Genetic Risk Scores Predictive of Diabetes Among Youth and Young Adults

Genetic risk scores were found to be consistent predictors for diabetes among an ethnically diverse population of youth in the United States, according to the results of a study published in Diabetes Care.

Data for this analysis were sourced from SEARCH, a multicenter, prospective cohort study conducted in the United States. Between 2002 and 2020, individuals diagnosed with incident diabetes at an age younger than 20 years were invited to participate in research visits at baseline and at 5 years. At each visit, the participants were evaluated for physical features and glycemic control, and they provided a blood sample. Genetic risk scores were generated for type 1 diabetes (T1D) and type 2 diabetes (T2D), and the distribution of genetic risk was compared with insulin sensitivity, insulin resistance, diabetes autoantibody status, and self-reported race and ethnicity among 2045 participants.

There have been 2 genetic risk scores published for T1D: 1 was based on 30 risk loci and the other on 67 risk loci. Overall, the 67-single nucleotide polymorphism (SNP) score was more discriminative (receiver operating characteristic [ROC] area under the curve [AUC], 0.903) than the 30-SNP score (ROC AUC, 0.852; P <.0001).

Stratified by ethnicity, the difference in scores was more evident among participants with Hispanic ancestry (ROC AUC, 0.935 vs 0.825; P <.0001) than those with Black ancestry (ROC AUC, 0.851 vs 0.807; P =.11).

Using the 67-SNP score and the T2D score, genetic risk scores were highest for T1D among study participants who were positive for diabetes autoantibodies and had insulin sensitivity. Genetic risk scores were lowest among study participants who were negative for diabetes autoantibodies and had insulin resistance. For T2D, the opposite pattern was observed.

In general, there was close clustering of T1D and T2D genetic risk scores among White participants, but Black and Hispanic participants had higher T2D genetic risk scores. The highest T1D genetic risk scores were observed among study participants with severe insulin deficiency, and the lowest T2D genetic risk scores were observed among those with persistent endogenous insulin.

Using a machine learning approach, an algorithm was trained using high T1D risk compared with high T2D risk. The group that included participants who were negative for diabetes autoantibodies and who had insulin sensitivity clustered with T1D risk (mean probability, 0.626). The group that included participants who were positive for diabetes autoantibodies and who had insulin resistance clustered with the group that included participants who were positive for diabetes autoantibodies and who had insulin sensitivity (median probability, 0.959).

In an analysis of etiologic subtypes, the investigators observed that children who were positive for diabetes autoantibodies and had persistent C-peptide were more likely to have T1D with less β-cell destruction. Children with persistent endogenous insulin had a low probability of developing T1D. Together, these findings indicate that diabetes autoantibody positivity may have the most clinical utility for evaluating T1D vs T2D risk.

This study may have been limited by relying on self-reported and not genetically evaluated race and ethnicity.

This study found that the 67-SNP T1D genetic risk score was discriminative across self-reported race and ethnicity among youth and young adults in the United States. In addition, these data suggested that diabetes autoantibody status was strongly correlated with disease progression. Combining genetic risk score with T1D and T2D information may improve prediction of progression to insulin deficiency.

Disclosure: An author declared affiliations with biotech, pharmaceutical, and/or device companies. Please refer to the original article for a full list of disclosures.

Reference

Oran RA, Sharp SA, Pihoker C, et al. Utility of diabetes type-specific genetic risk scores for the classification of diabetes type among multiethnic youth. Diabetes Care. Published online March 21, 2022. doi:10.2337/dc20-2872