The subjects consisted of the parents of children with classical AD accompanied by idiopathic mental retardation, and the parents of TDC. All of the autistic children were recruited from the pool of outpatients of the Developmental Disorders polyclinic, with a diagnosis of AD, established in the Child and Adolescent Psychiatry Department of Ege University, Izmir, Turkey. We selected probands in the age range of 4–18 years with AD accompanied by idiopathic mental retardation. We limited the sample to subjects with a core diagnosis of AD because of concerns about distinguishing between conditions within the autism spectrum. Turkish adaptations of instruments such as the Autism Diagnostic Interview–Revised and the Autism Diagnostic Observation Schedule–Generic are not yet available. We clinically diagnosed mental retardation, depending on the developmental and clinical symptoms and functioning levels of the children according to age. This information was obtained using the form developed by the authors. The child's age, gender, developmental history, parents' age, gender and educational years were noted on this form at the first evaluation. The children with AD who did not have any comorbid medical or neurological conditions such as Fragile-X syndrome, or phenylketonuria, were included in the study. A history of medical, neurological or genetic causes was the exclusion criterion for children.
Fifty-four probands with AD according to DSM-IV-TR criteria, and 84 TDC were included in this study. The mean age of the AD probands was 12.6 ± 3.2 years. The mean age of the control TDC was 11.4 ± 5.4 years. There was no statistically significant difference between the mean age of the AD children and the control TDC. A total of 100 parents of AD children and 100 parents of control TDC were needed. There was no significant difference between the mean age of the parents of the AD group (38.4 ± 6.9 years) and those of the TDC (36.7 ± 7.4 years; P = 0.09). The mean age of the mothers of the AD probands was 36.4 ± 5.8 years, and that of the fathers was 40.6 ± 6.7 years. The control group consisted of 100 parents (52 mothers and 48 fathers) of 84 children, who had a child without any significant behavioral problems. The mean age of the mothers of the control group was 35.7 ± 7.3 years, and that of the fathers was 37.8 ± 7.9 years. There was no significant difference between the ages of mothers (P = 0.61, t = 0.51); and fathers (P = 0,06, t = 1.91) mean age (Table 1). The control group was matched to index parents for age, gender and educational level (Table 1). It is suggested that education is one of the most powerful predictors of premorbid IQ. There was no significant difference between the mean age and between the mean years of education of the parents of ASD children and those of TDC. These control parents were recruited from among the hospital staff and their friends/relatives. As expected in the AD group there was a male predominance.
Table 1. Subject demographic characteristics
| ||AD group||TDC group||P|
|Age of children (years)||12.6 ± 3.2||11.4 ± 5.4||0.17|
|Gender of children, n (%)|| || || |
|Male||43 (79.6)||47 (55.9)||χ2 = 10.185|
|Female||11 (20.4)||37 (44.1)||P = 0.001|
|Age of parents (years)||38.4 ± 6.9||36.7 ± 7.4||0.09|
|Age of mothers (years)||36.4 ± 5.8||35.7 ± 7.3||0.61|
|Age of fathers (years)||40.6 ± 6.7||37.8 ± 7.9||0.06|
|Gender of parents, n|| || || |
|Male||47||48||χ2 = 0.020|
|Female||53||52||P = 0.89|
|Education years of parents||12.2 ± 3.3||12.7 ± 3.9||0.31|
As stated in the previous section, 100 parents of autistic children and 100 parents of control TDC were needed. All of the children with autism were recruited from the pool of outpatients of the Developmental Disorders polyclinic, with a diagnosis of AD accompanied by idiopathic mental retardation. We obtained the diagnosis information from their files. Initially, all of the parents (66 probands) of autistic children who did not have any comorbid genetic, medical or neurological conditions such as Fragile-X syndrome or phenylketonuria were interviewed by phone, email, or face to face during routine examinations of their child at the outpatient clinic. At the first stage, the children with autism, of whom the parents agreed to participate, were once again diagnosed for this study by a structured interview using the evaluation form developed by the authors and based on the DSM-IV-TR A-B, and C criteria of AD assessing the social functioning (five items), communication (four items), and stereotypic–ritualistic behavior/interests (four items) domains. The child's age, gender, developmental history, parents' age, gender and educational years were also noted on this form. The parents of 54 children of the 66 with AD agreed to participate. At the second stage, a second child psychiatrist who had 15 years' experience with autistic children conducted the second diagnostic interview. The researchers gave the diagnosis of AD for children according to DSM-IV-TR criteria, by consensus.
