What is the clinical evidence from this study?

Study design: Cross-Sectional. Population: Insulin resistance in adults without diabetes (n=61). Intervention: TyG-WHtR vs. HOMA-IR. Primary outcome: Area under the curve (AUC) for identifying clamp-defined insulin resistance (AUC 0.829, 95% CI 0.707-0.932, p=<0.001).

What question did this study set out to answer?

This research aims to assess the effectiveness of anthropometry-adjusted triglyceride-glucose indices in estimating insulin resistance in Japanese adults without diabetes.

April 8, 2026Open Access

Anthropometry-adjusted TyG indices improve insulin resistance estimation: an exploratory euglycemic–hyperinsulinemic clamp study in Japanese adults without diabetes

Q: What are the key findings of this study?

Anthropometry-adjusted TyG-derived indices, particularly TyG-WHtR (AUC 0.829) and TyG-BMI (AUC 0.810), demonstrated superior diagnostic performance for identifying clamp-defined insulin resistance compared to HOMA-IR.

Q: What question did this study set out to answer?

This research aims to assess the effectiveness of anthropometry-adjusted triglyceride-glucose indices in estimating insulin resistance in Japanese adults without diabetes.

Key Result

Anthropometry-adjusted TyG-derived indices, particularly TyG-WHtR (AUC 0.829) and TyG-BMI (AUC 0.810), demonstrated superior diagnostic performance for identifying clamp-defined insulin resistance compared to HOMA-IR.

Key Points

This research aims to assess the effectiveness of anthropometry-adjusted triglyceride-glucose indices in estimating insulin resistance in Japanese adults without diabetes.
Sixty-one Japanese adults without diabetes participated in an euglycemic-hyperinsulinemic clamp study to assess insulin sensitivity.
Surrogate indices of insulin resistance, including HOMA-IR and various TyG indices, were calculated and compared against the reference standard.
Receiver operating characteristic analysis and linear regression were utilized to determine the diagnostic accuracy of the indices.
TyG-BMI and TyG-WHtR showed strong correlations with insulin sensitivity index (ISI), with ρ = −0.544 and −0.546 respectively.
AUC values for TyG-BMI and TyG-WHtR were 0.810 and 0.829, indicating high diagnostic performance for identifying insulin resistance.
Anthropometry-adjusted TyG indices outperformed conventional markers like HOMA-IR, especially with optimal cutoff values around 203.7 for TyG-BMI.

Study Design

Type

Cross-Sectional (n=61)

Multicenter

Structured PICO

Do anthropometry-adjusted TyG-derived indices improve the diagnostic estimation of insulin resistance compared to conventional markers in Japanese adults without diabetes?

Population

61 Japanese adults aged ≥20 years without a prior diagnosis of diabetes, median age 34.0, 73.8% male. Excluded if receiving medications affecting glucose metabolism, pregnant, lactating, or having psychiatric disorders.

Intervention

Anthropometry-adjusted triglyceride-glucose (TyG)-derived indices (TyG-BMI, TyG-WC, TyG-WHtR)

Comparator

Conventional surrogate markers of insulin resistance (HOMA-IR, original TyG index, Matsuda index)

Outcome

Diagnostic performance (Area Under the Curve) for identifying insulin resistance, defined as the lowest quartile of the insulin sensitivity index (ISI) derived from the euglycemic-hyperinsulinemic clamp (EHC)surrogate

Anthropometry-adjusted TyG indices, particularly TyG-BMI and TyG-WHtR, are superior to conventional markers like HOMA-IR for estimating insulin resistance in adults without diabetes.

Main Result

Effect estimate: AUC 0.829 (95% CI 0.707-0.932)

Absolute Event Rate: 0.829% vs 0.667%

p-value: p=<0.001

Limitations

TyG-BMI and TyG-WC incorporate anthropometric components by design, and their discriminative ability for IR is therefore partly influenced by adiposity.
Relatively small sample size reflects the resource-intensive nature of clamp studies limiting statistical power, particularly for cutoff estimation.
Small number of female participants may limit the reliability of sex-specific analyses.
Key determinants of IR, including visceral fat area, adipokines, serum uric acid, and lifestyle factors, were not available.

Abstract

We evaluated the diagnostic performance of anthropometry-adjusted triglyceride-glucose (TyG)-derived indices for assessing insulin resistance (IR) in Japanese adults without diabetes, using the euglycemic–hyperinsulinemic clamp (EHC) as the reference standard. A total of 61 Japanese individuals without diabetes underwent insulin sensitivity assessment using the gold standard EHC. IR was defined as an insulin sensitivity index (ISI) below the 25th percentile. Surrogate IR indices, including Homeostasis Model Assessment of Insulin Resistance (HOMA-IR), the original TyG index, and anthropometry-adjusted TyG-derived indices (TyG-BMI, TyG-WC, TyG-WHtR), were calculated. Correlations between these indices and the ISI were evaluated using Spearman’s rank correlation. Receiver operating characteristic (ROC) analysis and linear regression were used to compare diagnostic accuracy and predictive ability. Pairwise DeLong tests were used to assess differences in the area under the curve (AUC) values. TyG-BMI and TyG-WHtR showed the stronger associations with ISI (ρ = − 0.544 and − 0.546, respectively) than both HOMA-IR and the original TyG index. TyG-BMI and TyG-WHtR had the highest AUCs (0.810 and 0.829, respectively) for identifying clamp-defined IR. The optimal cutoff value for HOMA-IR was approximately 2.5. The optimal cutoff for TyG-BMI was 203.7. Pairwise DeLong tests supported the higher diagnostic performance of selected anthropometry-adjusted TyG-derived indices compared with conventional markers. Anthropometry-adjusted TyG-derived indices, particularly TyG-BMI and TyG-WHtR, showed strong associations with IR as assessed by the gold standard EHC in Japanese adults without diabetes. These findings suggest that these indices may serve as practical surrogate markers for IR in settings where direct clamp measurements are not feasible.

Bookmark

View Full Paper