Auditory Discrimination-A Missing Piece of Speech and Language Development: A Study on 6-9-Year-Old Children with Auditory Processing Disorder

The paper examines how higher auditory functions, particularly auditory discrimination (perceiving differences in sound volume, duration, and frequency), relate to speech and language development in children. Auditory processing disorders (APD) are characterized by impaired central processing of sound despite normal peripheral hearing, leading to difficulties in understanding and memorizing sounds, with downstream impacts on language and literacy. The study aims to assess differences in frequency discrimination and phoneme discrimination between children with APD and typically developing (TD) peers aged 6–9 years, and to evaluate whether phoneme discrimination matures by early school age or continues developing. The authors hypothesize that children with APD perform worse on phoneme discrimination tasks and that phoneme discrimination development continues into school age. They also consider whether the phoneme discrimination test (PDT) should be included in standard APD diagnostic batteries.

Prior work indicates that deficits in frequency discrimination can contribute to difficulties in speech sound (phoneme) discrimination, potentially affecting speech perception, articulation, and literacy. The timeframe for maturation of phoneme discrimination is debated: some studies report early maturation, while others show continued refinement through school age and even adolescence. Studies have often used limited phonological contrasts or real words, potentially causing ceiling effects and masking subtle deficits. Recommendations from audiological associations emphasize assessing phoneme discrimination in APD workups. Comparative studies across languages (Portuguese, Japanese, Dutch, French) have used varying numbers of contrasts, with mixed findings regarding age effects. Nonsense word materials may better isolate auditory-perceptual skills by removing lexical/semantic cues and reducing ceiling effects.

Design and participants: Cross-sectional assessment of 366 Polish-speaking children aged 6–9 years with normal peripheral hearing: 220 diagnosed with APD and 146 typically developing (TD) controls. Exclusions: intellectual disability, developmental disorders (e.g., ASD), neurological problems (including epilepsy), and congenital genetic disorders. No SLI-diagnosed participants were included. Recruitment: patients referred for auditory difficulties/APD suspicion at the Audiology Department, Children's Memorial Health Institute (CMHI), Warsaw, and students from three Warsaw primary schools. Grouping: APD defined by more than one psychoacoustic test result below age norms (verbal and non-verbal), with normal cognition and hearing; TD had at most one result below norm. Sex distribution was similar (about two-thirds male) and not analyzed by sex.

Audiological screening: Pure-tone audiometry (125–8000 Hz; normal <20 dB HL), speech audiometry (AC40, Interacoustics; TDH39 headphones), tympanometry (type A per Jerger), and distortion product otoacoustic emissions (Titan, Interacoustics) confirmed normal peripheral and middle ear function.

Higher auditory function tests (ATS Neuroflow platform):

• Adaptive Speech in Noise Test (ASPN): speech perception in multitalker babble; SNR threshold derived from adaptive procedure.
• Dichotic Digits Test (DDT) with distracted attention: assesses central auditory maturity and hemispheric dominance.
• Frequency Pattern Test (FPT): 20 three-tone sequences (high 1020 Hz, low 880 Hz); child names order (e.g., high-low-high). Outcome: percent correct sequences. Assesses frequency discrimination, auditory short-term memory, and right-hemisphere function.

Phoneme Discrimination Test (PDT): Standardized test from the Educational Research Institute battery; 25 pairs of nonsense words with identical syllable structure: 18 pairs differ by one distinctive feature (single phoneme) across voicing (5), manner (4), place (7), and vowel nasality (2); 7 identical pairs as controls. One SLP presented pairs at natural rate and intonation; child judged “same” or “different.” Scoring: 1 point per correct, max 25; duration ~5 minutes. Nonsense words eliminate semantic influence.

Procedure: All auditory function tests administered in one day in an acoustically treated room by the same certified Neuroflow provider using an AC40 audiometer and TDH39 headphones at 60 dB sensation level. Caregivers completed a symptom questionnaire. Total higher-auditory testing time ~20 minutes, with fixed test order.

