European Child & Adolescent Psychiatry 5:44-51 (1996) © Steinkopff Verlag 1996 J.K. Buitelaar H. Swaab M. van der Wees M. Wildschut R.J. van der Gaag Neuropsychological impairments and deficits in theory of mind and emotion recognition in a non-autistic boy Received: 3 October 1994 Accepted: 25 May 1995 Dr. J.K. Buitelaar (~) • H. Swaab M. van der Wees • M. Wildschut R.J. van der Gaag Department of Child and Adolescent Psychiatry RO. Box 85500 3508 GA Utrecht, The Netherlands Abstract A 9-year-old non-autistic boy revealed marked deficits in visuo-spatial and visuo-motor skills, in planning and organizational capacities and in impulse inhibi- tion. Particular strengths were his verbal comprehension and reason- ing abilities. This neuropsychologi- cal pattern of assets and deficits fitted the nonverbal learning disability syndrome as described by Rourke (1989). On a battery of Theory of Mind (TOM) and emo- tion recognition tests he performed rather poor on several first-order TOM tasks and on all second-order TOM and emotion-matching tasks, compared to samples of autistic and normal subjects. It is suggested that his visuo-spatial and cognitive shifting deficits account for his social cognitive failures, while his superior verbal skills protect him from severe social handicaps. Key words Autism - theory of mind - neuropsychology - social behaviour Introduction Deficits in social cognition, particularly in Theory of Mind (TOM) functioning, have been posited as basic to the social impairments of autistic children (7, 17). Theory of Mind is the ability of normal children to attribute men- tal states, such as intentions, beliefs, desires, etc., to them- selves and other people and to understand and predict behavior accordingly. Across different studies 20-50% of the autistic children were found to pass first-order belief-understanding tests, compared to 65- 80% of non- autistic mentally retarded and 85-90% of normal con- trol children [for review see (6) and (32)]. Further, autistic subjects who passed first-order tests were reported to fail on more difficult second-order tests which ask for under- standing 'beliefs about beliefs' (4). Ozonoff and colleagues (28, 29) however measured both social cognitive and executive functions in autistic and control subjects and questioned the primacy of TOM deficits in autism. Executive function tests, in particular the results on the Tower of Hanoi, proved to be much more effective than TOM tests in separating autistic and non-autistic subjects. Moreover, executive function measures were significantly correlated with TOM perfor- mance in autistic subjects, whereas both type of measures were unrelated in the control group. It is of interest to examine not only why some autistic subjects pass TOM tests but also why some non-autistic subjects consistently fail on these tests. In a recent study on TOM and emotion recognition performance of chil- dren with autistic spectrum disorders and with non-autis- tic psychiatric disorders, a non-autistic boy appeared to have lower scores on some of the social-cognitive tests than even autistic subjects. We will present this case in more detail to illustrate the relation between social cogni- tive and neuropsychological performance and discuss the merits of the TOM hypothesis. Case description J.E., a 9-year-old boy and an only child, was referred to the outpatient Unit because of severe hyperactivity, inat- tentiveness and learning problems. The obstetric and