Pervasive developmental disorders in individuals with cerebral palsy


  • *

    North American usage: mental retardation.

  • Acknowledgments
    We thank the parents and children who participated in this study, Dr Meral Ozmen and Dr Burak Tatli from the Istanbul Medical Faculty Paediatric Neurology Department, and the staff of the Metin Sabanci Rehabilitation, Education and Productivity Centre for Children and Teenagers with Cerebral Palsy for their contribution in the collection of data.

Dr Ayse Kilincaslan at Universite Bulvari Binevler Mah, Kemal Ogucu Apt No 186 D:4 Sahinbey, Gaziantep, Turkiye. E-mail:


The aim of the present study was to describe the prevalence and associated factors of pervasive developmental disorders (PDD), including autistic disorder and PDD not otherwise specified (NOS), in a clinical sample of 126 children and adolescents (75 males, 51 females; age range 4–18y, mean 8y 8mo, SD 3y 8mo) with tetraplegic, hemiplegic, diplegic, dyskinetic, or mixed types of cerebral palsy (CP); 28% could not crawl or walk even with support, 29% could move with support, and 43% walked independently. Participants were examined for PDD in two stages. In the first stage, probable participants were determined by direct observation, Autism Behavior Checklist score, and medical reports. In the second stage, those with ‘probable’ symptoms underwent psychiatric examination and their autistic symptoms were scored on the Childhood Autism Rating Scale. The final diagnosis of autistic disorder or PDD-NOS was given according to DSM-IV criteria. Fourteen (11%) and five (4%) of the participants met the criteria for autistic disorder and PDD-NOS respectively. Children with CP and PDD differed from those without PDD in terms of type of CP (p=0.02), presence of epilepsy (p<0.001), intellectual level (p<0.001), and level of speech (p<0.001). PDD was more common in children with tetraplegic, mixed, and hemiplegic CP, and in children with epilepsy, learning disability,* and low level of speech. The findings corroborate the notion that CP is a complex disorder, often associated with additional impairments. PDD is not rare in CP and should be considered in patients with comorbid conditions such as epilepsy, learning disability, and language delay and in the presence of tetraplegic, mixed, and hemiplegic CP types.

List of abbreviations

Autism Behavior Checklist


Autistic spectrum disorder


Childhood Autism Rating Scale


Not otherwise specified


Pervasive developmental disorder

Cerebral palsy (CP) describes a group of permanent disorders of the development of movement and posture, which are attributed to non-progressive disturbances that occurred in the developing fetal or infant brain.1 CP is one of the most common disorders of childhood, with a prevalence of 2 to 5 per 1000 live births. Although motor dysfunction is the defining clinical feature of CP, sensory, cognitive, and verbal impairment in addition to learning difficulties and behavioural problems are common. People with CP may show increased rates of emotional lability, irritability, attention deficits, impulsiveness, and limited skills in problem solving.2 Although a wide variety of behavioural problems, including autistic behaviour and autism, has been reported in individuals with CP, few studies have been focused on the relationship between autism and CP. A review of the literature revealed three groups of studies on the coexistence of autism and CP.

The first group comprises studies that are focused on co-occurring medical disorders in populations with autistic disorder or other pervasive developmental disorders (PDD). Steffenburg3 and Barton and Volkmar4 found no children with CP in two populations with autism. However, other investigators5–8 reported rates between 0.8% and 4.8%, which are four to 24 times the general population rate for CP (0.2% in developed countries).

The second group includes studies of whether CP is associated with autism in various medical disorders. CP was found to be linked to autism in children with hydrocephalus,9 retinopathy of prematurity,10 and visual impairment.11 Nordin and Gillberg12 examined 177 children with heterogeneous clinical diagnoses causing learning or motor disability, or both, and found that four (10.5%) of the 38 children with CP had autistic spectrum disorders (ASD). Steffenburg et al.13 studied the comorbidity of ASD in 98 children with active epilepsy and learning disability. Thirty-seven of those children had concurrent CP and six of these had the diagnosis of ASD (16.2%). The authors of both of those studies reported no specific relationship between CP and autism but stated that the co-occurring learning disability or epilepsy was responsible for the high prevalence of autism in their populations. However, in both reports, the authors added that CP and other motor disabilities had been very poorly studied from the point of view of ASD.

