Estudo comparativo da habilidade de imitação no Transtorno Específico de Linguagem e no Transtorno do Espectro do Autismo Comparative study of the imitation ability in Specific Language Impairment and Autism Spectrum Impairment

Purpose: To compare abilities of imitating generic and sequential motion gesture schemes in family routines among children with Autism Spectrum Disorder (ASD) and Specific Language Impairment (SLI) and to analyze the relation between imitation index and verbal production in the ASD group. Methods: The sample was constituted by 2:1 pairing of 36 children, according to gender and age. All of them were diagnosed by a multidisciplinary team as belonging to the ASD group (n=24) or SLI group (n=12) and were under direct and indirect intervention in a school clinic. We have used the stage of imitation of the Assessment of Symbolic Maturity, which entails the imitation of nine generic and three sequential motion gesture schemes. Results: There was a tendency to a better performance of the SLI group at imitating both generic and sequential gesture schemes. As we have related the ability of imitation to the verbal production in the ASD group, a direct relation between the production of phrases and the imitation of sequential schemes was detected. Conclusion: The ability to imitate gesture and sequential schemes could be compared, and a more prominent impairment was identified in children with autism. Among them, a direct significant relationship between the ability of imitating sequential gesture schemes in family routine and verbal production of words and sentences was verified. DOI: 10.1590/2317-1782/20152014194 143 Imitação no Transtorno Específico de Linguagem e Transtorno do Espectro do Autismo CoDAS 2015;27(2):142-7 INTRODUÇÃO A imitação desempenha um papel central no desenvolvimento motor, linguístico e social da criança. Ela está intimamente relacionada ao desenvolvimento da linguagem, principalmente no que diz respeito ao desenvolvimento de habilidades sociocomunicativas. Sabese que a habilidade de imitar é responsável, em grande parte, por exemplo, pela produção e variância de vocabulário expressivo e receptivo nos primeiros anos de vida. Em crianças com transtornos que acometem o desenvolvimento, tais como o Transtorno do Espectro do Autismo (TEA) e o Transtorno Especifico de Linguagem e Fala (TEL), observamos falhas importantes nessa habilidade. No caso do TEA, condição caracterizada por prejuízos severos e persistentes na interação e comunicação social e pela presença de repertório restrito de interesses e atividades, a inabilidade relacionada à imitação é decorrente de prejuízos na cognição social, especialmente pela ausência de reciprocidade e engajamento social e dificuldade na atribuição de estados mentais. Portanto, essa inabilidade não pode ser explicada apenas por fatores como falhas de memória, de raciocínio espacial ou de controle motor. Já nos casos de TEL, condição na qual as modalidades de compreensão e expressão verbais estão comprometidas, sem que sejam atribuíveis a anormalidades neurológicas, sensoriais, cognitivas e/ou ambientais, as falhas na imitação não são decorrentes de prejuízos básicos de interação e comunicação social, e sim de fraco desempenho na memória de trabalho, atenção e no processamento visuoespacial. Sendo assim, a hipótese que consideramos neste estudo é a de haver diferença na habilidade de imitação nessas patologias do desenvolvimento e de que o desempenho de crianças acometidas por TEA deve estar mais prejudicado em relação ao TEL, por se tratar de uma condição cuja manifestação primária é o déficit de cognição social. Consideramos, ainda, que as relações entre as habilidades de imitação e de produção verbal são evidenciadas nos casos de crianças com diagnóstico compativel com o TEA. Assim, os objetivos deste estudo foram comparar as habilidades de imitação de esquemas gestuais simples e de sequências de ações em rotinas familiares de crianças com TEA e com TEL e analisar a relação entre o índice de imitação e a produção verbal das crianças do grupo com TEA.


INTRODUCTION
Imitation plays a central role in motor, language, and social development of children (1)(2)(3) .It is closely related to language development, especially with regard to the development of social communication skills (1)(2)(3) .
The ability to imitate is known to be fundamentally responsible for the production and variance of expressive and receptive vocabulary in the early years of life (3) .
Children with disorders that affect development, such as Autism Spectrum Disorder (ASD) and Specific Language Impairment (SLI), have major flaws as to this skill.
In the case of ASD, a condition characterized by severe and persistent impairments in social interaction and communication and a restricted range of activities and interests (4)(5)(6)(7) , the inability related to imitation results from impairments in social cognition, especially the lack of reciprocity and social engagement and difficulty in attribution of mental states.Therefore, this inability cannot be explained by factors such as memory, spatial reasoning, or motor control failures.
In cases of SLI, a condition in which verbal modes of comprehension and expression are compromised regardless of neurological, sensory, cognitive, and/or environmental abnormalities (4,(8)(9)(10)(11) , failure to imitate does not result from the lack of basic social interaction and communication but from poor performance of working memory, attention, and visualspatial orientation (8)(9)(10)(11) .
Thus, the hypothesis of this study is that there is a difference in the ability to imitate as related to these developmental disorders.Children whose performance is affected by ASD present disadvantage compared to those with SLI, for this is a condition whose primary manifestation is deficit of social cognition.We also considered that there is a relationship between imitation skills and verbal production among children diagnosed with ASD.
The objectives of this study were to compare abilities of imitating generic motion gesture schemes and sequential gestures in family routines of children with ASD and with SLI, and to analyze the relationship between imitation index and verbal production among children with ASD.

