Abstract
Background
Tinnitus is a prevalent auditory disorder that significantly affects quality of life, with emerging evidence suggesting a connection between metabolic dysregulation and tinnitus pathogenesis. The Atherogenic Index of Plasma (AIP), a marker of atherogenic lipid metabolism, has been associated with cardiovascular disease risk but remains underexplored in the context of tinnitus.
Objective
This study aimed to investigate the relationship between AIP and tinnitus prevalence in U.S. adults, and to further examine whether hypertension mediates this association.
Methods
We analyzed data from 2,464 participants in the National Health and Nutrition Examination Survey (NHANES, 2005–2012). Tinnitus status was determined by self-report, and AIP was calculated as log10(triglycerides/HDL-C). Multivariable logistic regression models were used to assess the association between AIP and tinnitus. Stratified analyses and mediation analysis were conducted to explore effect modification and the mediating role of hypertension. Hypertension was defined as a self-reported physician diagnosis or use of antihypertensive medication.
Results
Elevated AIP was significantly associated with an increased risk of tinnitus in a dose–response manner (fully adjusted OR = 1.73, 95% CI: 1.19–2.52, P = 0.01). Stronger associations were observed in younger participants, females, non-Hispanic Whites, and smokers. Mediation analysis showed that hypertension partially mediated the association between AIP and Tinnitus, accounting for 17.53% of the total effect.
Conclusions
Higher AIP is associated with greater tinnitus prevalence, with hypertension serving as a partial mediator. These findings highlight the potential value of AIP as a metabolic biomarker for tinnitus risk stratification and emphasize the importance of managing metabolic dysregulation in tinnitus prevention.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12944-025-02716-1.
Keywords: Atherogenic Index of Plasma, Tinnitus, National health and nutrition examination survey, Metabolic Dysregulation, Dyslipidemia
Introduction
Tinnitus is a prevalent condition characterized by the perception of sound in the absence of external stimuli, affecting approximately 10% of the global population at some point in their lives [1, 2]. This complex disorder profoundly impacts quality of life, with a notable proportion of sufferers enduring persistent, debilitating symptoms that markedly compromise psychosocial functioning and occupational capacity [3, 4]. Tinnitus significantly impacts not only auditory perception but also psychological health, with a well-established correlation between tinnitus and psychiatric symptoms, particularly anxiety and depression. These comorbidities further exacerbate the condition's detrimental effects on the quality of life, contributing to a vicious cycle of mental health deterioration [5]. The clinical heterogeneity spanning aetiological pathways, perceptual characteristics, and symptom severity poses substantial therapeutic challenges, compounded by current management strategies'frequent inadequacy in addressing patients'existential perception of therapeutic nihilism [6].
Emerging pathophysiological insights reveal intriguing connections between metabolic dysregulation and tinnitus pathogenesis, despite persistent knowledge gaps in mechanistic understanding [7]. Accumulating epidemiological evidence implicates obesity, dysglycaemia, atherogenic dyslipidaemia, and metabolic syndrome as potential modulators of tinnitus trajectory, necessitating rigorous examination of these metabolic interactions [8, 9]. The Atherogenic Index of Plasma (AIP), derived from the logarithmic triglyceride/HDL-cholesterol ratio, has emerged as a novel biomarker demonstrating superior predictive capacity for cardiovascular risk stratification compared to conventional lipid parameters [10, 11]. Notably, AIP's dual role as both atherogenic predictor and metabolic status indicator positions it as a compelling candidate for investigating tinnitus-metabolism interactions.
Tinnitus can be classified into subjective and objective types, with the former being the most common and diagnosed through self-reports. The primary causes of tinnitus include exposure to loud noises, age-related hearing loss, and underlying conditions such as metabolic disorders and hypertension. Understanding these causes is essential for developing effective treatment strategies [12]. Critical appraisal of current literature reveals a conspicuous knowledge gap regarding AIP-tinnitus associations, despite well-established metabolic syndrome comorbidity patterns [13]. This oversight underscores the imperative to investigate metabolic-inflammatory pathways potentially mediating tinnitus generation and perpetuation. Elucidation of these mechanisms could catalyse paradigm shifts in diagnostic approaches and therapeutic development.
Our investigation leverages the National Health and Nutrition Examination Survey (NHANES) cross-sectional analytical framework to systematically evaluate AIP-tinnitus correlations [14]. This methodological approach facilitates multivariable adjustment for demographic confounders and clinical covariates, ensuring methodological rigor while optimising translational relevance [15]. The primary study objective involves determining whether AIP elevation independently predicts tinnitus prevalence, with particular emphasis on disentangling age-related effects from genuine metabolic associations.
