Impact of Male Genital Tract Infections on Semen Quality: A Systematic Review and Meta-Analysis

GU infections were linked to measurable sperm parameter declines, but pregnancy outcomes mostly held — HPV was the outlier, with a tenfold antisperm antibody spike

Journal: Fertility and Sterility | Published: 2026-03-11 | Type: Systematic Review, Meta-Analysis | PMID: 41825760 Authors: Campbell KJ, Venkatesh A, Golan R, Tang Z, Shan G, Donelan W (University of Florida College of Medicine, Gainesville, FL) Funding/COI: Funding not listed. All authors declare no conflicts of interest.

Summary

Genitourinary infections impaired several semen parameters — sperm concentration dropped a mean of 6.65 million/mL, total sperm count by 47.15 million, progressive motility by 5.93%, and morphology by 0.76%. Despite those reductions, most pregnancy and live birth outcomes did not significantly differ between infected and uninfected men. The exception: HPV infection was associated with a tenfold increase in antisperm antibody prevalence, and hepatitis B was tied to a modest but significant uptick in miscarriage risk.

Claims

• Sperm concentration: MD −6.65 million/mL (95% CI −11.17 to −2.14), I² = 90%
• Total sperm count: MD −47.15 million (95% CI −91.42 to −2.88)
• Progressive motility: MD −5.93% (95% CI −9.07 to −2.78)
• Semen volume: MD −0.37 mL (95% CI −0.64 to −0.09)
• Morphology: MD −0.76% (95% CI −1.19 to −0.33)
• Total motility, vitality, and DNA fragmentation index: not significantly different between infected and control groups
• HPV infection: OR 10.63 (95% CI 1.49–75.93) for antisperm antibody prevalence — reported across five studies
• Hepatitis B infection: OR 1.43 (95% CI 1.05–1.93) for miscarriage risk
• ART-specific outcomes: limited data, generally nonsignificant
• Natural pregnancy and live birth outcomes: no significant difference between infected and control groups

Study Quality

This is a well-executed systematic review drawing from PubMed, Web of Science, and Embase. Fifty-one studies qualified for qualitative synthesis; 35 contributed to the quantitative meta-analysis. Using mean-difference pooling for continuous outcomes and odds ratios for binary outcomes is appropriate given the study designs captured. The funnel plot and forest plots presumably accompany the full text, though heterogeneity statistics do the heavy lifting here in signaling fragility.

The central problem is the heterogeneity, which the authors acknowledge plainly: I² values of 66–95% across outcomes mean these pooled estimates are describing a cloud of different studies rather than a stable effect. The underlying studies are predominantly observational, enrolled mostly infertile populations (which inflates apparent effect sizes), and used non-standardized diagnostic criteria for infection. That the pregnancy outcomes show no significant differences — despite meaningful parameter deficits — points either to residual confounding, underpowered binary outcome analyses (only five studies contributed), or genuine disconnect between lab parameters and reproductive outcomes.

Red Flags

• I² of 90% for sperm concentration renders the pooled MD nearly uninterpretable as a universal effect
• Binary outcome analyses (pregnancy, live birth, miscarriage) limited to five studies — severely underpowered for those endpoints
• Study population skewed toward infertile men, meaning effects may not generalize to the broader male population
• HPV antisperm antibody OR of 10.63 has a confidence interval spanning 1.49–75.93 — the upper bound is so wide this is essentially an exploratory signal, not a conclusion
• No standardized diagnostic criteria for "infection" across included studies — some definitions almost certainly varied substantially
• Funding source not reported; COI declarations are clean but the gap is worth noting

Strengths

• Three-database search, systematic design, pre-specified inclusion criteria with control-comparator requirement
• Covers a broad range of both semen parameters and reproductive outcomes — not just sperm count in isolation
• Transparent about heterogeneity rather than burying it
• Distinguishes between infertile and general populations in subgroup framing
• Clean COI disclosures across all six authors

Verdict

This meta-analysis is technically sound but the heterogeneity problem is severe enough that the pooled numbers should be treated as rough direction indicators, not reliable effect estimates. The headline finding — that GU infections modestly impair several semen parameters — is plausible and consistent with prior literature, but "I² = 90%" on sperm concentration means roughly that different infections in different populations have wildly different effects, and we've averaged them into a number that describes almost none of them well. The HPV-antisperm antibody association (OR 10.63) is the most clinically interesting signal here and the one most deserving of a dedicated prospective study. The paper is worth reading for its evidence-base summary and its explicit call for standardized diagnostic frameworks — those are the right takeaways, not the pooled mean differences.