Same Science, Different Access: Why the UK and US Disagree on Genetic Testing for Autism

Two Countries, One Question, Opposite Answers

If your child is diagnosed with intellectual disability or global developmental delay in the United States in 2026, the American Academy of Pediatrics now recommends that genetic testing, including whole exome or whole genome sequencing, should be offered as a first-tier investigation, alongside chromosomal microarray (1). This is a significant expansion from previous guidelines, which positioned sequencing as a second step after microarray. Although this particular report addresses developmental delay and intellectual disability rather than autism specifically, the overlap is enormous: many autistic children meet these criteria, and the AAP’s broader guidance on autism has long recommended that families be offered genetic testing such as chromosomal microarray.

If your child is diagnosed with autism in England, the picture is different. The NHS Genomic Medicine Service has narrowed its eligibility criteria. Children with isolated autism, meaning autism without moderate to severe intellectual disability or congenital anomalies, are no longer eligible for genetic testing through the NHS (2). Whole genome sequencing, which the NHS pioneered through the 100,000 Genomes Project, is reserved for cases where developmental delay is more significant.

Same science. Same technology. Different conclusions about who should receive it.

What Genetic Testing Actually Finds

To understand why this divergence matters, it helps to know what the tests can do.

Chromosomal microarray is the most established genetic test for autism. It detects deletions and duplications of DNA, that is, missing or extra pieces of chromosomes. The widely cited figure of “up to 42%” (3) refers to the proportion of autistic children in whom microarray detects a copy number variant of some kind. This is a ceiling figure from selected populations and includes variants of varying clinical significance. In unselected populations the pathogenic yield is lower, and it is higher in children with co-occurring intellectual disability or dysmorphic features.

Whole exome sequencing reads the protein-coding portions of every gene, roughly 1 to 2% of the genome, but the portion most likely to contain disease-causing variants. In children whose microarray was uninformative, exome sequencing identifies additional findings in approximately 20% of cases (4).

Whole genome sequencing reads the entire genome, including non-coding regions that regulate when and where genes are active. It is the most comprehensive test available, but the diagnostic yield for autism specifically, particularly in children without intellectual disability, remains modest.

Long-read sequencing, a newer technology not yet in routine clinical use, was shown in a March 2026 study to detect 33% more structural variants and 38% more tandem repeats than standard short-read sequencing when applied to the same samples from autism families (5). Some of what we have been calling “missing heritability” may simply have been hiding in genetic structures that our current tools cannot read well.

The direction of travel is clear: with each technological advance, more genetic variation is found. The question is whether the diagnostic yield in isolated autism, without intellectual disability, is high enough to justify population-level testing.

Why the US Expanded Access

The AAP’s July 2025 clinical report repositioned whole exome and whole genome sequencing as co-equal with chromosomal microarray for the initial genetic evaluation of children with intellectual disability or global developmental delay (1). The reasoning was straightforward: sequencing identifies variants that microarray misses, the cost has fallen dramatically, and a genetic diagnosis, when found, provides concrete benefits to families.

Those benefits include ending the diagnostic odyssey, since many families spend years pursuing multiple specialist referrals and tests, and a genetic diagnosis can consolidate this into a single finding. A molecular diagnosis also enables targeted medical surveillance, because some genetic conditions carry specific health risks (cardiac, renal, endocrine) that require monitoring and may go unrecognised without a genetic explanation. It informs recurrence risk counselling: for families considering further children, knowing whether a variant is de novo (new in the child) or inherited fundamentally changes the recurrence risk, from around 1% to up to 50%. And it opens the door to gene-specific research and treatment, because as precision medicine advances, a growing number of genetic subtypes have active clinical trials or emerging therapies. You cannot access a gene-specific trial if you do not know the gene.

The AAP’s position, read alongside its existing autism guidance, is that these benefits justify offering testing broadly, not only to children with the most severe presentations. The yield may be lower in milder cases, but when a finding is made, the clinical impact is the same.

Why the UK Narrowed Access

The NHS Genomic Medicine Service made a different calculation. In a health system with finite resources and a mandate for cost-effectiveness, the question is not just “can this test find something?” but “does finding something change management frequently enough to justify the cost at scale?”

For children with moderate to severe intellectual disability alongside autism, the diagnostic yield of whole genome sequencing is substantially higher, and the clinical actionability of results is clearer. For children with isolated autism and no intellectual disability, the yield drops, and the proportion of results that change immediate clinical management drops further.

There are also structural factors. The NHS GMS lead provider contracts expire on 31 March 2026, and a transition to a “Single NHS GMS” model is underway (6). During a period of organisational change and financial pressure, narrowing eligibility criteria is partly a pragmatic decision about where limited sequencing capacity is best deployed.