The only study in which children with CP were evaluated for the presence of ASD was conducted by Goodman and Graham,14 who included only children with hemiplegic CP. They examined 149 children, aged between 6 and 10 years, for the presence of psychiatric disorders, using parental and teacher reports and interviews with the children. They identified four children with autism (3%) and suggested that this rate was 10 times that in the general population. The report gave no information about the diagnostic criteria, clinical features of the cases, or coexisting conditions.

The aims of the present study were to assess (1) the prevalence of autistic disorder and PDD not otherwise specified (NOS) in a relatively large clinical population with CP; (2) probable associated factors such as type of CP, level of motility, epilepsy, and intelligence level; and (3) the clinical features of autism in individuals with CP.


The study involved 126 children and adolescents with a diagnosis of CP (75 males, 51 females; mean age 8y 8mo, SD 3y 8mo, range 4–18y). Between April and July 2006, they were attending the Istanbul Medical Faculty Paediatric Neurology Department Outpatient Clinic, the Paediatric Physiotherapy and Rehabilitation Clinic, or an association that provides assistance for individuals with CP in Istanbul, Turkey. Participants, all of whom were diagnosed by a paediatric neurologist, were selected from consecutive patients above 48 months of age (a cut-off chosen because of the difficulty in discriminating autism from severe learning disability in children younger than 4y) with spastic, dyskinetic, or mixed types of CP. Patients were not included if they had ataxic CP or progressive hereditary, neurological or metabolic disorders as the cause of the clinical presentation.

The study was approved by the Research Ethics Committee of Istanbul University. The aim, protocol, and forms to be completed in the study were explained to the caregivers. Written informed consent was obtained from all caregivers.


Participants were examined for the presence of PDD using a two-stage process. In the first stage, participants with a ‘probable’ diagnosis were identified through direct observation and scoring on the Autism Behavior Checklist (ABC)15 by the clinician using parental information, and through developmental and medical information gathered from the parents and medical reports. Each of the participants was observed and individuals who could communicate through speech were interviewed by the first author. Participants who scored over 35 on the ABC and showed autistic features during the observations and interviews were identified as ‘probable cases’.

In the second stage, the children with a ‘probable’ diagnosis underwent psychiatric examination and their autistic symptoms were scored on the Childhood Autism Rating Scale (CARS)16 by the two authors independently (interrater correlation r=0.84). The researchers, who had 5 and 17 years of experience working with children and adolescents with ASD, gave a diagnosis of PDD, including autistic disorder and PDD-NOS, according to DSM-IV17 criteria, by consensus. The second author, who is the founder of the Pervasive Developmental Disorders Clinic at the Istanbul University, also evaluated 10 randomly chosen children from the 97 participants who were not identified as ‘probable cases’ for the presence of PDD; none of these was considered to be in the PDD spectrum.


Autism Behaviour Checklist

The ABC consists of 57 behaviours that seem more common in children with autism than in children with other disabilities. It has been translated into Turkish and reliability and validity studies have been conducted.18 The total score generated on the ABC ranges from 0 to 158. A total score of 67 or above is considered to indicate autism with high probability and scores in the range of 53 to 67 are considered to indicate ‘questionable’ autism. Previous investigators11,12 of autistic disorder among children with disabilities reported lower cut-off scores on the ABC. Because of the results of those reports and the difficulty of scoring items related to motor function for some of the participants (items 1, 4, 11, 13, 20, 27, 30, 41, and 45), we counted participants with total scores over 35 as probable cases and included them in the second stage of the study.

Developmental and medical history form

This form was filled in by the first author using parental information and medical reports. It included information about socio-demographic features, psychomotor development, type of CP, level of motility, associated medical disorders (visual and hearing impairment, epilepsy, and other medical conditions), family history, and cognitive skills. Assessment of cognitive functions, which can be very challenging because of limitations of movement and motor control, was conducted by qualified psychologists experienced in assessing cognitive skills in physically disabled children at the facilities where the participants were being followed. The children were assigned to the following three levels of intelligence: moderate to severe learning disability (IQ ≤50), mild learning disability (IQ 51–70), and no learning disability (IQ ≥71) based on information from standardized intelligence tests (the Wechsler Intelligence Scales for Children, revised,19 and the Stanford–Binet Intelligence Scale20). Participants were also assigned to the following three levels of motility: cannot move by crawling or walking, even with support; crawls or walks with support but cannot walk independently; and walks independently.

