Marie Pigeyre1,2,3, Jennifer Sjaarda1,2,3, Shihong Mao1,2,3, Michael Chong1,2,3, Sibylle Hess4, Salim Yusuf1,5, Hertzel Gerstein1,5, Guillaume Paré6,2,3,5. 1. Population Health Research Institute, David Braley Cardiac, Vascular and Stroke Research Institute, Hamilton, Ontario, Canada. 2. Thrombosis and Atherosclerosis Research Institute, David Braley Cardiac, Vascular and Stroke Research Institute, Hamilton, Ontario, Canada. 3. Department of Pathology and Molecular Medicine, Michael G. DeGroote School of Medicine, McMaster University, Hamilton, Ontario, Canada. 4. R&D, Translational Medicine and Early Development, Biomarkers and Clinical Bioanalyses, Sanofi-Aventis Deutschland GmbH, Frankfurt, Germany. 5. Department of Clinical Epidemiology and Biostatistics, McMaster University, Hamilton, Ontario, Canada. 6. Population Health Research Institute, David Braley Cardiac, Vascular and Stroke Research Institute, Hamilton, Ontario, Canada pareg@mcmaster.ca.
Abstract
OBJECTIVE: Observations of a metabolically unhealthy normal weight phenotype suggest that a lack of favorable adiposity contributes to an increased risk of type 2 diabetes. We aimed to identify causal blood biomarkers linking favorable adiposity with type 2 diabetes risk for use in cardiometabolic risk assessments. RESEARCH DESIGN AND METHODS: A weighted polygenic risk score (PRS) underpinning metabolically favorable adiposity was validated in the UK Biobank (n = 341,872) and the Outcome Reduction With Initial Glargine Intervention (ORIGIN Trial) (n = 8,197) and tested for association with 238 blood biomarkers. Associated biomarkers were investigated for causation with type 2 diabetes risk using Mendelian randomization and for its performance in predictive models for incident major adverse cardiovascular events (MACE). RESULTS: Of the 238 biomarkers tested, only insulin-like growth factor-binding protein (IGFBP)-3 concentration was associated with the PRS, where a 1 unit increase in PRS predicted a 0.28-SD decrease in IGFBP-3 blood levels (P < 0.05/238). Higher IGFBP-3 levels causally increased type 2 diabetes risk (odds ratio 1.26 per 1 SD genetically determined IGFBP-3 level [95% CI 1.11-1.43]) and predicted a higher incidence of MACE (hazard ratio 1.13 per 1 SD IGFBP-3 concentration [95% CI 1.07-1.20]). Adding IGFBP-3 concentrations to the standard clinical assessment of metabolic health enhanced the prediction of incident MACE, with a net reclassification improvement of 11.5% in normal weight individuals (P = 0.004). CONCLUSIONS: We identified IGFBP-3 as a novel biomarker linking a lack of favorable adiposity with type 2 diabetes risk and a predictive marker for incident cardiovascular events. Using IGFBP-3 blood concentrations may improve the risk assessment of cardiometabolic diseases.
OBJECTIVE: Observations of a metabolically unhealthy normal weight phenotype suggest that a lack of favorable adiposity contributes to an increased risk of type 2 diabetes. We aimed to identify causal blood biomarkers linking favorable adiposity with type 2 diabetes risk for use in cardiometabolic risk assessments. RESEARCH DESIGN AND METHODS: A weighted polygenic risk score (PRS) underpinning metabolically favorable adiposity was validated in the UK Biobank (n = 341,872) and the Outcome Reduction With Initial Glargine Intervention (ORIGIN Trial) (n = 8,197) and tested for association with 238 blood biomarkers. Associated biomarkers were investigated for causation with type 2 diabetes risk using Mendelian randomization and for its performance in predictive models for incident major adverse cardiovascular events (MACE). RESULTS: Of the 238 biomarkers tested, only insulin-like growth factor-binding protein (IGFBP)-3 concentration was associated with the PRS, where a 1 unit increase in PRS predicted a 0.28-SD decrease in IGFBP-3 blood levels (P < 0.05/238). Higher IGFBP-3 levels causally increased type 2 diabetes risk (odds ratio 1.26 per 1 SD genetically determined IGFBP-3 level [95% CI 1.11-1.43]) and predicted a higher incidence of MACE (hazard ratio 1.13 per 1 SD IGFBP-3 concentration [95% CI 1.07-1.20]). Adding IGFBP-3 concentrations to the standard clinical assessment of metabolic health enhanced the prediction of incident MACE, with a net reclassification improvement of 11.5% in normal weight individuals (P = 0.004). CONCLUSIONS: We identified IGFBP-3 as a novel biomarker linking a lack of favorable adiposity with type 2 diabetes risk and a predictive marker for incident cardiovascular events. Using IGFBP-3 blood concentrations may improve the risk assessment of cardiometabolic diseases.
Authors: William D Thompson; Robin N Beaumont; Alan Kuang; Nicole M Warrington; Yingjie Ji; Jessica Tyrrell; Andrew R Wood; Denise M Scholtens; Bridget A Knight; David M Evans; William L Lowe; Gillian Santorelli; Rafaq Azad; Dan Mason; Andrew T Hattersley; Timothy M Frayling; Hanieh Yaghootkar; Maria Carolina Borges; Deborah A Lawlor; Rachel M Freathy Journal: Diabetologia Date: 2021-09-20 Impact factor: 10.122