Predictive Value of TyG-BMI, CTI, and SII in Identifying Metabolic Dysfunction-Associated Steatotic Liver Disease Among Patients with Type 2 Diabetes Mellitus

Why It Matters

Early, accurate detection of MASLD in type‑2 diabetes can guide timely interventions, reducing liver‑related morbidity and health‑care costs. The study demonstrates that readily available laboratory data can replace expensive imaging for large‑scale screening.

Key Takeaways

• TyG‑BMI, CTI, and SII independently predict MASLD in T2DM
• Combined model yields AUC 0.94, outperforming single markers
• External validation confirms high sensitivity (84.6%) and specificity (90.1%)
• Random forest confirms robustness with AUC up to 0.95
• Tool offers cost‑effective, non‑invasive screening for diabetic liver disease

Pulse Analysis

Metabolic dysfunction‑associated steatotic liver disease (MASLD) has emerged as a silent but serious comorbidity among patients with type‑2 diabetes mellitus (T2DM), affecting up to one‑third of this population. Traditional diagnosis relies on imaging modalities such as ultrasound or MRI, which are costly, operator‑dependent, and impractical for routine screening in busy outpatient settings. Consequently, clinicians have been searching for simple, inexpensive biomarkers that can flag high‑risk individuals before irreversible liver damage occurs. The growing prevalence of obesity, dyslipidemia, and chronic inflammation in diabetes creates a perfect storm for MASLD, making early detection a public‑health priority.

The study highlights three readily calculable indices: TyG‑BMI, which merges triglyceride levels with glucose and body‑mass index; the C‑reactive protein‑triglyceride‑glucose index (CTI), incorporating an inflammatory marker; and the systemic immune‑inflammation index (SII), derived from neutrophil, platelet, and lymphocyte counts. Each reflects a distinct pathophysiological facet of MASLD—insulin resistance, systemic inflammation, and immune dysregulation. When evaluated individually, all three showed significant predictive power, but their combination produced an AUC of 0.94, surpassing any single metric. External validation in an independent cohort confirmed the model’s reliability, delivering 84.6% sensitivity and 90.1% specificity, while random‑forest machine‑learning reinforced its robustness with an AUC of 0.95.

For healthcare systems, the implications are substantial. By leveraging routine lab results, providers can implement a cost‑effective, point‑of‑care screening algorithm without additional equipment or patient burden. Early identification enables lifestyle counseling, tighter glycemic control, and targeted pharmacotherapy—strategies known to slow hepatic steatosis progression. Moreover, integrating this composite index into electronic health records could trigger automated alerts, streamlining referral pathways to hepatology specialists. Future research should explore longitudinal outcomes, refine cut‑off thresholds across diverse ethnic groups, and assess how the index performs alongside emerging non‑invasive imaging techniques. If adopted widely, this approach could shift MASLD management from reactive treatment to proactive prevention, improving quality of life for millions of diabetics.