A type 1 genetic risk score can sharpen MODY testing, especially in young children

Design

Cross-sectional analysis of routine MODY panel testing

Population

1,129 insulin-treated people referred for suspected MODY

Primary outcome

Type 1 diabetes genetic risk score distribution and utility as a pre-testing triage tool

Effect

AUC 0.60 overall and 0.83 in children under 10; 16.1% of unsolved MODY identified as probable type 1 diabetes

Investigators added a type 1 diabetes genetic risk score, calculated from 10 variants, to the MODY gene panel in 1,129 insulin-treated people referred for suspected MODY from routine clinical practice.

The score in the referred group sat between reference populations with and without type 1 diabetes, consistent with about 20% (95% CI 14.9 to 25.2) having type 1 diabetes. People with genetically solved MODY showed no excess score. Those with unsolved MODY showed enrichment consistent with 26.0% type 1 diabetes, highest in children (76.2%) and lowest in adults (16.8%). Age-specific thresholds identified 16.1% of unsolved MODY as probable type 1 diabetes.

As a triage tool before testing, discrimination was limited overall (AUC 0.60) but good in children under 10 (AUC 0.83), where more than half of tests could be avoided without missing MODY.

This is an early step for laboratories and referral pathways rather than something an individual clinician can order tomorrow.

• Before referring a child under 10 for MODY testing, ask whether the clinical picture and autoantibodies already point to type 1 diabetes.
• In an unsolved MODY panel, consider that the diagnosis may be type 1 diabetes, particularly in children.
• Do not use the score as a screening test in adults; discrimination was weak (AUC 0.60 overall).
• Ask your genetics laboratory whether it reports a type 1 risk score with panel results.

Why it matters

It reframes an unsolved MODY result as a possible answer, not a dead end.

Don't overread it

The score was a research-informed triage aid in a referral population; it does not replace clinical features or islet autoantibodies.

The statistics, in plain English

An AUC of 0.5 is chance and 1.0 is perfect. The overall 0.60 is barely better than a coin toss, while 0.83 in under-10s is useful. The benefit is therefore concentrated in one age group.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.