Methods
Study design and participants
The data for this study were sourced from publicly available NHANES cycles conducted between 2005 and 2012. NHANES represents a long-term cross-sectional observational investigation designed to collect health-related data from a representative sample of the non-institutionalized U.S. population [16, 17]. The study protocol received approval from the National Center for Health Statistics (NCHS) Institutional Review Board, with written informed consent obtained from all participants. NHANES employs a sophisticated multistage probability cluster sampling design to ensure comprehensive and reliable data collection, maintaining rigorous methodological standards throughout the survey process.
This cross-sectional analysis utilized data from 7 NHANES cycles (2005–2012) involving 40,790 U.S. adults. After implementing exclusion criteria (age < 18 years, pregnancy, missing AIP/tinnitus data, or incomplete covariate records), 2,464 eligible participants were retained. Figure 1 illustrates the complete participant selection process.
Fig. 1.
Flowchart of the sample selection from NHANES 2005–2012
Assessment of tinnitus
The primary outcome of this study was tinnitus, assessed using self-reported data from the NHANES. Tinnitus status was determined through participant responses to question: “In the past year, have you been bothered by ringing, roaring, or buzzing in your ears or head that lasted for 5 min or longer?”. Using this criterion, participants were dichotomized into two mutually exclusive categories: no tinnitus (0) or present tinnitus (1). Among the study population, 778 individuals reported experiencing tinnitus during the reference period.
Calculation of atherogenic index of plasma
The AIP was calculated using participants'blood lipid parameters through the formula: AIP = log10(TG/HDL-C), where TG represents triglycerides (mmol/L) and HDL-C denotes high-density lipoprotein cholesterol (mmol/L) [18]. Recognized as a significant predictor of cardiometabolic risk, AIP values were stratified into quartile-based groups in this study to investigate their potential association with tinnitus.
The NHANES protocol received approval from the National Center for Health Statistics (NCHS) Institutional Review Board (Protocol #2005–06, approved in September 2005; renewed Protocol #2011–17, approved in October 2011). All participants provided written informed consent prior to participation.
Covariates
We included a comprehensive set of covariates in our analysis based on prior literature and clinical considerations [19–21]: age, sex, race, education level, smoking status, alcohol use, body mass index (BMI), waist circumference (measured in centimeters at the midpoint between the lower margin of the last palpable rib and the top of the iliac crest), fasting glucose, glycated hemoglobin (HbA1c), uric acid (UA), albumin, diabetes mellitus (self-reported physician diagnosis or use of antidiabetic medications) [22], hypertension (defined as a self-reported physician diagnosis of hypertension or current use of antihypertensive medications, consistent with NHANES protocols), and cardiovascular disease.
Statistical analysis
This study strictly followed the NHANES analytical protocols, accounting for complex sampling design features and incorporating mobile examination center (MEC) sample weights. Descriptive analyses characterized continuous variables using mean values with standard errors (SE) and categorical variables through frequency distributions with proportional percentages. The Shapiro–Wilk test was applied to assess the normality of continuous variables. For univariate comparisons, categorical variables were evaluated using Pearson's chi-square tests, while continuous variables were analyzed through either one-way analysis of variance (ANOVA) for normally distributed parameters or Kruskal–Wallis tests for non-normally distributed parameters, with a two-tailed significance threshold of P < 0.05. Multivariable weighted logistic regression models were applied to examine the association between AIP and Tinnitus across four sequential adjustment tiers. The crude model included only AIP, while Model 1 adjusted for core demographic confounders (age, sex, and race). Model 2 further incorporated body mass index (BMI), alcohol use, cardiovascular disease (CVD), marital status, hypertension, education level, waist circumference, smoking status, diabetes mellitus, fasting glucose, glycated hemoglobin (HbA1c), uric acid (UA), and albumin. Model 3 additionally accounted for noise exposure history and degree of hearing loss, as determined from NHANES pure-tone audiometry data. Sensitivity analyses were performed in specific subpopulations, such as after excluding individuals with severe hearing loss or restricting analyses to participants with confirmed noise exposure history. For trend analyses, AIP was modeled both as a continuous variable (median value) and as a categorical variable (quartiles coded as integers).
