Associations between child characteristics and parent response to emotion differ in young children with and without an autism diagnosis.

The study is cross-sectional, so it cannot show that child characteristics cause parent responses or that parent responses cause child characteristics. All measures were parent-report, which may create bias and shared variance across variables. The autism and non-autism groups differed on child age, race, ethnicity, sex, and parent education, and may have differed in other unmeasured ways. The overall group differences in parent responses were small to negligible. The predictors explained only 4.7% of variance in supportive responses and 5.1% in distress reactions, so many other factors likely influence these outcomes. The non-autism group included children whose parents reported developmental concerns, and some may later receive an autism diagnosis. Speaking ability and intellectual disability were not analysed because they were confounded with diagnosis and relied on unvalidated parent report. The sample was mainly United States, White, non-Hispanic, and college-educated, so results may not generalise widely. The Johnson-Neyman results at very low SCQ scores in the autism group were unreliable because only 8 children had scores below 4.01 and only 5 had scores below 3.19.

No author conflicts of interest are reported. The paper is labelled as NIH extramural research support, and the measures are listed in the NDA under R01 HD079512.

Do not conclude that receiving an autism diagnosis causes parents to become more supportive. The study is cross-sectional, uses only parent report, and the overall differences were small. It shows associations between child characteristics, diagnosis, and parent-reported responses, not a tested effect of diagnosis.