Early vs. Late Autism: Distinct Genetic Profiles Revealed

Summary: A large international study led by the University of Cambridge shows that autism diagnosed in early childhood differs in important genetic and developmental ways from autism diagnosed later in life. By combining behavioural and genetic data from tens of thousands of autistic people, researchers found distinct developmental trajectories and polygenic profiles linked to the age at diagnosis.

The research indicates that autism is not a single, uniform condition but a range of developmental pathways shaped by different genetic influences and life-course factors. Early-diagnosed children typically show social and communication differences from a very young age, while many who receive a diagnosis later tend to develop or manifest challenges during adolescence and have higher rates of co-occurring mental health conditions.

Key Facts

  • Distinct developmental pathways: Autism identified in early childhood and autism diagnosed later follow different socioemotional and behavioural trajectories.
  • Genetic distinctions: Later-diagnosed autism shows a greater genetic overlap with ADHD, depression and other psychiatric conditions than the genetic patterns observed in early-diagnosed autism.
  • Role of early support: Delays in recognition and lack of early support may contribute to poorer mental health outcomes for people diagnosed later, for example through increased vulnerability to bullying before diagnosis.

Source: University of Cambridge

An international team led by researchers at the University of Cambridge has demonstrated that autism diagnosed in early childhood differs in both developmental course and genetic architecture from autism diagnosed later in childhood or adolescence.

Challenging the assumption that autism represents a single unified biological condition, the study—published in Nature—combined behavioural data from longitudinal cohorts in the UK and Australia with genetic information from more than 45,000 autistic people across several large European and US cohorts.

The researchers observed that children diagnosed before roughly six years of age were more likely to display social and communication difficulties early in life, whereas people who received diagnoses later often showed increasing socioemotional and behavioural problems during adolescence. The later-diagnosed group also had an elevated incidence of mental health conditions such as depression.

Linking these patterns with genetics, the team found that the polygenic profiles associated with age at diagnosis differed substantially. On average, the genetic profile linked to later-diagnosed autism has a closer resemblance to the genetic risk factors for ADHD and certain psychiatric conditions than to the genetic profile commonly seen in early-diagnosed autism.

The study examined polygenic influences—composite effects of thousands of common genetic variants—and estimated that commonly heritable polygenic factors account for about 11% of the variation in age at autism diagnosis. This contribution is comparable to the influence of individual sociodemographic and clinical variables, which normally account for less than 15% of that variance.

The authors identify two modestly correlated polygenic autism factors (genetic correlation rg ≈ 0.38, s.e. ≈ 0.07). One factor is associated with earlier diagnosis and with reduced social and communication skills in early childhood, showing only modest genetic overlap with ADHD and mental-health conditions. The other factor associates with later diagnosis, increased socioemotional and behavioural difficulties in adolescence, and stronger genetic correlations with ADHD and psychiatric conditions.

Lead author Xinhe Zhang from the University of Cambridge said: “We found that, on average, individuals diagnosed with autism earlier and later in life follow different developmental pathways and, unexpectedly, have distinct underlying genetic profiles.” He and colleagues caution that these are average differences across a spectrum and that “earlier” and “later” diagnosed autism are not separate diagnostic categories but reflect variation across individuals.

Senior author Dr Varun Warrier added: “The term ‘autism’ likely encompasses multiple biological and developmental pathways. For the first time, we show that earlier- and later-diagnosed autism have different underlying profiles. The next important step is to understand how genetics interacts with social and environmental factors to influence mental health outcomes among those diagnosed later in life.”

These findings have practical implications for research, diagnosis and support. They suggest that earlier genetic influences can predispose some individuals to show autism-related traits very early, making diagnosis more likely in childhood, while other genetic and environmental combinations may lead to features that become apparent only later. Recognising these varied pathways could improve detection, timing of interventions, and tailored support across the lifespan.

Funding:

The study received support from the Wellcome Trust, UK Research and Innovation (UKRI), Horizon Europe (R2D2-MH), and the Simons Foundation for Autism Research Initiative. It was conducted in association with the NIHR CLAHRC for Cambridgeshire and Peterborough NHS Foundation Trust and the NIHR Cambridge Biomedical Research Centre.

Key Questions Answered:

Q: What did the study discover about the timing of autism diagnosis?

A: The study found that autism diagnosed early in childhood and autism diagnosed later follow different developmental trajectories and have different genetic profiles, indicating multiple pathways to an autism diagnosis.

Q: How do early and late autism diagnoses differ behaviourally?

A: Children diagnosed early tend to show social and communication difficulties from a very young age. Individuals diagnosed later more often develop socioemotional and behavioural challenges during adolescence and carry a higher risk of depression and related conditions.

Q: What did the genetic analyses reveal about early versus late diagnosis?

A: Genetic analysis identified two modestly correlated polygenic factors: one linked to earlier diagnosis and early social-communication differences, and another linked to later diagnosis, adolescent socioemotional difficulties, and greater overlap with ADHD and psychiatric risks.

About this autism and genetics research news

Author: Communications team
Source: University of Cambridge
Contact: Communications team, University of Cambridge
Image: Image credit: Neuroscience News

Original Research: Open access. “Polygenic and developmental profiles of autism differ by age at diagnosis” by Varun Warrier et al., Nature. DOI: 10.1038/s41586-025-09542-6 (citation only).


Abstract

Polygenic and developmental profiles of autism differ by age at diagnosis

While autism has commonly been thought to emerge in early childhood, many autistic people are diagnosed later. This study investigated whether earlier- and later-diagnosed autism follow different developmental courses and have distinct genetic profiles.

Using longitudinal data from four birth cohorts, the researchers identified two socioemotional and behavioural trajectories associated with age at diagnosis. Across independent cohorts, common genetic variants explained roughly 11% of the variance in age at diagnosis—an influence similar in size to individual sociodemographic or clinical factors.

The polygenic architecture of autism was decomposed into two modestly genetically correlated factors (rg ≈ 0.38). One factor associated with earlier diagnosis and lower social and communication abilities in early childhood, with only moderate genetic overlap with ADHD and mental-health conditions. The second factor associated with later diagnosis, increased socioemotional and behavioural difficulties in adolescence, and moderate to high genetic correlations with ADHD and psychiatric conditions.

These results indicate that earlier- and later-diagnosed autism represent different developmental trajectories and genetic backgrounds, offering a model to explain some of the heterogeneity observed across the autism spectrum.