Low-level and combined exposure to environmental metal elements affects male reproductive outcomes: Prospective MARHCS study in population of college students in Chongqing, China

Male fertility has shown a continuously declining tendency for decades. Over exposure to metal/metalloid elements has been proposed as associated with reproductive impairment. However, the hazard profile remained unclear in general public experiencing low-level and combined metal exposure. Based on the MARHCS cohort in Chongqing, China, 796 college students were recruited from June 2013 and 666 subjects were followed up next year. At each phase, semen and blood samples were collected for an assessment of semen quality and six sex hormones levels. Eighteen urinary metal/metalloid elements were quantified by ICP-MS as internal exposure biomarkers. Cluster analysis was conducted to characterize reproductive outcomes in the subgroups for different overall estimated exposure levels. Effects of each metal/metalloid element were analyzed using multiple statistical strategies: single-element mixed model, multiple-elements model and self before-after comparison design. The urine concentration for 18 metal/metalloid elements was at a typically lower level (far away from the exposure limits) and positively associated with each other. After adjustment of the potential confounders, a decrease of 11.53% (95% CI: −18.61, −3.84%) and 10.84% (95% CI: −17.93, −3.14%) in spermatid morphology was observed in the highest quantile groups of vanadium (V) and nickel (Ni), respectively. Urinary silver (Ag) was dose-dependent associated with an increase in total sperm number (6.91%, 95% CI: 1.14, 13.00%), sperm concentration (16.38%, 95% CI: 5.15, 28.81%) and semen volume (23.73%, 95% CI: 10.46, 38.60%). Further, hormone testosterone presented a significant decrease in subgroup with higher overall estimated exposure and a stable negative association with lithium (Li). The above relationships remained significant across different statistical strategies (all p values <0.05). Our study provided new evidences that exposure to metal/metalloid elements potentially exert bidirectional influences on semen quality at a relatively low level. And serum testosterone appears as a vulnerable index for metal exposure.