Precision diagnosis for monogenic diabetes requires ethnicity specific criteria for genetic testing

Abstract The detection of monogenic diabetes illustrates the potential of precision medicine, with treatments tailored to specific genes and diagnosis involving targeted genetic testing. Current detection criteria are derived from White populations. We investigated detection of monogenic diabetes in an unselected multiethnic cohort comprising 1,706 participants diagnosed with diabetes before the age of 30-years. Using broad biomarker criteria (triple pancreatic antibody negative and detectable C-peptide) to select for next generation sequencing of monogenic diabetes genes, we found a non-significantly different minimum cohort prevalence of monogenic diabetes of 2.1% in White, 2.0% in South Asian, 2.5% in African-Caribbean, and 3.6% in Mixed participants. The detection rate, however, varied significantly (17.7% in White, 5.3%in South Asian, 8.0% in African-Caribbean, and 15.2% in Mixed participants, p<0.001). Those without monogenic diabetes showed significant variations in BMI. No difference in phenotype of monogenic diabetes across ancestry groups was observed. Non-white ethnicity participants were significantly more likely to have undiagnosed monogenic diabetes than White with on average a 10-year duration before receiving a correct diagnosis. By applying ancestry-specific BMI cut-offs (White <30, South Asian <27, African-Caribbean and Mixed <35 kg/m²), the overall detection rate increased from 8.8 to 16%, reducing the number needed to test to identify one case from 11 to 6 and boosting detection rates to 39, 11, 9 and 26% in White, South Asian, African-Caribbean and Mixed-ethnicity participants, respectively. These findings were validated in an external real-world dataset. Applying broad biomarker criteria for initial selection, mitigates clinical biases leading to misclassification of monogenic diabetes in non-White ethnicities. However, further tailoring criteria with ethnic-specific BMI cut-offs doubled detection rates, improving cost-effectiveness by minimising unnecessary testing. Our study highlights the need to develop precision medicine approaches accounting for phenotypic variation across diverse populations, to ensure accurate diagnoses and cost-efficient healthcare provision.