A study of 479 autistic children in China identified four distinct patterns of social functioning, ranging from broad social difficulties to comparatively stronger performance across several social areas.
The researchers emphasized that the labels describe patterns found in this sample. They are not diagnostic categories or fixed autism subtypes, and the findings have not yet been replicated in a second group of children.
Four patterns, not a single scale of severity
The children were between 2 and 17 years old, with an average age of about 9. Parents completed the Mandarin version of the 71-item Stanford Social Dimensions Scale. The questionnaire measures social motivation, social affiliation, expressive social communication, social recognition and unusual approach.
Using latent profile analysis, the researchers compared models containing one to eight groups before selecting a four-profile model:
- Risky: 11.1% of the sample. This group had the lowest scores in social motivation, social affiliation, expressive social communication and social recognition, along with the highest unusual-approach score.
- Apathetic: 33.4%. Scores were moderate overall, but social motivation was as low as in the Risky group. This group had the lowest unusual-approach score.
- Active but Awkward: 40.5%. This was the largest group, with the second-highest scores across the four core social areas.
- Advantageous: 15.0%. This group had the highest scores across all four core social areas.
The names were chosen as shorthand for relative patterns in the study and should not be read as judgments about the children or as permanent descriptions of their development.
Differences in social motivation and behavior
Average social-motivation scores ranged from 27.5 in the Risky group and 27.6 in the Apathetic group to 38.5 in the Active but Awkward group and 50.9 in the Advantageous group.
Prosocial behavior scores followed a similar pattern: 6.4 for Risky, 6.8 for Apathetic, 8.7 for Active but Awkward and 10.9 for Advantageous.
Peer-problem scores showed a smaller difference in the opposite direction. Scores were 8.2 for Risky, 8.8 for Apathetic, 9.1 for Active but Awkward and 9.5 for Advantageous. The source describes this as a small but significant difference.
The groups did not differ in emotional symptoms, conduct problems, hyperactivity or total difficulties on the Strengths and Difficulties Questionnaire. They also had similar ages and boy-girl mixes.
What the other measures showed
On the Social Communication Questionnaire, the Risky group had the highest level of social-interaction difficulties and the Advantageous group had the lowest, with the other profiles between them.
The Social Responsiveness Scale-2 showed a different pattern in some respects. The Apathetic group had the highest social-communication, interaction and repetitive-behavior difficulties, while the Advantageous group had the lowest.
IQ information was available for only 110 of the 479 children. Average full-scale IQ was 41 in the Risky group, 51 in Apathetic, 68 in Active but Awkward and 88 in Advantageous. These averages should be treated cautiously because the information came from earlier tests reported by parents, and only nine children in the Risky group had IQ data.
Why the findings matter—and what they cannot show
The results suggest that social functioning may not be captured fully by one overall measure of difficulty. Children with similar challenges in one area may show different patterns in motivation, communication, recognition and social approach.
The researchers proposed that some autistic children may learn interaction strategies that bring them into social situations without leading to close, two-way friendships. They also raised the possibility that peer difficulties may reflect mutual misunderstanding between autistic and non-autistic children, sometimes called the double empathy problem. The study did not measure friendship quality or interaction style, so these explanations remain untested.
The research team at East China Normal University recruited families online through special schools, regular schools and autism intervention centers across China. Parents reported that a clinician had diagnosed their child with autism spectrum disorder. The study was published in Behavioral Sciences in September 2026.
A 2020 study by Uljarević and colleagues used the same five-area questionnaire and a similar profile method in 164 autistic children and teenagers. It identified five profiles that differed in patterns of social strengths and weaknesses rather than only overall severity. Differences between the studies could reflect culture, sample composition or measurement factors.
Important limits on interpretation
The four-profile model had a high entropy of 0.962, indicating clear separation within the selected model. However, one standard test did not favor four groups over three, meaning that the final number of profiles also involved researcher judgment.
The study used a single parent-report survey at one time point. Parents provided the diagnosis information, social ratings and symptom scores, and there was no direct clinical assessment reported. Because the families were recruited through an online convenience sample, children from families with limited internet access may be under-represented.
Children aged 2 to 17 were assessed using the same questionnaire, so the findings do not show how these patterns might change with development. The profiles came from one sample and need to be identified again in new groups before they can be considered stable subtypes.
AI tools were used to assist with the preparation of this article under our editorial standards.
Leave a Reply