Operational performance of the METS-IR score for metabolic syndrome and its association with cardiovascular risk in older adults

Background

Insulin resistance is the central pathophysiological driver of metabolic syndrome (MetS) and a major contributor to cardiovascular risk in older adults. The Metabolic Score for Insulin Resistance (METS-IR), which integrates fasting glucose, triglycerides, HDL-cholesterol, and body mass index, represents a pragmatic surrogate marker. However, its diagnostic performance in Latin American elderly populations remains insufficiently established.

Purpose

To evaluate the operational performance of METS-IR for identifying metabolic syndrome and to examine its association with estimated cardiovascular risk, in comparison with triglyceride–glucose (iTyG) and triglyceride–HDL (iTyHDL) indices.

Methods

We conducted a cross-sectional analytical study in adults aged ≥60 years from a nationally representative survey. From more than 23,000 participants, 4,690 individuals with complete data on glucose, triglycerides, HDL-cholesterol, and anthropometry were included. METS-IR was calculated as ln[(2×fasting glucose)+triglycerides]×(BMI/ln[HDL-C]). Metabolic syndrome was defined according to ATP III/IDF criteria.

Cardiovascular risk was estimated using a modified SCORE2-Older Persons (SCORE2-OP) model incorporating age, sex, systolic blood pressure, cholesterol, diabetes, and smoking. A normalized 0–100 index was derived as a relative risk estimator. Discriminatory performance was assessed using receiver operating characteristic (ROC) curves, and areas under the curve (AUC) were compared using DeLong’s method.

Results

Mean METS-IR was 37.9 ± 11.8, and the prevalence of metabolic syndrome was 58%. METS-IR demonstrated excellent discriminatory capacity for metabolic syndrome (AUC 0.884). The optimal cut-off (42.8) yielded a sensitivity of 83% and specificity of 76%.

A progressive increase in estimated cardiovascular risk was observed across METS-IR quartiles, with normalized SCORE2-OP values of 41.2, 49.6, 56.8, and 63.5 for Q1 through Q4, respectively.

In comparative ROC analysis, AUC values were 0.884 for METS-IR, 0.840 for iTyG, and 0.815 for iTyHDL. Pairwise comparisons using DeLong’s test showed no statistically significant differences between METS-IR and iTyG (p=0.183), METS-IR and iTyHDL (p=0.071), or iTyG and iTyHDL (p=0.640).

Conclusion

METS-IR demonstrates excellent operational performance for identifying metabolic syndrome and shows a clear, graded association with estimated cardiovascular risk in older adults. Together with other triglyceride-based indices, it represents a low-cost and scalable tool with potential utility for cardiometabolic risk stratification in elderly populations. Prospective studies are warranted to assess its prognostic value for hard cardiovascular outcomes.Figure 1 ROC Curve