evidence of a significant relation between metal(loids) at these environmentally relevant concentrations and pregnancy loss.