Critical windows for exposure to chemical composition of ambient particulate matter and human semen quality decline

Background: Critical windows for exposure to chemical components of particulate matter (PM <2.5 μm in diameter [PM_2.5]) associated with the human semen quality decline remain unclear.

Objectives: To address this gap, we developed a new analytical framework by integrating a Linear Mixed Model (LMM) with subject- and center-specific intercepts and a Distributed Lag Model (DLM) to fully account for correlations between finely vulnerable exposure windows based on complete profile of the spermatogenesis cycle.

Methods: We constructed a multicenter cohort involving 33,234 sperm donors with 78,952 semen samples covering 6 representative regions across China from 2014 to 2020 to investigate the week-scale critical windows for the exposure. Daily exposure to PM_2.5 chemical components of donors was derived from grid data based on 1-km spatial resolution surface measurements.

Results: Decreased sperm count was significantly associated with NO₃^- and SO₄^2- at 9-10 weeks (e.g., β: -0.05 %, 95%CI: [-0.10 %, -0.00 %] at the 9th week) and 0-2 weeks (e.g., β: -0.66 %, 95%CI: [-1.24 %, -0.07 %] at the 1st week), respectively. Critical windows of progressive motility decline were 0-10 weeks for BC (e.g., β: -0.07 %, 95%CI: [-0.11 %, -0.03 %] at the 5th week), Cl^- at 1-4 weeks (e.g., β: -2.21 %, 95%CI: [-3.77 %, -0.66 %] at the 2nd week), 0-6 weeks and 9-10 weeks for NO₃^- (e.g., β: -0.05 %, 95%CI: [-0.09 %, -0.01 %] at the 4th week), 1-3 weeks and the 8th week for NH₄⁺ (e.g., β: -0.06 %, 95%CI: [-0.11 %, -0.01 %] at the 2nd week). Total motility is significantly negatively associated with BC at entire windows, Cl^- at 0-3 weeks, the 5th week and 9-10 weeks.

Conclusions: There are week-scale vulnerable windows of exposure to PM_2.5 chemical components for human semen quality. This highlights the need for more targeted pollution control strategies addressing PM_2.5 and its chemical components.