Higher TyG index — a proxy for insulin resistance — correlated weakly with lower penile rigidity on objective testing (r = −0.22 at best)
Journal: The Aging Male | Published: 2026-02-11 | Type: Retrospective cohort | PMID: 41671684 Authors: Qian Xin et al. — all from the Department of Urology, Suining Central Hospital, China Funding/COI: Not listed for either
The triglyceride-glucose (TyG) index — calculated from fasting triglycerides and fasting glucose, used as a surrogate marker for insulin resistance — was negatively correlated with several objective penile rigidity measurements in men with ED. The correlations are statistically significant but weak. The study adds objective penile testing data to a literature that has mostly relied on questionnaire-based ED diagnosis.
This is a single-center retrospective study from one hospital in China. The 276 ED patients were evaluated using AVSS-RigiScan — audio-visual sexual stimulation combined with a device that continuously measures tumescence and rigidity at the penile tip and base, which is a more objective outcome than questionnaire scores like the IIEF. That methodological choice is the study's main selling point.
The control group, however, is 28 men drawn from health checkups and premarital exams — a wildly different clinical context than the ED cohort and far too small for meaningful comparison. The abstract does not report whether confounders (age, BMI, diabetes status, hypertension, smoking) were adjusted for in the correlation analysis, which is a significant omission given that TyG index is tightly linked to all of those variables.
The use of AVSS-RigiScan distinguishes this from the many ED-metabolic marker papers that rely solely on self-report. But the correlations are weak, the control group is too small and poorly matched to be meaningful, and the absence of any multivariate model makes it impossible to know whether TyG index is doing independent work or just proxying for age and BMI. This is a preliminary signal paper — useful as a reference for effect sizes if someone designs a properly controlled study, not useful on its own.