Enhancing Type 1 Diabetes Immunological Risk Prediction with Continuous Glucose Monitoring and Genetic Profiling
Authors' conclusion
A machine learning approach combining a 1-week CGM home test and T1D GRS reliably assesses T1D immunological risk, enabling early intervention.
Study context
- Population
- Autoantibodies (AB)-positive healthy relatives of people with T1D; mean age 22.1 ± 11.1 years, HbA1c 5.3 ± 0.3%, BMI 24.1 ± 5.8 kg/m²; 1 week of continuous glucose monitoring (CGM) with three standardized liquid mixed meals
- Participants
- 39 relatives
- Comparison
- Support vector machine (SVM) classifier with 2 recursive feature elimination (RFE)-selected features (T1D genetic risk scores (GRS) and incremental AUC) vs all features, classifying 1 vs ≥2 AB
- 1 ABn = 18reference
- ≥2 ABn = 21
- Outcome
- Classification of 1 vs ≥2 AB by AUC-ROC and AUC-PR
Methodological detailsNational Institutes of Health's (NIH) TrialNet ancillary study, linear SVM with recursive feature elimination, fivefold cross-validation
- NIH TrialNet ancillary study
- Linear SVM with recursive feature elimination
- Fivefold cross-validation
Performance
As reported in the abstract
Classification of 1 vs ≥2 AB
SVM with 2 RFE-selected features (T1D GRS and incremental AUC): AUC-ROC
0.93
95% CI 0.83–1.00
SVM with 2 RFE-selected features (T1D GRS and incremental AUC): AUC-PR
0.89
95% CI 0.71–0.99
SVM with all features: AUC-ROC
0.80
95% CI 0.46–1.00
SVM with all features: AUC-PR
0.82
95% CI 0.71–0.93
Further findings
| Outcome | P value |
|---|---|
| Differences between AB groups in post-SLMM % time glucose >180 mg/dL and GRSBetween 1 AB and ≥2 AB · significant difference | P = 0.020 |
| Differences between AB groups in post-SLMM % time glucose >180 mg/dL and GRSBetween 1 AB and ≥2 AB · significant difference | P = 0.001 |
Original publication
Enhancing Type 1 Diabetes Immunological Risk Prediction with Continuous Glucose Monitoring and Genetic Profiling. Diabetes Technol Ther. Published 17 December 2024. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →