Summary: A major meta-analysis finds that for most common diseases, inherited gene variants explain only a small fraction of risk—typically under 5–10 percent—while metabolic, environmental and lifestyle factors play a much larger role.
Source: University of Alberta
New research from the University of Alberta indicates that, in most cases, your genes account for only a small portion of the risk of developing common diseases.
Researchers performed the largest meta-analysis to date on two decades of genome-wide association study (GWAS) data, focusing on single nucleotide polymorphisms (SNPs) and their reported links to disease. The analysis shows that, for the majority of conditions examined, SNP-based genetic signals offer limited predictive power for disease risk.
David Wishart, professor in the University of Alberta’s Department of Biological Sciences and Department of Computing Science and co-author of the study, summarizes the finding simply: “DNA is not your destiny, and SNPs are duds for disease prediction.” He explains that for most diseases—including many cancers, forms of diabetes, and Alzheimer’s disease—the genetic contribution identified through common SNPs is generally in the range of 5 to 10 percent at most.
There are notable exceptions. A small subset of conditions, such as Crohn’s disease, celiac disease, and age-related macular degeneration, show much stronger genetic associations, with estimated genetic contributions of roughly 40 to 50 percent. Even so, these conditions are the minority among those evaluated.
The study’s authors argue that measuring biochemical and biological markers beyond DNA—such as metabolites, proteins, and the microbiome—provides more accurate and actionable assessments of health and disease risk. These molecular measurements reflect the combined effects of genes, environment, diet, infections, and other exposures, and so can be more powerful for diagnosis and risk stratification than SNP-based profiles alone.
“If you want an accurate picture of your current health or your propensity for disease, measuring metabolites, microbes, or proteins gives you far more useful information than a static readout of your genes,” Wishart says. He adds that the findings underline the importance of investigating environmental influences and the quality of food, air, and water that shape health outcomes.
Institution:
University of Alberta
Media contact:
Katie Willis – University of Alberta
Original research (open access):
“Assessing the performance of genome-wide association studies for predicting disease risk.” Jonas Patron, Arnau Serra-Cayuela, Beomsoo Han, Carin Li, David Scott Wishart. PLOS ONE. DOI: 10.1371/journal.pone.0220215.
Study summary and methods
The researchers built and applied a software package called G-WIZ to analyze summary-level GWAS data and compute receiver operating characteristic (ROC) curves and the area under the ROC curve (AUROC) for SNP-derived risk predictors. AUROC is a standard metric used to assess how well a model discriminates between cases and controls; values closer to 1.0 indicate strong predictive power, while a value near 0.5 indicates prediction no better than chance.
The team validated G-WIZ against AUROC values calculated from patient-level SNP data and published AUROC values, finding their method predicted AUROC with less than 3 percent error. They then applied G-WIZ to summary data from 569 GWAS covering 219 different conditions. While a few studies produced SNP-based predictors with relatively high AUROC values (>0.75), the typical multi-SNP predictor yielded an AUROC of around 0.55—only slightly better than random guessing and generally inferior to clinically based risk predictors.
Based on these results, the authors conclude that, for most diseases, the SNPs identified by current GWAS do not provide robust, clinically useful risk predictions. They assembled their AUROC calculations, ROC curves, and heritability estimates into a publicly accessible resource called GWAS-ROCS to support further research and comparison.
Implications
The study has practical implications for consumers, clinicians, and businesses that offer genetic risk assessments. While genetic testing can be informative for specific high-heritability conditions and for identifying rare pathogenic variants, relying on common SNP profiles for broad disease prediction appears limited. Integrative approaches that combine genetic data with metabolic markers, proteomic profiles, microbiome composition, and environmental exposure information are likely to yield more accurate and actionable health assessments.