Summary:
Researchers at Vanderbilt University Medical Center have shown that using detailed audiometric test results linked to genetic biobank data yields far better predictions of sensorineural hearing loss than relying on broad diagnostic billing codes. This precision phenotyping approach—using quantitative hearing thresholds rather than coarse ICD labels—strengthens genetic discovery, enables earlier risk screening, and lays the groundwork for personalized interventions before clinical hearing loss appears.
Key Facts:
- Precision Phenotypes Outperform Billing Codes: Polygenic risk scores built from quantitative audiometric measurements predicted sensorineural hearing loss substantially better than scores based on standard diagnostic billing codes.
- Large-Scale Biobank Integration: The study combined de-identified audiometric thresholds from about 16,000 individuals with Vanderbilt’s BioVU genomic repository to identify specific genetic variants associated with auditory impairment.
- Foundation for Proactive Care: High-resolution genetic profiling informed by audiometric patterns could enable clinicians to identify patients at elevated risk before symptoms arise and to tailor emerging gene and molecular therapies to particular frequency losses.
Source: Vanderbilt University Medical Center
Sensorineural hearing loss (SNHL) is among the most common sensory impairments worldwide, yet identifying individuals early enough to change clinical outcomes has been difficult. Traditional genetic studies of hearing loss often depend on large electronic health record (EHR) databases, which typically rely on diagnostic billing codes that give only a coarse view of auditory function.
Diagnostic codes such as ICD-9 and ICD-10 are convenient for large-scale analyses, but they mask important differences in the degree, pattern, and frequency-specific thresholds of hearing loss. Two patients with the same diagnostic code can have very different audiograms, and this heterogeneity blunts the power of genetic association studies when codes are used as the primary phenotype.
A multidisciplinary team of audiologists, otolaryngologists, and bioinformaticians at Vanderbilt addressed this limitation by applying “precision phenotyping”: they replaced crude case/control labels with exact audiometric measures, specifically pure-tone averages (PTAs), and integrated those measurements with genomic data from BioVU.
Moving Beyond Coarse Diagnostic Codes
In many biobank-based genome-wide association studies (GWAS), researchers define cases and controls using diagnostic codes because those codes are readily accessible across millions of EHRs. While practical, this approach treats hearing as a binary condition and overlooks its continuous, frequency-dependent nature.
Co-first author Andie DeFreese, AuD, explains that diagnostic codes introduce ambiguity: patients assigned the same code can differ substantially in severity and audiometric profile, which adds noise and reduces the ability to detect meaningful genetic signals.
The Power of BioVU and Audiometric Precision
To improve genetic discovery, the research team linked de-identified, full audiometric records—including hearing thresholds across standard frequencies—to whole-genome data from approximately 16,000 individuals in Vanderbilt’s BioVU biobank. They then ran parallel GWAS analyses: one using traditional diagnostic-code–based phenotypes and one using continuous PTA measures as the phenotype.
From those analyses, the polygenic risk scores generated from PTA-based GWAS had markedly better predictive performance when validated in an independent cohort. The PTA-based approach identified multiple genome-wide significant loci and produced higher heritability estimates than the ICD-based analysis, which identified no significant loci. These findings illustrate that quantitative audiometric data capture the genetic architecture of hearing sensitivity more effectively than billing codes.
DeFreese emphasized that pairing granular clinical tests with genomic resources improves risk prediction not only for hearing loss but for other medical domains where clinical measurements provide richer phenotypes than diagnostic codes alone.
Toward Early Detection and Precision Therapeutics
Accurate genetic risk prediction for hearing loss has important clinical implications. Identifying high-risk individuals early can guide preventive measures such as tailored hearing conservation, targeted monitoring, and timely auditory rehabilitation—steps that could slow or mitigate functional decline before significant hair cell loss occurs.
Precision genomic profiling also has growing relevance for emerging therapies. As gene and molecular treatments for genetic forms of hearing loss advance, detailed genotype-phenotype maps based on audiometric data can help match patients to therapies most likely to preserve or restore hearing across specific frequency ranges.
