Simple Insulin Resistance Score Outperforms Rivals in Predicting Metabolic Syndrome Severity in Chinese Diabetes Patients

Metabolic syndrome has long been treated as a yes-or-no diagnosis: a patient either meets the required cluster of criteria—central obesity, elevated blood pressure, dyslipidemia, and impaired glucose metabolism—or does not. But clinicians have grown increasingly uneasy with that binary framing, because two people who both technically qualify for the diagnosis can carry wildly different degrees of underlying metabolic dysfunction. A new study from China argues that the answer lies in converting metabolic syndrome into a continuous severity score, and in pairing that score with simple, blood-based measures of insulin resistance that any hospital laboratory can compute.

The research, published in BMC Endocrine Disorders by Jiying Zhu and Tao Chen of the Department of Endocrinology and Metabolism at Longyan First Affiliated Hospital of Fujian Medical University, followed 726 patients with confirmed type 2 diabetes mellitus who were hospitalized at the institution between June 2022 and May 2025. All participants provided written informed consent, and the study received approval from the hospital’s ethics committee. The team set out to answer a deceptively simple question: among three widely used insulin resistance indices—the metabolic score for insulin resistance (METS-IR), the homeostasis model assessment of insulin resistance (HOMA-IR), and the serum uric acid-to-high-density lipoprotein cholesterol ratio (UHR)—which one best captures both the presence and the depth of metabolic syndrome in a population already burdened by diabetes?

The choice of these three indices reflects a broader shift in metabolic research. The gold standard for measuring insulin resistance, the euglycemic hyperinsulinemic clamp, is technically demanding, expensive, and impractical for routine clinical use. Surrogate markers have therefore proliferated. HOMA-IR, derived from fasting insulin and fasting plasma glucose, has dominated the field for decades but depends on reliable insulin assays and loses accuracy as beta-cell function declines—a particular problem in type 2 diabetes, where insulin secretion is often impaired. METS-IR, a newer composite score built from fasting glucose, body mass index, triglycerides, and height, sidesteps the need for insulin measurements altogether. UHR, the ratio of serum uric acid to HDL cholesterol, draws on two routinely measured analytes and captures an oxidative and lipid-metabolic dimension of insulin resistance that the other indices ignore.

Using a retrospective case design, the investigators ran a battery of statistical tests on the cohort. Correlation analysis first established whether the three indices moved in tandem with the individual components of metabolic syndrome—waist circumference, blood pressure, triglycerides, HDL cholesterol, and glycemic measures. Multivariable regression models then tested whether each index remained independently associated with metabolic syndrome after accounting for confounders, with variance inflation factors used to check for problematic collinearity between predictors. Finally, receiver operating characteristic (ROC) curves quantified how well each index discriminated between patients with and without the syndrome, and decision curve analysis (DCA) translated those statistical properties into a more clinically meaningful question: which index would deliver the greatest net benefit if actually deployed as a screening tool across a range of risk thresholds?

The results were strikingly consistent. After full multivariable adjustment, all three indices were independently associated with the occurrence of metabolic syndrome, each with a P value below 0.001. When the team applied a China Age-Sex-Ethnicity-Specific Metabolic Syndrome Severity score—a continuous metric designed to grade metabolic burden rather than merely flag it—the positive associations held firm. METS-IR showed a beta coefficient of 0.12 per unit (95% confidence interval 0.11 to 0.12), HOMA-IR a beta of 0.14 (95% CI 0.12 to 0.17), and UHR a beta of 0.14 (95% CI 0.13 to 0.15), all statistically significant. In plain terms, the higher a patient’s insulin resistance score, the more severe their metabolic syndrome, in a graded and quantifiable relationship rather than a threshold effect.

