- Email: [email protected]
Who benefits from early intervention in autism spectrum disorders?
Research in Autism Spectrum Disorders 5 (2011) 345–350
Contents lists available at ScienceDirect
Research in Autism Spectrum Disorders Journal homepage: http://ees.elsevier.com/RASD/default.asp
Who benefits from early intervention in autism spectrum disorders? Esther Ben Itzchak a,b,*, Ditza A. Zachor b,c a
Department of Communication Disorders, Ariel University Center, Israel The Autism Center, Assaf Harofeh Medical Center, Israel c Tel Aviv University, Israel b
A R T I C L E I N F O
A B S T R A C T
Article history: Received 21 April 2010 Accepted 23 April 2010
Research in autism spectrum disorders (ASD) described individual differences in response to intervention. This study explored child and parental characteristics at baseline that predict outcomes in adaptive skills and acquisition of cognitive gains. Seventy-eight children aged 15–35 months diagnosed with ASD by standardized diagnostic tools were included. Evaluations of verbal and non-verbal abilities, adaptive skills and autism severity were obtained at pre-intervention (T1) and after one year of intervention (T2). At T2, children improved significantly in their verbal ability and the severity of autism symptoms was reduced. Outcome in adaptive skills was best predicted by baseline verbal ability and maternal age. Better verbal ability especially in those with severe autism symptoms, and older maternal age predicted better adaptive skills outcome. T1 autism severity, child’s age and maternal age and educational attainment best predicted cognitive gains. Less severe autism symptoms, younger child’s age at start of intervention, older maternal age and higher maternal education predicted greater cognitive gains with intervention. The study suggests biological factors including age, language abilities and autism severity and environmental factors including maternal age and education, impact the ability to benefit from early intervention in ASD. ß 2011 Elsevier Ltd. All rights reserved.
Keywords: Autism spectrum disorders Early intervention Predictors Outcome Adaptive skills Cognitive abilities Maternal age Maternal education
1. Introduction An important and unresolved question in autism intervention research is what specific child and family factors might be related to treatment outcome. Research in autism spectrum disorders (ASD) described individual differences in response to intervention. As ASD is a complex disorder that affects a myriad of developmental domains, post-intervention progress should be evaluated using various outcome measures (Matson, 2007). Those outcomes might be differentially affected by several preintervention variables. Previous studies looked at variables and predictors that might affect outcome in cognitive ability, adaptive skills, language acquisition and improvement in autism symptoms. Eaves and Ho (2004) reported that autism severity at baseline predicted overall cognitive ability outcome. Progress in adaptive socialization skills was predicted by the baseline ASD diagnosis and progress in adaptive communication skills by the baseline language ability (Szatmari, Bryson, Boyle, Streiner, & Duku, 2003). Specific aspects of neurocognitive abilities appeared to be good predictors of outcome in both communication and socializations adaptive skills (Munson, Faja, Meltzoff, Abbott, & Dawson, 2008). Most of the studies focused on verbal abilities outcome. Baseline cognitive ability (Ben Itzchak & Zachor, 2007; Gabriels, Hill, Pierce, Rogers, & Wehner, 2001), non-verbal cognitive level at age 2 (Eaves & Ho, 2004; Thurm, Lord, Lee, & Newschaffer, 2007), baseline social deficits
* Corresponding author at: Department of Communication Disorders, Ariel University Center, Ariel 40700, Israel. Tel.: +972 39066353; fax: +972 39066624. E-mail address: [email protected] (E. Ben Itzchak). 1750-9467/$ – see front matter ß 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.rasd.2010.04.018
346
E. Ben Itzchak, D.A. Zachor / Research in Autism Spectrum Disorders 5 (2011) 345–350
(Ben Itzchak & Zachor, 2007; Charman et al., 2003), communication adaptive skills at age 3 (Thurm et al., 2007) and motor imitation ability (Stone & Yoder, 2001) were found to be good predictors of verbal skills outcome. Change in autism diagnostic classification was predicted by baseline cognitive ability and autism symptoms severity (Turner & Stone, 2007; Zachor, Ben Itzchak, Rabinovitch, & Lahat, 2007), receptive and expressive language scores difference (Ben Itzchak & Zachor, 2009; Chawarska, Klin, Paul, Macari, & Volkmar, 2009), and by baseline non-verbal IQ level (Szatmari et al., 2003). Since ASD is now diagnosed early before the age of 2 years, understanding of the symptoms expression and the developmental trajectory with intervention is highly important for clinical practice and for research. The use of standardized diagnostic tools for ASD, specifically the recent changes in the autism Diagnostic Observation Schedule (ADOS) that specifies algorithms that include all the behavioral symptoms in ASD for young verbal and non-verbal children allow for better differentiation between different clinical profiles at this young age (Gotham et al., 2008; Gotham, Pickles, & Lord, 2009). In addition, the new calibrated autism severity metric measure based on ADOS raw totals offers a method of quantifying ASD severity with relative independence from individual characteristics such as age and verbal IQ (Gotham et al., 2009). Although previous research has examined cognitive outcome in ASD, only a paucity of studies focused on predictors of cognitive and adaptive skills outcomes. It is important to expand our understanding of who benefits from early intervention in these developmental realms. In the current research we focus on identifying child and parental characteristics at baseline that might predict acquisition of better cognitive gains and outcomes in adaptive skills. 2. Methods Participants: 78 (71 boys and 7 girls) aged 15–35 months were included in the current study. The diagnostic process included a clinical evaluation by a neurodevelopmental pediatrician, and the use of two standardized autism diagnostic tools, the Autism Diagnosis Interview-Revised (ADI-R) (Lord, Rutter, & LeCouteur, 1994) and the Autism Diagnostic Observation Schedule (ADOS) (Lord, Rutter, DiLavore, & Risi, 1999). All the professionals involved in the diagnosis process established reliability as required. All the participants met criteria for autism based on DSM-IV (APA, 1994) and the cut-off points on the ADI-R and the ADOS. After receiving a diagnosis of autism all the children were referred to either an applied behavior analysis (ABA) or eclectic (integration of several treatment approaches) center-based intervention programs. A detailed and comprehensive description of the treatment approaches was provided in our previous intervention study (Zachor & Ben Itzchak, 2010). 