Childhood Autism Rating Scale

The CARS is an autism diagnostic schedule covering 14 functional areas that may be compromised in autism, and a final general category referring to ‘degree of autism’. Turkish translation, reliability, and validity studies have been carried out by Sucuoglu et al.21 The CARS is widely used to determine the presence and degree of autism in the absence of more validated instruments. The total score range is between 15 and 60. Scores of 30 to 36 indicate mild to moderate autism and scores above 36 indicate severe autism.

Statistical analysis

Two groups (children with CP with PDD and children with CP without PDD) were compared with independent sample t-tests and χ2 tests for parametric and nonparametric data respectively. An arbitrary level of 5% statistical significance (two-tailed) was assumed. SPSS for Windows (version 10.0) was used for the statistical analysis.


Twenty-nine of the 126 participants were identified as ‘probable cases’ of PDD and proceeded to the psychiatric interview. Nineteen of these 29 (13 males, six females) had a final diagnosis of PDD (15% of the total sample; 95% confidence intervals [CI] 8.7–21.4) and the remaining 10 had problems with language, social contact, and stereotypical behaviours that did not fulfil the DSM-IV criteria for any PDD diagnosis. Fourteen (11%) and five (4%) of the 19 participants with a final diagnosis of PDD met the DSM-IV criteria for autistic disorder and PDD-NOS respectively. In this report, these 19 participants will be referred to as the ‘children with CP with PDD’ (CP + PDD), whereas the others will be referred to as the ‘children with CP without PDD’ (CP – PDD).

No statistically significant difference was found between the two groups in terms of age (CP + PDD: mean age 8y, SD 3y 9mo; CP – PDD mean age 8y 10mo, SD 3y 8mo; t=0.91, df=124, p=0.36), sex (χ2=0.36, df=1, p=0.55) and level of motility (χ2=2.24, df=2, p=0.33).

CP + PDD children showed a statistically significant difference from CP – PDD children in terms of type of CP (χ2=11.69, df=4, p=0.02), presence of epilepsy (χ2=12.77, df=1, p<0.001; odds ratio 6.62, 95% CI 2.33–18.80), intellectual level (χ2=28.27, df=2, p<0.001), and level of speech (χ2=43.93, df=2, p<0.001; Table I). The rates of PDD by type of CP were as follows: tetraplegic CP, seven out of 35 (20%); hemiplegic CP, five out of 32 (15.6%); diplegic CP, two out of 41 (4.9%); dyskinetic CP, no case out of 6 (0%); and mixed CP, five out of 12 (41.7%). Epilepsy affected 34 children with CP (27%; 95% CI 19.1–34.8). Infantile spasms occurred in three children with PDD and in two children without PDD. Two children in each group had Lennox–Gastaut syndrome. Two (10.5%) of the CP + PDD children and one (0.93%) of the CP – PDD children had autism in their family history (p=0.06).

Table I.   Associated factors in children with cerebral palsy (CP) with or without pervasive developmental disorder (PDD)
  Children with CP and PDD (n=19)Children with CP without PDD (n=107)Significancea
n (%)n (%)
  1. aχ2 test, significance level p<0.05. bSevere, cannot move by crawling or walking, even with support; moderate, crawls or walks with support but cannot walk independently; mild, walks independently.

 Male13 (68.4)62 (58)p=0.55
 Female6 (31.6)45 (42)
Type of CP
 Spastic tetraparesis7 (36.8)28 (26.2)p=0.02
 Spastic hemiparesis5 (26.3)27 (25.2)
 Spastic diparesis2 (10.5)39 (36.4)
 Dyskinetic0 (0)6 (5.6)
 Mixed5 (26.3)7 (6.5)
Motility impairmentb
 Severe8 (42)27 (25)p=0.33
 Moderate4 (21)33 (31)
 Mild7 (37)47 (44)
Epileptic seizures
 None7 (36.8)85 (79.4)p<0.001
 Epilepsy12 (63.2)22 (20.6)
Level of speech
 No phrasal speech14 (73.7)10 (9.3)p<0.001
 Phrases with 2–3 words3 (15.8)31 (28.9)
 Long sentences2 (10.5)66 (61.7)
IQ level
 Normal0 (0)66 (61.7)p<0.001
 Mild learning disability5 (26.3)19 (17.7)
 Moderate/severe learning disability14 (73.7)22 (20.6)
 Family history of autism2 (10.5)1 (0.9)p=0.06

