Resumen de Vocabulario expresivo en una muestra de niños preescolares con trastornos psiquiátricos y un grupo con desarrollo típico
Lilla Albores Gallo, Maria del Carmen Lara Muñoz, Dea del Carmen Melchor Contreras, David Ángeles Albores
- español
Introduccion Aun cuando los problemas de lenguaje expresivo son muy comunes tanto en la poblacion general (15-20%) como en la clinica (50-75%), su deteccion es insuficiente. Los problemas de lenguaje requieren atencion debido a su comorbilidad con problemas graves del desarrollo como el autismo, el trastorno de Asperger, el trastorno por deficit de la atencion e hiperactividad (TDAH) y el retraso mental. En teoria, el vocabulario asociado a estos trastornos psiquiatricos podria identificarse con un instrumento que midiera el vocabulario expresivo.
Objetivos 1. Determinar la frecuencia de atraso del lenguaje (AL) (SDL .50 palabras) en un grupo con desarrollo tipico de la comunidad. 2. Determinar el nivel de vocabulario para los subgrupos de: autismo, trastorno de Asperger (TA), TDAH y otros trastornos psiquiatricos (OTP) por medio del sondeo del desarrollo del lenguaje (SDL). 3. Analizar si el SDL puede discriminar entre los subgrupos clinicos.
Sujetos y metodo La muestra estuvo compuesta por: un grupo de la comunidad con desarrollo tipico (GDT) (n=302), y un grupo clinico (GC) (n=55), con un rango de edad de 2-5 anos. Se formaron cuatro subgrupos clinicos: autismo, trastorno de Asperger, TDAH y un grupo de OTP (enuresis, encopresis, ansiedad de separacion).
El SDL es una lista de palabras que identifica el padre sobre el vocabulario de los ninos que tiene un coeficiente de alpha de Cronbach de (.99), un test-retest de .97 a .99 y una sensibilidad y especificidad de 86-90%. Se utilizo la definicion de atraso de lenguaje (AL) basada en un punto de corte de .50 palabras.
Analisis estadistico. Los datos categoricos fueron analizados mediante la prueba de chi-cuadrada y para las medidas continuas como la edad, el MSE y el puntaje del SDL se usaron pruebas t de Student. Para el analisis del contraste de las medianas del SDL de los grupos se aplico una prueba de Kruskal-Wallis. Resultados Los grupos fueron semejantes para las variables como edad del nino, edad de los padres y la media del SDL. La frecuencia de AL (.50 palabras) fue de 21.2% para el GDT y de 23.6% para la poblacion clinica. Por sexo, los varones presentaron mayor frecuencia de atraso de lenguaje (GDT): 21.6% masculino vs. 20.7% femenino (p<0.926), GC: 26.2% masculino vs. 15.4% femenino (p<0.423). El vocabulario del grupo de autismo fue el menor de todos (Mdn=85, DE=78.68) seguido del grupo de OTP (Mdn=149, DE=121.0) que presento un desempeno muy semejante al grupo de la comunidad (GDT) (Mdn=179, DE=105.0). El grupo de Asperger (Mdn=259, DE=127) tuvo un puntaje cercano al grupo de TDAH (Mdn=286, DE=100.25). La prueba de Kruskal-Wallis para la diferencia en las medianas fue significativa (p<.002) pero solo el contraste entre el grupo de autismo y de TDAH (Mdn=85 vs. Mdn=286, p<.01) fue significativo.
Discusion La frecuencia de AL para el GDT fue de 21.6% y para el GC fue de 23.6%. El SDL fue sensible en la deteccion del nivel de vocabulario entre los grupos y los resultados fueron congruentes con el desempeno esperado con algunas excepciones. Los ninos con TDAH expresaron un mayor numero de palabras comparados con el GDT. El unico contraste significativo fue la comparacion entre el grupo de TDAH y el autismo. El vocabulario del grupo de Asperger fue mejor que el de autismo, pero esta diferencia no alcanzo significancia estadistica. Conclusiones La versión mexicana del SDL es un instrumento de tamizaje útil para identificar el atraso del lenguaje en los niños preescolares. Este estudio muestra que el atraso de lenguaje en un niño preescolar con TDAH es una indicación para profundizar en el diagnóstico del autismo. Tampoco deben pasarse por alto otros trastornos que pueden acompañar o no el TDAH como los trastornos del lenguaje específicos (pronunciación, expresión, comprensión). El SDL mide el vocabulario y no identifica alteraciones del lenguaje cualitativas más complejas asociadas al trastorno de Asperger.
- English
Introduction Expressive language problems are common amongst preschoolers both in the general population (15–20%) and in clinical settings (50–75%); furthermore, these problems are often not detected. Language problems require attention since they are associated with severe developmental disorders such as autism (Au), Asperger's syndrome (AS), attention–deficit hyperactivity disorder (ADHD) and mental retardation. In theory, language development, specifically expressive vocabulary, associated to psychiatric disorders could be identified with a scale that measures expressive language.
Objectives 1. To determine the frequency of language delay in a sample of Mexican children with typical development in the community. 2. To determine the vocabulary level for autism, Asperger's syndrome, ADHD and other psychiatric disorders through the use of the Language Development Survey (LDS). 3. To analyze if differences in vocabulary ratings among the clinical subgroups can be detected with this instrument.
Materials and methods The sample consisted of: A community group with typical development (TDG) (n=302) and a clinical group (CG) (n=55); both groups had an age range of 2–5 years. The clinical group was subdivided into 4 clinical subgroups based on DSM–IV criteria for: autism, Asperger's syndrome, ADHD and other psychiatric disorders (OPD) (enuresis, encopresis, separation anxiety). Exclusion criteria were: deafness, hypoacusia and other sensorial disorders and mental retardation.
