Linear and non-linear relationships between sulfur dioxide and semen quality: A longitudinal study in Anhui, China

Existing evidence indicates that ambient air pollutants pose a threat to human semen quality; however, these findings are sparse and controversial. Besides, their non-linear dose-response relationship has not yet been well investigated. This study aimed to explore the linear and non-linear associations of gaseous air pollutants exposure with semen quality based on a large longitudinal cohort. A total of 15,112 males (with 28,267 semen tests) from the Anhui prospective assisted reproduction cohort were analyzed. Individual air pollutants exposure before semen tests in four exposure windows (i.e., 0-9, 10-14, 70-90, and 0-90 days) were estimated by inverse distance weighting interpolation. Linear mixed-effects models, cubic spline analysis and piecewise regression were used to test the potential linear and non-linear dose-response relationships. Ambient SO2 exposure was negatively associated with all semen quality parameters (all p values < 0.05), except for the progressive motility in the 0-90 and 70-90 days exposure windows. There were 'J' or 'U' shaped dose-response relationships of ambient SO2 exposure with total sperm count, progressive motility, total motility, progressively motile sperm count, and total motile sperm count (p values for non-linearity < 0.05), but not sperm concentration. Piecewise regression analysis also indicated a negative association of SO2 exposure with semen quality only when SO2 exposure was below the cut-off points identified by cubic spline analyses, which were all smaller than 40 μg/m3, the 2021 updated WHO air quality guideline level for SO2 exposure. Overall, we found that SO2 exposure was negatively associated with semen quality. Ambient SO2 exposure could reach the maximum hazardous dose even below the WHO air quality guideline level for SO2 exposure, suggesting a refinement to the current guideline.