Mediation analysis was performed using the product-of-coefficients method (PROCESS macro in R) to quantify the proportion of the total effect of AIP on Tinnitus mediated by hypertension, diabetes, BMI, and smoke. Direct and indirect effects were estimated with 1,000 bootstrap resamples.
To further explore the association between the AIP and tinnitus prevalence, we performed stratified analyses to assess the potential effect modification by key demographic, clinical, and behavioral characteristics. Stratified analyses enable the identification of subgroups with differential susceptibility to the AIP-tinnitus association, providing insights into potential interactions between AIP and other factors.
All statistical analyses were performed using R software (version 4.1.0; R Core Team, Beijing, China) with key packages including"survey"(for weighted analyses), and"mice"(multiple imputation). A two-tailed alpha level of 0.05 was established for determining statistical significance.
Results
Participant demographics and clinical profiles
The cohort consisted of 2,464 participants stratified by AIP quartiles (Q1–Q4), revealing a dose-dependent increase in metabolic risk factors across ascending quartiles (Table 1). Significant associations were observed between AIP quartiles and various metabolic markers, such as BMI (Q1 = 26.01 vs Q4 = 31.79 kg/m2), waist circumference (Q1 = 90.68 vs Q4 = 108.84 cm), and glucose levels (Q1 = 5.48 vs Q4 = 6.49 mmol/L) (all P < 0.0001). Notably, the prevalence of tinnitus significantly increased across quartiles (Q1: 14.34%, Q4: 27.10%; P < 0.001), highlighting a clear association between higher AIP levels and tinnitus presence.
Table 1.
Baseline Characteristics by AIP Quartiles and Tinnitus Status (N = 2464)
| Characteristics | Total (n = 2464) | AIP(Q1) (n = 617) | AIP(Q2) (n = 616) | AIP(Q3) (n = 614) | AIP(Q4) (n = 617) | P-value (AIP) | Non-Tinnitus (n = 1991) |
Tinnitus (n = 473) |
P-value (Tinnitus) |
|---|---|---|---|---|---|---|---|---|---|
| Continuous | |||||||||
| Age (years) | 50.65(0.66) | 48.86(1.20) | 50.27(1.01) | 51.15(0.83) | 52.13(0.93) | 0.25 | 49.22(0.70) | 56.43(0.94) | < 0.0001 |
| BMI (kg/m2) | 28.81(0.26) | 26.01(0.42) | 27.74(0.35) | 29.40(0.28) | 31.79(0.44) | < 0.0001 | 28.51(0.24) | 30.01(0.47) | < 0.001 |
| Waist circumference (cm) | 99.57(0.64) | 90.68(0.94) | 96.64(0.84) | 101.21(0.63) | 108.84(1.17) | < 0.0001 | 98.62(0.54) | 103.43(1.27) | < 0.0001 |
| Glucose (mmol/L) | 5.87(0.06) | 5.48(0.05) | 5.67(0.06) | 5.79(0.04) | 6.49(0.19) | < 0.0001 | 5.83(0.06) | 6.04(0.12) | 0.04 |
| HbA1c (%) | 5.66(0.03) | 5.44(0.03) | 5.58(0.04) | 5.64(0.02) | 5.97(0.12) | < 0.0001 | 5.62(0.02) | 5.82(0.09) | 0.02 |
| Uric acid (µmol/L) | 331.58(3.34) | 296.00(3.91) | 316.50(4.53) | 344.17(6.62) | 366.17(4.09) | < 0.0001 | 328.96(3.65) | 342.27(5.21) | 0.02 |
| Albumin (g/dL) | 4.28(0.01) | 4.31(0.02) | 4.29(0.02) | 4.27(0.02) | 4.27(0.02) | 0.18 | 4.29(0.01) | 4.26(0.02) | 0.21 |
| AIP | - | - | - | - | - | - | −0.03(0.01) | 0.05(0.02) | < 0.0001 |
| Categorical | |||||||||
| Sex | < 0.0001 | 0.14 | |||||||
| Female | 1189(49.99) | 383(63.95) | 302(54.79) | 287(48.49) | 217(34.20) | 968(51.22) | 221(44.98) | ||
| Male | 1275(50.01) | 234(36.05) | 314(45.21) | 327(51.51) | 400(65.80) | 1023(48.78) | 252(55.02) | ||
| Race | < 0.0001 | 0.03 | |||||||
| Mexican American | 253(7.23) | 38(5.30) | 65(7.52) | 69(7.88) | 81(8.10) | 203(7.66) | 50(5.48) | ||
| Non-Hispanic Black | 504(8.81) | 198(14.56) | 129(9.32) | 115(7.45) | 62(4.40) | 421(9.17) | 83(7.36) | ||