3. Measures 3.1. Autism Diagnostic Interview-Revised (ADI-R) A semi-structured interview administered to parents was designed to make a diagnosis of autism according to both DSMIV (Lord et al., 1994) criteria. 3.2. Autism Diagnosis Observation Schedule (ADOS) A semi-structured, interactive schedule designed to assess social and communicative functioning. The new ADOS diagnostic algorithm that classifies children into categories of autism, ASD or non-spectrum was used (Gotham et al., 2008; Lord et al., 1999). For the severity of ASD the ADOS standardized measure of severity was used (Gotham et al., 2009). 3.3. Vineland Adaptive Behavior Scales The test assesses functioning in four adaptive skills domains: Communication, Daily Living skills, Socialization and Motor skills (Sparrow, Balla, & Cicchetti, 1984). Vineland composite standard scores were used for the statistical analyses. 3.4. Mullen Scales of Early Learning The test evaluates cognitive abilities in visual reception, fine motor, expressive language and language comprehension domains. We defined a non-verbal cognitive measure composed of visual reception plus fine motor standard scores [both domains were highly correlated (r = .63, p < .001)], and a verbal measure composed of expressive plus receptive language standard scores [both domains were highly correlated (r = .70, p < .001)] (Mullen, 1995). 4. Procedure Children underwent comprehensive evaluations at pre-intervention (T1) and after one year of intervention (T2). Of the 78 children who completed the ADOS at T1, 71 had the Vineland and the Mullen. At T2, 77 children completed the ADOS, of whom 75 had the Vineland and 69 the Mullen. Data on paternal and maternal ages (M = 36.5 years, SD = 6.1; M = 33.2 years, SD = 5.1) and educational attainment (Paternal M = 14.6, SD = 2.9; Maternal M = 14.6, SD = 2.4)
E. Ben Itzchak, D.A. Zachor / Research in Autism Spectrum Disorders 5 (2011) 345–350
347
were obtained from the medical chart of each participant. Informed consent was obtained from all parents for use of data from their child’s chart in accordance with the institutional Helsinki committee at Assaf Harofeh Medical Center. 5. Results In our previous study, we compared outcomes in cognitive, language and adaptive skills and in change in autism severity of two intervention groups, one based on behavioral principles (ABA) and the second received integration of several treatment approaches (eclectic). The results of this comparison revealed no significant differences between the two intervention approaches. Therefore, in the current study we treated both intervention groups as one group in all the analyses. We examined treatment outcome after one year of intervention in adaptive skills (Vineland) and in autism severity (ADOS severity scores) using two ANOVAs, and treatment outcome in verbal and non-verbal skills (MSEL) using MANOVA with repeated measures over time. The analyses yielded significant time-effect for outcome in verbal skills (MSEL) and in autism severity (ADOS severity scores) (Table 1). After one year of intervention children improved significantly in their verbal ability and their autism severity was reduced. 5.1. Predictors of intervention outcome The objective of this study was to identify child and parental variables at T1 that might explain the variance in intervention outcome at T2. For this purpose two hierarchical multiple regression analyses were conducted. The first analysis examined predictors for outcome in adaptive skills and used Vineland composite score at T2 as the dependent variable. T1 verbal MSEL and non-verbal T1 MSEL scores were entered in the first step, T1 ADOS severity scale scores in the second step and maternal education and maternal age in the third step. Since maternal and paternal education were highly correlated (r = .632, p < .001; r = .778, p < .01, respectively) only the maternal education was entered to the regression analysis. In addition interactions of the above variables were entered. As shown in Table 2 the child’s cognitive ability (MSEL) at T1 in the first step explained 40% of the variance in the Vineland composite score at T2, meaning that higher cognitive level at baseline resulted in better adaptive skills at T2. Only the verbal cognitive ability contributes significantly to the adaptive skills scores after one year of intervention. Autism severity at the second step and child’s age at the third step did not contribute significantly to the total explained variance. Maternal education and age at the forth step contributed 4% to the explained variance, however only maternal age contribution was significant. The older the mother was, the better the adaptive skills outcome was achieved. Interaction of T1 verbal MSEL and ADOS autism severity at the fifth stage was significant and contributed 5% to the explained variance. Altogether, the model explained 49% of the variance. To explain the significant interaction that was found in the regression analysis, the sample was divided into two groups: a high autism severity group (autism severity scores = 8–10) and a low autism severity group (scores = 6–7). In each group correlation between the T2 Vineland scores and the T1 verbal MSEL scores was performed. The correlation was significant for the high autism severity group (n = 52, r = .672, p < .001) but not for the low autism severity group (n = 15, r = .075, p > .05). Only for children with severe autism symptoms, higher verbal abilities at pre-intervention time correlated with better adaptive skills outcome. In the second analysis we looked for predictors of better cognitive gains with intervention. Therefore, we calculated the difference between T2 and T1 MSEL and used it as the dependent variable. For this analysis we used the following independent variables: in the first step the T1 ADOS severity score, in the second step the child’s age, and in the third step the maternal education and maternal age. As shown in Table 3, T1 autism severity explained 13% of the sample variance, meaning that cognitive gains were more pronounced in children with less severe autism symptoms at pre-intervention time. Child’s age in the second step contributed 3% to the explained variance. The additional explained variance was almost significant (p < .1) at the second step and reached significance at the third step (p < .05), meaning that younger children had better chance to gain from intervention. Maternal education and maternal age in the third stage contributed 14% to the
Table 1 Verbal and non-verbal MSEL scores, Vineland adaptive composite scores. T1 Verbal MSEL (n = 63) Nonverbal MSEL (n = 63) Vineland (n = 65) ADOS new algorithm (n = 78) ADOS severity score (n = 78) a * ***
df = 77–62. p < .05 p < .001.
60.9 73.9 67.4 20.5 8.4
Fa
T2 (24.4) (23.7) (6.4) (4.4) (2.0)
75.0 75.5 68.9 17.9 7.8
(27.0) (29.2) (13.0) (5.0) (1.9)
422.3*** .1 1.3 25.9*** 5.8*
.406 .001 .020 .252 .070
E. Ben Itzchak, D.A. Zachor / Research in Autism Spectrum Disorders 5 (2011) 345–350
348
Table 2 Summary of hierarchical regression analyses for variables predicting T2 Vineland composite scores. Variable
B
SE. B
***
Step 1 T1 MSEL-verbal T1 MSEL-nonverbal
.32 .13
.09 .10
.50*** .19
Step 2 T1 MSEL-verbal T1 MSEL-nonverbal T1 autism severity
.30 .12 1.38
.09 .10 .95
.46** .17 .16
Step 3 T1 MSEL-verbal T1 MSEL-nonverbal T1 autism severity Age
.31 .09 1.36 .30
.10 .10 .95 .42
.48** .13 .15 .07
Step 4 T1 MSEL-verbal T1 MSEL-nonverbal T1 autism severity Age Maternal education Maternal age
.33 .08 1.19 .38 .22 .74
.09 .10 .40 .39 .62 .30
.33* .17 .16 .09 .05 .24*
.41 .09 1.35 .41 .42 .86 4.29
.10 .10 .90 .39 .65 .29 1.79
.10*** .13 .15 .10 .06 .28** .26*
Step 5 T1 MSEL-verbal T1 MSEL-nonverbal T1 autism severity Age Maternal education Maternal age Autism severity X T1 MSEL-verbal * ** ***
Adjusted R2
b
DR2
.40
.42***
.41***
.02
.41***
.00
.45***
.05
.49***
.05*
p < .05. p < .01. p < .001.
Table 3 Summary of hierarchical regression analyses for variables predicting MSEL ECL score difference (T2 Variable
B
SE. B
T1). Adjusted R2
b
Step 1 T1 autism severity
2.89
.92
.38**
.13
Step 2 T1 autism severity Child’s age
2.76 .66
.91 .39
.36** .20
Step 3 T1 autism severity Child’s age Maternal education Maternal age
2.92 .79 1.33 1.02
.86 .36 .64 .29
.38*** .24* .24* .39***
**
DR2 .14**
.16**
.04
.29***
.16**
*
p < .05. p < .01. *** p < .001. **
explained variance. Children who had older and more educated mothers had better cognitive gains with intervention. Overall, the model explained 29% of the variance in this analysis. 6. Discussion The study aim was to identify child and parental variables prior to implementation of early intervention that predict outcome and best response to intervention. Outcome in adaptive skills after one year of intervention was best predicted by the baseline verbal ability. More specifically, only for children with severe autism symptoms at baseline, the verbal ability correlated with outcome. It is possible that in severe autism the verbal ability is a more crucial factor that enables the child to compensate for the significant social-communicative deficits with intervention. The predictive value of the baseline cognitive level has been noted in previous intervention studies, although intervention measures varied widely between
E. Ben Itzchak, D.A. Zachor / Research in Autism Spectrum Disorders 5 (2011) 345–350
349
studies (Gabriels et al., 2001; Harris & Handleman, 2000; Magiati, Charman, & Howlin, 2007; Turner & Stone, 2007). Several studies noted that the severity of the child’s baseline social deficits was related to a change in autism diagnostic category and to language outcome, but did not examine progress in adaptive skills (Ben Itzchak & Zachor, 2007; Turner & Stone, 2007). In the current study, autism severity at baseline did not predict outcome, although a high correlation was found between autism severity and outcome in adaptive skills. It is possible that the high correlation between cognitive scores and the ADOS measures masked the contribution of last measure. Of the examined maternal characteristics only maternal age predicted outcome in adaptive skills. Advanced maternal age resulted in better outcome. It is possible that more experienced and mature mothers may enable better implementation of their children’s potential in daily living skills. Several variables predicted gains in cognitive ability in response to intervention: autism severity, child’s age, and maternal age and education. Autism severity at baseline had the best predictive value, children with less severe autism symptoms showed better gains in cognitive abilities with intervention. Only a few studies looked at the predictive value of baseline autism severity on developmental gains following early intervention. The current study used autism severity as a predictor as only recently can this measure be quantified by standardized diagnostic tools. Magiati et al. (2007) reported that autism severity measured by the ADI total raw scores at baseline predicted ‘‘total progress rank’’ scores (magnitude of gains in cognitive, language, adaptive and autism severity domains). Turner and Stone (2007) found that the ADOS algorithm scores at baseline are associated with change in autism diagnostic category after two years. Charman and colleagues reported that non-verbal IQ, language, and autism symptom severity (ADI-R) at age 3 but not at age 2 were associated at age 7 with the same domains of functioning and with measures of symptom severity and adaptive behavior scores (Vineland) (Charman et al., 2005). For the first time, we found in our study that autism severity measured by the ADOS severity scale is a good predictor of cognitive gains with intervention. Child’s age was another significant predictor of cognitive gains in response to intervention, suggesting that the younger the child at the start of early intervention, the better the response in the cognitive domain. This finding is in accordance with previous studies that found that it is important to initiate intervention at a very young age when brain plasticity enables changes in central nervous system circuits (Harris & Handleman, 2000; Lovaas, 1987; Luiselli, Cannon, Ellis, & Sisson, 2000; Turner & Stone, 2007). Our study population consisted of very young children and the age range was quite narrow (15–35 months), therefore our finding strongly emphasizes the importance of early intervention in very young age in ASD. Maternal parameters of age and education predicted rate of cognitive gains in response to intervention. The older and more educated the mother was, the more accelerated was the cognitive gain with intervention. Other intervention studies in ASD found no effect of parental years of education and ages on progress with intervention (Magiati et al., 2007; Turner & Stone, 2007). Maternal education has long been known to be the single best predictor of child development outcome (DiPierto, 2000). The mechanisms through which maternal education affects the child’s IQ and other outcomes are not well known, although biological influences as well as experiential ones are likely (DiPietro, Costigan, Shupe, Pressman, & Johnson, 1998). Regarding the impact of maternal age on cognitive gains as mentioned before, it is