Test Equivalency of the Spanish and English AzBio Sentences among Bilingual Normal Hearers

Language influences speech perception in bilingual listeners, with non-native languages often requiring higher signal-to-noise ratios (SNRs) for comprehension. Structural features, such as phonology and syllabic rhythm, make Spanish easier to perceive in noise compared to English, which relies more heavily on high-frequency cues and complex consonant clusters. These differences have clinical implications, particularly for individuals with hearing loss, as speech-in-noise tasks guide treatment decisions, including amplification and cochlear implant candidacy. However, most assessment tools were developed for English speakers, limiting their accuracy for individuals with limited English proficiency. Spanish-language tests, such as the AzBio, often use low-context Spanish sentences presented with English babble, creating a linguistic mismatch that increases informational masking and cognitive load. This can reduce speech intelligibility, misrepresent actual hearing abilities, and potentially contribute to healthcare disparities. Addressing these limitations requires evaluation of how target language, masker language, and sentence structure influence speech perception in bilingual populations to ensure valid assessment and equitable access to hearing healthcare.

Twenty normal-hearing bilingual adults (mean age 32 years, 65% male) who were fluent in both English and Spanish participated in the study. Participants were classified as English- or Spanish-dominant based on self-report and proficiency measures. Speech-in-noise testing was assessed using the English and Spanish versions of the Hearing in Noise Test (HINT) and AzBio sentence test. An additional condition was included using the Spanish AzBio sentences presented with Spanish multi-talker babble.

Bilingual listeners required significantly higher SNRs for English than Spanish tests (HINT mean difference = 2.77 dB, 95% Highest Density Interval [HDI]: 1.64-3.88; AzBio mean difference = 2.5 dB, 95% HDI: -3.54 to -1.53). No overall differences emerged between Spanish AzBio with English versus Spanish babble; however, subgroup analyses revealed language-dominant effects. Spanish-dominant listeners performed worse when target and masker were both Spanish (mean difference = -1.64 dB, 95% HDI: -2.78 to -0.48), whereas English-dominant listeners showed greater difficulty with English stimuli. Higher self-reported proficiency in a given language correlated with lower SNRs on corresponding measures.

Language dominance and structural features of Spanish versus English significantly affect bilinguals' speech perception in noise. These findings underscore the need for linguistically appropriate test materials, as reliance on English-only tools may misrepresent communication abilities and contribute to disparities in hearing healthcare.