Authors: Navid Shahnaz¹, Ragul Loganathan¹, Mohammadreza Rahmani Manesh¹, Eytan A. David¹, Valter Ciocca¹
¹University of British Columbia
Background: Conventional audiometry and word recognition scores (WRS) in quiet often fail to explain why some individuals struggle with speech in noisy environments. Extended high-frequency (EHF; 9–16 kHz) hearing may provide a more sensitive index of auditory dysfunction that is not captured by conventional high-frequency pure-tone averages (HFPTA; 1–4 kHz) or standard speech testing.
Method: This retrospective study analyzed 1,095 ears from 593 adults (aged 18–82) with normal HFPTA (≤25 dB HL) and excellent WRS (≥88%). Participants were categorized by age: 18–39, 40–59, and ≥60 years. The primary predictors were the EHFPTA (average thresholds from 9–16 kHz) and their relationship to QuickSIN performance (modeled as both a continuous metric and categorically using ≥3 dB and ≥7 dB SNR loss thresholds). Bayesian regression was used to adjust for HFPTA, WRS, and age.
Results: Despite “normal” clinical results, speech-in-noise deficits were prevalent: 62.8% of ears showed an SNR loss ≥3 dB. Within the subset of ears with normal HFPTA and WRS ≥76%, every 10 dB worsening in EHFPTA was associated with an increased likelihood of mild QuickSIN SNR loss (OR = 1.12, 95% CrI [1.01, 1.24]). While the association with more severe SNR loss (≥7 dB) was weaker, adding EHFPTA to the regression models consistently improved predictive accuracy beyond conventional metrics alone.
Conclusion: Speech-in-noise deficits are common even when conventional audiograms and word recognition scores appear normal. EHF audiometry serves as a valuable clinical adjunct, helping to identify and explain mild listening difficulties that are otherwise overlooked by standard diagnostic batteries. These findings support the inclusion of EHF testing in the routine assessment of patients reporting difficulty in complex listening environments.
