ASHG Abstract Presented by Jeffrey Barrett, PhD, Nightingale Health Biobank Collaborative Group; Nightingale Health, Helsinki, Finland.
The following is a significant abstract released on Saturday, November 4, during the Annual American Association of Human Genetics meeting (ASHG 2023), being held in Washington, DC, November 1-5. In their abstract, Dr. Barrett and colleagues say the following:
“Life expectancy is increasing faster than healthy life expectancy, leading to individuals in high-income countries living more years restricted by chronic diseases. Healthcare systems face a cost crisis as they try to provide an increasingly wide range of transformative but expensive treatments to the older, sicker populations. Better prevention, to complement new treatments, is essential to provide healthier lives and sustainable healthcare. Polygenic scores have received most attention, but there is also an opportunity to combine genetics with other easily measured biomarkers.”
“Here, we present nuclear magnetic resonance metabolomics from half a million blood samples from three national biobanks. We built metabolomic risk scores that identify a high-risk group for each of 12 diseases that cause the most morbidity in high-income countries and show consistent cross-biobank replication of the relative risk of disease for these groups. For seven diseases (heart attack, stroke, lung cancer, type 2 diabetes, COPD, alcoholic liver disease, and liver cirrhosis) this high-risk group has >3-fold increased risk of disease (ranging up to 10-fold for diabetes and liver disease), and there are existing medical and lifestyle interventions available to reduce future risk.”
“We compared these metabolomic predictions to best available polygenic scores and found that for all diseases except Alzheimer’s and colon cancer, metabolomic scores are more strongly associated with future disease risk than polygenic scores. These two scores are additive, except for a weak negative interaction in diabetes, and in the cases with good performance of both scores (e.g., diabetes, heart attack) a combined score is significantly better than either alone. Metabolomic scores also show stable association to disease incidence for 10 years of follow up, demonstrating that scores composed of multiple biomarkers capture stable long-term risk, and not just immediate fluctuations.”
“Several factors suggest that metabolomic, or metabolomic+polygenic risk prediction could be used in the real world. First, because our measurements are in absolute units, no within-cohort normalization or batch correction is required, in contrast to some other -omic technologies.”
“Second, the absolute calibration of our predictions in separate biobanks/countries from those where we train the models is similar to the portability of existing risk tools, like Framingham or QRISK.”
“Finally, 28% of individuals were at high risk for one or more of the seven diseases with good prediction and available intervention, suggesting that population screening in a health-check setting could provide value.”