Rhythm, reading, and sound processing in the brain in preschool children

The study investigates whether preschool children's ability to synchronize to a beat is linked to foundational preliteracy skills and the precision of auditory neural processing. Prior research and theoretical frameworks (e.g., the Temporal Sampling Framework) suggest that deficits in auditory rhythm perception and neural entrainment may underlie reading difficulties, including dyslexia. Beat synchronization engages sensorimotor, auditory, and cognitive networks that overlap with language systems. The authors aim to establish, in a large and diverse preschool sample, whether better beat synchronization performance predicts stronger preliteracy skills and more precise frequency-following responses (FFRs) to speech, including under challenging listening conditions. They hypothesize that good synchronizers will outperform poor synchronizers on preliteracy tasks and exhibit more precise subcortical auditory encoding.

Multiple lines of evidence link rhythm skills with language and reading across ages. Children with dyslexia show atypical neural entrainment to the beat during passive listening, supporting an auditory rhythm perception deficit consistent with the Temporal Sampling Framework, which posits inefficient tracking of low-frequency modulation (1.5–10 Hz) relevant to speech envelope processing. Prior preschool studies reported that beat synchronization correlates with phonological awareness, auditory short-term memory, rapid naming, and FFR measures reflecting envelope encoding precision, response consistency, and phase-locking consistency. However, earlier work often involved small samples and focused on quiet listening and robust FFR components. The current study extends this literature with a larger cohort and assesses FFRs across stimulus polarities, time regions, frequencies, and in noise, better reflecting real-world listening and aspects crucial for phoneme identification.

Design and participants: Cross-sectional study of preschool children (N=156; ages ~3–5 years). Children were categorized by beat synchronization ability using a drumming task at two inter-stimulus intervals (ISI: 400 ms and 600 ms). Groups: Synchronizers (significant synchronization at both rates; N=63), Non-synchronizers (not significant at both rates; N=37), and Synchronizers-at-one-rate-only (significant at one rate only; N=56). Primary analyses contrasted Synchronizers vs Non-synchronizers; the one-rate group was examined descriptively and in secondary analyses. Sex differences in rhythm performance were observed; sex was included as a covariate in analyses. Beat synchronization task: Children drummed along to isochronous beats at 400 ms and 600 ms ISIs. Drumming consistency was quantified (R-value) with Rayleigh’s test; p<0.05 at an ISI indicated significant synchronization. Classification used performance across both ISIs (Fig. 8). Behavioral (preliteracy/music) measures: - Phonological awareness: CELF-P2 Phonological Awareness subtest (administered to children ≥4 years; n=108). - Auditory short-term memory: CELF-P2 Recalling Sentences subtest (all participants). - Rapid automatized naming (RAN): Objects (RAN-O) and Colors (RAN-C), Pro-Ed; lower times (ms) indicate better performance. - Music perception: Gordon’s AUDIE, Melody and Rhythm subtests; sensitivity indexed by d′. - Verbal and nonverbal IQ: WPPSI-III subtests (information; object assembly or matrix reasoning by age). FFR acquisition: - Stimuli: Synthesized consonant-vowel syllables [ba], [da], [ga] (170 ms: 5 ms burst, 60 ms formant transition, 60–170 ms steady-state vowel), plus [da] with multi-talker babble at −10 dB SNR relative to signal ([da]noise). Second formant trajectories: [ba] 900→1240 Hz; [da] 1700→1240 Hz; [ga] 2480→1240 Hz. - Presentation: Right ear via insert earphone (ER-3). 81 ms interstimulus interval (251 ms onset-to-onset), alternating polarity; 2100 trials per polarity per stimulus. Quiet and noise modes used; noise only with [da]. - Recording: BioSEMI Active 2. Electrodes: Cz active; earlobes referenced; forehead grounds (Fp1/Fp2). Children watched a quiet movie (<40 dB SPL) during passive listening. Digitization at 16.384 kHz; online band-pass 100–3000 Hz. Offline amplification per manufacturer, band-pass 70–2000 Hz (12 dB/oct zero-phase), epoch −40 to 210 ms, baseline corrected, artifact reject ±35 µV. FFR measures: - Envelope-encoding precision: Cross-correlation maximum between Hilbert-transformed, band-limited (70–2000 Hz then 200 Hz low-pass for envelope) stimulus and response envelopes within 5–12 ms lag; correlation z-transformed. - Response consistency (RC): Trial-to-trial neural stability as the average z-transformed correlation between 300 pairs of random 2000-sweep subaverages (0–170 ms). - Phase-locking consistency (PLC): Time-frequency phase-consistency around F0 (100 Hz) and harmonics to 1000 Hz via short-time FFT. Unit vectors averaged across 4000 trials; vector length (0–1) indicates phase consistency. PLC averaged across low (100–400 Hz; 200–400 Hz for subtracted polarity) and high (500–1000 Hz) ranges. Statistical analysis: SPSS. One-way ANOVAs for behavioral measures comparing Synchronizers vs Non-synchronizers; sex included as a covariate. Repeated-measures ANOVAs for neural measures with between-group factor: synchronization group; within-subject factors: time region (formant transition: 20–60 ms; steady-state: 60–170 ms), presentation mode (quiet, noise), polarity (added, subtracted; for RC and PLC), and frequency range (low, high; for PLC). Post hoc independent-samples t-tests used for follow-up. Additional analyses explored the Synchronizers-at-one-rate-only subgroup and differences between 400 ms-only vs 600 ms-only synchronizers.