possible that older mothers are more experienced in raising children. The impact of parental age on the outcome of young children with ASD has not been previously addressed. The current study suggests that both biological and environmental factors affect the response to intervention in very young children with ASD. Biological factors including age, language abilities and autism severity and environmental factors including maternal age and education, impact the ability to benefit from early intervention in ASD. There are several important clinical implications of these results. First, it is important to diagnose ASD and start intervention as early as possible. Second, the results offer the clinicians the ability to convey prognosis of future outcome based on construct of several major measures at baseline. This study encourages researchers to continue with investigation of other strong predictors of outcome in ASD. References American Psychiatric Association. (1994). Diagnostic and Statistical Manual of Mental Disorders DSM-IV. Ben Itzchak, E., & Zachor, D. A. (2007). The effects of intellectual functioning and autism severity on outcome of early behavioral intervention for children with autism. Research in Developmental Disabilities, 28, 287–303. Ben Itzchak, E., & Zachor, D. (2009). Change in autism classification with early intervention: Predictors and outcomes. Research in Developmental Disabilities, 3, 967–976. Charman, T., Baron-Cohen, S., Swettenham, J., Baird, G., Drew, A., & Cox, A. (2003). Predicting language outcome in infants with autism and pervasive developmental disorder. International Journal of Language and Communication Disorders, 38, 265–285. Charman, T., Taylor, E., Drew, A., Cockerill, H., Brown, J. A., & Baird, G. (2005). Outcome at 7 years of children diagnosed with autism at age 2: predictive validity of assessments conducted at 2 and 3 years of age and pattern of symptom change over time. Journal of Child Psychology and Psychiatry, 46, 500–513. Chawarska, K., Klin, A., Paul, R., Macari, S., & Volkmar, F. (2009). A prospective study of toddlers with ASD: short-term diagnostic and cognitive outcomes. Journal of Child Psychology and Psychiatry, 50, 1235–1245. DiPierto, J. A. (2000). Baby and the brain: Advances in child development. Annual Review of Public Health, 21, 455–471. DiPietro, J. A., Costigan, K. A., Shupe, A. K., Pressman, E. K., & Johnson, T. R. B. (1998). Fetal neurobehavioral development: Associations with socioeconomic class and fetal sex. Developmental Psychobiology, 33, 79–91. Eaves, L. C., & Ho, H. H. (2004). The very early identification of autism: outcome to age 4 1/2–5. Journal of Autism and Developmental Disorders, 34, 367–378. Gabriels, R. L., Hill, D. E., Pierce, R. A., Rogers, S. J., & Wehner, B. (2001). Predictors of treatment outcome in young children with autism: a retrospective study. Autism, 5, 407–429. Gotham, K., Pickles, A., & Lord, C. (2009). Standardizing ADOS scores for a measure of severity in autism spectrum disorders. Journal of Autism and Developmental Disorders, 39, 693–705.
350
E. Ben Itzchak, D.A. Zachor / Research in Autism Spectrum Disorders 5 (2011) 345–350
Gotham, K., Risi, S., Dawson, G., Tager-Flusberg, H., Joseph, R., Carter, A., et al. (2008). A replication of the Autism Diagnostic Observation Schedule (ADOS) revised algorithms. Journal of American Academy of Child & Adolescent Psychiatry, 47, 642–651. Harris, S. L., & Handleman, J. S. (2000). Age and I.Q. at intake as predictors of placement for young children with autism: A four- to six-year follows up. Journal of Autism and Developmental Disorders, 30, 137–142. Lord, C., Rutter, M., DiLavore, P. C., & Risi, S. (1999). Autism Diagnostic Observation Schedule-WPS (ADOS-WPS). Los Angeles, CA: Western Psychological Services. Lord, C., Rutter, M., & LeCouteur, A. (1994). Autism Diagnostic Interview Revised: A revised version of a diagnostic interview for caregivers of individuals with possible pervasive developmental disorder. Journal of Autism and Developmental Disorders, 24, 659–668. Lovaas, O. I. (1987). Behavioral treatment and normal educational and intellectual functioning in young autistic children. Journal of Consulting and Clinical Psychology, 55, 3–9. Luiselli, J. K., Cannon, B. O., Ellis, J. T., & Sisson, R. W. (2000). Home-based behavioral intervention for young children with autism/pervasive developmental disorder. Autism, 4, 427–437. Magiati, I., Charman, T., & Howlin, P. A. (2007). Two-year prospective follow-up of community-based early intensive behavioural intervention and specialist nursery provision for children with autism spectrum disorders. Journal of Child Psychology and Psychiatry, 4, 803–812. Matson, J. L. (2007). Determining treatment outcome in early intervention programs for autism spectrum disorders: A critical analysis of measurement issues in learning based interventions. Research in Developmental Disabilities, 28, 207–218. Mullen, E. (1995). The Mullen Scales of Early Learning. Circle Pines, MN: American Guidance Service. Munson, J., Faja, S., Meltzoff, A., Abbott, R., & Dawson, G. (2008). Neurocognitive predictors of social and communicative developmental trajectories in preschoolers with autism spectrum disorders. Journal of International Neuropsychology Society, 14, 956–966. Sparrow, S., Balla, D., & Cicchetti, D. (1984). Vineland Adaptive Behavior Scales. Circle Pines, MN: American Guidance Services. Stone, W. L., & Yoder, P. J. (2001). Predicting spoken language level in children with autism spectrum disorders. Autism, 5, 341–361. Szatmari, P., Bryson, S. E., Boyle, M. H., Streiner, D. L., & Duku, E. (2003). Predictors of outcome among high functioning children with autism and Asperger syndrome. Journal of Child Psychology and Psychiatry, 44, 520–528. Thurm, A., Lord, C., Lee, L. C., & Newschaffer, C. (2007). Predictors of language acquisition in preschool children with autism spectrum disorders. Journal of Autism and Developmental Disorders, 37, 1721–1734. Turner, L. M., & Stone, W. L. (2007). Variability in outcome for children with an ASD diagnosis at age 2. Journal of Child Psychology and Psychiatry, 48, 793–802. Zachor, D., & Ben Itzchak, E. (2010). Treatment approach, autism severity and intervention outcomes in young children. Research in Autism Spectrum Disorders, 4(3), 425–432. Zachor, D. A., Ben Itzchak, E., Rabinovitch, A., & Lahat, E. (2007). Change in autism core symptoms after intervention—a comparative study. Research in Autism Spectrum Disorders, 1, 304–317.