The CARS score for the CP + PDD children ranged between 30.5 and 47; 11 had scores between 30 and 36, indicating mild to moderate autism, and eight had scores between 37 and 47, showing severe autism. The other 10 participants who satisfied the cut-off point on the ABC but not on the CARS had scores between 19.5 and 24.5. The mean scores of each CARS item for the CP + PDD children are shown in Figure 1.

Figure 1.

 Mean scores for each item on the Childhood Autism Rating Scale (CARS) for children with cerebral palsy and pervasive developmental disorder: 1, relationship to people; 2, imitation; 3, emotional response; 4, body use; 5, object use; 6, adaptation to change; 7, visual response; 8, listening response; 9, taste/smell/touch response and use; 10, fear and nervousness; 11, verbal communication; 12, nonverbal communication; 13, activity level; 14, level and consistency of intellectual response; 15, general impressions.


The prevalence of PDD in a clinical sample of children and adolescents with CP was found to be 15% (11% autistic disorder and 4% PDD-NOS). The presence of epilepsy, learning disability, language delay, and some types of CP were found to be associated factors for the coexistence of PDD. Because there is no other study on the prevalence of PDD in a similar population, our findings will be compared with those of the following two studies.

Goodman and Graham14 examined the prevalence of psychiatric problems in 149 children with hemiplegia aged between 6 and 10 years. They found that conduct, emotional, and hyperactivity disorders dominated, and 3% of the participants (95% CI 0.7–7.7) were defined as having autistic disorder. In our study, 15.6% (95% CI 6–30) of the participants with hemiplegia had a diagnosis of PDD, which is five times the prevalence in the study by Goodman and Graham. The rate of learning disability was similar in both groups with hemiplegia (40.6% in our study vs 35.5% in Goodman and Graham’s). The rate of epilepsy was 31.7% in our participants with hemiplegia, but there is no information on epilepsy in the other study. It is difficult to compare the results of these studies because of differences in assessment instruments and sample characteristics. Goodman and Graham aimed to assess the rate of all psychiatric disorders using instruments designed mostly for emotional and behavioural problems and gave no information about the classification system used to make the autism diagnosis. In our study we used the ABC and CARS, both of which are widely used and validated instruments for autism, especially in individuals with severe learning disabilities,12,22 and the diagnosis was given according to DSM-IV criteria.

Nordin and Gillberg12 studied the prevalence of PDD in children with mental or physical disabilities, or both. Children with various neurological and neurodevelopmental disorders, including CP, were assessed with the ABC, the CARS, and detailed neuropsychiatric examination and were diagnosed according to DSM-III-R23 criteria. Among 38 children with CP, four (10.5%; 95% CI 3.4–22.7) had an ASD. The rate of PDD found in our study is higher than in that study (15% vs 10.5%) but the difference is not statistically significant (p>0.05). The rate of epilepsy in both CP populations was similar (27% in our study and 29% in Nordin and Gillberg’s, p>0.05) but the rate of learning disability in our population was twice that of the other study (47.7% vs 23.7%, p<0.001). In both studies, all of the participants identified as having ASD had learning disabilities. Because our study and previous studies24–26 show a close relationship between learning disability and PDD, a higher rate of participants with learning disability might be associated with a higher rate of PDD in our study.

In contrast to the above studies, the present study is not population based. The participants were recruited from tertiary clinics. The distribution of the CP types in our sample differed from the Turkish population,27 with a higher rate of tetraplegic CP (27.8% vs 19.9%, p<0.05). Also the rate of tetraplegic CP in our study was higher than in population studies conducted in Norway28 and Sweden29 (15% and 6.3% respectively, p<0.001) but not in Iceland30 (22%, p>0.05). In terms of epilepsy, our study was similar to the above studies (27% vs 28%, 33% and 26.7% respectively, p>0.05). The rate of learning disability in our study differed from that in the Norwegian study (47% vs 31%, p<0.001) whereas it was similar to that of the Swedish and Icelandic studies (both 40%, p>0.05). Although there is no study of the Turkish CP population with which we can compare our sample in terms of severity of CP and additional impairments, it is possible that we included more severe cases with higher rates of tetraplegic CP and learning disability. Therefore, our findings may be somewhat overestimated and should not be generalized to the whole population with CP.

