Tosi · Human reproduction (Oxford, England) 2017 · retrospective diagnostic validation study · n=375

Insulin resistance in a large cohort of women with polycystic ovary syndrome: a comparison between euglycaemic-hyperinsulinaemic clamp and surrogate indexes.

Cited 157 times in the scientific literature.

Level 3 - non-randomized controlled study

Retrospective diagnostic accuracy study comparing index tests to a gold-standard reference

PubMed 29040529 · doi:10.1093/humrep/dex308 · record verified 2026-08-29

What was done

Researchers conducted a retrospective diagnostic study of 375 women with polycystic ovary syndrome (PCOS) diagnosed by Rotterdam criteria (from a cohort of 406 evaluated between 1998 and 2015). All participants completed a hyperinsulinemic-euglycemic clamp (M-clamp) as the gold-standard reference. The diagnostic accuracy of six surrogate indexes of insulin sensitivity—three fasting (HOMA, glucose/insulin ratio, QUICKI) and three post-oral glucose load (Gutt, Stumvoll0-120, Matsuda)—was evaluated against clamp-derived insulin resistance.

What was found

Insulin resistance by M-clamp was identified in 74.9% of women overall, occurring in 59.3% of normal-weight, 77.5% of overweight, and 93.9% of obese participants. All surrogate indexes correlated with M-clamp measurements, with ROC AUC values ranging between 0.782 and 0.817. However, sensitivity was limited across indexes, particularly in normal-weight individuals, leading to false-negative classifications.

Why it matters

Surrogate fasting and post-load indexes can rule in insulin resistance when abnormal, but they cannot reliably rule it out in women with PCOS, especially lean patients.

Limits

The single-center retrospective design introduces potential referral bias toward more severe cases. The abstract does not report specific cut-off thresholds, exact sensitivity/specificity percentages for individual markers, or whether clamp measurements and surrogate index calculations were interpreted blindly.

Cited by