Boys with hypospadias have shorter anogenital distance (SMD −1.24), with the gap widening as severity increases across 6,905 infants
Journal: European Journal of Pediatrics | Published: 2026-03-21 | Type: Systematic Review, Meta-Analysis | PMID: 41863616 Authors: Hu Shaohua, Zhao Zhenli, Wan Zhisheng, Bu Weizhen, Chen Songqiang (Hainan Women and Children's Medical Center); Lu Yiqun (Children's Hospital of Fudan University, Shanghai) Funding/COI: Hainan Province Joint Health Science & Technology Program; Hainan Province Science and Technology Special Fund; Hainan Women and Children's Medical Center "1126 Plan." No competing interests declared.
Anogenital distance (AGD) — the perineal measurement from anus to genitalia, a proxy for fetal androgen exposure — is shorter in boys born with hypospadias than in those without, and shorter still as the defect moves proximally. This meta-analysis of 13 studies and 6,905 male infants is the first to pool this association by severity subgroup. The authors frame the finding as supporting the fetal androgen disruption hypothesis, though they are careful to note that causation cannot be established from observational data.
This is a meta-analysis of observational studies, which is the appropriate design for aggregating an association signal that individual studies have reported inconsistently. Random-effects modeling was correctly chosen given the expected heterogeneity across populations, measurement protocols, and hypospadias classification schemes. The five-database international search plus three Chinese-language databases is a genuine methodological strength — it reduces the language bias that plagues meta-analyses built only on English-language literature. PROSPERO pre-registration is a positive signal.
The heterogeneity, however, is severe enough to raise real questions about the pooled estimates. The confidence intervals are extremely wide — spanning 1.5 to 3.4 SMD units in the subgroup analyses — which is more consistent with mixing meaningfully different populations than with genuine variance around a stable effect. The distal hypospadias subgroup (the most common form) has a CI that crosses zero (−2.23 to 0.13), meaning the association with the mildest phenotype is not statistically significant. The paper's own conclusion appropriately flags this, but the headline finding deserves that caveat prominently.
This is a competent meta-analysis that consolidates a noisy literature into a usable signal: AGD and hypospadias severity track together, and the gradient across proximal/middle/distal phenotypes is the paper's novel contribution. The statistical architecture is appropriate and the search is thorough. What limits it is not the authors' choices but the state of the underlying evidence — high heterogeneity and non-standardized AGD measurement across source studies means the pooled SMDs are approximations rather than precise estimates. The distal subgroup's non-significant result is buried in the severity gradient narrative and deserves more attention. Worth reading as a reference for the AGD-hypospadias literature base; not worth citing as definitive quantification of the effect size.