Audiovisual speech perception in Brazilian Portuguese-speaking children with and without Speech Sound Disorder (SSD)

This study compared single-modality and audiovisual (crossmodally congruent and incongruent) speech perception in Brazilian Portuguese-speaking children with and without Speech Sound Disorders (SSD). Thirty-six children (17 with and 19 without SSD; average age = 6;7) participated in an adapted syllable identification test. The auditory and visual stimuli consisted of monosyllabic consonant-vowel utterances involving [p,b,t,d,k,g,f,v,s,z,ʃ,ʒ,m,n,ɲ,l,ʎ,ɾ,χ,w], produced in the context of [a] vowel. The stimuli were presented in different modality conditions: auditory-only (A_O), visual-only (V_O), audio-visual congruent (AV₊), and audio-visual incongruent (AV-). As expected, children with SSD showed lower perceptual accuracy than TD children across all modalities (A_O, V_O, and AV₊), indicating that their difficulties are not restricted to the visual domain. The prediction that SSD children would show reduced audiovisual integration was not confirmed. No differences were observed between groups in terms of visual gain from congruent audiovisual stimuli or susceptibility to the McGurk illusion, although both groups exhibited the effect. Exploratory analyses revealed that age accounts for part of the group differences, particularly in the V_O condition, while differences in A_O and AV₊ remained partially independent of age. Within the SSD group, PCC-R showed no significant multivariate effect, although univariate results indicated positive associations with A_O and AV₊, but not V_O. Although SSD children show reduced accuracy in speech perception across sensory modalities, audiovisual speech integration appears preserved. The variability observed within the SSD group suggests a complex and heterogeneous developmental profile that cannot be fully explained by age or severity, underscoring the need for integrated and individualised intervention approaches.