Study challenges cerebral palsy genetic links, finds only 89 of 515 proposed genes have real evidence

Cerebral palsy, the most common childhood disability worldwide, has long puzzled medicine. Birth complications—premature delivery, infection, oxygen deprivation, perinatal stroke—explain only some cases. Over the past decade, as gene sequencing became faster and cheaper, researchers began hunting for genetic answers. They identified hundreds of genetic variants in children with cerebral palsy and declared them the cause. For families, this offered something precious: an end to the diagnostic odyssey, a concrete explanation, a path forward.

But a new study published in The American Journal of Human Genetics has upended that narrative. Researchers led by Peter N. Robinson, a computational biologist at the Jackson Laboratory for Genomic Medicine, combed through 21 previously published genomic studies involving 5,440 individuals with cerebral palsy. Across these cohorts, the original researchers had identified 515 different genes as direct causes of the patients' motor disabilities. When Robinson's team applied rigorous statistical analysis to this historical data, the number collapsed to 89. Only 89 of the 515 genes showed convincing evidence of a true association with cerebral palsy.

The team then sequenced genomes from 460 patients at Shriners Children's hospital network in the United States. They found disease-causing mutations in 60 genes across 15.8 percent of the children. But when they applied their stricter statistical standard, only 16 genes remained defensible. The gap between what genetic panels report and what the evidence actually supports is vast.

The problem stems from a fundamental confusion in how we interpret genetic findings. When a child with cerebral palsy is sequenced, researchers often find rare mutations. But rarity does not equal causation. Roughly 5 percent of the general population carries rare genetic disease mutations entirely unrelated to their health. A child with cerebral palsy might carry a mutation in the LIPH gene—which actually causes a rare form of hair loss with no known neurological connection—simply by chance. The mutation and the disability coexist, but one did not cause the other. This is coincidence masquerading as diagnosis.

Yet some genes do show up far more often than random chance would predict. CTNNB1 mutations appeared repeatedly across both historical literature and the newly sequenced cohort, providing genuine statistical proof of association. The challenge is distinguishing the real signals from the noise.

Kuntal Sen, a clinical geneticist and pediatric neurologist at Children's National Hospital, explained the stakes plainly: a genetic test report might show "just a slight association rather than causal." If a doctor accepts an incidental mutation as the final answer, they may stop investigating prematurely. A child with a treatable metabolic disorder or rare genetic mimic—conditions that masquerade as cerebral palsy but respond to specific interventions—might never receive the correct diagnosis. The false genetic answer becomes a barrier to the real one.

Cerebral palsy itself is not a single disease but a clinical umbrella, lumping together conditions with vastly different origins. Some stem from birth injury. Others are mimics: rare metabolic or genetic disorders that produce similar motor symptoms. This heterogeneity is why precision matters. Between 8.5 and 24 percent of individuals who receive a valid genetic diagnosis have access to specific, available treatments—dietary changes for arginase deficiency, medications for vitamin-dependent epilepsies, targeted interventions for hereditary dystonia. For these families, getting the genetics right means access to treatments that can alter a child's developmental trajectory.

For countries like India, where three in every 1,000 live births are affected by cerebral palsy and genetic testing is rapidly expanding, this study offers a practical advantage. A severe mutation that breaks a specific protein does so universally, regardless of ethnicity. By identifying only the true genetic causes, clinics can skip the trial-and-error phase of bloated sequencing panels and move directly to accurate testing and precise treatment. Robinson acknowledged that his team's strict mathematical model might actually underestimate the true number of genes involved—absence of statistical significance does not prove absence of association. The path forward requires large, globally shared datasets of detailed patient profiles, allowing researchers to calculate the exact degree of risk each genetic variant carries. Until then, physicians must integrate genetic findings cautiously with clinical judgment, treating genomic testing as one tool among many rather than a final answer.