A NEW prediction model was shown to accurately identify patients with Type 2 diabetes mellitus at higher risk of developing thyroid dysfunction, offering a practical tool for earlier detection and prevention.
Researchers developed and validated a thyroid dysfunction prediction model using clinical data from 1,853 patients with Type 2 diabetes mellitus treated between 2019–2024. The study addressed a growing clinical challenge, as co-existing thyroid dysfunction has been linked to worse metabolic control, increased cardiovascular risk, and higher healthcare costs in diabetes care.
Why Thyroid Dysfunction Matters in Type 2 Diabetes
Thyroid dysfunction is more common in patients with Type 2 diabetes mellitus than in the general population, yet it often remains underdiagnosed. Thyroid hormones play a key role in glucose metabolism, lipid regulation, and cardiovascular health. When thyroid dysfunction occurs alongside diabetes, patients face a higher risk of complications such as coronary heart disease and hypertension, increasing the burden on already strained healthcare systems.
Early identification of individuals at greatest risk is therefore clinically important. However, routine thyroid screening for all patients with Type 2 diabetes mellitus is not always feasible, highlighting the need for targeted risk stratification.
Thyroid Dysfunction Prediction Model Shows Strong Performance
Using univariate and multivariate logistic regression, the investigators identified eight independent predictors of thyroid dysfunction: high-density lipoprotein cholesterol, blood urea nitrogen, gender, fasting glucose, hypertension, hyperuricaemia, coronary heart disease, and liver disease. These variables were integrated into a thyroid dysfunction prediction model presented as a nomogram for use in clinical practice.
The model demonstrated good discrimination in both training and validation cohorts, with area under the curve analyses confirming reliable predictive accuracy. Calibration curves showed strong agreement between predicted and observed risk, while decision curve analysis indicated meaningful clinical benefit across a range of threshold probabilities.
Importantly, all predictors included in the nomogram are routinely measured in standard diabetes care, supporting real-world applicability.
Clinical Implications and Limitations
This thyroid dysfunction prediction model may help clinicians identify patients with Type 2 diabetes mellitus who would benefit most from closer thyroid monitoring or early intervention. By focusing screening efforts on higher-risk individuals, the approach could improve prognosis while reducing unnecessary testing.
The single-centre design and retrospective data collection may affect generalisability. Hence, external validation in diverse populations will be needed before widespread adoption. Nevertheless, the findings support a move toward more personalised thyroid dysfunction risk assessment in diabetes management.
Reference
Niu Y et al. Development and validation of a nomogram prediction model for thyroid dysfunction in patients with Type 2 diabetes mellitus. Sci Rep. 2026; DOI:10.1038/s41598-026-36582-3.