A semi–structured interview based on DSM–IV criteria was designed ad hoc to diagnose: autism, Asperger's syndrome, ADHD (inattentive, combined or hyperactive impulsive varieties), specific phobia disorder, tics (transitory, chronic and Tourette's syndrome), dysthymic disorder, depression, enuresis, separation anxiety disorder based on parent information. The clinical evaluation included a semi–structured play session with age–appropriate didactic material. Discrepancies in diagnosis were resolved by consensus. All interviews were conducted by an experienced clinician. The number of bulbs in the household was used to measure socioeconomic status (SES).
The LDS is a list of words that explores children's vocabulary based upon parental report. The original survey has a Cronbach's alpha coefficient of 0.99, test–retest coefficient of 0.97–0.99, and a sensitivity and specificity of 86–90%. Language delay (LD) was defined as ≤50 words, as recommended by several researchers. All parents signed an informed consent form and answered the LDS.
Statistical analysis. Categorical data was analyzed using a χ2 analysis; continuous data such as age, socioeconomic status, and LDS score, were analyzed using t–tests. To analytically compare the LDS group medians, a Kruskal–Wallis test was used, since the variable distribution violated the normality distribution requirements for parametric tests. For the post hoc tests, a Tamhane analysis was used for groups of different sizes. Differences were considered statistically significant if they had a p<0.05.
Results The groups were similar for variables such as child's age, parents' age and the LDS median between the normal development group and the clinical group t(355)=1.12, p=.26. The proportion of male children was greater in the clinical group (CG) than in the TDG, 76.4% vs. 53%, χ2(1,N=357)=10.63, p<.001. SES was higher for the TDG (M=7.2, SD=4.2) than for the CG (M=5.8, SD=3), p<.005. The father's age (r=.15, p<.009), the mother's age (r=.16, p<.003) and the SES (r=.13, p<.01) were correlated to the LDS score. Additionally, father's and mother's age were strongly correlated (r=.72, p<.0001) and the mother's age showed small correlations with the socioeconomic status (r=.15, p<.004). The mother's age was correlated with the child's vocabulary for both sexes (males: r=.16, p<.04, females: r=.16, p<.02), and vocabulary was significantly correlated with the SES, only for the males.
Language delay frequency in the TDG was 21.2%, and 23.6% for the CG, χ2(1,N=352)=1.03, p<0.59. By sex, males in both groups exhibited a greater frequency of LD [TDG: 21.6% males vs. 20.7% females, χ2(1,N=302)=.154, p<0.926; CG: 26.2% males vs. 15.4% females, χ2(1,N=55)=.642, p<0.423].
The autism subgroup had the lowest vocabulary rating (M=85, SD=78.68), followed by the OPD subgroup (M=149, SD=121), whose rating was very similar to the typically development group (M=179, SD=105). The Asperger group (M=259, SD=27) had a similar score to the ADHD group (M=286, SD=100.2), which had the highest vocabulary score of all. The Kruskal–Wallis test for median differences was significant [H(4)=17.47, p<.002]. Multiple contrast comparisons and Tamhane's post hoc analysis showed that only the contrast between the autism and the ADHD subgroups (means: 85 vs. 286, respectively) was significant (ANOVA Tamhane post hoc, p<.01).
Discussion Language delay (LD) measured through LDS in the TDG (21.6%) was very similar to other studies (15–20%). Nevertheless, LD in the clinical sample was inferior (23.6%) to what has been reported elsewhere (35–75%). It is possible that exclusion of mental retardation, in the clinical sample, could explain these differences. The greater proportion of males suffering from language delay in the clinical group is due to the fact that children with developmental problems occur more frequently among males. Some studies have found a marked and significant female superiority in vocabulary at an early age, but our results found that the girls' vocabularies in both groups were very similar to the boys'. This result is not explained by differences in the instruments used, since other studies carried out with the LDS also report greater scores for girls in both community and clinical samples. Therefore, it is possible that these results are due to cultural differences in the quantity and quality of the verbal stimulus towards Mexican girls and boys. Our scores were also inferior to the Caucasian American population scores.
The LDS was sensitive to detection of vocabulary level between groups and the results were congruent with the expected observations, with a few exceptions: Mainly, the community group had a nonsignificant lower score than the clinical group.
The subgroups analysis showed that, generally, the vocabulary level corresponded with the expected levels; nevertheless, we found great variability in the scores for the different subgroups. The ADHD had the highest score of all, which determined that the LDS mean for the CG was higher than TDG's. Children with ADHD generally express a greater number of words compared to children in the general population at younger ages. Different studies show that this vocabulary advantage is lost between 8–12 years of age. The group with the next highest score was the Asperger subgroup, which also had the greatest variability, along with the OPD subgroup. This makes identification of Asperger's syndrome through the LDS impossible, due to confounding with other groups. This is unfortunate because differential diagnosis between ADHD and the autism spectrum disorders is very important. Vocabulary was better in the Asperger than the autism subgroup, although these differences were not significant in the post hoc tests. The only significant comparison was between ADHD and autism groups, which is consistent with other studies through other measures.
Conclusions The Mexican version of the LDS is a useful tool for screening language delays in preschoolers. It may aid in the identification of autism and other common causes of language delays such as: transitory delay, late language emergence, specific language impairment and selective mutism. The LDS is a quantitative screening instrument and does not measure other complex alterations in other language domains such as: narrative, expository and conversational discourse; and other pragmatic language abilities individually or in any combination. Language impairment should be diagnosed based on multiple sources of information, including naturalistic observation and standardized, culturally appropriate psychometric measures.