METHODS
This is a case-control study (CEP No. 110442).All parents or guardians were aware of the methodological procedures of the study and signed the informed consent, complying with the rules of the Research Ethics Committee.

Sample
The sample was formed by 2:1 pairing of 36 children, according to gender (male and female), age group (6-10 years), and speech therapy intervention time.
All children were diagnosed by a multidisciplinary team -composed of a child psychiatrist, a clinical neuropsychologist, and a speech language pathologist, according to ICD-10 (4) and DSM RTVI criteria, and compatible with the DSM-5 (5) -with ASD (group ASD, n = 24) or SLI (group SLI, n = 12) assisted at the Center for Investigation of Children and Adolescents with Global Developmental Disorders (NIFLINC-TGD) and at the Outpatient Clinic for Child Language Therapy, both pertaining to the Speech Language Pathology and Audiology Department, Universidade Federal de São Paulo (UNIFESP).
Group ASD comprised 22 boys and 2 girls aged 6-10 years.These children had been in a speech therapy intervention program for 2 years and 6 months, on average, and regularly enrolled in a child or elementary public school.
Group SLI comprised 11 boys and 1 girl aged 6-10 years.These children had also been in a speech therapy intervention program for 2 years and 6 months, on average, and regularly enrolled in a child or elementary public school.
Inclusion criteria were multidisciplinary diagnosis of both disorders (ASD and SLI), age, and length of therapeutic intervention.Exclusion criteria were the presence of associated malformations or genetic syndromes, as well as motor, physical, and sensory (auditory or visual) impairment, for both groups.
Rates of intelligence quotient (IQ) were obtained by applying the Brazilian version of the Wechsler Scale (WISC) (12) by the team of psychologists.
Among 24 children with ASD, 10 were classified with unspecified degree of mental retardation, a term used when there is strong presumption of mental retardation but the individual cannot be adequately tested by the usual intelligence measurement instruments.The other individuals in this group had a score of 50-70, average 61, and were classified with mild mental retardation.
The IQ indices of 12 children in group SLI, obtained in psychological assessment, were more than or equal to 85, with an average of 87, were classified as normal.
All subjects were evaluated by speech language pathologists at NIFLINC-TGD.To characterize oral language, as to the presence of isolated words and phrases in expressive language, we used the information obtained through Lista de Avaliação do Vocabulário Expressivo (LAVE) (13) , an adaptation for the Brazilian Portuguese language of the Language Development Survey (14) .Parents or caregivers were interviewed by the language pathologist so they could sign the words their children spoke spontaneously in their daily lives, among the 307 from the list, and were asked to provide examples of spontaneous phrases used by children in everyday life at the time of the survey.To classify phrases, we considered the issuing of two or more words combined and produced systematically, not echolalic (13,14) .
Of the children with ASD, 10 issued vocalizations only (vowel sounds, sometimes decontextualized and without naming function) and 2 single words (identified as naming function and containing at least 75% of the Brazilian Portuguese phonemes).Other 12 children issued spontaneous phrases.
As for verbal production, all children diagnosed with SLI systematically issued spontaneous phrases (two or more words combined and produced systematically).

Procedures
To assess the ability to imitate, we used the imitation stage of the Symbolic Maturity Assessment (9)(10)(11) , which comprises the imitation of nine generic motion gesture schemes (e.g., drinking and smelling) and three sequential gesture schemes in family routine (feeding a doll, putting it to bed, and bathing it) (Appendix 1).
The children were evaluated individually during a session of 45 minutes by a language pathologist from NIFLINC-TGD and at the Outpatient Clinic for Child Language Therapy, both pertaining to the Speech Language Pathology and Audiology Department of UNIFESP.They were all familiar with the material and cases studied.The recordings were made in the presence of an adult familiar with children (language pathologist) and/ or their mothers to avoid interaction with a stranger to interfere with children's performance during the activity.
Initially, nine generic motion gesture schemes were presented to children, as they should be reproduced by them immediately at the language pathologist's orientation.The maximum score was 9, each imitated gesture being considered right or wrong only.
The nine generic motion gesture schemes were drinking using a toy, eating using a toy spoon, putting a hat, pushing a cart, making a toy airplane fly, brushing teeth, hugging a doll, washing hands with soap, and smelling a plastic flower.
They were then presented three sequential motion gesture schemes composed of four actions.Each action reproduced by the child scored 1 point; therefore, the maximum score was 12.
The sequential motion gesture schemes present in family routines were the following: • feeding a doll: putting the doll in a toy chair, putting a bib on it, feeding it, and cleaning its mouth; • putting a doll to sleep: putting the doll in a toy crib, covering it, reading it a story, and kissing it; • bathing a doll: taking off the doll's clothes, putting it in the tub, washing it with soap, and wiping it with a towel.
For data analysis, scores obtained in the imitation tasks of generic motion and sequential motion gesture schemes in each group were considered.
Finally, we analyzed the correlation between imitation index and verbal production of isolated words and phrases by children in group ASD.Significance level was set at 5%, especially in the analysis of variance.

