Type 1 diabetes prediction in autoantibody-positive individuals: performance, time and money matter
Authors' conclusion
Cost-performance model analyses identified trade-offs between cost and performance models, and identified cost-minimising options to tailor risk-screening strategies.
Study context
- Population
- Autoantibody-positive first-degree relatives of individuals with T1D, TrialNet Pathway to Prevention study; 49% female, 14.9 ± 12.1 years of age1,311 developed T1D
- Participants
- 3,967 relatives
- Comparison
- T1D predictive models compared for performance, cost and participant time
- Outcome
- Predictive performance, estimated cost and participant time of models predicting clinical T1D
Methodological detailsCox model, 1943 predictive models, split-sample training and validation
- Cox model
- 1943 predictive models
- Split-sample training and validation
Performance
As reported in the abstract
Model performance, cost and participant time · median follow-up 4.7 years (IQR 2.0-8.1)
Models with metabolic measures had best performance; most showed small performance differences
less than 3% and p>0.05
Cost of measuring metabolic variables
US$56 to US$293
Participant time for measuring metabolic variables
10-165 min
Further findings
Temporal variability of model performance
- Highest genetic risk score (GRS2) influence and discriminative power in the earliest preclinical stages
Original publication
Type 1 diabetes prediction in autoantibody-positive individuals: performance, time and money matter. Diabetologia. Published 10 May 2025. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →