Background. Subjective tinnitus is an auditory percept unrelated to an external sound source. The lack of curative treatments and limited understanding of its risk factors complicate the prevention and management of this distressing symptom. This study seeks to identify socio-demographic, psychological, and health-related risk factors predicting tinnitus presence (how often individuals perceive tinnitus) and severity separately, and their evolution over time. Methods Using the UK Biobank dataset which encompasses data on the socio-demographic, physical, mental and hearing health from more than 170,000 participants, we trained two distinct machine learning models to identify risk scores predicting tinnitus presence and severity separately. These models were used to predict tinnitus over time and were replicated in 463 individuals from the Tinnitus Research Initiative database. Finding Machine learning based approach identified hearing health as a primary risk factor for the presence and severity of tinnitus, while mood, neuroticism, hearing health, and sleep only predicted tinnitus severity. Only the severity model accurately predicted the evolution over nine years, with a large effect size for individuals developing severe tinnitus (Cohen's d = 1.10, AUC-ROC = 0.70). To facilitate its clinical applications, we simplified the severity model and validated a five-item questionnaire to detect individuals at risk of developing severe tinnitus. Interpretation This study is the first to clearly identify risk factors predicting tinnitus presence and severity separately. Hearing health emerges as a major predictor of tinnitus presence, while mental health plays a crucial role in its severity. The successful prediction of the evolution of tinnitus severity over nine years based on socio-emotional, hearing and sleep factors suggests that modifying these factors could mitigate the impact of tinnitus. The newly developed questionnaire represents a significant advancement in identifying individuals at risk of severe tinnitus, for which early supportive care would be crucial.

The authors have declared no competing interest.

LH was funded by the Horizon Europe Framework Programme (HORIZON) under the Marie Sklodowska-Curie Postdoctoral Fellowship (grant No. 101146406) and the Fondation des Gueules Cassées. LH and SS were supported by the Fondation pour l'Audition (grant No. FPA RD-2019-10). MF was funded by the Louise and Alan Edwards Foundation. CTS was funded by the Canadian Institutes of Health Research, the Institut TransMedTech and the Canada First Research Excellence Fund.

I confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained.

Yes

The details of the IRB/oversight body that provided approval or exemption for the research described are given below:

The dataset from UK Biobank analyzed in the study is available via application to the Access Management System athttps://www.ukbiobank.ac.uk . The TRI dataset is accessible upon requests, see https://tinnitusresearch.net/index.php/for-researchers/tinnitus-database . Ethical approval for the collection of the TRI database was obtained from the Ethics Committee of the University of Regensburg (protocol number 08/046).

I confirm that all necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived, and that any patient/participant/sample identifiers included were not known to anyone (e.g., hospital staff, patients or participants themselves) outside the research group so cannot be used to identify individuals.

Yes

I understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance).

Yes

I have followed all appropriate research reporting guidelines, such as any relevant EQUATOR Network research reporting checklist(s) and other pertinent material, if applicable.

Yes

The dataset from UK Biobank analyzed in the study is available via application to the Access Management System athttps://www.ukbiobank.ac.uk . The TRI dataset is accessible upon requests, see https://tinnitusresearch.net/index.php/for-researchers/tinnitus-database .