Two Kinds of Repetition, Two Kinds of Control

What the research found

Restricted and repetitive behaviours are one of the two defining features of autism. But the term covers a wide range. On one end sit rituals, insistence on sameness, and deep, narrow interests. On the other sit repetitive movements and strong sensory preoccupations. Researchers call the first group higher-order behaviours and the second sensorimotor behaviours. This study asked whether the two are supported by different mental skills.

Susan Faja and Laura Nelson studied 102 autistic children aged seven to eleven. All had a diagnosis confirmed through standard assessments, and all had an IQ of at least 80, so none had an intellectual disability. Only eleven were girls. The children completed tasks measuring two mental abilities: inhibitory control, the capacity to hold back an automatic response, and shifting, the capacity to move flexibly from one rule or idea to another. Repetitive behaviours were rated from the diagnostic interviews and observations, and parents completed a questionnaire.

The central result is that the two kinds of repetition tracked different skills. Better performance on the Stroop task, a classic measure of resisting an interfering response, went with more higher-order behaviours, not fewer (r = .258, p = .013). Shifting told a similar story. On the parent-rated shifting scale, where a higher score means more difficulty, greater difficulty was linked to more rituals and intense interests (r = .295, p = .003). When the two measures were combined in a single model, each still added something on its own. Neither of these measures was related to sensorimotor behaviours at all.

Sensorimotor behaviours lined up with a different skill entirely. The researchers measured proactive inhibition, a child’s tendency to slow down in advance when they expect they might have to stop. Children who slowed more had fewer repetitive movements and sensory preoccupations (r = -.301, p = .003). This measure, in turn, had nothing to say about rituals and interests.

Two points deserve emphasis. First, these are correlations, and modest ones. A coefficient near .3 means the mental skill accounts for well under a tenth of the differences between children. Most of the variation is left unexplained. Second, the direction of the higher-order findings is not the intuitive one. Stronger interference control and, in the combined model, the executive picture as a whole were associated with more rituals and interests, not fewer. The authors suggest that pursuing an elaborate interest or maintaining a routine may actually draw on intact control, rather than reflecting a failure of it.

What it means for you and your child

The most useful idea to take from this study is that repetitive behaviour is not one thing. A child’s absorbing interest in trains and a child’s hand-flapping may look like members of the same family, but they appear to rest on different foundations. That matters, because it cautions against assuming that one explanation, or one response, will fit every repetitive behaviour your child shows.

It is worth sitting with the counter-intuitive finding. A rich, rule-governed interest is not obviously a sign that something is going wrong with a child’s self-control. In this sample, the children with stronger control tended to have more of these higher-order behaviours. Deep interests can be a place where a child’s attention and capability show themselves. This study does not tell you whether to encourage or redirect any particular behaviour, but it should soften the instinct to treat all repetition as a problem to be reduced.

Be careful, though, about how far you carry any of this. The relationships here are weak. Knowing how a child performs on a lab task would let you predict very little about that individual child’s behaviour at home. Averages across a hundred children do not describe the child in front of you. Nothing here is a diagnostic tool or a basis for deciding what a specific child needs.

Causation is the sharper caution. This is a snapshot taken at one point in time. It cannot tell us whether mental skills shape repetitive behaviour, whether repetitive behaviour shapes those skills, or whether some third factor drives both. The authors are explicit about this: the design cannot reveal how these patterns develop over the childhood years. A finding like this points to questions worth asking, not to a mechanism you can act on.

The authors name other limits honestly, and they are worth knowing. The sample was narrow. It included very few girls and no children with intellectual disability, so the findings may not extend to the full range of autistic children. The battery of mental tasks was deliberately small, covering only two skills. And they note that the parent shifting questionnaire may partly measure repetitive behaviour itself, since some items resemble one another, which could inflate that particular link.

So what is left? A careful, single study, with real numbers behind it, suggesting that the machinery underneath rituals and interests differs from the machinery underneath repetitive movement. That is a genuine contribution to how researchers think about autism. For a parent, it is best held as a way of seeing rather than a plan of action: your child’s repetitions may be more varied, and less uniformly worrying, than a single label implies.