Tinnitus and COVID-19: effect of infection, vaccination, and the pandemic

The study addresses whether and how tinnitus in the general population is associated with acute COVID-19 infection, long COVID, or COVID-19 vaccination, and how co-occurring otologic (hearing loss, hyperacusis) and psychological (depression, anxiety) factors relate to tinnitus onset and severity. Motivated by mixed clinical reports and limited population-level evidence, the authors aim to: (1) estimate the prevalence of self-reported tinnitus believed to develop after COVID infection, long COVID, or vaccination (COVID-associated tinnitus); (2) determine how otologic or psychological symptoms are linked to COVID-associated tinnitus and to changes in pre-existing tinnitus during the pandemic; and (3) compare tinnitus severity among groups and identify predictors of severity.

Early case reports and hospital-based retrospective studies documented otologic manifestations (hearing loss, tinnitus, vertigo) during or after COVID-19 infection, with variable tinnitus prevalence (1.2–23.2%) in infected individuals and uncertainty about the roles of infection versus treatment medications. Long COVID, defined by persistent symptoms beyond two months without alternative diagnoses, has been associated with tinnitus in limited, often observational studies with potential biases and restricted geography, indicating a need for broader population-level data. As vaccination campaigns progressed, common immediate side effects were noted, followed by reports of audiovestibular events (SSNHL, tinnitus) in surveillance systems (MHRA, VAERS), although these sources lack detail on co-occurring factors. Clinic-based reports suggested 1.0–6.2% tinnitus after vaccination, complicating attribution among people who had both infection and subsequent vaccination. Beyond physiological factors, pandemic-related stressors (social isolation, economic hardship, sleep disturbance, depression, anxiety) have been linked to increased tinnitus bother, though some studies found no significant changes in pre-existing tinnitus, highlighting heterogeneity and the importance of psychological sequelae.

Design: Cross-sectional online survey (Qualtrics), conducted October 19, 2022–August 29, 2023, under UIUC IRB protocol #23379. Recruitment via emails, social media, and tinnitus support organizations; geo-tracking disabled and no identifiable information collected; respondents confirmed non-residency in China. Participants: Adults ≥18 years with pre-existing tinnitus prior to the pandemic or tinnitus onset believed associated with COVID infection, long COVID, or COVID vaccination (including resolved cases). Consent obtained online. Measures: 53 survey questions plus the Tinnitus Functional Index (TFI; 25 items, 8 subscales). Sections: (1) demographics (age, gender, country, race/ethnicity); (2) tinnitus (current status, duration, laterality, sound type; self-rated loudness 0–10; annoyance 0–10; TFI); classification into four groups by self-judged etiology: COVID infection, long COVID, COVID vaccination, pre-existing tinnitus. Group-specific follow-ups probed onset timing, test results, healthcare consultations, treatments, and effectiveness; pre-existing group queried pandemic-related changes and attributions. (3) Hearing loss (presence, laterality, degree, duration, cause; for vaccination-attributed cases, vaccine brand and onset timing). (4) Comorbidities: hyperacusis identified via self-report of “hypersensitivity to sounds”; depression and anxiety screened via PHQ-4 (two items each, scored 0–3; sum >2 indicates condition). Statistical analysis: R (v4.2.1). Descriptive statistics (Table 1). Group comparisons: Chi-square tests with Bonferroni-corrected pairwise proportion tests for categorical variables; Shapiro-Wilk tested normality of age, followed by Kruskal–Wallis and Dunn’s tests with Bonferroni correction for non-parametric age comparisons. Logistic regression: age- and gender-adjusted odds ratios (ORs, 95% CIs) for factors associated with COVID-associated tinnitus; in pre-existing tinnitus, ORs for factors associated with reported changes after COVID infection, long COVID, or vaccination. Tinnitus severity: TFI non-normality confirmed (Shapiro–Wilk p<0.05); Kruskal–Wallis with Dunn’s post-hoc for group differences. Predictors of TFI: multiple linear regressions per group with dichotomous predictors (gender, hearing loss, hyperacusis, depression, anxiety) and continuous age; examined outliers/influential observations; sensitivity analyses assessed robustness. Significance threshold p<0.05.

