Air pollution and male infertility: A meta-umbrella study of evidence on semen quality

Four meta-analyses pooled: air pollution correlates with modestly lower sperm motility, concentration, and morphology

Journal: Medicine | Published: 2026-Mar-06 | Type: Systematic Review, Meta-Analysis | PMID: 41790664 Authors: Margiana R et al. (Universitas Indonesia; University of Anbar, Iraq; Al-Zahrawi University College, Iraq; Universidad Yachay, Ecuador; Symbiosis International University, India; Guilan University of Medical Sciences, Iran) Funding/COI: No funding or conflicts of interest disclosed

Summary

This "meta-umbrella" — a meta-analysis of existing meta-analyses — pooled results from four eligible studies to quantify how ambient air pollution affects semen quality. It found statistically significant but modest negative associations for total motility, progressive motility, morphology, and sperm concentration. Sperm count trended downward but failed to reach statistical significance. The authors conclude that air pollution is a "modifiable risk factor" for male infertility — a causal claim that four pooled observational meta-analyses cannot support.

Claims

• Total motility: SMD −0.24 (95% CI −0.48 to −0.01; P = .04) — higher pollution associated with lower total motility
• Progressive motility: SMD −0.06 (95% CI −0.10 to −0.01; P < .001)
• Morphology: SMD −0.09 (95% CI −0.13 to −0.04; P < .001)
• Sperm concentration: SMD −0.14 (95% CI −0.21 to −0.06; P < .001)
• Sperm count: SMD −0.12 (95% CI −0.26 to 0.01; P = .06) — did not reach statistical significance
• All effect sizes fall in the small range (SMD < 0.25 by convention); clinical translation to infertility risk is not established

Study Quality

This is a meta-umbrella: a systematic review of previously published meta-analyses, adding a third layer of abstraction over the underlying observational studies. Only four meta-analyses met inclusion criteria — a strikingly thin base from which to draw cross-pollutant conclusions. Methodological quality was assessed using AMSTAR 2, the appropriate tool for this design. Random-effects models were used, which is correct given expected heterogeneity across pollutant types, exposure measurement methods, and populations.

The deeper problem is definitional: "air pollution" is not a single exposure. PM2.5, PM10, nitrogen dioxide, ozone, and polycyclic aromatic hydrocarbons have distinct mechanisms, sources, and potencies. Pooling across all of them without pollutant-stratified analysis obscures more than it reveals. Search was systematic (PubMed, Web of Science, Scopus through January 2025), which is a point in the paper's favor.

Red Flags

• Only 4 meta-analyses included — one rogue constituent study can swing the pooled SMD substantially at this scale
• Total motility CI nearly crosses zero (−0.48 to −0.01; P = .04) — borderline significant; a fragile finding
• Sperm count does not reach significance (P = .06) — underemphasized in the conclusion
• "Modifiable risk factor" is a causal claim from purely observational association data
• Heterogeneous exposures pooled — chemically distinct pollutants with different mechanisms are treated as one variable
• Six institutions across five countries with no evident prior publication history together; warrants scrutiny of coordination and data handling quality
• Published in Medicine (Wolters Kluwer) — legitimate peer-reviewed journal, but broad scope with minimal selectivity

Strengths

• Systematic search across three major databases
• AMSTAR 2 quality assessment applied appropriately
• Multiple semen parameters analyzed separately rather than collapsed into a composite
• Random-effects models correctly chosen over fixed-effects
• No declared funding or commercial conflicts

Verdict

A meta-umbrella built on four meta-analyses is a shaky tower. The findings — modest negative associations between air pollution and sperm parameters — are biologically plausible given oxidative stress and endocrine disruption mechanisms, but the effect sizes are small, sperm count significance doesn't cross the line, and the exposure heterogeneity is enormous. Calling pollution a "modifiable risk factor" in the conclusion is an epidemiological overclaim this study design cannot support. The paper's main value is as a signal: the primary literature needs better longitudinal studies with pollutant-specific exposures, not another layer of abstraction over the same weak evidence base.