Associations between routine systemic biochemical indices and semen quality: a cross-sectional study in men undergoing health examination

To evaluate associations between routinely measured systemic biochemical indices and semen quality parameters in a real-world health-check population. In this retrospective cross-sectional analysis of routinely collected data, 264 men aged 21–60 years who underwent health examination between January 2023 and December 2024 were included. Nine continuous semen outcomes were correlated with 42 clinical and biochemical indices using Spearman rank correlation (378 tests). Benjamini–Hochberg false discovery rate (FDR) correction was applied within each outcome (primary analysis) and across all 378 tests (sensitivity analysis). The association between calculated serum osmolality and normal morphology was further examined in a post hoc supportive multivariable model. Men with all semen parameters within World Health Organization (WHO) 2010 reference limits were compared with men with at least one parameter below a limit, with a WHO 2021 sensitivity analysis. Correlations were weak (|ρ| ≤ 0.21). Six associations met the outcome-specific FDR threshold: magnesium with progressive motility (ρ = 0.210; FDR P = 0.025); globulin (ρ = 0.197), albumin-to-globulin ratio (ρ = −0.193), and CO₂ combining power (ρ = 0.196) with normal morphology (FDR P = 0.023 each); and sodium (ρ = 0.192) and calculated serum osmolality (ρ = 0.187) with DNA fragmentation index (DFI) (FDR P = 0.048 each). None remained significant in the global FDR sensitivity analysis (FDR P = 0.123 for each). No clinical or biochemical variable differed significantly between reference-status groups after FDR correction (WHO 2010: all q ≥ 0.415; WHO 2021: all q ≥ 0.897). The post hoc supportive model for osmolality and normal morphology is reported in Supplementary Table S2 . In this health-check population, routine systemic biochemical indices showed weak correlations with semen parameters, and no association was robust to global multiplicity correction. Given the retrospective cross-sectional design, findings should be interpreted as associative rather than causal. These results are hypothesis-generating and warrant longitudinal validation to clarify temporal relationships and biological mechanisms.