Among the parents, 46 couples and eight single parents (seven mothers, one father) agreed to participate. None of the parents had a history of psychiatric treatment. The final index group consisted of 100 parents (53 mothers and 47 fathers) of 54 probands with AD according to DSM-IV-TR criteria.
All included parents completed the AQ-TR. The AQ is a brief, easy to use and score, self-reported questionnaire for assessing the broader phenotype in adults of normal IQ. It consists of 50 questions, made up of 10 questions assessing five different areas: social skill; communication skills; imagination abilities; attention switching; and attention to details. Each of the items listed here scores 1 point if the respondent records the abnormal or autistic-like behavior either mildly or strongly. The higher the score on the AQ, the more autistic traits the individual has. The AQ was translated into Turkish (AQ-TR) by Köse et al. The AQ-TR was completed by 406 university students (58% female, 42% male). To show the reliability of the Turkish version of the AQ (AQ-TR), Cronbach's alpha and test–retest reliability were evaluated for the university students. Factor analysis was used to test the construct validity of the scale. Cronbach's alpha for the scale was 0.64. Test–retest reliability of the scale was satisfactory (0.72). All subscales were in correlation with the total score of the scale. The total score and subscale scores of AQ-TR did not correlate with age (r = −0.10–0.07, P = 0.06–0.84) or duration of education (r = −0.10–0.02, P = 0.06–0.99). The total score of the AQ-TR (P = 0.003) and the scores of social skills (P = 0.001), communication (P = 0.034), and imagination (P = 0.002) subscales were higher in men. AQ-TR was found to be a reliable instrument. Parents were encouraged to fill in any missing questions, which were detected during the collection of questionnaires.
Control subjects were interviewed using the Child Behavior Checklist (CBCL)[35, 36] to exclude those parents who had a child with behavioral or developmental problems.[37-40] The Turkish version of the CBCL is a self-administered parent-report questionnaire that contains 20 competence and 118 problem items. The competence scales are Activity, Social, and School. The problem scales are Aggression, Anxiety/Depression, Attention Problems, Delinquency, Social Problems, Somatization, Thought Problems, and Withdrawn. The 118 problem items describe a wide array of problems that are rated on a 3-point scale. Parents score each item 0, 1, or 2 (not true, somewhat true, or very true). The test–retest reliability of the Turkish form was 0.84 for Total Problems, and the internal consistency was adequate (Cronbach's alpha = 0.88).[35, 36] The children who score 0–66 are considered to be without behavioral or developmental problems. A history of psychotic disorders and history of psychiatric treatment were the exclusion criteria for all subjects. Control subjects had no family history for autism/AS and schizophrenia. As stated, the control group consisted of 100 parents (52 mothers and 48 fathers) of 84 TDC. The research protocol was approved by the Ege University Faculty of Medicine, Child and Adolescent Psychiatry Department Academic Committee. The participants were individually briefed on the aim of the study and signed an informed consent form in accordance with the Declaration of Helsinki.
Statistical analysis was conducted using SPSS 16.0 (SPSS, Chicago, IL, USA). Between-group differences were examined using univariate analysis of variance. For the AQ-TR and its subscales, two-way anova (group and gender) was performed because previous evidence suggest that there might be gender differences for the AQ-TR. If a tendency for a gender–proband diagnosis interaction effect was observed, pairwise t-tests were also conducted. Pearson correlation test was used to determine the relationship between variables. We used multiple regression analysis to explore the predictors of the AQ-TR total and five subscale scores. We considered P < 0.05 to be statistically significant.