- Preliteracy and music perception: Synchronizers outperformed Non-synchronizers on all preliteracy measures and rhythm discrimination. • Phonological awareness: F=11.680, p<0.001, ηp²=0.175. • Auditory short-term memory: F=4.773, p=0.031, ηp²=0.050. • Rapid automatized naming (objects and colors combined): F=4.881, p=0.030, ηp²=0.053 (lower ms is better; Synchronizers faster). • Music perception—Rhythm: F=5.083, p=0.027, ηp²≈0.052; Melody: n.s. (F=0.177, p=0.675, ηp²=0.002). - Envelope-encoding precision (all stimuli [ba], [da], [ga], [da]noise): • Main effect of group: Synchronizers > Non-synchronizers, F(1,87)=10.860, p=0.001, ηp²=0.111. • Quiet > Noise: F(1,87)=10.867, p=0.001, ηp²=0.111. • Steady-state > Transition: F(1,87)=6.602, p=0.012, ηp²=0.071. • Three-way interaction (group × stimulus mode × time region): F(1,87)=4.314, p=0.041, ηp²=0.047. - Response consistency (RC): • Main effect of group: Synchronizers > Non-synchronizers, F(1,79)=4.593, p=0.035, ηp²=0.055. • Quiet > Noise: F(1,79)=12.545, p=0.001, ηp²=0.137. • Added polarity > Subtracted: F(1,79)=23.197, p<0.001, ηp²=0.227. • Group × Time region: F(1,79)=8.091, p=0.006, ηp²=0.093 (larger steady-state benefit in Synchronizers). • Group × Mode × Polarity: F(1,79)=9.860, p=0.002, ηp²=0.111. • Group × Time region × Polarity: F(1,79)=4.316, p=0.041, ηp²=0.052. • Four-way interaction (Group × Mode × Time region × Polarity): F(1,79)=11.068, p=0.001, ηp²=0.123. - Phase-locking consistency (PLC): • Main effect of group: Synchronizers > Non-synchronizers, F(1,76)=4.498, p=0.037, ηp²=0.056. • Quiet > Noise: F(1,76)=8.927, p=0.004, ηp²=0.105. • Steady-state > Transition: F(1,76)=22.238, p<0.0005, ηp²=0.226. • High frequency (500–1000 Hz) > Low frequency (100–400 Hz): F(1,76)=10.585, p=0.002, ηp²=0.122. • No main effect of polarity: F(1,76)=0.559, p=0.457, ηp²=0.007. • Group × Time region: F(1,76)=9.182, p=0.003, ηp²=0.108 (greater transition-to-steady-state increase in Synchronizers). • Group × Time region × Mode: F(1,76)=5.217, p=0.025, ηp²=0.064. • Group × Time region × Frequency: F(1,76)=7.037, p=0.010, ηp²=0.085. - Robustness in noise: Synchronizers showed less degradation of FFR measures under noise, particularly in steady-state and higher-frequency ranges, indicating more resilient subcortical encoding under adverse conditions. - Synchronizers-at-one-rate-only subgroup (N=56; younger on average by ~0.4 years, p=0.001): Behavioral differences vs other groups were limited (trends for phonological awareness p=0.064; RAN p=0.004 vs Synchronizers). FFR differences reached significance only for envelope-encoding precision (p=0.036). Within this subgroup, 600 ms-only synchronizers had better envelope-encoding precision than 400 ms-only (F(1,49)=4.507, p=0.039, ηp²=0.084). Patterns were suggestive but inconclusive, warranting longitudinal follow-up.

Findings support that preschoolers’ rhythm synchronization ability is tightly linked to preliteracy skills and the precision and stability of subcortical auditory encoding of speech. Better synchronizers had superior phonological awareness, auditory short-term memory, and rapid naming, alongside higher envelope-encoding precision, response consistency, and phase-locking consistency. These relationships held and were accentuated in challenging encoding contexts (steady-state segments, higher harmonics, and noisy presentation), indicating that robust auditory brainstem synchrony and envelope tracking underpin individual differences in rhythm and emerging language skills. Results align with theories (e.g., Temporal Sampling Framework) positing that inefficient low-frequency envelope tracking contributes to language/reading difficulties, and extend prior small-sample reports by demonstrating effects in a larger, diverse cohort and under noise. The observation that synchronizers exhibit less FFR degradation in noise resonates with models linking early-life auditory-in-noise processing to later literacy outcomes and suggests shared mechanisms between auditory-motor entrainment and resilient auditory encoding.

This study establishes, in a large preschool sample, that rhythm synchronization ability is associated with stronger preliteracy skills and more precise, consistent subcortical auditory encoding of speech, including under noisy conditions. Rhythmic drumming performance and FFR metrics may serve as early, complementary, objective windows into children’s readiness for reading and potential risk for later literacy difficulties. Future research should include longitudinal tracking to determine developmental trajectories (particularly for children synchronizing at only one rate), evaluate whether rhythm or auditory training can strengthen FFRs and preliteracy skills, examine attentional contributions more deeply, and test generalization across additional stimuli and real-world listening environments.

- The phonological awareness measure was only administered to children aged ≥4 years (108/156), reducing sample size for that analysis. - Noise presentation was limited to the [da] stimulus; generalization to other phonemes in noise was not tested. - The Synchronizers-at-one-rate-only subgroup showed inconsistent patterns; their classification and developmental implications remain uncertain and require longitudinal follow-up. - Observed sex differences in rhythm performance necessitated controlling for sex; residual confounding by demographic or experiential factors (e.g., musical exposure, SES) cannot be fully excluded. - Cross-sectional design precludes causal inferences about rhythm ability leading to changes in auditory encoding or preliteracy skills.