No Causal Effect Between Prostatitis and Erectile Dysfunction: Insights from a 2-Sample Bidirectional Mendelian Randomization Study

Genetic analysis of FinnGen data found no causal link between prostatitis and ED in either direction (OR=1.132, P=.199)

Journal: Medicine | Published: 2026-04-17 | Type: Journal Article | PMID: 41995490 Authors: Dai Xiaowei (Jilin University Second Norman Bethune Hospital); Huang Jiaguo, Ding Hongxiang, Xie Yinfang (Hangzhou Normal University Affiliated Xiaoshan Hospital) Funding/COI: No funding or conflicts of interest declared

Summary

Observational studies have long suggested prostatitis and erectile dysfunction travel together, but correlation in clinical populations is notoriously confounded by shared risk factors—pain, depression, poor metabolic health. This Mendelian randomization study used genetic variants as natural instruments to test whether that association is causal. In both directions, it isn't: prostatitis does not genetically predict ED, and ED does not genetically predict prostatitis.

Claims

• No causal effect of prostatitis on ED: OR = 1.132 (95% CI 0.937–1.367, P = .199) by inverse variance weighted method
• No causal effect of ED on prostatitis: OR = 1.044 (95% CI 0.960–1.135, P = .311)
• Sensitivity analyses (MR-Egger, weighted median) consistent with primary IVW results
• No significant pleiotropy detected by MR-Egger intercept test
• No influential outlier SNPs identified by leave-one-out or MR-PRESSO

Study Quality

Mendelian randomization is methodologically superior to observational studies for causal inference because genetic variants are assigned at conception, before lifestyle or disease confounders accumulate. The authors used the FinnGen consortium—one of the better-powered biobank datasets—and ran a bidirectional design, which tests both proposed directions of causality rather than assuming one. The three complementary MR methods (IVW, MR-Egger, weighted median) each handle violations of the instrumental variable assumptions differently; concordance across all three strengthens confidence in the null result.

The critical unaddressed limitation is phenotype heterogeneity. "Prostatitis" in FinnGen almost certainly collapses chronic bacterial prostatitis, chronic pelvic pain syndrome (CPPS), and asymptomatic inflammatory prostatitis into a single ICD code. These are biologically distinct entities with different pathophysiology. A null result across this lumped category doesn't rule out a causal relationship in a specific subtype—particularly CPPS, which has the strongest clinical overlap with sexual dysfunction.

Red Flags

• European ancestry only — FinnGen is Finnish; results may not generalize to other populations
• Phenotype lumping — "prostatitis" is a heterogeneous diagnosis; ICD-based case definitions in biobanks are noisy
• Sample sizes not reported in abstract — effect estimate precision depends entirely on FinnGen's prostatitis and ED case counts, which vary widely
• Null findings need adequate power — the confidence intervals, particularly for prostatitis → ED (0.937–1.367), are wide enough that a modest true effect cannot be excluded
• No subgroup analysis by prostatitis type — CPPS vs. bacterial vs. inflammatory subtypes have different plausible mechanisms

Strengths

• Bidirectional MR design tests causality in both directions simultaneously
• Three sensitivity analyses with different assumption sets all agree
• Pleiotropy formally tested with MR-Egger intercept and MR-PRESSO
• Leave-one-out analysis rules out single-SNP dominance
• No declared funding or conflicts of interest
• Uses publicly available FinnGen data, making findings reproducible

Verdict

This is a competent MR study that does what it sets out to do: challenge a widely assumed clinical association with genetic causal inference. The null result is credible and matters—it suggests that the prostatitis-ED correlation seen in clinical cohorts is likely confounded, not causal, at least at the population-genetic level. The paper's main weakness isn't the statistics; it's that "prostatitis" is doing too much heavy lifting as a single exposure. A well-powered MR specifically targeting chronic prostatitis/CPPS—the subtype most plausibly linked to sexual dysfunction via pain and pelvic floor mechanisms—would be far more informative. Read this as useful evidence against simple causality, not as a clean bill of health for the full hypothesis.