Contents lists available at ScienceDirect
Research in Autism Spectrum Disorders Journal homepage: http://ees.elsevier.com/RASD/default.asp
Who benefits from early intervention in autism spectrum disorders? Esther Ben Itzchak a,b,*, Ditza A. Zachor b,c a
Department of Communication Disorders, Ariel University Center, Israel The Autism Center, Assaf Harofeh Medical Center, Israel c Tel Aviv University, Israel b
A R T I C L E I N F O
A B S T R A C T
Article history: Received 21 April 2010 Accepted 23 April 2010
Research in autism spectrum disorders (ASD) described individual differences in response to intervention. This study explored child and parental characteristics at baseline that predict outcomes in adaptive skills and acquisition of cognitive gains. Seventy-eight children aged 15–35 months diagnosed with ASD by standardized diagnostic tools were included. Evaluations of verbal and non-verbal abilities, adaptive skills and autism severity were obtained at pre-intervention (T1) and after one year of intervention (T2). At T2, children improved significantly in their verbal ability and the severity of autism symptoms was reduced. Outcome in adaptive skills was best predicted by baseline verbal ability and maternal age. Better verbal ability especially in those with severe autism symptoms, and older maternal age predicted better adaptive skills outcome. T1 autism severity, child’s age and maternal age and educational attainment best predicted cognitive gains. Less severe autism symptoms, younger child’s age at start of intervention, older maternal age and higher maternal education predicted greater cognitive gains with intervention. The study suggests biological factors including age, language abilities and autism severity and environmental factors including maternal age and education, impact the ability to benefit from early intervention in ASD. ß 2011 Elsevier Ltd. All rights reserved.
Keywords: Autism spectrum disorders Early intervention Predictors Outcome Adaptive skills Cognitive abilities Maternal age Maternal education
1. Introduction An important and unresolved question in autism intervention research is what specific child and family factors might be related to treatment outcome. Research in autism spectrum disorders (ASD) described individual differences in response to intervention. As ASD is a complex disorder that affects a myriad of developmental domains, post-intervention progress should be evaluated using various outcome measures (Matson, 2007). Those outcomes might be differentially affected by several preintervention variables. Previous studies looked at variables and predictors that might affect outcome in cognitive ability, adaptive skills, language acquisition and improvement in autism symptoms. Eaves and Ho (2004) reported that autism severity at baseline predicted overall cognitive ability outcome. Progress in adaptive socialization skills was predicted by the baseline ASD diagnosis and progress in adaptive communication skills by the baseline language ability (Szatmari, Bryson, Boyle, Streiner, & Duku, 2003). Specific aspects of neurocognitive abilities appeared to be good predictors of outcome in both communication and socializations adaptive skills (Munson, Faja, Meltzoff, Abbott, & Dawson, 2008). Most of the studies focused on verbal abilities outcome. Baseline cognitive ability (Ben Itzchak & Zachor, 2007; Gabriels, Hill, Pierce, Rogers, & Wehner, 2001), non-verbal cognitive level at age 2 (Eaves & Ho, 2004; Thurm, Lord, Lee, & Newschaffer, 2007), baseline social deficits
* Corresponding author at: Department of Communication Disorders, Ariel University Center, Ariel 40700, Israel. Tel.: +972 39066353; fax: +972 39066624. E-mail address: [email protected] (E. Ben Itzchak). 1750-9467/$ – see front matter ß 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.rasd.2010.04.018
346
E. Ben Itzchak, D.A. Zachor / Research in Autism Spectrum Disorders 5 (2011) 345–350
(Ben Itzchak & Zachor, 2007; Charman et al., 2003), communication adaptive skills at age 3 (Thurm et al., 2007) and motor imitation ability (Stone & Yoder, 2001) were found to be good predictors of verbal skills outcome. Change in autism diagnostic classification was predicted by baseline cognitive ability and autism symptoms severity (Turner & Stone, 2007; Zachor, Ben Itzchak, Rabinovitch, & Lahat, 2007), receptive and expressive language scores difference (Ben Itzchak & Zachor, 2009; Chawarska, Klin, Paul, Macari, & Volkmar, 2009), and by baseline non-verbal IQ level (Szatmari et al., 2003). Since ASD is now diagnosed early before the age of 2 years, understanding of the symptoms expression and the developmental trajectory with intervention is highly important for clinical practice and for research. The use of standardized diagnostic tools for ASD, specifically the recent changes in the autism Diagnostic Observation Schedule (ADOS) that specifies algorithms that include all the behavioral symptoms in ASD for young verbal and non-verbal children allow for better differentiation between different clinical profiles at this young age (Gotham et al., 2008; Gotham, Pickles, & Lord, 2009). In addition, the new calibrated autism severity metric measure based on ADOS raw totals offers a method of quantifying ASD severity with relative independence from individual characteristics such as age and verbal IQ (Gotham et al., 2009). Although previous research has examined cognitive outcome in ASD, only a paucity of studies focused on predictors of cognitive and adaptive skills outcomes. It is important to expand our understanding of who benefits from early intervention in these developmental realms. In the current research we focus on identifying child and parental characteristics at baseline that might predict acquisition of better cognitive gains and outcomes in adaptive skills. 2. Methods Participants: 78 (71 boys and 7 girls) aged 15–35 months were included in the current study. The diagnostic process included a clinical evaluation by a neurodevelopmental pediatrician, and the use of two standardized autism diagnostic tools, the Autism Diagnosis Interview-Revised (ADI-R) (Lord, Rutter, & LeCouteur, 1994) and the Autism Diagnostic Observation Schedule (ADOS) (Lord, Rutter, DiLavore, & Risi, 1999). All the professionals involved in the diagnosis process established reliability as required. All the participants met criteria for autism based on DSM-IV (APA, 1994) and the cut-off points on the ADI-R and the ADOS. After receiving a diagnosis of autism all the children were referred to either an applied behavior analysis (ABA) or eclectic (integration of several treatment approaches) center-based intervention programs. A detailed and comprehensive description of the treatment approaches was provided in our previous intervention study (Zachor & Ben Itzchak, 2010). 