ADHD and autistic traits run in the same families

This is one of the studies our AuDHD pages cite most often. It helped move the field from “these cannot occur together” toward a question that now feels obvious: if ADHD and autism often appear in the same person and the same family, what might they share?

The paper is also a useful lesson in layers of evidence. Family and twin studies supported overlap. The molecular studies available in 2010 mostly could not explain it. Those are different findings, and collapsing them into “scientists found the AuDHD genes” would badly overstate the paper.

The authors asked whether the co-occurrence of ADHD and autistic traits could arise partly from shared familial and genetic influences. They reviewed family studies, twin studies, candidate-gene work, linkage studies, and the first genome-wide association studies.

At the time, DSM-IV rules prevented an ADHD diagnosis alongside autism. Much of the literature therefore measured traits on rating scales rather than recruiting people with both diagnoses. The review’s question was ahead of the diagnostic system that supplied its data.

The paper is a narrative review, not a meta-analysis with one pooled estimate. It moves through several evidence types and asks what each can contribute.

Family studies looked for autistic traits among children with ADHD and their siblings. Twin studies estimated how much of the relationship between ADHD and autistic traits could be attributed to shared inherited influences. Candidate-gene and linkage studies looked for particular variants or chromosomal regions. Early genome-wide studies scanned more broadly but were still too small for firm cross-condition conclusions.

The family signal was present but uneven. In one sample of 821 children with ADHD, 1,050 siblings, and 149 controls, ADHD and autistic-trait scores correlated and a small sibling cross-correlation suggested shared family influences, especially in boys. A second, smaller family analysis also found more autistic traits in ADHD families but did not reproduce that cross-trait sibling correlation. The review says the disagreement could reflect sample size, different measures, or a genuinely more complicated pattern.

The twin evidence was clearer. A study of 6,771 eight-year-old twins found genetic correlations above .50 between ADHD and autistic-trait measures. A study of 674 young-adult twins estimated the genetic correlation at .72. These numbers describe overlap between sources of variation in a population. They do not say that 72% of one person’s ADHD is “caused by autism”.

The molecular evidence lagged behind. Individual candidate genes produced negative, mixed, or at most cautious signals. Linkage peaks for the two conditions sometimes sat near one another, but most came from separate samples. Genome-wide work had generated possible leads, not replicated answers. The authors repeatedly call these studies underpowered.

The review’s own summary was appropriately split: family and twin findings supported partly shared familial or genetic influences, while the evidence for particular genes was too early for firm conclusions.

Not the cause of AuDHD in one person. Heritability and genetic correlation are population statistics. They do not identify an individual’s route to a trait or diagnosis.

Not a clean estimate of co-diagnosis. DSM-IV discouraged or prohibited diagnosing both, and some family studies excluded children with a full autism diagnosis. The review often tracks questionnaire scores, not present-day AuDHD identification.

Not much about adults. Most of the evidence concerns children. One twin study included young adults, but the paper cannot tell us how the overlap appears across adult life or after late diagnosis.

Not a current gene map. The candidate-gene and early genome-wide sections are a snapshot of 2010. The authors themselves say the studies were too small and that direct ADHD-plus-autism designs were largely missing.

Not genes without context. The paper explicitly warns that environmental influences and gene–environment interactions should not be overlooked. Shared heritability is not genetic destiny.

The strongest part of this review is conceptual. It took a diagnostic boundary that said ADHD and autism were separate and tested whether family data respected that boundary. They did not. Traits crossed it, clustered in the same families, and shared some inherited influence.

That does not make ADHD and autism one condition. It means the old either-or rule was a poor map of how people and families vary. The most useful phrase in the paper is “partly similar”. It leaves room for overlap and difference at the same time.

The weakest part, viewed now, is the long search through individual candidate genes. The paper is honest about how little those findings could establish. Readers should keep the family and twin signal and treat the molecular shortlist as a record of where the field was, not where it ended.

The central evidence comes from children and families recruited under older diagnostic rules. The largest twin study involved eight-year-olds; family samples were heavily male, and one analysis was 88% boys. Measures and results for girls were less stable because there were fewer of them.

The paper says little about adults, people identified late, racial and cultural variation, or how support and environment shape the expression of traits. It also predates the contemporary language of AuDHD and community-led research on living with both.

The paper’s own best recommendation still makes sense: study whole families with the same ADHD, autism, cognitive, and environmental measures for everyone. Modern work can add larger and more diverse samples, direct recruitment of people with both diagnoses, and methods that do not force traits into an either-or box.

Genetic findings should also reconnect to lived experience. Shared statistical influence matters most when it helps explain why support designed for one label can miss the push and pull of both.