It is worth noting that this is an economic and logistical choice, not a scientific one. The NHS is not saying that genetic testing is unhelpful for children with isolated autism. It is saying that, within current resource constraints, it prioritises cases where the yield is highest.

What Families Should Know Before Testing

Both the US and UK positions tend to emphasise the benefits or the yield statistics. Neither dwells on the complications, and these matter for families making decisions.

Variants of Uncertain Significance

When a genome is sequenced, the results are not simply “yes, we found the cause” or “no, nothing there.” A substantial proportion of results fall into a third category: variants of uncertain significance, or VUS. These are genetic changes that have been detected but whose clinical meaning is unclear. They might be pathogenic. They might be benign. There is not yet enough evidence to say.

For families, a VUS can be worse than no result at all. It creates uncertainty without resolution, a finding that cannot be acted upon but cannot be forgotten. A 2025 systematic review in the Journal of the American Academy of Child and Adolescent Psychiatry found that parents experienced stress and anxiety regardless of what type of result they received, and that residual uncertainty about the meaning and actionability of results was a recurring theme (7).

The Information Gap

The same review found a striking disconnect: 89.6% of parents rated understanding test capabilities and limitations as important, but only 40% reported receiving adequate information about what testing could and could not do (7). In other words, most families are consenting to genetic testing without fully understanding what the results might, or might not, tell them.

This matters because genetic counselling capacity is limited in both countries. In the UK, access to genetic counsellors varies significantly by region. In the US, the expansion of testing recommendations risks outstripping the workforce available to explain results. A test without adequate counselling is a partial intervention at best.

Uninformative Results and Disappointment

When testing returns no finding, which happens in the majority of cases, some families feel relief (“at least we ruled things out”), but others feel disappointment or frustration. A sense that the investigation has reached a dead end. This is compounded by the fact that “no finding” does not mean “no genetic contribution.” It means the current technology did not detect one. As the long-read sequencing study demonstrates, there are categories of genetic variation that standard tests simply cannot see yet (5).

For families who have invested emotional energy in the hope that testing would provide answers, an uninformative result can feel like a door closing, when in reality it is a limitation of the technology, not a statement about their child’s biology.

Few Genotype-Specific Treatments Exist, Yet

A genetic diagnosis can guide medical surveillance and inform reproductive counselling, but for most genetic subtypes of autism, it does not currently change treatment. There are no approved therapies for the vast majority of autism-associated genes. Gene therapy trials are underway for a small number of conditions: SHANK3 (Phelan-McDermid syndrome) is the most advanced, with Jaguar Gene Therapy having completed dosing of its first patient cohort in a paediatric clinical trial of JAG201 (8). For SCN2A, a CRISPR activation approach has shown rescue of neurological phenotypes in mouse models and human stem-cell-derived neurons, though this remains preclinical (9). Both are early-stage, and no genotype-specific therapy for an autism-associated gene has yet reached approval.

This is changing. The identification of druggable convergent pathways, such as the nitric oxide/TSC2/mTOR mechanism confirmed across multiple autism models in February 2026 (10), suggests that genotype-specific and pathway-specific treatments are moving closer. But for a family receiving a genetic result today, the honest answer to “what does this change for my child’s treatment?” is usually: not much, yet.

This does not mean the information is valueless. It means its value is prospective: it positions the family to benefit from future developments. Whether that prospective value justifies the emotional and financial cost of testing is a decision each family must make with adequate information.

What Is Coming Next

The American College of Medical Genetics and Genomics (ACMG) has identified “Next-generation sequencing for the genetic diagnosis of autism spectrum disorder” as a topic for new evidence-based clinical practice guideline development (11). When published, these guidelines will carry significant weight internationally and may influence the NHS’s position.

Meanwhile, the NHS itself is evolving. The “Single NHS GMS” model launching in 2026 may eventually standardise and potentially expand access. The UK government’s 10 Year Health Plan and Life Sciences Sector Plan signal an ambition to create a genomic population health service accessible to people in England by the end of the decade (12). Whether autism genetic testing will be part of that expansion remains to be seen.

Long-read sequencing, currently a research tool, is expected to enter clinical use within five to ten years as costs continue to fall. When it does, stored genome data from children tested today may be re-analysable with newer methods, potentially revealing variants that were invisible at the time of the original test.

What This Means for Families

If you are a parent in the UK and your child has autism without moderate to severe intellectual disability, the NHS is unlikely to offer genetic testing at present. This does not mean testing would not be informative for your child. It means the system has made a resource allocation decision.

