Evaluating the TyG Index in Diabetes Prediction for COVID-19 Patients

A new cross-sectional pilot shows the triglyceride–glucose (TyG) index strongly associates with newly diagnosed diabetes among patients hospitalized with moderate–severe COVID-19, and the association is amplified when combined with BMI; the study identifies TyG as a rapid, low-cost marker suitable for admission risk stratification.

Compared with insulin-based indices such as HOMA-IR, TyG is operationally simpler: it requires only fasting triglyceride and glucose, avoiding insulin assays. This practicality shortens turnaround time and facilitates deployment at scale in resource-limited settings.

In this cohort of 84 hospitalized moderate–severe COVID-19 patients (predominantly Black African), 45 were newly diagnosed with diabetes. Per the study data, TyG ≥5.08 showed a univariate crude OR of 11.25 and an adjusted OR of 6.83 for newly diagnosed diabetes; model AUCs ranged from 0.73 for TyG alone to 0.91 for the model including TyG, BMI, IL-10 and IL-1β.

Because systemic inflammation and adiposity were prominent, adding inflammatory markers and BMI materially improved discrimination: TyG plus BMI increased AUC to 0.86, and the full model with IL-10 and IL-1β reached an AUC of 0.91. These results derive from cross-sectional, in‑hospital data and may not generalize beyond similar cohorts or acute-care settings.

Operationally, the findings support pragmatic adoption of TyG-based screening in constrained environments, provided targeted validation and workflow integration at admission are undertaken.

Taken together, the paper provides a clear rationale for using routinely available laboratory tests to enhance early case-finding in similar inpatient populations while underscoring the need for prospective, multicenter validation and longer-term follow-up to distinguish transient stress hyperglycaemia from persistent diabetes.