Can Hearing Loss Be Predicted Before Symptoms Appear? New Research Explores Early Detection

By combining detailed audiometric testing data with genetic biobanks rather than relying on generic medical billing codes, researchers have significantly improved their ability to predict sensorineural hearing loss. This precision phenotyping approach paves the way for early genetic risk screening and personalized therapeutic interventions before clinical hearing loss manifests. Precision Phenotypes Outperform Billing Codes: Polygenic risk scores derived from quantitative audiometric data predicted sensorineural hearing loss far more accurately than models built using standard diagnostic billing codes.

Large-Scale Biobank Integration: The team integrated de-identified audiometric hearing thresholds from 16,000 individuals with Vanderbilts BioVU genomic repository to map specific genetic variants linked to auditory impairment. Foundation for Proactive Interventions: Earlier, high-resolution genetic profiling could soon allow clinicians to identify at-risk patients prior to symptom onset and tailor emerging gene therapies to specific frequency configurations. Sensorineural hearing loss is one of the most common sensory deficits globally, yet detecting it early enough to alter its clinical trajectory has remained a significant hurdle in modern medicine.

What we know

Traditionally, uncovering the genetic underpinnings of complex sensory conditions has relied heavily on massive electronic health record (EHR) databases. However, these repositories predominantly track patient conditions through diagnostic billing codes—coarse labels that often fail to capture the subtle, subjective, and physiological variations inherent in auditory impairment. Now, a multidisciplinary team of clinical audiologists, otolaryngologists, and bioinformaticians at Vanderbilt University Medical Center (VUMC) has demonstrated that replacing generic diagnostic codes with granular clinical test results, a concept known as “precision phenotyping”, substantially enhances the discovery of genetic variants associated with hearing loss.

What we know

The findings, published in JAMA Otolaryngology–Head & Neck Surgery, point toward a future where patients predisposed to hearing loss can be reliably identified before noticeable auditory decline occurs. In typical genetic biobank studies, researchers rely on International Classification of Diseases (ICD) diagnostic codes to divide individuals into “cases” or “controls.” While expedient for processing millions of patient charts, these billing codes act as an imprecise proxy for sensory function. “Diagnostic codes are very common because thats the data thats available in large genetic biobanks that facilitate this research,” explained co-first author Andie DeFreese, AuD, a clinical audiologist and doctoral candidate in the Department of Hearing and Speech Sciences at Vanderbilt Health.

“Using them leads to a kind of gray area in which were not able to accurately define who has hearing loss and who doesnt.” Because two patients sharing an identical diagnostic code might possess drastically different degrees, patterns, or frequency thresholds of hearing loss, genetic associations derived from these codes tend to be blunted by noise in the data. To overcome this limitation, the Vanderbilt team leveraged BioVU, Vanderbilt Healths extensive biobank containing de-identified genetic information linked to electronic health records. The investigators de-identified and cross-referenced complete clinical audiometric records, including comprehensive hearing threshold measurements across various audio frequencies, with the genomic data of 16,000 consenting individuals.

Using this rich dataset, the researchers constructed polygenic risk scores (PRS) based on precise quantitative audiometric measurements and tested them in an independent validation cohort. The polygenic risk scores established via precision phenotyping demonstrated significantly superior predictive power for hearing loss compared to risk scores modeled strictly on standard diagnostic codes. “In pairing these resources, we can use a more precise phenotype for hearing loss instead of diagnostic codes that can be inaccurate,” DeFreese noted.

Key details

“This better prediction power tells us the importance of precision phenotyping not just for hearing loss but for all disciplines in medicine that take this risk prediction approach.” The ability to forecast hearing loss at the genetic level carries vital clinical implications. Proactive identification can inform early auditory rehabilitation, noise-protection strategies, and monitoring protocols well before hair cell degradation translates into functional communication deficits. Furthermore, precision genomics is increasingly critical as biological and molecular therapeutics enter the auditory health landscape.

Why it matters

“Genetics is becoming increasingly relevant for precision therapy, especially now that the Food and Drug Administration has approved its first gene therapy for genetic hearing loss,” said corresponding author Taha Jan, MD, Assistant Professor of Otolaryngology–Head and Neck Surgery at VUMC. “This work is an example of how our world-class clinicians and scientists at Vanderbilt Health are pushing the boundaries of precision medicine.” Looking ahead, the research group plans to refine these genomic models to predict not just the presence or absence of hearing loss, but also the specific auditory configurations and frequencies likely to deteriorate, providing an unprecedented roadmap for lifelong hearing health. Funding: The research was funded by a Vanderbilt Lacy-Fischer Interdisciplinary Grant, an American Otological Society Fellowship Grant and by the National Institute on Deafness and Other Communication Disorders, part of the National Institutes of Health (grants R03DC021550, R21DC021276, R21DC023019, K08DC019683 and R21DC022058).