Effect of air quality on assisted human reproduction.

BACKGROUND

Air pollution has been associated with reproductive complications. We hypothesized that declining air quality during in vitro fertilization (IVF) would adversely affect live birth rates.

METHODS

Data from US Environmental Protection Agency air quality monitors and an established national-scale, log-normal kriging method were used to spatially estimate daily mean concentrations of criteria pollutants at addresses of 7403 females undergoing their first IVF cycle and at the their IVF labs from 2000 to 2007 in the Northeastern USA. These data were related to pregnancy outcomes.

RESULTS

Increases in nitrogen dioxide (NO(2)) concentration both at the patient's address and at the IVF lab were significantly associated with a lower chance of pregnancy and live birth during all phases of an IVF cycle from medication start to pregnancy test [most significantly after embryo transfer, odds ratio (OR) 0.76, 95% confidence interval (CI) 0.66-0.86, per 0.01 ppm increase]. Increasing ozone (O(3)) concentration at the patient's address was significantly associated with an increased chance of live birth during ovulation induction (OR 1.26, 95% CI 1.10-1.44, per 0.02 ppm increase), but with decreased odds of live birth when exposed from embryo transfer to live birth (OR 0.62, 95% CI 0.48-0.81, per 0.02 ppm increase). After modeling for interactions of NO(2) and O(3) at the IVF lab, NO(2) remained negatively and significantly associated with live birth (OR 0.86, 95% CI 0.78-0.96), whereas O(3) was non-significant. Fine particulate matter (PM(2.5)) at the IVF lab during embryo culture was associated with decreased conception rates (OR 0.90, 95% CI 0.82-0.99, per 8 microg/m(3) increase), but not with live birth rates. No associations were noted with sulfur dioxide or larger particulate matter (PM(10)).

CONCLUSIONS

The effects of declining air quality on reproductive outcomes after IVF are variable, cycle-dependent and complex, though increased NO(2) is consistently associated with lower live birth rates. Our findings are limited by the lack of direct measure of pollutants at homes and lab sites.