Routine Blood Markers Show Limited Cardiovascular Signal

A retrospective study of 7034 diabetes patients found modest differences in routine blood-derived indices and weak machine-learning discrimination.

Updated on Nov. 17, 2026 in Diabetes

Bold flat-color editorial illustration of a row of plain glass blood tubes in deep red, one set slightly apart, on a cream field.
A retrospective study of 7,034 diabetes patients found routine blood markers offer only modest, weak signals of cardiovascular comorbidity. AI Illustration. Upload story photo >

Researchers analyzed routine laboratory data from 7034 diabetes patients, comparing those with and without documented cardiovascular comorbidity. Most parameter differences were negligible or small, and exploratory classifiers showed poor to modest discrimination between the groups.

Why it matters

Routine blood tests are inexpensive and widely available, so knowing whether they can flag cardiovascular risk in diabetes patients matters for clinical screening. The results suggest these markers alone are not yet reliable for that purpose.

The analysis used retrospective data from 7034 patients, with machine-learning ROC AUC values between 0.544 and 0.668 and balanced accuracy between 0.526 and 0.599. Whether these routine indices add clinical value beyond existing risk tools has not been established.

The details

Investigators compared routine laboratory variables and derived indices across diabetes patients with and without documented cardiovascular comorbidity, using robust effect-size estimates and FDR-adjusted regression. Higher PDW, TyG, MHR, THR and MLR were associated with higher odds of cardiovascular comorbidity, while higher MPV, SIRI and NLR were associated with lower odds. A complementary analysis excluding hypertension-only patients was conducted because hypertension predominated in the comorbidity group.

Timeline

  1. Article published as early access on November 17, 2026, ahead of the final version of record.

Health Landscape

The study tests whether the triglyceride-glucose (TyG) index, a previously proposed marker of insulin resistance, holds up alongside other routine blood indices for predicting cardiovascular comorbidity in diabetes. Its association was attenuated after excluding hypertension-only patients, which tempers claims for standalone clinical use.

For now, routine blood results should not be used alone to judge cardiovascular risk in people with diabetes, and clinical decisions should follow established risk assessment. Patients should continue discussing cardiovascular screening with their clinicians, since these findings are exploratory and not yet validated for practice.

The takeaway

Routine blood markers may add some information about cardiovascular risk in diabetes, but current evidence is too weak to replace established screening. Watch for prospective validation before these indices influence clinical care.

Further reading

For broader context on how diabetes is managed and monitored, explore the Diabetes section.

Source note: This article includes information reported by Nature.