Higher severity of autistic traits, co-occurring ADHD, internalising and maladaptive behaviours were significantly associated with lower normalised hair cortisol concentration (HCC), while higher sleep anxiety and IQ were associated with higher HCC, indicating the clinical value of HCC as a viable biomarker to identify subgroups based on co-occurring medical and mental health conditions.
Key Findings
Results
Higher severity of autistic traits overall, and social affect issues in particular, were significantly associated with lower normalised hair cortisol concentration (HCC).
Analysis was conducted using multivariable linear regression after controlling for key sociodemographic covariates including child's IQ
The study included 307 autistic children, 158 non-autistic siblings, and 124 unrelated non-autistic children
Children aged 2-17 years were drawn from the Australian Autism Biobank
Social affect was specifically identified as a sub-domain driving the association with lower HCC
Results
Co-occurring ADHD was significantly associated with lower normalised HCC in autistic children.
The association was identified through multivariable linear regression controlling for sociodemographic covariates including IQ
ADHD was among the commonly reported co-occurring conditions in the sample
The finding suggests HCC may help differentiate autistic subgroups based on psychiatric co-occurrence
Results
Internalising behaviours and maladaptive behaviours were significantly associated with lower normalised HCC.
Both internalising and maladaptive behaviour measures were included as part of the primary multivariable linear regression analyses
These associations were identified after controlling for child's IQ and key sociodemographic covariates
The direction of association (lower HCC with higher behavioural severity) was consistent across these behavioural domains
Results
Higher sleep anxiety was significantly associated with higher HCC.
Sleep disturbances were examined as a commonly co-occurring condition in autistic children
The association between sleep anxiety and HCC was in the opposite direction compared to autistic trait severity and ADHD
This finding distinguishes sleep anxiety from other co-occurring conditions in its relationship to cortisol levels
Results
Higher IQ was significantly associated with higher HCC.
IQ was included both as a covariate and examined as a predictor in the analyses
The study sample included children with a range of co-occurring conditions including intellectual disability (6.1%) and global developmental delay (8.5%)
The positive association between IQ and HCC was identified within the multivariable regression framework
Results
Older age and higher family income were associated with lower HCC.
These were identified as significant sociodemographic factors in the multivariable linear regression analyses
The study sample spanned children aged 2-17 years, providing a wide age range for detecting age-related effects
Family income was among the key sociodemographic covariates controlled for and examined in the primary analyses
Results
The most commonly reported co-occurring conditions in the autistic sample were global developmental delay, intellectual disability, and otitis media.
Global developmental delay was reported in 8.5% of the sample
Intellectual disability was reported in 6.1% of the sample
Otitis media was reported in 6.1% of the sample
The study included 307 autistic children from the Australian Autism Biobank
Methods
The study used a cross-sectional design with data from the Australian Autism Biobank comprising clinical and biological samples from children aged 2-17 years.
Total sample included 307 autistic children, 158 non-autistic siblings, and 124 unrelated non-autistic children
Hair cortisol concentration (HCC) was used as the primary biomarker reflecting chronic stress over a period of weeks to months
Primary analysis used multivariable linear regression to identify associations with HCC after controlling for sociodemographic covariates including IQ
The cross-sectional design limits causal inference
What This Means
This research suggests that a biological marker of chronic stress — cortisol measured from hair samples (hair cortisol concentration, or HCC) — is meaningfully linked to the severity of autism-related traits and commonly co-occurring conditions in children. Specifically, children with more severe autistic traits (especially difficulties with social communication), co-occurring ADHD, more internalising problems, and more maladaptive behaviours tended to have lower HCC levels. Interestingly, children with higher sleep anxiety had higher HCC, pointing to sleep-related stress as a distinct pathway. These patterns were found after accounting for factors like the child's age, IQ, and family income.
The finding that more severe presentations were linked to lower rather than higher cortisol may seem counterintuitive, but it is consistent with research suggesting that prolonged or chronic stress can eventually lead to a 'blunted' or dysregulated stress response system. This research suggests that HCC could be a useful tool to help identify subgroups of autistic children who may have different underlying biological stress profiles, potentially reflecting different support needs.
The practical implication is that measuring cortisol in hair — a non-invasive method that reflects stress over weeks to months — could complement existing behavioural assessments for autism, particularly for young children who may not be able to communicate their distress. The authors call for further research to better understand what individual, family, and environmental factors drive these cortisol differences, with the goal of offering more targeted and personalised supports for autistic children and their families.
Chan N, John J, Mathew N, Masi A, Ong L, Eapen V, et al.. (2026). The relationship between chronic stress, co-occurring conditions, sleep, and autistic features including severity using hair cortisol concentration.. Psychoneuroendocrinology. https://doi.org/10.1016/j.psyneuen.2026.107769