O-075 Particulate matter (PM2.5 and PM10) exposure prior to oocyte collection is associated with decreased live birth rates in subsequent frozen embryo transfers

Abstract Study question Does exposure to particulate matter air pollution prior to oocyte collection (OPU) or frozen embryo transfer (FET) impact FET live birth rates? Summary answer Increasing PM10 and PM2.5 exposure prior to OPU is associated with decreased odds of live birth in subsequent FET, independent of the conditions at FET. What is known already Ambient air pollution is increasing, and exposure is associated with adverse reproductive health outcomes, including infertility, miscarriage, preterm birth and low birth weight. Animal studies have shown that air pollutants can impair spermatogenesis, oogenesis, ovarian reserve and embryogenesis. These findings are corroborated by human epidemiological studies. The impact of ambient air pollution on IVF outcomes is uncertain. Additionally, whether pollutants exert a greater impact during oocyte development or early pregnancy is uncertain. Therefore, studies of FETs offer an opportunity to delineate different phases of reproduction that may be adversely affected by pollution. Study design, size, duration A retrospective analysis of 3,657 FETs in a single Australian centre from January 2013 – December 2021. All FETs during the study period were included; data were collected from the clinic database. Air pollutant concentrations for each day during the study were obtained from the Governmental monitoring body, including SO2, CO, O3, NO2, PM10 and PM2.5. Evaluation of FETs allowed separation of environmental effects during oocyte development and during early pregnancy. Participants/materials, setting, methods Quartiles were created for pollutants averaged over 24 hours, 2 weeks, 4 weeks and 3 months prior to OPU and prior to FET. Multi-pollutant models were created to account for co-exposures and were adjusted for variables at the time of OPU and FET. Bonferroni correction was made to account for the 4 timepoints tested on each outcome. Linear trends were assessed for each model, and pairwise comparisons against the first quartile were performed. Main results and the role of chance Overall, air quality was excellent, with pollutant levels within published safe limits on > 99% of days during the study period. Increasing PM10 exposure in the 2 weeks prior to OPU was associated with decreased odds of livebirth (linear trend p = 0.009); the odds of livebirth decreased by 38% when PM10 concentrations were in the highest quartile compared with the lowest (OR 0.62, p = 0.010). Increasing PM2.5 exposure in the 3 months prior to OPU was also associated with decreased odds of livebirth (linear trend p = 0.011). No significant linear trends were observed between livebirths and other pollutants. The summer seasonal association with livebirth that we have previously demonstrated consistently held in the multi-pollutant models. No significant linear trends were observed between air pollutants and biochemical pregnancy, clinical pregnancy, miscarriage, gestation at birth or embryological outcomes. Limitations, reasons for caution The retrospective design is a limitation. Whilst corrections were made for multiple pollutant co-exposures, other environmental conditions, patient factors and repeated cycles, we were unable to account for all potential covariates. To reduce the risk of type 1 error due to the multiple timeframes, we utilised a Bonferroni correction. Wider implications of the findings Even when PM2.5 and PM10 exposures are within so-called “safe” ranges, increasing exposure in the weeks and months prior to OPU are associated with decreasing livebirth rates, independent of the conditions at the time of FET. Trial registration number 2022/ET000980