Children with neurological disorders are reported to have a high rate of problems associated with autism spectrum disorders (ASD), as well as an increased risk of receiving an ASD diagnosis . ASD refers to a group of developmental/psychiatric disorders characterized by severe social and communication difficulties, as well as by a restricted pattern of repetitive and stereotyped behaviors . For the present study, we will use the Autism Spectrum Screening Questionnaire (ASSQ)  to assess symptoms associated with ASD in a population based sample of children with neurological disorders.
Previous studies assessing ASD symptoms in children with specific neurological disorders have used a range of different assessment methods. One recent study showed that at 24 months of age, prematurely born children diagnosed with cerebral palsy were more likely than preterm babies without this diagnosis to screen positively for ASD on the Modified Checklist for Autism in Toddlers (M-CHAT) , when assessed at 24 months of age . In studies of hydrocephalus, ASD as measured by the Autistic Behavior Checklist (ABC)  were present in 23% of children aged 6–17 , and autism as rated by the Childhood Autism Rating Scale (CARS)  was reported in 13% of children aged 5–12 . In a study using the parent version of the Autism Screening Questionnaire (ASQ) in assessing symptoms of ASD , 32% of children aged 2–18 with epilepsy included from a tertiary care epilepsy clinic were found to fulfill the ASQ criteria for having an ASD . In a questionnaire-based study using parent report, 3.1% of males with Duchenne Muscular Dystrophy were reported to have an ASD .
Other clinical studies have used diagnostic instruments to estimate prevalence rates of clinical diagnoses within the autism spectrum. According to the Autism Diagnostic Interview-Revised (ADI-R) , 49% of children under the age of 18 with myotonic dystrophy type 1 had an ASD , while 15% of children aged 4–18 with cerebral palsy have been reported to meet DSM-IV criteria for an ASD .
In spite of an increasing awareness in child psychiatry and developmental medicine that various conditions, such as neurological disorders and ASD, often co-exist and share symptoms [16–19], health services for children largely remain specialized. Children with neurological disorders are most commonly treated by specialist somatic health care services in pediatric clinics, where psychiatric problems, such as ASD, may remain unrecognized [19, 20]. A study from 2006 showed that only 8% of pediatricians routinely screened for ASD, partly due to unfamiliarity with ASD screening tools . Awareness and use of adequate screening instruments to identify symptoms of ASD would thus be of benefit to pediatric clinics.
The ASSQ  has, to our knowledge, not previously been used in studies focusing on children with neurological disorders. The questionnaire has been validated as a screening tool for ASD symptoms by studies in northern Europe and UK since the late 1990s [3, 22–26], more recently in China , and in Norway using data from the first wave of the Bergen Child Study (BCS) . The BCS sub-study, which included a sample of children with neurological disorders, showed that more than 90% of the children who received an ASD diagnosis according to the Diagnostic Interview for Social and Communication Disorders (DISCO)  were also scored above the 98th percentile on the ASSQ by parents and/or teachers, corresponding to a sensitivity of 0.91 and specificity of 0.86 . Furthermore, a factor analysis of the ASSQ items in the first wave of the BCS revealed a stable three-factor structure within both parent and teacher ASSQs. These factors were labeled “social difficulties”, “motor/tics/OCD”, and “autistic style” . For validation purposes, the factors were correlated with the five subscales of the Strengths and Difficulties Questionnaire (SDQ): emotional problems, conduct problems, hyperactivity-inattention, peer problems, and prosocial behavior [31, 32]. The ASSQ total score and factors showed the highest correlation with the SDQ peer problems subscale . Another sub-study from the BCS on mental health problems in children with chronic illness found SDQ peer problems to be one of two areas scored particularly highly for children with neurological disorders . These BCS findings motivated us to assess the utility of the ASSQ in detecting symptoms associated with ASD in children with neurological disorders.
The ASSQ is short, and unlike many other ASD screening instruments, it may be completed by parents and teachers alike. Multiple informants are considered to be important when screening children for psychiatric symptoms, as several studies have revealed only low to moderate agreement between raters [34–36]. For the ASSQ, the parent-teacher correlation was r = 0.66 in a clinical study of ASD assessment by Ehlers and collaborators , but only 15% of high-scorers were identified by both parents and teachers in the first wave of the BCS . For ASD more generally, a Finnish population study found that only 24% of ASD cases were identified by both parents and teachers . In a study of children with epilepsy, however, parent-teacher correlations for behavior ratings were moderate to high .
The aim of the present study was to assess symptoms associated with ASD in children with neurological disorders, as reported by parents and teachers on the ASSQ. Our study will thus be the first to investigate use of the ASSQ in this group of children. Based on results from earlier studies of children with neurological disorders, we expected to find the highest ASSQ scores and a higher frequency of children with a score associated with ASD in children with neurological disorders, as compared with control groups of children with other chronic illnesses or no chronic illness. Secondly, we wished to examine parent-teacher agreement on ASSQ-based assessments of children with neurological disorders, by looking at overall scores as well as three sub scores based on a factor analysis, and single item scores.