How do siblings influence theory of mind development in children with autism?

Guess which one is me...

It goes without saying that brothers and sisters play an important role in a child's development, particularly when it comes to their social skills. For young children especially, the family is their social environment and the main opportunity to learn about other people and what makes them tick.

Research conducted in the past 15 years or so has consistently shown that children with siblings of a similar age tend to pass tests of "theory of mind" at a younger age than those without siblings. The implication is that the experience of interacting with siblings helps children to develop the concept that other people have minds and that their thoughts and beliefs are sometimes different from their own.

Children with autism typically struggle on tests of theory of mind. An interesting question, then, is what effect siblings have on theory of mind development in autism. Based on the literature on typically developing kids, we might expect siblings of autistic children to have a beneficial effect. We could even make a case that, because autistic kids may have fewer interactions with non-family members, siblings may be even more important than normal. Counterintuitively, however, a new study published in the Journal of Child Psychology and Psychiatry suggests that having older siblings can have a detrimental effect on autistic children's theory of mind development.

The study was conducted by Karen O'Brien, a PhD student from the University of Queensland and involved 60 kids with autism aged between three and thirteen years. The children each completed a battery of six "theory of mind" tests, that assessed their awareness of the separation between mental states and reality. Tests involved understanding that another person can have a false belief; that an object can look like something else; and that someone can pretend something that isn't true. Each child received a score out of 6 depending on how many tests they passed [1]. 

Here are the results for the children, divided up according to the siblings in their family. The striking finding is that autistic kids who had older siblings did worse than those with no older siblings. Children with only older siblings (ie those who were the youngest in their family) performed worst of all.

Number of "theory of mind" tests passed

Given the unexpected nature of their results, O'Brien and her co-authors tried to find other differences between the groups that might explain their performance on the theory of mind tasks. They looked at their age, language skills, planning skills, and measures of autism severity, but the four groups were pretty well matched on all of these. So what else might explain the sibling effect?

The authors speculate that well-meaning older siblings may over-compensate for the autistic child's difficulties. By treating them with kid gloves, they may somehow limit their development. Younger siblings might be less likely to do this and so have a more benign influence. However, it's not clear why having older siblings would be worse than having none at all.

As O'Brien et al. also point out, children without older siblings are by definition first born children. As such, they probably benefit from more one-to-one time with their parents. It may be that, for an autistic child, this advantage outweighs the lack of input from siblings.

The authors are refreshingly circumspect in discussing the implications and limitations of their study, noting that the results really need to be replicated in another study before we can have full confidence in them. They also urge caution in generalising beyond the middle-class Western culture to which the families in the study belonged.

To this I would add that performance on these kinds of theory of mind tasks is a fairly narrow measure of social functioning. Even if it's true that siblings can unwittingly hold back the theory of mind development of their younger autistic siblings, they may well have beneficial effects on other aspects of social development. And, as in typical development, the effect of siblings on the development of social skills is going to vary considerably depending on the individuals involved and the family circumstances - factors that aren't captured by the measures used in the current study.

Clearly, these are preliminary findings but they point towards an important new direction for research into theory of mind in autism. If nothing else, this study highlights the relevance of the social environment in the development of autistic children's social skills - something that has been ignored in a research literature that has traditionally viewed theory of mind as a component that is simply missing from the autistic child's brain. Despite the acknowledged limitations of the study, I think it represents an important step in the right direction.


Notes:

[1] To me, the way the tasks were scored was a bit weird. It's common practice to include control questions to check that subjects have a basic understanding of the task requirements. If they fail the control question then their performance on the theory of mind question is pretty much impossible to interpret. However, O'Brien et al. used the control questions to check that the kids weren't just guessing. So whether or not they got the theory of mind question right, they were scored as failing if they got any of the control questions wrong. This means that kids with low scores may have had specific problems with theory of mind questions or with the general demands of the question - we just don't know. That said, it's not clear how this unusual scoring method might explain their pattern results.

Reference:

O'Brien K, Slaughter V, & Peterson CC (2011). Sibling influences on theory of mind development for children with ASD. Journal of child psychology and psychiatry, and allied disciplines PMID: 21418062