Predicting language ability of HL children (Zhai et al., 2025)

<p dir="ltr"><b>Purpose: </b>Children with congenital hearing loss (HL) have auditory impairments that may place them at increased risk for delays or variability in language development. However, obtaining reliable brain markers for early classification of young children with HL versus those with normal hearing (NH), as well as for precise assessment of HL children’s language ability, remains a challenge due to limitations in traditional neuroimaging techniques and theoretical frameworks. To address this gap, we propose the maternal mirror hypothesis, which suggests that brain activities of mothers might mirror or indirectly reflect children’s auditory language ability, offering an additional and useful approach for obtaining brain markers of HL children in clinical assessment.</p><p dir="ltr"><b>Method: </b>Children aged 2–5 years with HL (<i>n</i> = 105) and NH (<i>n</i> = 89), along with their mothers, participated in the study. Brain activity in each mother–child dyad was simultaneously measured using functional near-infrared spectroscopy (fNIRS) while they watched a silent video together. From these data, we derived maternal and child intrapersonal brain functional connectivity (FC), as well as mother–child intersubject correlation (ISC). Children’s language comprehension and production ability were assessed at baseline with a follow-up of their changes over 6 months.</p><p dir="ltr"><b>Results and Conclusions: </b>We found that maternal brain FC or mother–child ISC outperformed child-based FC in predicting HL children’s language comprehension and production, as well as their plastic changes across 6 months. Moreover, brain markers predicting HL children’s language ability did not differ between groups of HL and NH, whereas those brain markers that classified HL versus NH group status were not correlated with HL children’s language ability. This dissociation suggests distinct neural mechanisms underlying HL pathology with brain deficits versus the compensatory mechanisms with the functional recovery of HL children. These findings support the maternal mirror hypothesis, having the potential to address traditional challenges in early functional assessment and prediction of HL children by providing a novel neuroimaging approach and an original theoretical framework.</p><p dir="ltr"><b>Supplemental Text S1. </b>Additional study details.</p><p dir="ltr"><b>Supplemental Table S1.</b> Demographic information of the participants.</p><p dir="ltr"><b>Supplemental Table S2.</b> The MNI coordinates of each channel.</p><p dir="ltr"><b>Supplemental Table S3.</b> The language ability in HL and NH groups.</p><p dir="ltr"><b>Supplemental Table S4.</b> The language ability pre- and post-tests in the HL group.</p><p dir="ltr"><b>Supplemental Table S5.</b> Correlations between behavioral factors and the most contributed features in classification or prediction models.</p><p dir="ltr"><b>Supplemental Figure S1.</b> Principal Component Analysis (PCA) on the language scores.</p><p dir="ltr"><b>Supplemental Figure S2.</b> Comparative analysis of language comprehension (A) and production (B) abilities between HL and NH groups and the plastic changes from pre- to post-test (C-D).</p><p dir="ltr"><b>Supplemental Figure S3.</b> The consistency between subjective (maternal questionnaires: CAP and SIR) and objective behavioral assessments (SMAALAHIC), using the Pearson correlation analyses at both pre- and post-test in the HL group.</p><p dir="ltr"><b>Supplemental Figure S4.</b> The pipeline of fNIRS data prerocessing.</p><p dir="ltr"><b>Supplemental Figure S5.</b> fNIRS quality check and bad CHs distribution.</p><p dir="ltr"><b>Supplemental Figure S6.</b> Relationship of HL mothers’ movement statistics with child’s language measures (HL group only) at both pre- and post-test.</p><p dir="ltr"><b>Supplemental Figure S7.</b> Schematic of nested cross-validation model framework.</p><p dir="ltr"><b>Supplemental Figure S8.</b> SVM classification probability for individuals.</p><p dir="ltr"><b>Supplemental Figure S9. </b>Child’s intrapersonal functional connectivity significantly contributed to the prediction on language ability of children with HL.</p><p dir="ltr"><b>Supplemental Figure S10.</b> Mother’s intrapersonal functional connectivity significantly contributed to the prediction on language ability of children with HL.</p><p dir="ltr"><b>Supplemental Figure S11.</b> Child’s intrapersonal functional connectivity significantly contributed to the prediction on change in language of children with HL.</p><p dir="ltr"><b>Supplemental Figure S12.</b> Mother-child ISC that significantly contributed to the prediction on change in language of children with HL.</p><p dir="ltr">Zhai, Y., Xing, Y., Zhao, J., He, X., Jiang, K., Zhang, T., & Lu, C. (2025). Predicting the auditory language ability of young children with hearing loss using their mothers’ brain activity. <i>Journal of Speech, Language, and Hearing Research,</i><i> </i><i>68</i>(10), 4996–5020. <a href="https://doi.org/10.1044/2025_JSLHR-25-00008" target="_blank">https://doi.org/10.1044/2025_JSLHR-25-00008</a></p>