FT3 combined with the TyG index for risk stratification of mild cognitive impairment in euthyroid patients with type 2 diabetes.
Cognitive impairment is increasingly recognized as an important central nervous system complication of type 2 diabetes mellitus (T2DM). Thyroid hormones (THs) and insulin resistance are both implicated in brain metabolism; however, the associations of free triiodothyronine (FT3) and the triglyceride-glucose (TyG) index with cognitive impairment in euthyroid patients with T2DM remain incompletely understood.
This retrospective cross-sectional study enrolled 525 euthyroid hospitalized patients with T2DM between September 2023 and March 2025. Cognitive function was assessed using the Montreal Cognitive Assessment (MoCA), and mild cognitive impairment (MCI) was defined as a MoCA score of 19-25. Logistic regression analyses were used to evaluate the associations of FT3 and the TyG index with MCI. Restricted cubic spline (RCS) models were applied to explore dose-response relationships and potential nonlinearity. Receiver operating characteristic (ROC) curve analyses were performed to assess the discriminative ability of FT3, the TyG index, and their combined model. The DeLong test was used to compare differences in AUCs between models. Calibration analysis and decision curve analysis (DCA) were further conducted to evaluate the predictive agreement and potential clinical utility of the combined FT3 - TyG model.
Lower FT3 levels were independently associated with a higher odds of MCI, even within the euthyroid range, exhibiting a significant linear negative dose-response relationship. The TyG index showed a nonlinear, asymmetric inverted J-shaped association with the odds of prevalent MCI, with a clear inflection point at TyG = 9.24, above which the estimated odds increased more rapidly. ROC analysis demonstrated good discriminative performance for FT3 alone (AUC = 0.761, 95% CI: 0.721-0.802), whereas the TyG index alone showed limited discriminative value (AUC = 0.571, 95% CI: 0.520-0.622). The combined FT3-TyG model yielded an AUC of 0.766 (95% CI: 0.726-0.807), which was not significantly higher than that of FT3 alone. However, within individuals with higher TyG levels, FT3 provided additional granularity for distinguishing those with a higher likelihood of MCI, supporting its role in risk stratification across heterogeneous metabolic profiles.
In euthyroid patients with T2DM, physiological variation in FT3 remains independently associated with cognitive status, while the TyG index is nonlinearly associated with cognitive impairment, with an identifiable metabolic threshold. The combined FT3-TyG model demonstrated good discriminative ability, suggesting that these readily accessible, noninvasive metabolic indicators may be useful for identifying individuals with a higher likelihood of MCI and for cognitive risk stratification in T2DM.
This retrospective cross-sectional study enrolled 525 euthyroid hospitalized patients with T2DM between September 2023 and March 2025. Cognitive function was assessed using the Montreal Cognitive Assessment (MoCA), and mild cognitive impairment (MCI) was defined as a MoCA score of 19-25. Logistic regression analyses were used to evaluate the associations of FT3 and the TyG index with MCI. Restricted cubic spline (RCS) models were applied to explore dose-response relationships and potential nonlinearity. Receiver operating characteristic (ROC) curve analyses were performed to assess the discriminative ability of FT3, the TyG index, and their combined model. The DeLong test was used to compare differences in AUCs between models. Calibration analysis and decision curve analysis (DCA) were further conducted to evaluate the predictive agreement and potential clinical utility of the combined FT3 - TyG model.
Lower FT3 levels were independently associated with a higher odds of MCI, even within the euthyroid range, exhibiting a significant linear negative dose-response relationship. The TyG index showed a nonlinear, asymmetric inverted J-shaped association with the odds of prevalent MCI, with a clear inflection point at TyG = 9.24, above which the estimated odds increased more rapidly. ROC analysis demonstrated good discriminative performance for FT3 alone (AUC = 0.761, 95% CI: 0.721-0.802), whereas the TyG index alone showed limited discriminative value (AUC = 0.571, 95% CI: 0.520-0.622). The combined FT3-TyG model yielded an AUC of 0.766 (95% CI: 0.726-0.807), which was not significantly higher than that of FT3 alone. However, within individuals with higher TyG levels, FT3 provided additional granularity for distinguishing those with a higher likelihood of MCI, supporting its role in risk stratification across heterogeneous metabolic profiles.
In euthyroid patients with T2DM, physiological variation in FT3 remains independently associated with cognitive status, while the TyG index is nonlinearly associated with cognitive impairment, with an identifiable metabolic threshold. The combined FT3-TyG model demonstrated good discriminative ability, suggesting that these readily accessible, noninvasive metabolic indicators may be useful for identifying individuals with a higher likelihood of MCI and for cognitive risk stratification in T2DM.
Authors
Su Su, Chen Chen, Huang Huang, Lyu Lyu, Feng Feng, Zhan Zhan, Li Li, Ye Ye, Liu Liu, Ying Ying
View on Pubmed