Background Information

This tool predicts the endotypes of Type 2 Diabetes in four groups such as:

  • Severe Insulin Deficient Diabetes (SIDD)
  • Insulin Resistant Obese Diabetes (IROD)
  • Combined Insulin Resistant and Deficient Diabetes (CIRDD)
  • Mild Age-Related Diabetes (MARD)

Information Required for Cluster Assignment

To assign a person with type 2 diabetes to one of the identified subtypes, the following clinical parameters are required:

  • Age at diagnosis (years)
  • Body Mass Index (BMI)
  • Waist circumference
  • HbA1c (%)
  • Triglycerides
  • HDL cholesterol
  • C-peptide (stimulated) – CPS
  • C-peptide (fasting) – CPF

Information Required for Risk Prediction

The following parameters are used to estimate the risk of diabetes-related complications or disease progression:

  • Age
  • Age at diagnosis
  • Waist circumference
  • Gender
  • Body Mass Index (BMI)
  • HbA1c (%)
  • Triglycerides
  • HDL cholesterol
  • Total cholesterol
  • Systolic and Diastolic Blood Pressure
  • Serum Creatinine

These parameters are analyzed using machine learning models trained on longitudinal data to predict individualized risk scores and disease progression profiles.

Note: The clustering model was originally developed for individuals with newly onset diabetes patients; please consider this context while interpreting the results

References



developed by Madras Diabetes Research Foundation

Legal Information

Contact

If you have any questions or feedback regarding this tool, please contact us at: baskarv@mdrf.in

Author Information

Department of Data Science
Madras Diabetes Research Foundation
Plot No. 20, Golden Jubilee Biotech Park for Women Society,
SIPCOT - IT Park, Siruseri,
Chennai, Tamil Nadu 603103, India

Disclaimer

This tool is developed solely for research purposes to assist in classifying individuals with type 2 diabetes into specific clusters, based on published methodologies.

It is not intended to replace medical advice, diagnosis, or treatment. The results generated by this tool must be interpreted in conjunction with professional clinical evaluation and judgment.

By using this tool, you acknowledge and agree that:

  • The outputs do not constitute medical advice or a diagnosis.
  • No personal medical decisions should be made solely based on this tool.
  • The authors, developers, and affiliated institutions are not liable for any direct or indirect damages resulting from the use or misuse of this tool or the data provided within.

Important Note

We track the number of visits to the DIANA Tools page, but no personal data entered is stored