Association between outdoor air pollution during in vitro culture and the outcomes of frozen-thawed embryo transfer

STUDY QUESTION Does outdoor air pollution differentially affect the outcomes of frozen-thawed embryo transfer (FET) and fresh transfer in IVF treatment? SUMMARY ANSWER Increased SO2 and O-3 levels at the site of IVF unit were significantly associated with lower live birth rates following FET but did not affect the contemporary fresh transfer outcomes. WHAT IS KNOWN ALREADY Ambient air pollution has been associated with human infertility and IVF outcomes. However, most of the studies excluded FET cycles. STUDY DESIGN, SIZE, DURATION A retrospective cohort study of 11148 patients contributing to 16290 transfer cycles between January 2013 and December 2016 was carried out. PARTICIPANTS/MATERIALS, SETTING, METHODS The average age of the cohort was 31.51 4.48 years and the average BMI was 21.14 2.37 kg/cm(2). Inverse distance weighting interpolation was used to estimate the daily ambient exposures to six pollutants (PM2.5, PM10, SO2, NO2, CO, O-3) at an IVF clinical site, according to the data from fixed air quality monitoring stations in the city. The exposures of each cycle were presented as average daily concentrations of pollutants from oocyte retrieval to embryo transfer/cryopreservation. Exposures were analyzed in quartiles. A generalized estimating equation was used to evaluate the association between pollutants and IVF outcomes, adjusted for important confounding factors including maternal age, infertility diagnosis, BMI, endometrial status and embryo transfer policy. MAIN RESULTS AND THE ROLE OF CHANCE The clinical pregnancy rate and live birth rate of the cycles was 55.1% (8981/16290) and 47.1% (7672/16290), respectively. Among the included cycles, 4013 patients received 5299 FET cycles, resulting in 2263 live births (42.7% per ET), whereas 9553 patients received 10991 fresh transfer cycles, resulting in 5409 live births (49.2% per ET). SO2 and O-3 levels were significantly associated with live birth rates in FET cycles, whereas none of the pollutants were significantly associated with IVF outcomes in contemporary fresh transfer cycles. The FET cycles in the highest quartile of SO2 and O-3 exposure had significantly lower live birth rates (adjusted odds ratio (OR) 0.63, 95%CI 0.53-0.74; 0.69, 95% CI 0.58-0.82, respectively) in comparison with those in the lowest quartile. Models involving all transfer cycles and interaction terms (FETxexposures) suggested that FET significantly enhanced the effects of SO2 and O-3 exposure on IVF outcomes (P < 0.001). Multi-pollutant models gave consistent results for the association between SO2 and live birth in FET cycles. Accounting for all six pollutants, women in the highest quartile of SO2 still had the lowest live birth rates (OR 0.61, 95%CI 0.47-0.80). LIMITATIONS, REASONS FOR CAUTION The study was limited by its retrospective nature. The exposure data were estimated according to monitoring data rather than measured directly from the IVF unit. Unknown confounding factors may skew the results. WIDER IMPLICATIONS OF THE FINDINGS Our data implied that embryos undergoing FET may be more vulnerable to a suboptimal environment than those undergoing fresh transfer. In heavily polluted sites or seasons, fluctuation in FET outcomes may be partially explained by the dynamic changes of ambient gaseous air pollutant. STUDY FUNDING/COMPETING INTEREST(S) National Natural Science Foundation (81302454). The authors have no conflicts of interest to declare. TRIAL REGISTRATION NUMBER N/A.