Epilepsy, learning disability, language delay and tetraplegic, hemiplegic, and mixed types of CP were found to be associated factors for the coexistence of autistic disorder in our study. Several investigators have described a connection between autism and the severity of brain dysfunction.6,9 In our study, autism was found to be more common in children with signs of a more extensive brain lesion, resulting in epilepsy, learning disability, or tetraplegic or mixed types of CP. Higher rates of autism in the family history of children with CP + PDD also suggests the role of genetic factors in the aetiology of autism. The level of motility did not differ between groups, which disproves the possible role of under-stimulation due to low levels of ambulation as a possible contributor to the aetiology of autism. However, the results of our study are far from determining whether the same brain lesion leading to CP is also responsible for the autistic pattern or whether the co-occurring learning disability or epilepsy is the reason for the high prevalence of autism in this population.

When the children with PDD were assessed for autistic symptoms using the CARS, the mean scores were in the range of mild to moderate severity for all items except ‘adaptation to change’. This item questions the child’s interest in routines and his or her reactions when interrupted. Insistence on routines and ritualistic behaviours have been reported to occur more commonly in higher-functioning individuals with autism.22 All of our participants with PDD were low-functioning, with mild to severe learning disabilities, and a lower level of resistance to change is compatible with our group.

Clinically, the full range of autistic behaviour was observed in our children with CP and PDD, but the presentation of autistic symptoms was affected by the child’s cognitive and motor development. Social interaction problems mostly included problems with eye contact, lack of social smile, not reaching out when reached for, having no friends, and not responding to social and environmental cues. Stereotypical behaviour was quite common and differed according to motor development. Children who could walk often whirled, but the others rocked, flapped, clapped, and watched their hands. Playing with parts of objects such as spinning wheels (in nine children) and fiddling with bits of string or plastic bags were common. Sensitivity to smells and sounds was common in both groups.

Assessment of impairment in communication was the most difficult. Because of bilateral corticobulbar dysfunction in many CP syndromes, anarthric or dysarthric speech caused by oromotor dysfunction is common. The rate of severely impaired speech or no speech at all is reported to be around 20% in CP.31 Also, children with CP are often described as passive communicators, taking a largely respondent role while adults introduce topics and start most conversations.32 In our PDD group, 11 participants had no words, three spoke with individual words, a further three used two- to three-word sentences, and the remaining two used longer sentences. We counted mutism as an autistic feature when it was accompanied by no compensatory alternative modes of communication such as gestures or mime. Most of the nonverbal children were reported as behaving as if they could not hear, although hearing assessments revealed no problems. These children did not respond to their names being called but were responsive when offered something that they wanted. They did not point or gaze at the objects that they wanted spontaneously. For those who could talk, five children had echolalia and one had pronoun reversal. It is important to note that motor and cognitive impairments should be taken into consideration when PDD is being assessed in children with CP in order not to over-diagnose.


The finding of a high number of children with PDD should alert professionals working with children with CP. In children diagnosed with a physical or neurological disorder such as CP in the early stages of their lives, problems with social interaction, verbal and nonverbal communication, and behaviour may be overlooked or thought of as a part of the disorder, so a diagnosis of autistic disorder or PDD-NOS may go unnoticed (only three of the 19 children with PDD in our study were diagnosed as having autistic disorder before the study). It is known that early diagnosis and intervention are of critical importance in the prognosis for children with ASD,33–35 so health care professionals need to be more sensitive and knowledgeable about the symptoms of ASD, which can cause more suffering than motor deficit or epilepsy for the children and their families. The results of our study also point to the need to be more suspicious when CP is accompanied by epilepsy, learning disability, speech delay, or family history of autism, or in the case of tetraplegic, hemiplegic, or mixed types of CP.