Corresponding author Taha Jan, MD, notes that this work showcases how clinical and research teams can move precision medicine forward in auditory health by integrating clinical audiology with high-quality genomic data.
Future plans for the research group include refining genomic models to predict not just overall risk but the specific frequency configurations and progression patterns most likely to deteriorate, offering clinicians a more actionable framework for lifelong hearing care.
Funding: Supported by a Vanderbilt Lacy-Fischer Interdisciplinary Grant, an American Otological Society Fellowship Grant, and the National Institute on Deafness and Other Communication Disorders (NIH) under grants R03DC021550, R21DC021276, R21DC023019, K08DC019683, and R21DC022058.
Editorial Notes:
- This article was edited by a Neuroscience News editor.
- The journal paper was reviewed in full by the editorial team.
- Additional context and editorial adjustments were provided by staff writers.
About this Neurology and Neuropharmacology Research:
- Media Contact: Evan Dorian
- Source: Vanderbilt University Medical Center
- Image Credit: Image credited to Neuroscience News
- Original Research (Open Access): JAMA Otolaryngology–Head & Neck Surgery (Oct 1, 2026). Title: “Precision Phenotyping With Audiometric Data and Gene Discovery for Sensorineural Hearing Loss.” Authors: Andrea J. DeFreese, AuD; Tanguy Rubat du Mérac, MSc; Quanhu Sheng, PhD; Srishti Nayak, PhD; and Taha A. Jan, MD.
- DOI: 10.1001/jamaoto.2026.3089
Abstract
Precision Phenotyping With Audiometric Data and Gene Discovery for Sensorineural Hearing Loss
Importance
Understanding the genetic architecture of hearing sensitivity as a continuous quantitative trait is essential for advancing precision medicine in auditory health. Despite many GWAS efforts, few robust genetic drivers of hearing sensitivity have been consistently identified when relying on coarse phenotypes.
Objective
To compare the effectiveness of precision phenotyping using pure-tone averages (PTAs) versus common diagnostic-code–based phenotyping for detecting genetic variants associated with sensorineural hearing loss (SNHL).
Design, Setting, and Participants
This genetic association study performed two GWAS analyses using different phenotyping strategies: (1) a diagnostic-code–based case/control GWAS relying on ICD-9 and ICD-10 codes related to hearing loss, and (2) a precision phenotyping GWAS using continuous PTA measurements (three-frequency PTA of the better-hearing ear) from audiograms. Analyses used whole-genome sequencing data from individuals of European genetic ancestry drawn from a single tertiary center biobank. Data analysis occurred from August 2025 to April 2026.
Main Outcomes and Measures
GWAS summary statistics were used to estimate single-nucleotide variant (SNV)–based heritability for PTA and ICD phenotypes, map genetic risk loci to candidate genes through functional annotation, evaluate genetic correlations with existing GWAS, and calculate polygenic risk scores validated in an independent cohort from the NIH All of Us Research Program.
Results
The diagnostic-code GWAS included 10,164 cases and 51,305 controls, while the PTA-based GWAS included 16,057 participants with audiometric data. The PTA-based analysis identified three genome-wide significant loci mapped to four genes (EML6, SPTBN1, ARHGEF28, and EYA4) and produced a higher SNV-based heritability estimate (h2 ≈ 11.78%). The ICD-based GWAS found no significant loci and had a lower heritability estimate (h2 ≈ 2.90%). PTA-derived polygenic risk scores correlated strongly with hearing-related traits and were significantly associated with self-reported deafness in an independent cohort, whereas ICD-derived scores were not.
Conclusions and Relevance
Precision phenotyping using quantitative audiometric PTAs enhances genetic discovery for sensorineural hearing loss compared with diagnostic-code–based methods. Higher heritability, identification of genome-wide significant loci, and stronger polygenic risk prediction indicate that quantitative hearing measures better reflect the underlying genetic architecture, supporting the integration of precision phenotyping into hearing-health research and clinical practice.
