In June 2026, researchers at the Seaver Autism Center for Research and Treatment at Mount Sinai published a finding in Autism Research that says as much about testing as it does about genetics.
They examined data from nearly 180,000 autistic people who had undergone genetic testing, looking for Phelan-McDermid syndrome — a condition caused by deletion or mutation of the SHANK3 gene on chromosome 22. After adjusting for undiagnosed cases, the limits of the tests themselves, and people with the syndrome who do not meet autism criteria, they arrived at an estimate of 13.7 cases per 100,000 people — roughly 1 in 7,300.
That would mean more than 45,000 people in the United States are living with Phelan-McDermid syndrome. It had been classed as rare. The researchers named the main reason the old numbers were so low: genetic testing is badly underused.
The real headline is about testing, not one syndrome
Phelan-McDermid syndrome is specific and most families reading this will not have it. The transferable point is that a condition can look twenty times rarer than it is simply because nobody looked.
If you have never been offered genetic testing after your child's autism diagnosis, you are in the majority — and that is a gap in practice, not a judgement about whether your child needs it.
What the tests actually are
Chromosomal microarray
Usually the first-line test. It looks for missing or duplicated stretches of chromosome that are too small to see under a microscope. This is the test that finds Phelan-McDermid syndrome and a number of other copy-number conditions. A blood draw.
Fragile X testing
A targeted test for the most common inherited cause of intellectual disability, which frequently co-occurs with autism. Often ordered alongside the microarray because it looks for a type of change that microarray does not detect.
Exome or genome sequencing
Reads the protein-coding genes, or the whole genome, letter by letter. It finds single-letter changes that a microarray cannot see, and increasingly it is offered earlier rather than as a last resort. More expensive, slower, and much more likely to return findings of uncertain meaning.
What a result can genuinely change
Be clear-eyed about this. A genetic result does not change your child. It can change four other things.
Medical monitoring. This is the strongest argument. Many genetic syndromes carry known risks — seizures, kidney or heart differences, particular feeding or growth patterns — that are worth watching for deliberately rather than discovering late. A diagnosis converts vague vigilance into a specific checklist.
Access to research and trials. Trials increasingly recruit by genetic subtype rather than by behavioural diagnosis. Without a genetic result, your child cannot be considered.
Community. Syndrome-specific family organisations are often remarkably well-informed, and other parents a few years ahead of you are a resource nothing else replaces.
Information for the family. Some findings are inherited and some arise new in the child. That distinction matters to parents thinking about having more children, and sometimes to siblings later.
What it cannot tell you
It cannot predict how your child will turn out. Two people with the same finding can differ enormously. Nobody can read a report and tell you whether your child will talk, read, or live independently.
Most families get no answer. Depending on the test and the child, a clear cause is found in a minority of cases. Going in expecting an explanation and receiving "nothing identified" is a real and common disappointment, and worth preparing for.
"Uncertain significance" is a frequent result. A variant is found; nobody yet knows whether it means anything. These get reclassified over the years, sometimes to harmless. Families are often unprepared for how unsatisfying this is.
There is no treatment at the end of it. Not yet. Which brings us to the second piece of 2026 news.
The protein map, and why it is promising rather than useful
Also in September 2026, researchers published the largest map yet of the proteins encoded by around 100 autism risk genes and the molecular interactions between them. The interesting result was that these proteins are not scattered — they converge on shared hubs.
The hope that follows is real: if many different genetic routes to autism pass through a small number of molecular junctions, a future treatment might target a junction rather than needing to be built separately for every gene.
The distance between that sentence and a medicine is measured in years and in trials that have not started. Treat it as a reason for cautious optimism about the field, and as no reason at all to change anything you are doing this year. Be especially wary of anyone selling a treatment that cites this kind of research as justification.
How to decide, and what to ask
1. Ask whether it has already been done. Some children are tested during evaluation without it being clearly explained. Request the records before paying for anything.
2. Ask for a referral to genetics, not just a test. A genetic counsellor or clinical geneticist will help you decide which test fits, and — more importantly — interpret a result that is rarely as clean as families expect.
3. Ask what it costs and what insurance covers before the sample is taken. Coverage varies by plan and by the reason documented for ordering it. This is the stage where families get caught out.
4. Ask what happens with an uncertain result. Who reviews it, whether you will be contacted if it is reclassified, and how long results are kept.
5. Ask yourself what you would do differently with each possible answer. If the honest answer is "nothing", that is a legitimate reason to decline. Declining is a reasonable choice, not a failure of diligence.
The bottom line
Genetic testing is worth asking about — the Mount Sinai finding is a fair warning about how much goes unfound when nobody looks. It is most valuable when there are additional features alongside autism: seizures, significant intellectual disability, unusual physical characteristics, a family history, or regression.
It is least valuable when pursued in the hope of an explanation that will make the diagnosis easier to bear. That is a completely understandable reason to want it, and the one most likely to leave you disappointed.
Your child is the same person before and after the blood draw. The test is a tool for their medical care, not a verdict on their future.
A note on this article. This is general information for families, not medical advice, and it does not recommend for or against testing your child. Decisions about genetic testing should be made with a clinician who knows your child's history, ideally a genetic counsellor or clinical geneticist.