| Non-Hispanic White | 1193(72.75) | 261(68.98) | 296(72.28) | 295(73.10) | 341(76.29) | 923(71.33) | 270(78.55) | ||
| Other | 514(11.21) | 120(11.15) | 126(10.88) | 135(11.57) | 133(11.22) | 444(11.84) | 70(8.61) | ||
| Education level | < 0.0001 | 0.16 | |||||||
| High school | 545(21.16) | 118(17.08) | 135(21.94) | 153(22.84) | 139(22.50) | 439(21.03) | 106(21.67) | ||
| < High school | 607(17.78) | 105(10.42) | 150(18.17) | 164(20.66) | 188(21.36) | 464(16.85) | 143(21.57) | ||
| > High school | 1312(61.06) | 394(72.49) | 331(59.90) | 297(56.49) | 290(56.14) | 1088(62.12) | 224(56.76) | ||
| Marital status | 0.8 | 0.57 | |||||||
| Married | 1304(55.51) | 290(54.97) | 323(53.66) | 349(56.40) | 342(56.92) | 1060(56.00) | 244(53.52) | ||
| Non-Married | 1160(44.49) | 327(45.03) | 293(46.34) | 265(43.60) | 275(43.08) | 931(44.00) | 229(46.48) | ||
| Alcohol user | 0.001 | < 0.001 | |||||||
| Former | 501(16.82) | 89(10.37) | 128(17.91) | 124(15.76) | 160(22.66) | 363(15.16) | 138(23.58) | ||
| No | 365(10.73) | 82(10.56) | 94(10.94) | 104(12.80) | 85(8.71) | 302(10.74) | 63(10.69) | ||
| Yes | 1598(72.45) | 446(79.08) | 394(71.15) | 386(71.44) | 372(68.63) | 1326(74.10) | 272(65.73) | ||
| Hypertension | < 0.0001 | < 0.0001 | |||||||
| No | 1298(57.44) | 372(67.01) | 337(60.14) | 310(57.28) | 279(46.33) | 1110(60.57) | 188(44.68) | ||
| Yes | 1166(42.56) | 245(32.99) | 279(39.86) | 304(42.72) | 338(53.67) | 881(39.43) | 285(55.32) | ||
| Diabetes mellitus | < 0.0001 | < 0.0001 | |||||||
| DM | 557(17.23) | 65(7.82) | 119(12.44) | 152(17.29) | 221(30.24) | 418(15.32) | 139(24.99) | ||
| No | 1433(64.24) | 458(77.73) | 387(71.28) | 328(64.49) | 260(45.10) | 1196(66.75) | 237(54.06) | ||
| PreDM | 474(18.53) | 94(14.45) | 110(16.29) | 134(18.22) | 136(24.66) | 377(17.94) | 97(20.95) | ||
| Smoking status | < 0.001 | 0.23 | |||||||
| Never | 1358(55.71) | 389(65.22) | 353(57.18) | 322(51.80) | 294(49.42) | 1135(57.16) | 223(49.80) | ||
| Former | 682(25.80) | 151(22.77) | 172(27.08) | 176(26.93) | 183(26.26) | 518(25.10) | 164(28.63) | ||
| Current | 424(18.49) | 77(12.01) | 91(15.74) | 116(21.27) | 140(24.32) | 338(17.73) | 86(21.56) | ||
| CVD | 0.01 | 0.1 | |||||||
| No | 2134(88.83) | 554(92.29) | 541(89.96) | 526(88.15) | 513(85.27) | 1756(89.66) | 378(85.46) | ||
| Yes | 330(11.17) | 63(7.71) | 75(10.04) | 88(11.85) | 104(14.73) | 235(10.34) | 95(14.54) | ||
| Tinnitus | < 0.001 | - | |||||||
| No | 1991(80.28) | 523(85.66) | 499(82.26) | 497(80.91) | 472(72.90) | - | - | ||
| Yes | 473(19.72) | 94(14.34) | 117(17.74) | 117(19.09) | 145(27.10) | - | - | ||
Data are presented as mean ± SD for continuous variables or n (%) for categorical variables. Statistical tests used: ANOVA for continuous variables, chi-square test for categorical variables. P-values are shown for comparisons between AIP quartiles and tinnitus status. Statistically significant p-values (p < 0.05) are highlighted. Alcohol use is defined as regular consumption of alcohol in the past month
Comparative analysis: tinnitus vs non-tinnitus cohorts
The tinnitus group (n = 473) exhibited higher metabolic markers than the non-tinnitus group (n = 1,991), including elevated BMI (30.01 vs 28.51 kg/m2, P < 0.001), waist circumference (103.43 vs 98.62 cm, P < 0.0001), and glucose levels (6.04 vs 5.83 mmol/L, P = 0.04) (Table 2). The AIP values were also significantly different between the two groups (Tinnitus: 0.05 ± 0.02 vs Non-Tinnitus: −0.03 ± 0.01; P < 0.0001), confirming that higher AIP levels were associated with the presence of tinnitus.
