New study reveals self-injurious behaviours in young autistic children are more dynamic than previously thought

Research from UC Davis MIND Institute shows that self-harm behaviours in autistic children are subject to change and influenced by environmental factors, emphasising the importance of observation and tailored support in schools.

Self-injurious behaviours such as head banging, self-biting and skin picking are not unusual in young autistic children, and a new study suggests they may be more common and more changeable than many school teams assume. Researchers at the UC Davis MIND Institute followed autistic children from the preschool years into early school age and found that these behaviours often shifted in intensity rather than staying fixed. The study, published in Autism Research, offers a clearer picture of how schools might observe and respond to them.

The research tracked 325 children aged two to five and 161 children aged four to nine, including 122 assessed at both stages. Using caregiver questionnaires, the team grouped children by severity from no self-injury to severe self-injury, while also measuring developmental ability, sensory features and broader emotional and behavioural dysregulation. The findings showed that 74% of children had some self-injurious behaviour at the earlier stage and 67% at the later one.

Children with more severe self-injurious behaviours were more likely to show stronger sensory differences and greater dysregulation at both time points, according to the paper. Lower cognitive scores only appeared to separate the more severe groups later on. The study did not find that early scores on sensory, dysregulation or cognitive measures predicted which children would worsen or improve, but it did find that children whose sensory features improved over time, and those with higher IQ at the later assessment, were more likely to show less severe self-injury.

That pattern matters for schools because it suggests these behaviours are not static traits. About a third of children became less severe over time, while a similar share became more severe or stayed roughly the same. For teachers, therapists and families, that reinforces the value of watching for triggers, changes and settings where behaviour becomes more likely. It also points to the classroom environment itself, including noise, lighting, transitions and crowding, as part of the picture.

The practical message is that observation should come before assumption. Schools can document what happens before, during and after a behaviour, note sensory and regulation concerns, and share those observations with families and related professionals. The study cannot prove cause and effect, but it does suggest that sensory support, communication help and regulation strategies may be useful parts of a wider response, especially when teams revisit plans as children develop.