RESULTS
Figure 1 shows the comparison of performances by both groups at imitating generic and sequential motion gesture schemes present in family routines.
The performances at imitating generic schemes were very similar in both groups (p = 0.382).We noticed better performance with sequential schemes present in family routines in group SLI (p = 0.053).
Figure 2 shows the distribution of performances related to verbal production in both groups.
Sentence production was predominant in group SLI (p = 0.003) and vocalization, in group ASD (p = 0.009).
Table 1 shows the relationship between imitation of schemes and verbal production in group ASD.
We found a significant direct relationship between imitation of sequential schemes and verbal performance.That is, children who performed better in sequential schemes presented verbal production of words and phrases (p = 0.007).

DISCUSSION
Looking at each group individually, group SLI performed better in imitation tasks of sequential schemes compared to generic motion gestures.Previous studies have pointed out that the context of everyday actions in this population can be a facilitator and booster of the development of the ability to imitate because it allows integration of relevant information to the child's context (8)(9)(10)(11) .
Children with SLI present social interaction abilities and preserved shared attention, which can further facilitate observation of symbolic action and their sequential and contextualized reproduction (8)(9)(10)(11) .Furthermore, group ASD showed better performance in imitation tasks of generic compared to sequential gestures.This finding suggests that children with autism have difficulties in integrating actions comprising sequences not only due to their difficulty of short-and long-term memory but also due to the difference in their information-processing system.These children have difficulty in processing information to form a whole structure with meaning, that is, they lack central coherence, one of the most striking features in autism spectrum (15)(16)(17) .
Other studies have also indicated a significant relationship between imitation deficit and social interaction deficit, especially when it comes to social reciprocity and shared attention, which are evident in children with autism (15)(16)(17) .
When comparing the groups, children with SLI had a tendency to perform better in both isolated and sequential actions, which confirmed our hypothesis that the skill of imitation is more preserved in this group compared to those in ASD group.
We know that the ability to imitate is a predictor of social relations, communication, and learning throughout life.Some studies have suggested that the difficulty of motor planning and motor is also intrinsically related to the inability to imitate in ASD (18)(19)(20)(21) .Therefore, this inability may also be a reflection of the primary difficulty in formulating and executing an intentional motor plan, although these individuals have preserved motor skills (18)(19)(20)(21)(22)(23)(24)(25) .
Regarding verbal production, all children in group SLI were able to produce sentences, whereas only half of the children from group ASD were able to do it.
Interestingly, we found significant correlation between the ability to imitate and verbal production among children with ASD.That is, children in this group had better performance in testing the imitation of sequential schemes and also more prominent sentences in verbal production, which confirm the intrinsic relationship between these skills.
Several studies had already pointed imitation as predictor of social and cognitive-linguistic skills (1)(2)(3)6,17,18) . Directedand motor imitation without the use of objects can be responsible for the variance in production of expressive and receptive vocabulary (3,17,18) .
The results of this study provided important contribution as to the ability to imitate in SLI and ASD; nevertheless, it is important to point out some limitations of one study.The influence of IQ on verbal and motion performances seems evident, especially among children with ASD; however, we could not confirm this premise because some of the children did not obtain IQ index and were classified with unspecified mental retardation.
Therefore, we suggest that further studies with larger samples should be carried out, aiming at analyzing the effect of IQ in the ability to imitate and verbal production.
Interestingly, the speech pathologist maintains, repairs, complements, and extends the meaning of verbal and nonverbal communication used by children.Thus, when assessing performance and managing therapeutic intervention, one must consider the relationship between verbal and nonverbal skills.
The Language Pathology Clinics is aimed at achieving goals through mediation, verbal, and nonverbal strategies in contextualized activities where spontaneous or directed imitation is part of resources for the expansion of the repertoire of exploration activities, including games and social communication skills.

CONCLUSION
We were able to compare the ability to imitate motion gesture schemes present in family routine of children with ASD and SLI, and to check impaired performance among children with autism.
A significant relationship was also observed between imitation of family routine gestures and verbal production of words and phrases among children with ASD.

Figure 1 .Figure 2 .
Figure 1.Percentage of correct imitation of generic and sequential motion gesture schemes in both groups

Table 1 .
Relationship between imitation of schemes and verbal production in autism spectrum disorder *Statistical significance; **tendency to statistical significance AS = autism spectrum; IA = index of agreement