Table 2.
Analysis between AIP and Tinnitus
| Characteristic | Crude model | P-Vaule | Model1 | P-Vaule | Model2 | P-Vaule | Model 3 | P-Value | Sensitivity Analysis | P-Value |
|---|---|---|---|---|---|---|---|---|---|---|
| OR(95%CI) | OR(95%CI) | OR(95%CI) | OR(95%CI) | OR(95%CI) | ||||||
| AIP | 2.41(1.80,3.23) | < 0.0001 | 2.21(1.57,3.12) | < 0.0001 | 1.73(1.19,2.52) | 0.01 | 1.65(1.12,2.44) | 0.02 | 1.61(1.08,2.40) | 0.03 |
| AIP group | ||||||||||
| Q1 | Ref | Ref | Ref | Ref | Ref | |||||
| Q2 | 1.29(0.83,1.99) | 0.25 | 1.24(0.80,1.94) | 0.33 | 0.86(0.64,1.15) | 0.59 | 0.83(0.61,1.13) | 0.62 | 0.82(0.60,1.11) | 0.65 |
| Q3 | 1.41(0.85,2.32) | 0.17 | 1.33(0.79,2.23) | 0.27 | 0.87(0.64,1.19) | 0.57 | 0.85(0.62,1.17) | 0.6 | 0.84(0.61,1.16) | 0.62 |
| Q4 | 2.22(1.62,3.05) | < 0.0001 | 2.03(1.42,2.89) | < 0.001 | 0.72(0.54,0.95) | 0.04 | 0.69(0.51,0.93) | 0.03 | 0.67(0.49,0.91) | 0.03 |
| p for trend (character2integer) | < 0.0001 | < 0.001 | 0.03 | 0.04 | 0.04 | |||||
| p for trend (Median value) | < 0.0001 | < 0.001 | 0.03 | 0.04 | 0.04 | |||||
OR odds ratio, CI confidence interval, Ref reference
Crudel model: AIP
Model 1: AIP, age, sex, race
Model 2: Model 1 + BMI, alcohol use, CVD, marital status, hypertension, education level, waist circumference, smoking status, diabetes mellitus, glucose, HbA1c, uric acid, albumin
Model 3: Model 2 + noise exposure history + degree of hearing loss (from NHANES pure-tone audiometry data)
Sensitivity Analysis: Analysis conducted in specific subpopulations (e.g., excluding individuals with severe hearing loss, analyzing only those with confirmed noise exposure history)
AIP-Tinnitus risk associations
Multivariable logistic regression demonstrated robust associations between AIP and Tinnitus across three models. In the crude model, each unit increase in AIP was associated with a 2.41-fold higher likelihood of tinnitus (OR = 2.41, 95% CI: 1.80–3.23; P < 0.0001) (Table 3). After adjusting for demographic factors (Model 1), the odds ratio was 2.21 (95% CI: 1.57–3.12; P < 0.0001), and further adjustments in the fully adjusted model (Model 2) yielded an OR of 1.73 (95% CI: 1.19–2.52; P = 0.01). The dose–response trend remained significant across all models (P-trend < 0.05), underscoring the strength of the association between elevated AIP and increased tinnitus risk.
Table 3.