3. Measures 3.1. Autism Diagnostic Interview-Revised (ADI-R) A semi-structured interview administered to parents was designed to make a diagnosis of autism according to both DSMIV (Lord et al., 1994) criteria. 3.2. Autism Diagnosis Observation Schedule (ADOS) A semi-structured, interactive schedule designed to assess social and communicative functioning. The new ADOS diagnostic algorithm that classifies children into categories of autism, ASD or non-spectrum was used (Gotham et al., 2008; Lord et al., 1999). For the severity of ASD the ADOS standardized measure of severity was used (Gotham et al., 2009). 3.3. Vineland Adaptive Behavior Scales The test assesses functioning in four adaptive skills domains: Communication, Daily Living skills, Socialization and Motor skills (Sparrow, Balla, & Cicchetti, 1984). Vineland composite standard scores were used for the statistical analyses. 3.4. Mullen Scales of Early Learning The test evaluates cognitive abilities in visual reception, fine motor, expressive language and language comprehension domains. We defined a non-verbal cognitive measure composed of visual reception plus fine motor standard scores [both domains were highly correlated (r = .63, p < .001)], and a verbal measure composed of expressive plus receptive language standard scores [both domains were highly correlated (r = .70, p < .001)] (Mullen, 1995). 4. Procedure Children underwent comprehensive evaluations at pre-intervention (T1) and after one year of intervention (T2). Of the 78 children who completed the ADOS at T1, 71 had the Vineland and the Mullen. At T2, 77 children completed the ADOS, of whom 75 had the Vineland and 69 the Mullen. Data on paternal and maternal ages (M = 36.5 years, SD = 6.1; M = 33.2 years, SD = 5.1) and educational attainment (Paternal M = 14.6, SD = 2.9; Maternal M = 14.6, SD = 2.4)
E. Ben Itzchak, D.A. Zachor / Research in Autism Spectrum Disorders 5 (2011) 345–350
347
were obtained from the medical chart of each participant. Informed consent was obtained from all parents for use of data from their child’s chart in accordance with the institutional Helsinki committee at Assaf Harofeh Medical Center. 5. Results In our previous study, we compared outcomes in cognitive, language and adaptive skills and in change in autism severity of two intervention groups, one based on behavioral principles (ABA) and the second received integration of several treatment approaches (eclectic). The results of this comparison revealed no significant differences between the two intervention approaches. Therefore, in the current study we treated both intervention groups as one group in all the analyses. We examined treatment outcome after one year of intervention in adaptive skills (Vineland) and in autism severity (ADOS severity scores) using two ANOVAs, and treatment outcome in verbal and non-verbal skills (MSEL) using MANOVA with repeated measures over time. The analyses yielded significant time-effect for outcome in verbal skills (MSEL) and in autism severity (ADOS severity scores) (Table 1). After one year of intervention children improved significantly in their verbal ability and their autism severity was reduced. 5.1. Predictors of intervention outcome The objective of this study was to identify child and parental variables at T1 that might explain the variance in intervention outcome at T2. For this purpose two hierarchical multiple regression analyses were conducted. The first analysis examined predictors for outcome in adaptive skills and used Vineland composite score at T2 as the dependent variable. T1 verbal MSEL and non-verbal T1 MSEL scores were entered in the first step, T1 ADOS severity scale scores in the second step and maternal education and maternal age in the third step. Since maternal and paternal education were highly correlated (r = .632, p < .001; r = .778, p < .01, respectively) only the maternal education was entered to the regression analysis. In addition interactions of the above variables were entered. As shown in Table 2 the child’s cognitive ability (MSEL) at T1 in the first step explained 40% of the variance in the Vineland composite score at T2, meaning that higher cognitive level at baseline resulted in better adaptive skills at T2. Only the verbal cognitive ability contributes significantly to the adaptive skills scores after one year of intervention. Autism severity at the second step and child’s age at the third step did not contribute significantly to the total explained variance. Maternal education and age at the forth step contributed 4% to the explained variance, however only maternal age contribution was significant. The older the mother was, the better the adaptive skills outcome was achieved. Interaction of T1 verbal MSEL and ADOS autism severity at the fifth stage was significant and contributed 5% to the explained variance. Altogether, the model explained 49% of the variance. To explain the significant interaction that was found in the regression analysis, the sample was divided into two groups: a high autism severity group (autism severity scores = 8–10) and a low autism severity group (scores = 6–7). In each group correlation between the T2 Vineland scores and the T1 verbal MSEL scores was performed. The correlation was significant for the high autism severity group (n = 52, r = .672, p < .001) but not for the low autism severity group (n = 15, r = .075, p > .05). Only for children with severe autism symptoms, higher verbal abilities at pre-intervention time correlated with better adaptive skills outcome. In the second analysis we looked for predictors of better cognitive gains with intervention. Therefore, we calculated the difference between T2 and T1 MSEL and used it as the dependent variable. For this analysis we used the following independent variables: in the first step the T1 ADOS severity score, in the second step the child’s age, and in the third step the maternal education and maternal age. As shown in Table 3, T1 autism severity explained 13% of the sample variance, meaning that cognitive gains were more pronounced in children with less severe autism symptoms at pre-intervention time. Child’s age in the second step contributed 3% to the explained variance. The additional explained variance was almost significant (p < .1) at the second step and reached significance at the third step (p < .05), meaning that younger children had better chance to gain from intervention. Maternal education and maternal age in the third stage contributed 14% to the
Table 1 Verbal and non-verbal MSEL scores, Vineland adaptive composite scores. T1 Verbal MSEL (n = 63) Nonverbal MSEL (n = 63) Vineland (n = 65) ADOS new algorithm (n = 78) ADOS severity score (n = 78) a * ***
df = 77–62. p < .05 p < .001.
60.9 73.9 67.4 20.5 8.4
Fa
T2 (24.4) (23.7) (6.4) (4.4) (2.0)
75.0 75.5 68.9 17.9 7.8
(27.0) (29.2) (13.0) (5.0) (1.9)
422.3*** .1 1.3 25.9*** 5.8*
.406 .001 .020 .252 .070
E. Ben Itzchak, D.A. Zachor / Research in Autism Spectrum Disorders 5 (2011) 345–350
348
Table 2 Summary of hierarchical regression analyses for variables predicting T2 Vineland composite scores. Variable
B
SE. B
***
Step 1 T1 MSEL-verbal T1 MSEL-nonverbal
.32 .13
.09 .10
.50*** .19
Step 2 T1 MSEL-verbal T1 MSEL-nonverbal T1 autism severity
.30 .12 1.38
.09 .10 .95
.46** .17 .16
Step 3 T1 MSEL-verbal T1 MSEL-nonverbal T1 autism severity Age
.31 .09 1.36 .30
.10 .10 .95 .42
.48** .13 .15 .07
Step 4 T1 MSEL-verbal T1 MSEL-nonverbal T1 autism severity Age Maternal education Maternal age
.33 .08 1.19 .38 .22 .74
.09 .10 .40 .39 .62 .30
.33* .17 .16 .09 .05 .24*
.41 .09 1.35 .41 .42 .86 4.29
.10 .10 .90 .39 .65 .29 1.79
.10*** .13 .15 .10 .06 .28** .26*
Step 5 T1 MSEL-verbal T1 MSEL-nonverbal T1 autism severity Age Maternal education Maternal age Autism severity X T1 MSEL-verbal * ** ***
Adjusted R2
b
DR2
.40
.42***
.41***
.02
.41***
.00
.45***
.05
.49***
.05*
p < .05. p < .01. p < .001.
Table 3 Summary of hierarchical regression analyses for variables predicting MSEL ECL score difference (T2 Variable
B
SE. B
T1). Adjusted R2
b
Step 1 T1 autism severity
2.89
.92
.38**
.13
Step 2 T1 autism severity Child’s age
2.76 .66
.91 .39
.36** .20
Step 3 T1 autism severity Child’s age Maternal education Maternal age
2.92 .79 1.33 1.02
.86 .36 .64 .29
.38*** .24* .24* .39***
**
DR2 .14**
.16**
.04
.29***
.16**
*
p < .05. p < .01. *** p < .001. **
explained variance. Children who had older and more educated mothers had better cognitive gains with intervention. Overall, the model explained 29% of the variance in this analysis. 6. Discussion The study aim was to identify child and parental variables prior to implementation of early intervention that predict outcome and best response to intervention. Outcome in adaptive skills after one year of intervention was best predicted by the baseline verbal ability. More specifically, only for children with severe autism symptoms at baseline, the verbal ability correlated with outcome. It is possible that in severe autism the verbal ability is a more crucial factor that enables the child to compensate for the significant social-communicative deficits with intervention. The predictive value of the baseline cognitive level has been noted in previous intervention studies, although intervention measures varied widely between
E. Ben Itzchak, D.A. Zachor / Research in Autism Spectrum Disorders 5 (2011) 345–350
349
studies (Gabriels et al., 2001; Harris & Handleman, 2000; Magiati, Charman, & Howlin, 2007; Turner & Stone, 2007). Several studies noted that the severity of the child’s baseline social deficits was related to a change in autism diagnostic category and to language outcome, but did not examine progress in adaptive skills (Ben Itzchak & Zachor, 2007; Turner & Stone, 2007). In the current study, autism severity at baseline did not predict outcome, although a high correlation was found between autism severity and outcome in adaptive skills. It is possible that the high correlation between cognitive scores and the ADOS measures masked the contribution of last measure. Of the examined maternal characteristics only maternal age predicted outcome in adaptive skills. Advanced maternal age resulted in better outcome. It is possible that more experienced and mature mothers may enable better implementation of their children’s potential in daily living skills. Several variables predicted gains in cognitive ability in response to intervention: autism severity, child’s age, and maternal age and education. Autism severity at baseline had the best predictive value, children with less severe autism symptoms showed better gains in cognitive abilities with intervention. Only a few studies looked at the predictive value of baseline autism severity on developmental gains following early intervention. The current study used autism severity as a predictor as only recently can this measure be quantified by standardized diagnostic tools. Magiati et al. (2007) reported that autism severity measured by the ADI total raw scores at baseline predicted ‘‘total progress rank’’ scores (magnitude of gains in cognitive, language, adaptive and autism severity domains). Turner and Stone (2007) found that the ADOS algorithm scores at baseline are associated with change in autism diagnostic category after two years. Charman and colleagues reported that non-verbal IQ, language, and autism symptom severity (ADI-R) at age 3 but not at age 2 were associated at age 7 with the same domains of functioning and with measures of symptom severity and adaptive behavior scores (Vineland) (Charman et al., 2005). For the first time, we found in our study that autism severity measured by the ADOS severity scale is a good predictor of cognitive gains with intervention. Child’s age was another significant predictor of cognitive gains in response to intervention, suggesting that the younger the child at the start of early intervention, the better the response in the cognitive domain. This finding is in accordance with previous studies that found that it is important to initiate intervention at a very young age when brain plasticity enables changes in central nervous system circuits (Harris & Handleman, 2000; Lovaas, 1987; Luiselli, Cannon, Ellis, & Sisson, 2000; Turner & Stone, 2007). Our study population consisted of very young children and the age range was quite narrow (15–35 months), therefore our finding strongly emphasizes the importance of early intervention in very young age in ASD. Maternal parameters of age and education predicted rate of cognitive gains in response to intervention. The older and more educated the mother was, the more accelerated was the cognitive gain with intervention. Other intervention studies in ASD found no effect of parental years of education and ages on progress with intervention (Magiati et al., 2007; Turner & Stone, 2007). Maternal education has long been known to be the single best predictor of child development outcome (DiPierto, 2000). The mechanisms through which maternal education affects the child’s IQ and other outcomes are not well known, although biological influences as well as experiential ones are likely (DiPietro, Costigan, Shupe, Pressman, & Johnson, 1998). Regarding the impact of maternal age on cognitive gains as mentioned before, it is