Prediction of glycaemic control and quality of life in people with type 2 diabetes using glucose‐lowering drugs with machine learning—The Maastricht study
Authors' conclusion
Prediction of glycaemic control after 1 year in T2D is feasible with good model performance. However, the prediction of deterioration in QoL remains a challenge and needs further work.
Study context
- Population
- People with T2D using glucose-lowering drugs, The Maastricht Study; 842 with HbA1c data and 964 with quality of life (QoL) data
- Comparison
- Machine learning algorithms for predicting inadequate glycaemic control and QoL deterioration at 1 year
- Outcome
- Prediction of inadequate glycaemic control and of deterioration in QoL
- Follow-up
- 1 year
Methodological detailsmachine learning, fivefold cross-validation, feature selection, hyperparameter tuning
- Machine learning
- Fivefold cross-validation
- Feature selection
- Hyperparameter tuning
Performance
As reported in the abstract
Prediction of inadequate glycaemic control
| Test or model | AUC-ROC | Accuracy | Precision | Recall | F1-score |
|---|---|---|---|---|---|
| Support vector machine | 0.85 (95% CI 0.80-0.89) | 0.76 (95% CI 0.71-0.79) | 0.79 (95% CI 0.71-0.83) | — not stated in the abstract | — not stated in the abstract |
| Multi-layer perceptron | — not stated in the abstract | — not stated in the abstract | — not stated in the abstract | 0.72 (95% CI 0.64-0.79) | 0.73 (95% CI 0.64-0.79) |
AUC-ROC
Support vector machine0.85 (95% CI 0.80-0.89)
Multi-layer perceptron— not stated in the abstract
Accuracy
Support vector machine0.76 (95% CI 0.71-0.79)
Multi-layer perceptron— not stated in the abstract
Precision
Support vector machine0.79 (95% CI 0.71-0.83)
Multi-layer perceptron— not stated in the abstract
Recall
Support vector machine— not stated in the abstract
Multi-layer perceptron0.72 (95% CI 0.64-0.79)
F1-score
Support vector machine— not stated in the abstract
Multi-layer perceptron0.73 (95% CI 0.64-0.79)
— not stated in the abstract
Further findings
Prediction of deterioration in QoL · 1 year
- Inadequate performance; did not seem feasible
Original publication
Prediction of glycaemic control and quality of life in people with type 2 diabetes using glucose‐lowering drugs with machine learning—The Maastricht study. Diabetes Obes Metab. Published 17 July 2025. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →