Children with specific language impairments not a distinct group
MedWire News: Children with specific language impairments (SLIs) fall at the lower end of a continuous distribution of language ability rather than forming a distinct category, results of a taxonomic study show.
The researchers say the next step will be to determine whether different cutoff values are associated with substantive differences in symptoms, adverse effects on academic and social success, responsiveness to treatment, or long-term outcomes.
The diagnostic category of SLI has a long history and there is general agreement that affected children have language deficits that are not readily explained by deficits in other areas of development.
However, there is little consistency across research studies and clinical sites in the specific indicators and criteria by which the condition is diagnosed.
"One of the most basic and longstanding questions about SLI is whether children with SLI represent a distinct group with language skills that differ qualitatively or nonarbitrarily from those of other children or, alternatively, they represent those at the lower end of a continuous distribution, with language skills that fall below some arbitrary threshold but are not otherwise unique," Christine Dollaghan (University of Texas at Dallas, USA) and colleagues comment.
To investigate, the team performed language analysis on 601 children aged 6 years who were free of neonatal risk factors. The researchers focused on two measures: number of words produced in a language sample and average percentage phonemes repeated correctly in three- and four-syllable nonwords.
They then performed taxometric analysis, which is based on examining the relationship between scores on two or more diagnostic indicators at various points along a large distribution of scores, from a sample that includes members of the hypothesized category and members of the complement class.
From this they calculated mean above minus below a cut (MAMBAC) values, which were plotted graphically.
A graph with particular area of the distribution where MAMBAC values increase substantially, yielding an upwardly convex graph with a single peak, is consistent with the presence of a discrete subgroup within the distribution of scores.
By contrast, a graph of MAMBAC values that is flat, concave, or decreasing suggests that the relationship between scores is consistent throughout the distribution and that the condition is dimensional.
Dollaghan and colleagues found that the data in the present study fitted most consistently with the latter scenario.
"Although studies of the latent structure of SLI are in their infancy, at present it appears that the preponderance of evidence, including results from MAMBAC analyses, supports a dimensional view of SLI in which otherwise typical children with language deficits at ages 3, 4, and 6 years have language skills that differ quantitatively rather than qualitatively from those of their peers," they comment.
The research is published in the Journal of Speech, Language, and Hearing Research.
By Andrew Czyzewski