possible that older mothers are more experienced in raising children. The impact of parental age on the outcome of young children with ASD has not been previously addressed. The current study suggests that both biological and environmental factors affect the response to intervention in very young children with ASD. Biological factors including age, language abilities and autism severity and environmental factors including maternal age and education, impact the ability to benefit from early intervention in ASD. There are several important clinical implications of these results. First, it is important to diagnose ASD and start intervention as early as possible. Second, the results offer the clinicians the ability to convey prognosis of future outcome based on construct of several major measures at baseline. This study encourages researchers to continue with investigation of other strong predictors of outcome in ASD. References American Psychiatric Association. (1994). Diagnostic and Statistical Manual of Mental Disorders DSM-IV. Ben Itzchak, E., & Zachor, D. A. (2007). The effects of intellectual functioning and autism severity on outcome of early behavioral intervention for children with autism. Research in Developmental Disabilities, 28, 287–303. Ben Itzchak, E., & Zachor, D. (2009). Change in autism classification with early intervention: Predictors and outcomes. Research in Developmental Disabilities, 3, 967–976. Charman, T., Baron-Cohen, S., Swettenham, J., Baird, G., Drew, A., & Cox, A. (2003). Predicting language outcome in infants with autism and pervasive developmental disorder. International Journal of Language and Communication Disorders, 38, 265–285. Charman, T., Taylor, E., Drew, A., Cockerill, H., Brown, J. A., & Baird, G. (2005). Outcome at 7 years of children diagnosed with autism at age 2: predictive validity of assessments conducted at 2 and 3 years of age and pattern of symptom change over time. Journal of Child Psychology and Psychiatry, 46, 500–513. Chawarska, K., Klin, A., Paul, R., Macari, S., & Volkmar, F. (2009). A prospective study of toddlers with ASD: short-term diagnostic and cognitive outcomes. Journal of Child Psychology and Psychiatry, 50, 1235–1245. DiPierto, J. A. (2000). Baby and the brain: Advances in child development. Annual Review of Public Health, 21, 455–471. DiPietro, J. A., Costigan, K. A., Shupe, A. K., Pressman, E. K., & Johnson, T. R. B. (1998). Fetal neurobehavioral development: Associations with socioeconomic class and fetal sex. Developmental Psychobiology, 33, 79–91. Eaves, L. C., & Ho, H. H. (2004). The very early identification of autism: outcome to age 4 1/2–5. Journal of Autism and Developmental Disorders, 34, 367–378. Gabriels, R. L., Hill, D. E., Pierce, R. A., Rogers, S. J., & Wehner, B. (2001). Predictors of treatment outcome in young children with autism: a retrospective study. Autism, 5, 407–429. Gotham, K., Pickles, A., & Lord, C. (2009). Standardizing ADOS scores for a measure of severity in autism spectrum disorders. Journal of Autism and Developmental Disorders, 39, 693–705.
350
E. Ben Itzchak, D.A. Zachor / Research in Autism Spectrum Disorders 5 (2011) 345–350
Gotham, K., Risi, S., Dawson, G., Tager-Flusberg, H., Joseph, R., Carter, A., et al. (2008). A replication of the Autism Diagnostic Observation Schedule (ADOS) revised algorithms. Journal of American Academy of Child & Adolescent Psychiatry, 47, 642–651. Harris, S. L., & Handleman, J. S. (2000). Age and I.Q. at intake as predictors of placement for young children with autism: A four- to six-year follows up. Journal of Autism and Developmental Disorders, 30, 137–142. Lord, C., Rutter, M., DiLavore, P. C., & Risi, S. (1999). Autism Diagnostic Observation Schedule-WPS (ADOS-WPS). Los Angeles, CA: Western Psychological Services. Lord, C., Rutter, M., & LeCouteur, A. (1994). Autism Diagnostic Interview Revised: A revised version of a diagnostic interview for caregivers of individuals with possible pervasive developmental disorder. Journal of Autism and Developmental Disorders, 24, 659–668. Lovaas, O. I. (1987). Behavioral treatment and normal educational and intellectual functioning in young autistic children. Journal of Consulting and Clinical Psychology, 55, 3–9. Luiselli, J. K., Cannon, B. O., Ellis, J. T., & Sisson, R. W. (2000). Home-based behavioral intervention for young children with autism/pervasive developmental disorder. Autism, 4, 427–437. Magiati, I., Charman, T., & Howlin, P. A. (2007). Two-year prospective follow-up of community-based early intensive behavioural intervention and specialist nursery provision for children with autism spectrum disorders. Journal of Child Psychology and Psychiatry, 4, 803–812. Matson, J. L. (2007). Determining treatment outcome in early intervention programs for autism spectrum disorders: A critical analysis of measurement issues in learning based interventions. Research in Developmental Disabilities, 28, 207–218. Mullen, E. (1995). The Mullen Scales of Early Learning. Circle Pines, MN: American Guidance Service. Munson, J., Faja, S., Meltzoff, A., Abbott, R., & Dawson, G. (2008). Neurocognitive predictors of social and communicative developmental trajectories in preschoolers with autism spectrum disorders. Journal of International Neuropsychology Society, 14, 956–966. Sparrow, S., Balla, D., & Cicchetti, D. (1984). Vineland Adaptive Behavior Scales. Circle Pines, MN: American Guidance Services. Stone, W. L., & Yoder, P. J. (2001). Predicting spoken language level in children with autism spectrum disorders. Autism, 5, 341–361. Szatmari, P., Bryson, S. E., Boyle, M. H., Streiner, D. L., & Duku, E. (2003). Predictors of outcome among high functioning children with autism and Asperger syndrome. Journal of Child Psychology and Psychiatry, 44, 520–528. Thurm, A., Lord, C., Lee, L. C., & Newschaffer, C. (2007). Predictors of language acquisition in preschool children with autism spectrum disorders. Journal of Autism and Developmental Disorders, 37, 1721–1734. Turner, L. M., & Stone, W. L. (2007). Variability in outcome for children with an ASD diagnosis at age 2. Journal of Child Psychology and Psychiatry, 48, 793–802. Zachor, D., & Ben Itzchak, E. (2010). Treatment approach, autism severity and intervention outcomes in young children. Research in Autism Spectrum Disorders, 4(3), 425–432. Zachor, D. A., Ben Itzchak, E., Rabinovitch, A., & Lahat, E. (2007). Change in autism core symptoms after intervention—a comparative study. Research in Autism Spectrum Disorders, 1, 304–317.