A landmark genetic study reveals distinct developmental trajectories in early- and late-diagnosed autism — and a surprising link to ADHD.
[This is not the first time I step a little outside of ADHD proper, but this study caught my eye because (1) it confirms some popular rumblings that Autism is not a unified diagnosis, (2) late diagnosis is not necessarily scope creep as is often suggested, and (3) there is a genetic link to ADHD that does not presuppose the now popular AuDHD.]
A new study suggests that early- and late-diagnosed autism follow different genetic and developmental paths — and that might change how we understand the spectrum itself.
The Short Version
Autism isn’t just something spotted early or missed until later. According to a massive new Nature study, it may actually develop differently depending on when it first shows up.
Using data from nearly 50,000 autistic people, researchers found two broad patterns:
Early-emerging autism – differences in social and communication skills appear young and stay fairly consistent.
Late-emerging autism – childhood looks typical, but emotional and behavioural struggles increase during adolescence before diagnosis.
These two paths had distinct genetic fingerprints, suggesting that timing matters — biologically as well as developmentally.
The Research in Plain English
The study combined several long-term birth cohorts with two of the world’s largest autism genetic datasets — SPARK (US) and iPSYCH (Denmark). Together they reveal that:
Genes explain about 11% of when a person is diagnosed — roughly equal to social or clinical factors like family income or access to care.
Two partly overlapping genetic profiles emerged:
The early-diagnosed profile correlated with early social-communication differences but not strongly with other mental-health traits.
The late-diagnosed profile correlated with adolescent emotional difficulties and showed strong genetic overlap with ADHD, depression, PTSD, and self-harm.
The two profiles were only moderately related, suggesting that early- and late-diagnosed autism may have partly different genetic roots.
Why This Shifts the Lens
For years, late diagnosis was blamed on poor screening or subtle early traits. This study challenges that view: late-diagnosed autism might not simply be “missed” — it might emerge differently.
It also reframes common assumptions:
Co-occurring mental-health issues in late-diagnosed people may reflect shared biology, not just diagnostic delay.
Research that treats “autism” as one unified condition risks flattening important developmental and genetic differences.
Early screening should look beyond early childhood, since some pathways only become visible in adolescence.
What This Doesn’t Mean (Yet)
This is the part most headlines will skip — but it matters.
The authors are clear: these findings describe tendencies, not hard categories. The differences between early and late diagnosis are gradients, not “two kinds of autism.”
Here’s what to keep in mind:
Genetic modesty:
Common genetic variants explain only around 11% of diagnostic timing. That means almost 90% still depends on environment, culture, gender bias, and access to care — all notoriously uneven across countries and communities.Measurement bias:
The behavioural data came from a general tool (the Strengths and Difficulties Questionnaire), not an autism-specific measure. Part of what looks like “late-emerging autism” could actually reflect increasing visibility of emotional distress, not delayed autism itself.Demographic narrowness:
All genetic data came from people of European ancestry. That’s a problem, given how cultural context and diagnostic bias shape who even gets recognised as autistic in the first place.About that ADHD connection:
The link between late-diagnosed autism and ADHD is strong and replicated — genetic correlation around 0.5–0.7, which is substantial. Even after controlling for ADHD genetics, the relationship held.
But the researchers don’t claim this as a new “AuDHD” subtype. What they’re really showing is that some of the same genes contribute to both conditions — especially in people diagnosed with autism later.
In short: the AuDHD overlap many people describe may not just be clinical coincidence; it could be the genetic echo of one developmental pathway within autism itself.
The Bigger Message
Autism may be better understood as a family of related pathways — each shaped by a mix of genetics, development, and life context. Recognizing those differences could transform not just how we diagnose, but how we support people across the lifespan.
It also reminds us that diversity within autism isn’t noise — it’s the signal we should be listening to.
Key Takeaways
Autism’s timing and traits are influenced by partly distinct genetic profiles.
Early- and late-diagnosed groups show different developmental patterns.
Mental-health overlap in late-diagnosed autism is likely biological, not incidental.
Diagnosis timing reflects developmental wiring, not just detection gaps.
The “autism spectrum” may be more like a map of parallel routes than a single line.
Something to ponder
If late-diagnosed autism often travels hand in hand with ADHD, does that make “AuDHD” a distinct reality — or just a name for how these paths intertwine?
