Simple Blood Sugar-and-Fat Score Tracks Severity of Clogged Leg Arteries in Hospitalized Patients

A routine blood test that costs almost nothing may reveal how badly the arteries in a patient’s legs are clogged, according to a new cross-sectional study of 656 hospitalized patients in China. Researchers at the General Hospital of Ningxia Medical University found that the triglyceride–glucose index, or TyG index — a number calculated simply from fasting blood sugar and triglyceride levels — rose steadily in step with the clinical severity of lower extremity arterial disease, from mild claudication all the way to gangrene. In the most severe category, patients with the highest TyG values faced dramatically higher odds of tissue loss than those with the mildest disease.

The study, published in Frontiers in Endocrinology, examined patients whose leg artery disease was confirmed by digital subtraction angiography, the gold-standard imaging technique, between January 2023 and December 2024. All patients were graded on the Rutherford classification, a six-point scale that runs from mild intermittent claudication (grade 1) through moderate and severe claudication (grades 2 and 3), rest pain (grade 4), and finally minor tissue loss or ulceration (grade 5) and major tissue loss or gangrene (grade 6). Nearly a third of the cohort — 205 patients, or 31.2 percent — had already reached the tissue-loss stages when they were admitted.

The TyG index itself is a logarithmic transformation of the product of fasting triglycerides and fasting glucose. It has become popular in recent years as a cheap surrogate for insulin resistance, avoiding the insulin measurements required by tools such as HOMA-IR or the hyperinsulinemic–euglycemic clamp. Prior meta-analytic evidence had already linked elevated TyG values to an increased risk of developing lower extremity arterial disease, with a relative risk of 1.42, and a stronger association among people with diabetes than among those without. What remained unclear was whether the index tracked the severity of established disease, and whether that relationship differed depending on a patient’s underlying glucose metabolic state.

To probe that question, the team divided the 656 patients into four groups: 150 with normal glucose tolerance, 148 with impaired glucose regulation, 174 with diabetes of less than ten years’ duration, and 184 with diabetes lasting ten years or longer. The mean TyG index climbed across these categories, from 8.79 in the normal-glucose group to 9.15, 9.57, and 9.67 in the progressively more affected groups. Median Rutherford grade followed the same trajectory, sitting at 2.0 in both non-diabetic groups but rising to 4.0 and 5.0 in the early- and long-duration diabetes groups respectively. A Spearman correlation analysis confirmed that the TyG index was the variable most strongly associated with severity, with a correlation coefficient of 0.677.

The regression results were striking. In multinomial logistic models adjusted for age, sex, smoking, hypertension, body mass index, and fatty liver disease, each 0.5-unit increase in the TyG index was associated with adjusted odds ratios of 2.51 for moderate claudication, 2.83 for severe claudication, 7.56 for rest pain, 9.62 for ulceration, and 20.45 for gangrene, compared with mild claudication as the reference. The authors caution, however, that the model’s very large intercepts for the advanced categories signal sparse-data instability, meaning the absolute magnitude of these odds ratios should not be read as clinically meaningful effect sizes. Sensitivity analyses that excluded implausibly low fasting glucose values or added liver enzymes to the adjustment set left the estimates largely unchanged.

The most conceptually interesting finding concerned effect modification by glucose metabolic state. A formal interaction test was statistically significant, and stratum-specific analyses suggested the TyG–severity association was weakest in impaired glucose regulation (odds ratio 1.51 per 0.5 unit), intermediate in early diabetes (2.50), and strongest in long-standing diabetes (8.14). The estimate in the normal-glucose-tolerance stratum was statistically unstable and could not be interpreted. On its face, this gradient fits a plausible biological story in which prolonged hyperglycemia and insulin resistance synergistically damage endothelial function, promote oxidative stress, and accelerate plaque progression.

But the researchers themselves dismantled part of that narrative. Because the TyG index mathematically contains fasting glucose, and the four groups were defined partly by glycemic status, a degree of mathematical coupling between exposure and grouping is unavoidable. When the team replaced the categorical groups with continuous fasting glucose in a sensitivity analysis, the interaction remained significant but reversed direction: the TyG–severity association was actually strongest at the lowest glucose levels (an odds ratio of 4.81 at 4.8 mmol/L) and weakened as glucose rose (3.46 at 8.3 mmol/L). The authors conclude that the apparent four-group gradient is at least partly an artifact of the grouping itself and range restriction, though they note that continuous glucose cannot capture diabetes duration or treatment history, so duration-related biological effects cannot be excluded.

The study also quantified how well the TyG index discriminated prevalent ischemic tissue loss. The area under the receiver operating characteristic curve was 0.833, and bootstrap internal validation with 1,000 resamples confirmed this estimate with negligible optimism. The Youden-optimal cutoff was 9.42, a value that remained stable across resamples, yielding a sensitivity of 85.9 percent and a specificity of 71.0 percent. The positive predictive value was a more modest 57.3 percent — meaning nearly half of patients flagged by the cutoff did not have tissue loss — while the negative predictive value reached 91.7 percent. Decision-curve analysis showed a net benefit over treat-all or treat-none strategies only across a limited band of threshold probabilities, roughly 0.09 to 0.57, supporting a role as an adjunctive triage variable rather than a stand-alone test.

The authors are explicit about the study’s limits. Because fasting labs and severity assessments were collected during the same hospitalization, the findings describe cross-sectional association and concurrent discrimination, not prediction of future events; reverse causation is plausible, since active ulceration, infection, and physiological stress can all perturb glucose and triglyceride levels. The cohort consisted entirely of hospitalized vascular surgery patients with angiographically confirmed disease, so the results cannot be generalized to primary care or early-stage disease. Key covariates — including HbA1c, glucose-lowering and lipid-lowering treatment, renal function, ankle–brachial index, and toe pressures — were unavailable, and the Rutherford system alone cannot distinguish ischemic from neuropathic or infected ulcers. The proposed cutoff of 9.42 was derived and evaluated in the same sample and requires external validation before any clinical use.

Even with those caveats, the work adds a technically careful data point to a growing literature on metabolic markers and limb threat. The TyG index costs pennies, requires only two routine fasting measurements, and is widely available in settings where vascular imaging capacity is limited. In this selected population, a value above roughly 9.4 was associated with a substantially higher concurrent probability of tissue loss. The authors emphasize that the index should be understood as one element of metabolic risk assessment — never a substitute for ankle–brachial index measurement, duplex ultrasonography, or angiography — and that whether TyG-guided assessment can actually prevent ulceration or improve limb outcomes is a question only prospective, multicenter studies can answer.

Subject of Research: Association between the triglyceride–glucose index and the severity of lower extremity arterial disease across glucose metabolic states

Article Title: Association of the triglyceride–glucose index with the severity of lower extremity arterial disease across glucose metabolic states: a single-center cross-sectional study of 656 patients with angiographically confirmed disease

Article References: Zhou, X., Zhao, D., Lei, H., Zhao, G., Sha, L., Lu, X., & Liang, H. (2026). Association of the triglyceride–glucose index with the severity of lower extremity arterial disease across glucose metabolic states: a single-center cross-sectional study of 656 patients with angiographically confirmed disease. Frontiers in Endocrinology, 17, Article 1957246. https://doi.org/10.3389/fendo.2026.1957246

Image Credits: AI Generated

DOI: 10.3389/fendo.2026.1957246

Keywords: triglyceride–glucose index, lower extremity arterial disease, insulin resistance, Rutherford classification, ischemic tissue loss, diabetes, cross-sectional study, peripheral arterial disease, fasting glucose, triglycerides, diabetic foot, risk stratification