Autism with ADHD, anxiety or depression: what the genes show

Many autistic children also have other conditions, such as attention deficit hyperactivity disorder (ADHD), anxiety or depression. Parents often ask whether a genetic test could show which of these conditions their child will have. A study led from the Autism Research Centre in Cambridge, published in Molecular Psychiatry in September 2026, looked at the genetic side of this question using polygenic scores (1). Polygenic scores are research tools. They are not a clinical test for an individual child.

What a polygenic score is

Much of the genetic contribution to autism, ADHD, anxiety and depression comes from many common genetic differences. Each is carried by many people and has a tiny effect on its own.

A polygenic score (PGS) is a number that adds up these small differences for one person. Each difference is weighted by how strongly it was linked to a condition in a very large study called a genome-wide association study (GWAS). A GWAS compares the DNA of many thousands of people who have a condition with many thousands who do not.

Polygenic scores describe average differences between large groups of people. They cannot diagnose autism or any other condition. They cannot predict whether one child will develop ADHD, anxiety or depression, and they cannot rule these conditions out. They are not offered in NHS clinical practice for autism.

What the study did

Adeniran Okewole and colleagues used data from SPARK (Simons Foundation Powering Autism Research), a large research cohort of autistic people and their families in the United States (1).

They studied 74,204 autistic people of European ancestry. The median age when they joined SPARK was 9 years, and 26% were female. Information about other conditions came from parents, or from the autistic person, reporting diagnoses made by a healthcare professional. Genetic data were available for 17,582 of them.

The researchers first looked at which conditions tend to occur together. Six conditions fell into three groups:

Behavioural: ADHD and disruptive behaviour disorders (oppositional defiant disorder and conduct disorder).
Mood and anxiety (the authors call this “cothymic”): anxiety and depression.
Thought disorder: schizophrenia and bipolar disorder.

Each person was given a score for each group. The researchers tested whether these scores were linked with polygenic scores for autism, ADHD, depression, anxiety, schizophrenia, bipolar disorder and some other traits. This part used 13,290 unrelated autistic people. A polygenic score for hair colour was included as a check, because it should not be linked with any of these conditions.

They also looked at rare genetic changes, such as missing or extra pieces of chromosomes (copy number variants). In 5,236 families where the child and both parents had genetic data, they compared the child’s polygenic scores with the parents’.

The main question was whether these conditions are linked to the same common genetic differences as autism, or to their own genetic differences, the same ones linked to them in the general population.

What they found

Other conditions were common. In this group, 56% had at least one co-occurring psychiatric condition, and 16.3% had intellectual disability or cognitive impairment. Reported rates were: ADHD 37%, anxiety 28%, depression 12.4%, disruptive behaviour disorders 8.6%, bipolar disorder 3.5% and schizophrenia 1.3%.

All three groups were partly heritable. With one method, common genetic differences accounted for about 21% of the variation in the behavioural score, 10% in the thought disorder score and 9% in the mood and anxiety score. A second method gave lower figures.
Higher polygenic scores for ADHD and for depression were linked with higher scores in all three groups. A higher polygenic score for educational attainment (built from studies of how long people stay in education) was linked with lower scores in all three groups. These links held when all the polygenic scores were tested together.
The polygenic score for autism was not linked with any of the three groups. Nor was the hair colour score. A second method, called genetic correlation, also found no shared genetic signal between autism and any of the groups.
Genetic differences linked to ADHD, disruptive behaviour, depression and post-traumatic stress disorder (PTSD) in the general population were also linked with the behavioural and the mood and anxiety groups in autistic people.
Rare genetic changes were linked with cognitive impairment, as expected. The authors found no clear link between rare changes and any of the three groups.
In the family analysis, the parents’ polygenic scores were not linked with the child’s group scores. The child’s own scores for depression and educational attainment were linked with the behavioural and the mood and anxiety groups, and the child’s ADHD score with the behavioural group.

Put simply, these co-occurring conditions seem to come with their own genetic differences, of the same kind seen in people who are not autistic. They do not seem to be explained by the common genetic differences linked to autism. The authors call this the additivity model.

The family results suggest the link runs through the genes the child inherited, rather than through the parents’ genes shaping the home environment. These are associations across thousands of people. They do not show that any one genetic difference causes a condition.

What this does and does not mean for one child

This study is about groups, not a test. It does not report how well polygenic scores could sort individual children, and it was not designed to.

The authors list several limits:

The genetic analyses included only people of European ancestry. The authors say it is not known whether the findings apply to other groups. They also note that polygenic scores built from European-ancestry studies work particularly poorly in people of African ancestry. Many UK families have South Asian, African, Caribbean, mixed or other ancestry, so the results may not apply to them.
In a check in 2,476 autistic people of African ancestry, the mood and anxiety group held up, but the other two were less stable because numbers were small.
All the data came from one US research cohort.
Diagnoses were reported by parents or by the autistic person. They were not checked against medical records in this study.
The cohort was young. Few people had schizophrenia or bipolar disorder, so these estimates are less precise.
Social and economic factors were not included, although the authors recognise that environment matters a great deal.
The study did not look at epilepsy. Intellectual disability was used as a background factor, not studied as one of the conditions.

What this means in clinic

Assess co-occurring conditions on clinical grounds: history, observation, and information from home and school. NICE (the National Institute for Health and Care Excellence) guidance on autism in children and young people already asks clinicians to look for coexisting conditions.
Do not assume that attention difficulties, anxiety or low mood are simply part of the autism.
Do not use polygenic scores to diagnose, predict or rule out autism or co-occurring conditions. This includes scores in commercial genetic reports that families may bring to clinic.
NHS genetic testing in autism is different. Where a child meets NHS criteria, for example because of learning disability, unusual physical features or a relevant family history, tests may include chromosomal microarray, fragile X testing, and for some children exome or genome sequencing. These look for rare genetic changes. Polygenic scores are not part of this testing.

What this means for families

ADHD, anxiety and depression are common in autistic children. This study suggests they are not simply part of autism. They seem to have their own genetic background, as in people who are not autistic, alongside social and environmental factors.

Each condition deserves its own assessment and support. If you are worried about your child’s attention, mood, anxiety or behaviour, tell your GP, paediatrician or your child’s school.

You do not need a polygenic score for your child. No genetic test can tell you for certain which of these conditions your child will or will not have. The study found no sign that the links ran through parents’ genes shaping the home, but it did not measure parenting or home life directly.

References

Okewole A, Bourque VR, Koko M, et al. Common genetic variants are associated with increased likelihood of latent co-occurring neurodevelopmental and mental health factors among autistic individuals. Molecular Psychiatry. Published online 12 September 2026. doi:10.1038/s41380-026-03880-1

Dr Odet Aszkenasy is a Consultant Community Paediatrician and the author of The Genetics of Autism: A Guide for Parents and Professionals.