Assessment of Intellectual Functioning at Age of Diagnosis: Testability and Comparison of Available Tests

Background:  DSM-5 autism diagnosis now requires specifying the level of intellectual functioning. An assessment of adaptive behavior and IQ is therefore necessary at time of diagnosis. However, few IQ tests  are available at this age and assessment of young autistic children bears many challenges (Akshoomoff, 2006; Rutter, 1986). In addition to this, there is a lack of studies regarding the use of IQ tests in preschool autistic children.

Objectives:  To document the testability of young autistic, developmentally delayed and typically developing children. To compare the cognitive profile of these 3 groups on available IQ tests.

Methods:  40 autistic, 40 typically developing (TD) and 25 developmentally delayed (DD) children aged from 2 to 5 years old will be tested using the MSEL, Wechsler Preschool and Primary Scales of Intelligence (WPPSI-IV), Leiter-3, Raven Color Progressive Matrices (puzzle form) (RCPM). To date, 27 autistic, 27 TD and 12 DD children were assessed. Assessment sessions typically lasted 1h, but where shortened if the child could not remain attentive throughout the whole hour. We computed a testability ratio (number of tests completed/ number of assessment sessions) and registered the number of sessions necessary to complete each test. Finally, the performance on the completed tests was compared within subjects to determine the cognitive profile of each group depending on the test used.

Results:  Within groups comparisons indicated that there was no significant difference between the number of sessions necessary to complete any of the tests in any of the three groups. Between group comparison indicated that tests including verbal assessment (MSEL and WPPSI-IV) required significantly more assessment sessions in autistic compared to TD children (all ps<.05), while it was not the case for the RCPM and Leiter-3 (all ps> .05). Neither Autistic nor TD group differed from DD group on this measure (all ps >.05). Also, the testability ratio was significantly lower in the autistic group compared to the TD group (p <.05), thus indicating that the number of tests completed by assessment session was lower in the autistic group. Other group comparisons were non-significant. Furthermore, the same ratio was significantly higher among participants who did not complete the study compared to those who did in the three groups (p<.0005). Finally, MSEL was significantly lower than Leiter-3 (p<.05) and RCPM (p<.01) in autistics, Leiter-3 was significantly lower than RCPM (p<.05) in DD and MSEL was significantly lower than RCPM (p<.05) in TD. All other tests did not differ (all ps>.05). Performance on the MSEL was the lowest in all groups.

Conclusions:  MSEL, a test that is often used to assess intelligence level at age of diagnostic, seem to underestimate the potential of all groups. Even though non-verbal tests might increase testability and be easier to administer, assessment of young autistic children remains particularly challenging. Both the lack of knowledge about the validity of the available tests at age of early diagnosis and the difficulty to assess young autistics children leads us to question the relevance of assessing intellectual potential at such a young age.