Abstract:Background: There is limited information on how polygenic scores (PSs) — based on variants from genome-wide association studies (GWAS) of T2D — add to clinical variables in predicting T2D incidence. Methods: For participants in a longitudinal study in an Indigenous population from the Southwest United States (US) with high T2D prevalence, we analyzed 10 constructions of PS using publicly available GWAS summary statistics. T2D incidence was examined in three cohorts of individuals without T2D at baseline. The adult cohort, comprised of 2333 participants followed from age≥20y, had 640 T2D cases. The youth cohort was comprised of 2229 participants followed from age 5-19y (228 cases). The birth cohort was comprised of 2894 participants followed from birth (438 cases). We assessed contributions of the PS and clinical variables in predicting T2D incidence. Results: Of the 10 PS constructions, a PS using 293 genome-wide significant variants from a large T2D GWAS meta-analysis in European-ancestry populations performed best. In the adult cohort, area under the receiver operating characteristic curve (AUC) for clinical variables for prediction of T2D incidence was 0·728; with the PS, 0·735. The PS’s hazard ratio (HR) was 1.27 per SD (p=1·6×10-8; 95% CI 1·17-1·38). In the youth cohort, the analogous model’s AUC was 0·805; with the PS, 0·812. The PS’s HR was 1·49 (p=4·3×10-8; 95% CI 1·29-1·72). In the birth cohort, AUC for the clinical variables was 0·614; with the PS, 0·685. The PS’s HR was 1·48 (p=2·8×10-16; 95% CI 1·35-1·63). To further assess the potential impact of including PS for assessing individual T2D risk, the net reclassification improvement (NRI) was calculated: NRI for the PS was 0·264, 0·249, and 0·309 for adult, youth, and birth cohorts, respectively. For comparison, NRI for HbA1c was 0·292 and 0·150 for adult and youth cohorts, respectively. Across all cohorts, as estimated using decision curve analyses, the net benefit of including the PS in addition to clinical variables was most pronounced at moderately stringent threshold probability (pt) values for instituting a preventive intervention. Discussion: This study demonstrates that the PS contributes significantly to prediction of T2D incidence in addition to the information provided by clinical factors; however, overall improvement of the prediction model was modest. Discriminatory power of the PS for predicting incident T2D was similar to that of other commonly measured clinical factors (e.g. HbA1c). Including T2D PS in addition to clinical factors may be clinically beneficial for identifying individuals at higher risk for T2D, especially at younger ages. Funding Statement: This study used computational resources of the Biowulf system at the National Institutes of Health (NIH), Bethesda, MD. The National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) Intramural Research Program provided support for NIDDK-based coauthors. The NIH Oxford-Cambridge Scholars Program provided funding support for L.E.W.’s doctoral programme. M.I.M. was a Wellcome Senior Investigator and an NIHR Senior Investigator. This work was funded in Oxford by the Wellcome Trust (098381, 106130, 203141), NIH (U01- DK105535; U01-DK085545) and National Institute for Health Research (NIHR) (NF-SI-0617- 10090) and Oxford Biomedical Research Centre (BRC). Declaration of Interests: The views expressed in this article are those of the author(s) and not necessarily those of NIH, NHS, NIHR, or UK Department of Health. M.I.M. has served on advisory panels for Pfizer, Novo Nordisk, and Zoe Global; has received honoraria from Merck, Pfizer, Novo Nordisk, and Eli Lilly; and research funding from Abbvie, Astra Zeneca, Boehringer Ingelheim, Eli Lilly, Janssen, Merck, NovoNordisk, Pfizer, Roche, Sanofi Aventis, Servier, and Takeda. As of June 2019, M.I.M. and A.M. are employees of Genentech and holders of Roche stock. Ethics Approval Statement: Protocols were approved by the institutional review board of the National Institute of Diabetes and Digestive and Kidney Diseases.

An Improved Genome-Wide Polygenic Score Model for Predicting the Risk of Type 2 Diabetes

Evaluation of polygenic risk scores to differentiate between type 1 and type 2 diabetes

Integrated clinical risk prediction of type 2 diabetes with a multifactorial polygenic risk score

Genome-Wide Polygenic Risk Score Predicts Incident Type 2 Diabetes in Women With History of Gestational Diabetes

Development and validation of a trans-ancestry polygenic risk score for type 2 diabetes in diverse populations

Polygenic Prediction of Type 2 Diabetes in Africa

Risk prediction and interaction analysis using polygenic risk score of type 2 diabetes in a Korean population

Genetic Risk Score Increased Discriminant Efficiency of Predictive Models for Type 2 Diabetes Mellitus Using Machine Learning: Cohort Study.

Associations of Polysocial Risk Score, Lifestyle and Genetic Factors with Incident Type 2 Diabetes: a Prospective Cohort Study

Prediction of type 1 diabetes using a genetic risk model in the Diabetes Autoimmunity Study in the Young.

Incremental Prognostic Value of a Coronary Heart Disease Polygenic Risk Score in Type 2 Diabetes

A Deep Learning-based Genome-wide Polygenic Risk Score for Common Diseases Identifies Individuals with Risk

Metabolic Polygenic Risk Scores for Prediction of Obesity, Type 2 Diabetes, and Related Morbidities

Optimization of Multi-Ancestry Polygenic Risk Score Disease Prediction Models

An early prediction model for type 2 diabetes mellitus based on genetic variants and nongenetic risk factors in a Han Chinese cohort

A Machine Learning Model for Disease Risk Prediction by Integrating Genetic and Non-Genetic Factors

The Utility of a Type 2 Diabetes (T2D) Polygenic Score in Addition to Clinical Variables for Prediction of T2D Incidence in Birth, Youth, and Adult Cohorts

Risk assessment for colorectal cancer via polygenic risk score and lifestyle exposure: a large-scale association study of East Asian and European populations

Development of a Polygenic Risk Score to Improve Detection of Peripheral Artery Disease

Genome-wide polygenic risk score, cardiometabolic risk factors, and type 2 diabetes mellitus in the Chinese population