Can the triglyceride-glucose index predict cardiovascular autonomic neuropathy in type 2 diabetes?

Aims: To investigate if baseline triglyceride-glucose (TyG) index and changes in TyG index over time are associated with the incidence of cardiovascular autonomic neuropathy (CAN) in type 2 diabetes (T2D).

Methods: Data from 10,251 participants in the ACCORD trial were analyzed, excluding those with missing data or baseline CAN. The TyG was calculated as ln[triglycerides [mg/dL] × fasting blood glucose [mg/dL] /2], while CAN was diagnosed based on abnormal heart-rate variability (HRV) time-domain indices from 10-second ECG recordings (SDNN (< 8.2 ms) and rMSSD (< 8.0 ms). Multivariate logistic regression assessed the association between TyG quartiles and CAN prevalence. Restricted cubic splines (RCS) analysis evaluated the relationship between baseline TyG index and incident CAN, and the area under the ROC curve (AUC) assessed the diagnostic accuracy of TyG index for CAN. A latent growth curve model (LGCM) identified subgroups (“classes”) with similar TyG trajectories; their associations with incident CAN were examined using Kaplan-Meier curves and Cox proportional hazards models and classified by standard or intensive glycemic treatment.

Results: Participants in the highest TyG quartile had a greater CAN risk compared to those in the lowest quartile (OR = 1.29, P = 0.027). RCS revealed a linear increase in CAN risk with higher TyG values (P = 0.030). The TyG index demonstrated an AUC of 0.636 for predicting CAN (95% CI 0.620–0.651; P < 0.001). Over a follow-up period of 7 years, three TyG index trajectories were identified: class 1, class 2, and class 3. Participants in class 1 had lower TyG index values and a reduced incidence of CAN compared to those in classes 2 and 3. Class 3 exhibited the highest incidence of CAN (log-rank P < 0.001). In Cox regression, class 3 had 1.52-fold higher risk of CAN than class 1 (P < 0.001). Among participants on standard glycemic treatment, classes 2 or 3 had higher CAN risk than class 1. On intensive glycemic treatment, class 2 had reduced CAN risk (HR: 0.83, P = 0.014), whereas class 3 had increased risk (HR: 1.35, P = 0.018).

Conclusions: Higher baseline TyG index and increases of TyG over time were associated with a greater incidence of CAN. Intensive glycemic treatment appeared to modify the relationship between TyG index trajectories and the risk of developing CAN.

Comments: The TyG index has been proposed as an easily accessible surrogate marker of insulin resistance and has been associated with numerous cardio-metabolic outcomes, including metabolic syndrome, T2D, hypertension, diabetic retinopathy and chronic kidney disease. This study analyzing the data of the ACCORD trials reveals the potential of both baseline and longitudinal TyG index measures for predicting the risk of HRV-based CAN in individuals with T2D. Moreover, it shows that the impact of intensive glycemic treatment (targeting HbA1c < 6%) on CAN risk may vary according to TyG index trajectories, thus identifying TyG as a possible modifier of this effect.  However, given the limited diagnostic accuracy of baseline TyG index for CAN, its combination with other screening methods should be considered. Furthermore, assessment of changes in TyG index over time could support patient risk stratification. Large-scale prospective studies are warranted to confirm the findings of this study and examine their impact on clinical decision-making.

Pietro Pertile

Reference: Huang Q, Nan W, He B, Xing Z, Peng Z. Association of baseline and trajectory of triglyceride-glucose index with the incidence of cardiovascular autonomic neuropathy in type 2 diabetes mellitus. Cardiovasc Diabetol. 2025 Feb 7;24(1):66. doi: 10.1186/s12933-025-02622-x. PMID: 39920656; PMCID: PMC11806751.

🔗 https://cardiab.biomedcentral.com/articles/10.1186/s12933-025-02622-x