Speech disfluencies in children with developmental dyslexia: How do they differ from typical development?

Area of Science:

• Linguistics
• Developmental Psychology
• Speech-Language Pathology

Background:

• Disfluency is a complex linguistic phenomenon affecting language production in both typical and atypical populations.
• Limited research exists on disfluency patterns in developing language systems, particularly in children with neurodevelopmental disorders like developmental dyslexia (DD).
• Previous studies show language-specific variations in disfluency rates in typical development (TD).

Purpose of the Study:

• To analyze disfluency types and their evolution in French-speaking children aged 8-12 years.
• To compare disfluency patterns between children with DD and TD.
• To investigate the correlation between disfluencies and reading difficulties.

Main Methods:

• A total of 25 children with DD and 21 children with TD, aged 8-12.6 years, produced autobiographical narratives.
• Seven types of disfluencies were coded: part-word repetitions, monosyllabic word repetitions, other repetitions, filled pauses, substitutions, additions, and abandoned utterances.
• Proportions of each disfluency type were compared between groups; Spearman correlations assessed relationships with reading performance and age.

Main Results:

• Both DD and TD children predominantly exhibited filled pauses, monosyllabic word repetitions, and substitutions.
• Both groups displayed high overall disfluency rates exceeding 10%.
• Disfluency rates did not correlate with age, but correlations with reading performance were significant only in the TD group.

Conclusions:

• Developmental dyslexia (DD) is not characterized by a distinct disfluency pattern in French-speaking children aged 8-12 years.
• Disfluency rates are similar between DD and TD children, with both groups exhibiting high rates (>10%), which is notable for the French language.
• Interpreting disfluency in DD requires caution, as high rates are also present in TD; pathological thresholds should be language-adapted to avoid overestimation of deficits.