Summary: A new predictive model combining genetics with developmental milestones may help forecast which autistic children are at risk of developing intellectual disabilities. In a large study of over 5,600 children, researchers were able to correctly predict 10% of such cases and differentiate between low and high-risk groups twice as well as current methods.
This marks a shift from “wait and see” approaches toward proactive, tailored support for families. Although still early in development, the model shows promise as genetic data and computational power continue to improve.
Key Facts:
- Predictive Breakthrough: The model accurately predicted 10% of intellectual disability cases in autistic children.
- Improved Risk Differentiation: It doubled the ability to distinguish between low and high-risk cases compared to current methods.
- Towards Early Support: Findings support a shift to proactive care instead of waiting for disabilities to become apparent.
Source: University of Montreal
Will a child who’s evaluated for autism later develop an intellectual disability? Can this be accurately predicted? Early-childhood experts in Quebec say they’ve have come up with a way to better find out.
In a study of 5,633 children drawn from three North American cohorts, clinician-researchers affiliated with Université de Montréal developed a new predictive model that combines a wide range of genetic variants with data on each stage of a young child’s development.
Their goal? To obtain reliable information as early as possible to predict the children’s developmental trajectory and thus offer more proactive support to those who may need it – namely, parents trying to better understand and anticipate their child’s needs.
Published today, the study marks an important turning point in the use of genomics data: instead of being used to explain a developmental condition that was previously observed, with the new method genetic testing is used to predict future development.
‘Difficult to foresee the future’
“In some children, initial signs of autism can be observed as early as 18 months of age,” said the study’s first author Dr. Vincent-Raphaël Bourque, a PhD candidate at the Centre de recherche Azrieli du CHU Sainte-Justine, the mother-and-child hospital affiliated with UdeM.
But “it remains difficult to foresee the child’s developmental trajectory over the medium and long term,” said Bourque, who’s also a senior medical resident in child and adolescent psychiatry at the Montreal Children’s Hospital.
“Many parents wonder whether their child will develop an intellectual disability, which occurs in 10 to 40 per cent of autistic children, and only becomes apparent around the age of 6 to 8 years,” he said.
Currently, doctors rely primarily on monitoring a child’s early developmental milestones, such as acquiring language and walking, to predict their later development. However, these predictions are limited, especially for younger children. It’s also uncertain how accurate these predictions are.
As a result, most doctors now simply wait and see how the child develops. Unfortunately, this can result in broadening the gap between a child’s unique strengths and challenges, their increasing social and academic demands, and the support they and their family are offered, and cause high levels of stress.
A better approach, the study’s authors say, would be to offer children earlier, more personalized attention and support. But to do that, doctors would need to foresee the challenges a child will experience, not wait until they’re overwhelmed.
That’s where the new model comes in.
10 percent correctly predicted
Taking into account a large number of genetic variations—including some that are usually considered to have little predictive value—and combining them with data on each stage of language and motor development, the study was able to correctly predict 10 per cent of the occurrences of intellectual disabilities in the participating children.
Additionally, the researchers were twice as able to differentiate between low and high probabilities of intellectual disability in children with significant developmental delays.
These findings establish a benchmark and show promise for future predictions, the study’s authors say.
“Research teams are constantly discovering new genetic variations, and computational models are becoming more and more powerful,” said senior author Dr. Sébastien Jacquemont, a geneticist and researcher at CHU Sainte-Justine.
“So we can be sure that in the coming years, with cheaper and more readily available genetic testing, our ability to make accurate predictions will increase, allowing for more tailored medical care and support that will be accessible to more children.”
The study pays special attention to measuring the degree of uncertainty – that is, how often predictions were accurate or inaccurate, which is key information when communicating with the children’s families.
“As a medical practitioner, I think it is essential to explain to families the degree of certainty (and uncertainty) of our predictions,” said Bourque. “It helps them to make better decisions with and for their child.”
About this genetics and autism research news
Author: Jeff Heinrich
Source: University of Montreal
Contact: Jeff Heinrich – University of Montreal
Image: The image is credited to Neuroscience News
Original Research: Open access.
“Genomic and Developmental Models to Predict Cognitive and Adaptive Outcomes in Autistic Children” by Vincent-Raphaël Bourque et al. JAMA Pediatrics
Abstract
Genomic and Developmental Models to Predict Cognitive and Adaptive Outcomes in Autistic Children
Importance
Although early signs of autism are often observed between 18 and 36 months of age, there is considerable uncertainty regarding future development. Clinicians lack predictive tools to identify those who will later be diagnosed with co-occurring intellectual disability (ID).
Objective
To predict ID in children diagnosed with autism.
Design, Setting, and Participants
This prognostic study involved the development and validation of models integrating genetic variants and developmental milestones to predict ID. Models were trained, cross-validated, and tested for generalizability across 3 autism cohorts: Simons Foundation Powering Autism Research (SPARK), Simons Simplex Collection, and MSSNG. Autistic participants were assessed older than 6 years of age for ID. Study data were analyzed from January 2023 to July 2024.
Exposures
Ages at attaining early developmental milestones, occurrence of language regression, polygenic scores for cognitive ability and autism, rare copy number variants, de novo loss-of-function and missense variants impacting constrained genes.
Main Outcomes and Measures
The out-of-sample performance of predictive models was assessed using the area under the receiver operating characteristic curve (AUROC), positive predictive values (PPVs), and negative predictive values (NPVs).
Results
A total of 5633 autistic participants (4574 male [81.2%]) were included in this analysis. On average, participants were diagnosed with autism at 4 (IQR, 3-7) years of age and assessed for ID at 11 (8-14) years of age, with 1159 participants (20.6%) being diagnosed with ID.
The model integrating all predictors yielded an AUROC of 0.653 (95% CI, 0.625-0.681), and this predictive performance was cross-validated and generalized across cohorts. This modest performance reflected that only a subset of individuals carried large-effect variants, high polygenic scores, or presented delayed milestones.
However, combinations of genetic variants that are typically not considered clinically relevant by diagnostic laboratories achieved PPVs of 55% and correctly identified 10% of individuals developing ID.
The addition of polygenic scores to developmental milestones specifically improved NPVs rather than PPVs. Notably, the ability to stratify ID probabilities using genetic variants was up to 2-fold higher in individuals with delayed milestones compared with those with typical development.
Conclusions and Relevance
Results of this prognostic study suggest that the growing number of neurodevelopmental condition–associated variants cannot, in most cases, be used alone for predicting ID.
However, models combining different classes of variants with developmental milestones provide clinically relevant individual-level predictions that could be useful for targeting early interventions.
