Assessing global developmental delay across instruments in minimally verbal preschool autistic children: The importance of a multi‐method and multi‐informant approach

Intellectual assessment in preschool autistic children bears many challenges, particularly for those who have lower language and/or cognitive abilities. These challenges often result in underestimation of their potential or exclusion from research studies. Understanding how different instruments and definitions used to identify autistic preschool children with global developmental delay (GDD) affect sample composition is critical to advance research on this understudied clinical population. This study set out to examine the extent to which using different instruments to define GDD affects sample composition and whether different definitions affect resultant cognitive and adaptive profiles. Data from the Mullen Scales of Early Learning and the Vineland Adaptive Behavior Scales—Second Edition, a parent‐report tool, were analyzed in a sample of 64 autistic and 73 neurotypical children (28–69 months). Our results highlight that cognitive assessment alone should not be used in clinical or research practices to infer a comorbid diagnosis of GDD, as it might lead to underestimating autistic children's potential. Instead, using both adaptive and cognitive levels as a stratification method to create subgroups of children with and without GDD might be a promising approach to adequately differentiate them, with less risk of underestimating them. Lay summary It is difficult for examiners to conduct valid intellectual assessment in autistic children at preschool age, which often results in underestimation of their potential or their exclusion from studies. However, some of these children have specific cognitive abilities that often do not show up using conventional intellectual measures, which suggest that they do not necessarily present a cognitive delay. It would be important to properly distinguish autistic children with and without cognitive delay to orient them toward interventions adapted to their needs and to inform methodological choices in research. In this study, we discussed how the choice of intellectual measure (1) impacts our conclusions as whether they should be classified as cognitively delayed or not and (2) leads to different results in terms of patterns of strengths and weaknesses in preschool autistic children.