Sample: 1,511 respondents. Group distribution: COVID infection n=134 (8.87%), long COVID n=136 (9.00%), COVID vaccination n=562 (37.19%), pre-existing tinnitus n=679 (44.94%). Recovery among COVID-associated groups: infection 8 respondents, long COVID 5, vaccination 39 (recovery rates ~3.68–6.94%). Demographics: Pre-existing group had older median age (63 years; H(3)=166.9, p<0.001) and highest male proportion (55.82%) and hearing loss (64.65%), but lowest depression (18.7%) and anxiety (17.38%). Long COVID group had highest hyperacusis (54.41%), depression (33.09%), anxiety (28.68%). Group differences significant for gender, hearing loss, hyperacusis, depression, anxiety; no significant differences among the three COVID-associated groups. COVID-associated tinnitus and factors (age- and gender-adjusted): Infection group had decreased odds of hyperacusis (OR 0.68; 95% CI 0.46–0.98). Vaccination group had decreased odds of hearing loss (OR 0.53; 95% CI 0.43–0.67) and increased odds of anxiety (OR 1.44; 95% CI 1.12–1.84). Long COVID group showed no significant associations with examined factors. Pre-existing tinnitus changes: Of 679 with pre-existing tinnitus, 294 (43.3%) reported changes (worse 283, 41.68%; better 11, 1.62%); none of the examined factors significantly altered odds of reporting changes (Table 3). Tinnitus severity: Median TFI scores differed among groups (H(3)=95.41, p<0.001); medians—long COVID 60.8, vaccination 56.4, infection 50.8, pre-existing 42. All COVID-associated groups had significantly higher TFI than pre-existing. Predictors of TFI (multiple regression): Infection group (n=125): R²=0.271; significant predictors—gender (male, β=-0.225, p=0.006), hearing loss (β=0.193, p=0.019), anxiety (β=0.352, p=0.002). Long COVID (n=131): R²=0.279; significant—hyperacusis (β=0.19, p=0.019), depression (β=0.438, p<0.001). Vaccination (n=523): R²=0.261; significant—age (β=0.128, p<0.001), hearing loss (β=0.115, p=0.003), hyperacusis (β=0.099, p=0.012), depression (β=0.273, p<0.001), anxiety (β=0.255, p<0.001). Pre-existing (n=674): R²=0.256; significant—age (β=0.076, p=0.034), gender (male, β=-0.067, p=0.048), hearing loss (β=0.169, p<0.001), hyperacusis (β=0.15, p<0.001), depression (β=0.276, p<0.001), anxiety (β=0.214, p<0.001). Sensitivity analyses indicated most effects were robust; variance explained increased >45% in infection and long COVID models after removing outliers, reflecting small sample size sensitivity.

The study demonstrates that self-reported COVID-associated tinnitus differs from pre-existing tinnitus in demographic and comorbidity profiles and is associated with higher tinnitus severity. COVID infection-associated tinnitus was less likely to co-occur with hyperacusis, while vaccination-associated tinnitus was less likely to co-occur with hearing loss and more likely with anxiety—potentially reflecting population differences from clinic-based SSNHL reports and the role of vaccination-related stress or nocebo effects. Long COVID-associated tinnitus did not show significant associations with the examined factors, possibly due to smaller sample size or lower population prevalence. Among those with pre-existing tinnitus, reported changes during COVID-related events were common but not significantly linked to the studied otologic or psychological factors. Across all groups, depression and/or anxiety were strong predictors of tinnitus severity, affirming the central role of psychological states in tinnitus distress. The findings support the need for comprehensive assessment of otologic and psychological symptoms and tailored management for individuals with COVID-associated tinnitus.

Several otologic (hearing loss, hyperacusis) and psychological (anxiety, depression) factors are correlated with self-reported COVID-associated tinnitus, and COVID-associated groups exhibit higher tinnitus severity than those with pre-existing tinnitus. Depression and/or anxiety consistently and substantially contribute to tinnitus severity across groups. The results advocate for patient-centered, multidisciplinary care—incorporating audiology, otolaryngology, psychology, and neurology—and for clinical assessments that better quantify hearing and psychological disorders in COVID-associated tinnitus. Future research should include longitudinal designs with control groups and comprehensive audiological/physiological measures to clarify causality, mechanisms, and long-term effects, and to inform evidence-based interventions for COVID-associated tinnitus.

• Tinnitus was defined broadly without subtype differentiation (e.g., somatosensory tinnitus), potentially oversimplifying associations. - Potential sample and recall biases due to online dissemination and self-report; unequal group sizes may affect statistical power. - Hyperacusis and hearing loss were assessed via single survey items; PHQ-4, while validated, is brief relative to comprehensive instruments. - Cross-sectional design precludes causal inference and assessment of long-term effects. - Sensitivity to outliers in smaller groups; limited variance explained suggests unmeasured contributors (e.g., tinnitus loudness, personality, sleep). - Physiological mechanisms underlying increased tinnitus incidence after infection or vaccination remain unresolved; longitudinal studies with controls and detailed audiological/physiological assessments are needed.