Expressive Language Sampling May Aid Research in Fragile X

A procedure for measuring language skills, called expressive language sampling (ELS), could be used to assess treatment outcomes in people with fragile X syndrome and other conditions characterized by intellectual disability, a new study suggests.

The study, “Expressive language sampling as a source of outcome measures for treatment studies in fragile X syndrome: feasibility, practice effects, test-retest reliability, and construct validity,” was published in the Journal of Neurodevelopmental Disorders.

Better understanding of the causes of intellectual disability had led to the development of treatments tailored for specific disorders.

Testing such treatments requires carefully designed clinical trials. In the case of fragile X, the leading inherited cause of intellectual disability, trials of potential treatments have generally been lackluster, with more than two dozen failing to show effectiveness in patients, the study notes.

One reason for these failures is a “lack of adequate outcome measures,” Leonard Abbeduto, PhD, director of UC Davis MIND Institute and the study’s lead author, said in a press release.

A particular obstacle has been the so-called practice effect, in which individuals get higher scores on tests over time, simply due to having more practice. In the context of a clinical trial, this could make it unclear whether an observed improvement is due to a treatment working as intended, or just the result of repeating a test.

Researchers evaluated ELS as a way to measure language proficiency (which is closely related to intellectual disability) and create language outcome measures for research in fragile X. This approach has been gaining support as an accurate measure of language abilities in this patient population.

The ELS procedure has two parts: conversation and narration. In the conversation part, which lasted around 12 minutes in this study, the person administering the test talks to the person being tested. First, the conversation focuses on a topic the participant enjoys (such as favorite activities, fictional characters, and friends), then moves to one of several pre-designed topics. During these conversations, the researcher follows a script designed to encourage the patient to do most of the talking.

In the narration part, the investigator shows the participant a picture book (“Frog Goes to Dinner” or “Frog, Where Are You?”). Over 10 to 15 minutes, the participant then tells the story in their own words with minimal help.

Both parts are recorded and analyzed with computer algorithms across a number of metrics, such as talkativeness, vocabulary, intelligibility (whether the words are comprehensible), and complexity of syntax.

A total of 106 young people with fragile X (ages 6–23) were evaluated on two occasions, one month apart (test and re-test). Most participants (85) were males. All had intellectual disability, defined as an IQ of 70 or bellow, and all used speech as their primary way of communicating.

The noncompliance rate — the percentage of participants unwilling or unable to complete ELS — was under 15% for both the conversation and narration portions. Noncompliance rates were higher among participants who were younger, had lower IQ, and had more pronounced autism symptoms.

Results showed no statistically significant difference between test and re-test. This suggests no practice effect — if there had been one, scores would be higher on the re-test.

“The lack of meaningful practice effects for the ELS measures makes them well suited to treatment studies,” the researchers wrote.

They also compared scores on the ELS with those obtained through standardized questionnaires. ELS scores related to intelligibility, syntax, and vocabulary correlated well with standard measurements. Assessments of talkativeness and dysfluency (interruptions in speech) showed no such correlation.

Although the reasons for this lack of correlation remain unclear, the researchers recommended against using talkativeness or dysfluency as outcome measures in studies with fragile X patients.

Overall, this study supports ELS as an outcome measure for people with fragile X and intellectual disability, the investigators said.

“Most importantly, the measures are functional for the individual,” Abbeduto added. “If a treatment can improve language, it can have far reaching benefits for the individual at school, in the workplace, with peers and beyond.”