You have options. Ask your paediatrician or clinical geneticist directly. Eligibility criteria have exceptions, and individual clinicians can make the case for testing if there are additional features, such as unusual physical findings, a family history of genetic conditions, or regression, that strengthen the clinical indication. Stay connected to your child’s genetics service, because eligibility criteria change. The ACMG guidelines, when published, may shift the conversation. Stored DNA can be re-tested as technology improves. Consider private testing if it is financially feasible and you have access to genetic counselling to help interpret the results. A test without counselling is only half an intervention. And above all, understand what the test can and cannot do before you pursue it. The most important preparation is realistic expectations. A genetic result may provide clarity, or it may introduce new uncertainty. Both outcomes are common.

If you are a parent in the US, the expanded recommendations mean testing should be offered, but “offered” is not the same as “explained.” Ensure you understand the range of possible results, including VUS and uninformative findings, before consenting. Ask about access to genetic counselling, both before and after testing.

Wherever you are, the underlying science is the same. The tools are improving. The yield is increasing. And the gap between what we can detect and what we can act on is narrowing, slowly, but measurably.

The question of who gets tested should not depend on geography. But until policy catches up with science, it does. Being informed is the best tool any parent has.

Many of the topics discussed in this post, including variants of uncertain significance, recurrence risk, whole genome sequencing, and how to prepare for a genetics appointment, are covered in greater detail in my book The Genetics of Autism: A Guide for Parents and Professionals.

---

References

1. Rodan LH, Stoler J, Chen E, Geleske T; Council on Genetics. Genetic Evaluation of the Child With Intellectual Disability or Global Developmental Delay: Clinical Report. Pediatrics. 2025;156(1):e2025072219. https://publications.aap.org/pediatrics/article/156/1/e2025072219/202230/
2. NHS Genomic Medicine Service — East Genomics Rare Disease Directory Update, September 2024. https://www.eastgenomics.nhs.uk/for-healthcare-professionals/updates/rare-disease-update-sept24/
3. Schaefer GB, Mendelsohn NJ; Professional Practice and Guidelines Committee. Clinical genetics evaluation in identifying the etiology of autism spectrum disorders: 2013 guideline revisions. Genetics in Medicine. 2013;15(5):399-407.
4. Srivastava S, Love-Nichols JA, Dies KA, et al; NDD Exome Scoping Review Work Group. Meta-analysis and multidisciplinary consensus statement: exome sequencing is a first-tier clinical diagnostic test for individuals with neurodevelopmental disorders. Genetics in Medicine. 2019;21(11):2413-2421.
5. Mortazavi M, Guevara J, Diaz J, et al. Long-read genome sequencing improves detection and functional interpretation of structural and repeat variants in autism. Cell Genomics. 2026;6:101186. https://pubmed.ncbi.nlm.nih.gov/40778130/
6. NHS Genomic Medicine Service Procurement Update. East Genomics. https://www.eastgenomics.nhs.uk/about-us/news-and-events/nhs-genomic-medicine-service-procurement-update/
7. Wills BC, Matthews MM, Johnston J, Bolo I, Ottman R, Appelbaum PS. Systematic Review: The Psychosocial Impacts of Autism-Related Genetic Testing. Journal of the American Academy of Child & Adolescent Psychiatry. 2025. https://pmc.ncbi.nlm.nih.gov/articles/PMC12303367/
8. Jaguar Gene Therapy — JAG201 for SHANK3 haploinsufficiency clinical trial. https://jaguargenetherapy.com/press-release/jaguar-gene-therapy-to-initiate-inaugural-pediatric-clinical-trial-targeting-a-genetic-form-of-autism-spectrum-disorder-and-phelan-mcdermid-syndrome/
9. Tamura S, Nelson AD, Spratt PWE, et al. CRISPR activation for SCN2A-related neurodevelopmental disorders. Nature. 2025;646(8086):983-991. https://pubmed.ncbi.nlm.nih.gov/40963013/
10. Ojha SK, Kartawy M, Hamoudi W, Tripathi MK, Aran A, Amal H. Nitric Oxide-Mediated S-Nitrosylation of TSC2 Drives mTOR dysregulation across Shank3 and Cntnap2 Models of Autism Spectrum Disorder. Molecular Psychiatry. 2026. https://www.nature.com/articles/s41380-026-03514-6
11. ACMG Evidence-Based Clinical Practice Guidelines — topics in development. https://www.acmg.net/ACMG/Medical-Genetics-Practice-Resources/Practice-Guidelines.aspx
12. NHS England — Genomics. https://www.england.nhs.uk/long-read/genomics/