Context-dependent encoding in the human auditory brainstem relates to hearing speech in noise: implications for developmental dyslexia

Abstract

We examined context-dependent encoding of speech in children with and without developmental dyslexia by measuring auditory brainstem responses to a speech syllable presented in a repetitive or variable context. Typically developing children showed enhanced brainstem representation of features related to voice pitch in the repetitive context, relative to the variable context. In contrast, children with developmental dyslexia exhibited impairment in their ability to modify representation in predictable contexts. From a functional perspective, we found that the extent of context-dependent encoding in the auditory brainstem correlated positively with behavioral indices of speech perception in noise. The ability to sharpen representation of repeating elements is crucial to speech perception in noise, since it allows superior "tagging" of voice pitch, an important cue for segregating sound streams in background noise. The disruption of this mechanism contributes to a critical deficit in noise-exclusion, a hallmark symptom in developmental dyslexia.

Figure 1. Stimulus characteristics and experimental design

(Top) The spectrogram of the stimulus /da/. The boundary of the consonant-vowel formant transition and the steady-state vowel portion of the syllable is marked by a dashed white line. The spectrogram was generated via frequency analyses over 40ms bins starting at time 0 and the midpoint of each bin is plotted. The stimulus /da/ is presented in variable (middle) and repetitive (bottom) contexts. As seen in the spectrograms, the stimuli in the variable context differed from /da/ in a number of spectral and temporal features. Responses to /da/ are event-matched between the two conditions to control for presentation order.

Figure 2. Experiment 1. Human auditory brainstem responses are sensitive to stimulus context

Difference in H2 magnitude between repetitive and variable contexts correlates with speech in noise scores. (A) The grand averages of the time-amplitude responses in the repetitive (red) and variable (black) conditions are overlaid. The black box demarcates formant transition period (7-60 ms). Context did not affect measures of peak latency or response amplitude. (B) Grand-average spectra for the repetitive (red) and variable (black) conditions show enhanced encoding of the second (H2) and the fourth harmonics (H4) in the repetitive condition (left). Mean spectral amplitudes of the second (H2) and fourth (H4) harmonics are shown in the repetitive (red) and variable (black) conditions (right). Error bars represent one standard error of the mean. (C) The normalized difference in H2 magnitude between the two conditions (repetitive minus variable) is related to speech in noise perception measures (HINT-RIGHT, left, HINT-COMPOSITE, right).

Figure 3. Experiment 2. Context effects are seen for good readers but not for poor readers

(A) The grand averages of the responses in the repetitive (left) and variable (right) conditions are overlaid for the good (left) and poor (right) readers. (B) Grand average spectra over the formant transition period for the good (left) and poor (right) readers show enhanced harmonic encoding in the good readers in the repetitive (red) condition and enhanced encoding of harmonics in the poor readers in the variable (black) condition. (C-left) Bar plots of H2 amplitude support the response spectra, with greater H2 amplitude in good (left) than poor (right) in the repetitive (red) condition and the opposite effect in the variable (black) condition. (C-right) The normalized difference in H2 magnitude between the two conditions is again related to speech in noise perception measures (HINT-COMPOSITE) across the whole group. Good readers are plotted as open diamonds and poor readers as filled stars. Overall, poor readers show inferior speech in noise perception relative to good readers (t(28) = −4.00, p < 0.001; see inset).

Figure 4. Experiment 2. Good readers show superior encoding in the repetitive context while the opposite pattern is seen in poor readers

Good readers (left) have greater H2 and H4 amplitudes (200 and 400 Hz, respectively) in the repetitive condition than the variable (signified by warm colors) while the poor readers (right) show greater H2 and H4 amplitudes in the variable condition than the repetitive condition (signified by cool colors). Figures were created by first generating response spectrograms for both conditions, and then subtracting spectral amplitudes elicited in the variable context condition from those elicited in the repetitive condition for each group.