Discrimination told an even sharper story. METS-IR achieved an area under the ROC curve of 0.893 for identifying metabolic syndrome—the strongest of the three indices. An AUC approaching 0.9 is generally considered excellent, indicating that the score separates patients with the syndrome from those without it with high fidelity. The DeLong test, a formal statistical comparison of correlated ROC curves included in the study’s supplementary analyses, was used to confirm that differences in discriminative performance between the indices were meaningful rather than artifacts of sampling. For a metric that requires nothing more than a fasting blood draw and a few anthropometric measurements, that level of performance is notable.

Perhaps the most clinically consequential finding came from the decision curve analysis. ROC curves measure statistical discrimination, but they say little about what happens when a test is actually used to make decisions—which patients get intensified follow-up, which get lifestyle intervention, which get closer pharmacologic attention. DCA addresses this by calculating net benefit across a spectrum of threshold probabilities, weighing the cost of false positives against the cost of missed cases. Across a broad range of thresholds, METS-IR conferred the highest clinical net benefit of the three indices, suggesting that it is not merely the best statistical performer but the most useful practical screening instrument for real-world metabolic risk assessment in this population.

The study’s embrace of a continuous severity framework deserves particular attention. The China Age-Sex-Ethnicity-Specific severity scoring approach recognizes that metabolic risk is not distributed uniformly across a population: the same cluster of abnormalities may carry different prognostic weight depending on a patient’s age, sex, and ethnic background. Chinese cohorts, for example, tend to develop central obesity and metabolic complications at lower waist circumferences and body mass indices than many Western reference populations, and ethnicity-specific calibration helps correct for that. By converting the diagnosis into a graded score, the framework allows clinicians and researchers to track small but meaningful changes in metabolic burden over time—changes that a binary diagnosis would simply erase.

Why does this matter for the millions of people living with type 2 diabetes? Because metabolic syndrome in a diabetic patient is not a redundant label; it marks a substantially elevated risk of cardiovascular events, fatty liver disease, and chronic kidney disease, and it signals that the patient’s overall metabolic machinery is under strain beyond the glucose disorder itself. Insulin resistance sits at the mechanistic heart of this cluster: it drives hepatic fat accumulation, promotes very-low-density lipoprotein secretion, impairs vascular function, and links adipose tissue inflammation to systemic metabolic derangement. A cheap, reproducible index that tracks this underlying driver—and that correlates with the graded severity of the syndrome—gives clinicians a quantitative handle on risk that the current diagnostic checklist cannot provide.

The authors conclude that METS-IR is strongly associated with both the development and the severity of metabolic syndrome and plays a significant role in its risk assessment and clinical management. The study is not without the usual limits of a single-center retrospective design, and the findings will need validation in prospective, multi-ethnic cohorts before practice changes. But the direction of travel is clear. As medicine moves away from binary disease labels toward continuous, personalized measures of metabolic health, tools like METS-IR—simple enough to calculate in any clinic, yet powerful enough to grade the full spectrum of metabolic dysfunction—may become a routine part of how physicians judge and manage risk in patients with type 2 diabetes, in China and well beyond.

Subject of Research: Insulin resistance indices and metabolic syndrome severity in Chinese patients with type 2 diabetes

Article Title: The association between insulin resistance and the occurrence of metabolic syndrome in Chinese patients with type 2 diabetes, and the application of China Age-Sex-Ethnicity-Specific of Metabolic Syndrome Severity scoring

Article References: Zhu, J., & Chen, T. (2026). The association between insulin resistance and the occurrence of metabolic syndrome in Chinese patients with type 2 diabetes, and the application of China Age-Sex-Ethnicity-Specific of Metabolic Syndrome Severity scoring. BMC Endocrine Disorders. https://doi.org/10.1186/s12902-026-02580-9

Image Credits: AI Generated

DOI: 10.1186/s12902-026-02580-9

Keywords: metabolic syndrome, insulin resistance, type 2 diabetes, METS-IR, HOMA-IR, UHR, metabolic syndrome severity score, ROC analysis, decision curve analysis, BMC Endocrine Disorders, China, cardiometabolic risk