Statistical analysis: Normality checked with Kolmogorov–Smirnov and Lilliefors tests (and Shapiro–Wilk in appendices). Due to non-normal distributions, non-parametric Kruskal–Wallis ANOVA with post hoc multiple comparisons tested group and age effects (significance p<0.05). FPT results also expressed as percentage of age norms (reference values in Appendix). Correlations between FPT and PDT assessed with Spearman’s rank and Pearson’s correlation (significance p<0.05).

• Frequency Pattern Test (FPT): Across ages 6–9, APD participants scored markedly lower than TD peers. Overall median FPT was 20% in APD vs 50% in TD (p<0.05). Means (SD): APD 24.91% (19.25) vs TD 50.16% (27.27). Differences between APD and TD were significant (Kruskal–Wallis p=0.0000). Within-group age differences for FPT were not significant (APD p=0.5316; TD p=0.3937).
• FPT relative to age norms: Median for APD ≈45% of age standard vs 100% for TD; APD results were more than twice as low as TD across all ages. Notably, 9-year-old APD participants had worse FPT medians than TD 6-year-olds (30% vs 40%; p<0.05).
• Phoneme Discrimination Test (PDT): TD outperformed APD at all ages (Kruskal–Wallis p=0.0000). Overall means (SD): APD 20.4 (3.37) vs TD 23.0 (2.1); medians 21 vs 24. Age effects within groups: APD showed significant differences between ages 6 and 8 (p=0.000) and 6 and 9 (p=0.000); TD showed a significant difference between ages 6 and 9 (p=0.029). Nine-year-old APD children had a median PDT of 22 vs 23 for TD 6-year-olds.
• Correlations: FPT and PDT results were significantly correlated in both APD and TD groups (Spearman and Pearson; p<0.05), with at least fair-strength correlations and slightly higher coefficients in the TD group.
• Developmental trajectory: Both FPT and PDT scores increased with age in both groups, indicating ongoing maturation of auditory discrimination and phoneme discrimination during school age. In TD, PDT variability (SD) decreased with age, suggesting convergence among peers; APD variability did not linearly decrease, reflecting heterogeneous development and potential intervention effects.

Findings confirm that children with APD exhibit significant deficits in both non-speech frequency pattern discrimination (FPT) and speech phoneme discrimination (PDT) relative to TD peers. The association between large FPT deficits and APD supports FPT as an indicator of central auditory dysfunction. Significant, fair correlations between FPT and PDT in both groups suggest shared or related auditory processing mechanisms underpinning non-speech and speech discrimination. Importantly, phoneme discrimination did not plateau at phoneme acquisition but continued to improve through ages 6–9, with TD children not yet fully equalized by age 9. Using nonsense words minimized lexical/semantic cues and reduced ceiling effects seen in word-based tests, enabling detection of subtle but clinically relevant PDT deficits. These results align with prior literature indicating continued maturation of categorical speech perception into later childhood and adolescence and support incorporating phoneme discrimination assessment into APD diagnostics to inform individualized therapy.

The study demonstrates that 6–9-year-old children with APD have pronounced deficits in both frequency and phoneme discrimination compared with TD peers. Even at age 9, children with APD did not reach the performance level of TD 6-year-olds on FPT and PDT. Key conclusions: (1) phoneme discrimination development continues through school age rather than ending at phoneme acquisition; (2) the depth of FPT deficits corresponds with APD occurrence, with APD results about half those of TD peers; (3) nonsense-word PDT is sensitive and language-proficiency independent, making it suitable for APD assessment. The authors recommend including the PDT alongside FPT in APD diagnostic batteries to guide individualized therapeutic programs. Future research should examine working memory and cognitive influences, compare APD with SLI populations, and apply more advanced statistical models (e.g., linear mixed-effects).

• Working memory capacity, which may affect FPT performance, was not assessed.
• No dedicated group with specific language impairment (SLI) was tested; comparisons between APD and clearly diagnosed SLI are needed.
• More complex statistical models common in neuropsychology (e.g., linear mixed-effects models) were not applied; the authors plan to use them in future analyses.