Researchers Urge Caution Over Large-Scale Newborn Genome Screening, Citing Overdiagnosis Risks
Key Insights
University of Exeter researchers found that genetic variant disease risk is often lower in the general population than previously reported, raising concerns about overdiagnosis in newborn genome screening programs.
Analysis of nearly one million volunteers across UK Biobank and All of Us datasets showed weaker links between genetic variants and disease compared to studies conducted in high-risk populations.
The team focused on more than 50 genes across 15 diseases included in large-scale newborn screening trials, finding the biggest overdiagnosis risk when disease stems from loss of just one gene copy.
New research from the University of Exeter is raising significant concerns about the rapid expansion of newborn genome screening programs, warning that current genetic risk estimates may lead to substantial overdiagnosis and unnecessary harm to families. The findings, presented at the European Human Genetics Conference and published in the European Journal of Human Genetics, draw on population-scale datasets to challenge assumptions about how strongly genetic variants predict disease.
The research team, funded by the Medical Research Council and the National Institute for Health and Care Research (NIHR) Exeter Biomedical Research Centre, analyzed genetic variants previously classified as disease-causing using data from nearly a million volunteers in UK Biobank and the All of Us research program. Their focus spanned more than 50 genes across 15 diseases that are included in large-scale trials for newborn genome screening.
Weaker Links in the General Population
A central finding of the research is that the relationship between genetic variants and disease—known as penetrance—is frequently weaker in the general population than earlier studies suggested. Most genetic research to date has been conducted in individuals who already have a condition or belong to high-risk families, inflating the apparent risk.
"We found that in most cases, although there is a link between the genetic variants and the disease, it's weaker than previous research has suggested," said Dr. Leigh Jackson of the University of Exeter Medical School. "This is because most research is conducted in people who already have a disease or are at high risk. The actual risk is often lower if you're looking at the general population, as would be the case if you applied this risk to newborn screening."
The team's published research identified that the greatest risk of overdiagnosis arose when disease risk stems from a loss of just one of the two copies of a gene. However, the picture was mixed: upcoming research from the same group indicates that where both copies of a gene must be lost to cause disease, the evidence suggests a lower risk of overdiagnosis.
Implications for Families and Policy
Professor Caroline Wright, of the University of Exeter Medical School, emphasized the stakes involved as countries including the UK increasingly pilot newborn genome screening programs. "The UK is leading the way on newborn genome screening and the world is watching. We urgently need the best evidence to ensure we get this right," she said. "The benefits of a successful programme will be profound, but we need to be aware of the risk of overdiagnosis based on current evidence. Being told your baby is at high risk of developing disease has a massive impact on families—we need to ensure we use the most accurate risk estimates possible in order to provide the best care."
Newborn genome screening involves determining the entire sequence of a baby's genetic code and interrogating it for hundreds of potentially treatable conditions. Pilot studies are increasingly being rolled out internationally, yet the Exeter team cautions that the evidence base underpinning these programs requires strengthening before widespread implementation.
The three papers published so far in the European Journal of Human Genetics address penetrance in the genomic screening era, reduced penetrance of COL1A1 (search)/COL1A2 (search) pathogenic variants linked with osteogenesis imperfecta (search), and incomplete penetrance of TSC1 (search)/TSC2 (search) variants with implications for newborn screening. Two additional papers are available on the medRxiv preprint server.