Men whose mothers were obese during pregnancy had 5.9× higher odds of low testosterone — but only 12 of 418 men had that history
Journal: Andrology | Published: 2025-05-14 | Type: Cross-sectional study | PMID: 40366358 Authors: Negri F, Boeri L, Pozzi E, et al. (IRCCS Ospedale San Raffaele & IRCCS Fondazione Ca' Granda, Milan, Italy) Funding/COI: Funding not listed. Authors declare no conflicts of interest.
A single-center Italian study of 418 primary infertile men found that a history of maternal obesity during pregnancy was independently associated with low total testosterone (OR 5.9) and lower calculated free testosterone (β = −4.1). The effect is striking on paper, but the exposure group is tiny: only 12 men in the entire cohort reported maternal obesity, which limits how much weight the odds ratio can carry. The proposed mechanism — fetal androgen reprogramming via maternal adipose hormonal activity — is biologically plausible but unproven in humans.
This is a retrospective cross-sectional study from a single academic tertiary referral center — a design that establishes association, not causation, and is subject to substantial referral bias (these are men sick enough to present to a subspecialty center, not a population sample). The maternal obesity exposure was self-reported and defined retrospectively as BMI ≥ 30 kg/m² in early pregnancy, which introduces recall bias with no validation against medical records. The regression models adjust for age, BMI, and testicular volume, which is reasonable, but the effective sample size for the key exposure group is just 12 men. An OR of 5.9 derived from 12 exposed individuals has extremely wide confidence intervals and is highly sensitive to one or two misclassifications.
The exclusion criteria were aggressive: 221 men with azoospermia, 104 with genetic abnormalities, 35 with genitourinary infections, and 29 with recent treatment history were all removed, leaving a narrower and arguably healthier infertile population than typical clinical cohorts. This shapes who the findings apply to.
The biological hypothesis here is genuinely interesting — that prenatal exposure to an obesogenic hormonal environment during the male reprogramming window could permanently alter testosterone production — and the OR of 5.9 sounds alarming. But this paper cannot support the weight placed on that number. Twelve exposed men across a single referral center is not a foundation for an odds ratio; it's a pilot observation. The retrospective self-reported exposure, single-center design, and absence of replication make this hypothesis-generating at best. It is worth watching if a prospective multi-center study picks this up with maternal medical records and larger exposed samples. Until then, treat the 5.9 as a signal, not a finding.