Association of ambient air pollution and pregnancy rate among women undergoing assisted reproduction technology in Fujian, China: A retrospective cohort study

Abstract

Background: Previous studies have reported the impact of ambient air pollutants on assisted reproduction. They concentrated on highly polluted environments and individual pollutants. It is unclear whether these effects continue at lower levels and as mixed effects. We aimed to study the influence of lower pollutant concentrations on pregnancy rates and identify vulnerable populations.

Methods: We conducted a retrospective cohort study involving 9465 patients with infertility who received treatment from a local hospital between 2015 and 2021. Daily average levels of six pollutants (PM_2.5, PM₁₀, NO₂, CO, SO₂, and O₃) were collected from air quality monitoring stations. We employed generalized linear regression models (logistic, linear, and lasso), weighted quantile sum (WQS) regression, and Bayesian kernel machine regression (BKMR) to assess the impact of pollutants on pregnancy rates. Additionally, stratified analyses were performed to identify potentially vulnerable populations.

Results: Findings from the generalized linear models revealed a significant negative correlation between interquartile range increment exposure to PM_2.5 (OR = 1.17, 95 % CI = 1.09-1.26), PM₁₀ (OR = 1.18, 95 % CI = 1.11-1.26), NO₂ (OR = 1.21, 95 % CI = 1.13-1.30), CO (OR = 1.02, 95 % CI = 1.00-1.03), SO₂ (OR = 1.11, 95 % CI = 1.05-1.17) and pregnancy rate when considering the effects of individual pollutants. The WQS index exhibited a negative correlation with pregnancy rates and the number of oocytes retrieved (aOR = 1.20, 95 % CI = 1.08-1.34). BKMR analyses indicated an overall significant trend of decreasing pregnancy rates as pollutant concentrations increased across percentiles. Stratified analysis unveiled heightened sensitivity to pollutants among individuals aged ≥35 years.

Conclusions: By comparing results obtained from diverse models, we observed that exposure to lower levels of air pollutants led to decreased pregnancy rates. Notably, PM₁₀, NO₂, SO₂, and CO emerged as the four most prominent pollutants in this context. Moreover, stratified analyses highlighted that individuals aged ≥35 years exhibited heightened susceptibility to pollutants.

Keywords: Ambient air pollution; Assisted reproduction technology; Bayesian kernel machine regression (BKMR); Mixed effect.