Stratified analysis of the association between AIP and Tinnitus by potential effect modifiers
| Character | 95% CI | P-value | P for interaction | Character | 95% CI | P-value | P for interaction |
|---|---|---|---|---|---|---|---|
| Age | 0.215 | Alcohol user | 0.787 | ||||
| >= 60 | 1.650(0.723,3.765) | 0.228 | yes | 2.375(1.595,3.537) | < 0.0001 | ||
| < 60 | 3.171(2.101,4.785) | < 0.0001 | former | 2.501(0.755,8.284) | 0.130 | ||
| Sex | 0.61 | no | 1.370(0.360,5.216) | 0.636 | |||
| Male | 2.060(1.132,3.748) | 0.019 | Hypertension | 0.33 | |||
| Female | 2.676(1.423,5.032) | 0.003 | yes | 2.492(1.682,3.692) | < 0.0001 | ||
| Race | 0.212 | no | 1.613(0.819,3.179) | 0.162 | |||
| Non-Hispanic White | 2.241(1.509,3.328) | < 0.001 | Diabetes mellitus | 0.927 | |||
| Non-Hispanic Black | 1.343(0.670,2.693) | 0.389 | DM | 2.033(1.001,4.130) | 0.050 | ||
| Other | 8.393(1.335,52.784) | 0.025 | no | 2.069(1.171,3.655) | 0.014 | ||
| Mexican American | 1.580(0.271,9.230) | 0.593 | preDM | 1.665(0.644,4.304) | 0.284 | ||
| BMI | 0.162 | Smoking status | 0.235 | ||||
| < 25 | 4.004(1.629,9.844) | 0.003 | never | 2.186(1.248,3.830) | 0.007 | ||
| >= 30 | 1.234(0.617,2.469) | 0.544 | former | 1.578(0.601,4.141) | 0.346 | ||
| 25–30 | 2.885(1.203,6.918) | 0.019 | now | 4.669(2.429,8.973) | < 0.0001 | ||
| Marital status | 0.096 | CVD | 0.35 | ||||
| Married | 1.627(0.922,2.872) | 0.091 | no | 2.511(1.927,3.272) | < 0.0001 | ||
| Non-Married | 4.030(2.078,7.814) | < 0.001 | yes | 1.531(0.516,4.539) | 0.433 |
Stratified and mediation analyses
Stratified analyses revealed that the relationship between AIP and tinnitus was strongest in certain subgroups. The risk of tinnitus was highest in participants aged < 60 years (OR = 3.17, 95% CI: 2.10–4.79; P < 0.0001), females (OR = 2.68, 95% CI: 1.42–5.03; P = 0.003), and non-Hispanic Whites (OR = 2.24, 95% CI: 1.51–3.33; P < 0.001) (Table 4). Additionally, smokers were at the highest risk (OR = 4.67, 95% CI: 2.43–8.97; P < 0.0001).
Mediation analysis revealed that hypertension partially mediated the relationship between AIP and Tinnitus. The mediation effect was significant, with hypertension explaining 17.53% of the total effect of AIP on tinnitus (P < 0.01). The estimated indirect effect was 0.04 (95% CI: 0.02–0.09, P < 0.01), while the direct effect of AIP on tinnitus remained significant after controlling for hypertension (OR = 1.20, 95% CI: 1.04–1.33, P = 0.03). These results suggest that part of the AIP-tinnitus association is mediated by hypertension, underscoring its role as a key factor in the relationship between metabolic dysregulation and tinnitus (Fig. 2). In this study, hypertension was defined as a self-reported physician diagnosis or the current use of antihypertensive medications, consistent with NHANES methodology.
Fig. 2.
Mediation analysis of hypertension and association between atherogenic index of plasma and tinnitus
Practical considerations
The data suggest that AIP, as a modifiable biomarker, could serve as a useful tool for identifying individuals at risk for tinnitus, particularly in those with elevated metabolic markers such as BMI and waist circumference. These findings emphasize the potential for early intervention strategies targeting metabolic dysregulation, including weight management and blood pressure control, to reduce tinnitus risk. Moreover, stratified analyses point to the importance of considering demographic factors such as age, sex, and race in predicting tinnitus risk, which could help in tailoring prevention strategies.
Discussion
Tinnitus, a complex neurological condition that affects the auditory system, is associated with a significant impairment in quality of life, with varying degrees of severity ranging from mild annoyance to debilitating distress [1, 2, 6]. Emerging evidence suggests a strong link between metabolic dysregulation and tinnitus [23], yet the role of the Atherogenic Index of Plasma (AIP)—a marker of atherogenic lipid metabolism—has not been well explored in this context. This study provides novel insights into the relationship between elevated AIP levels and tinnitus prevalence, demonstrating that AIP is independently associated with increased tinnitus risk in a dose-dependent manner.
Our findings suggest that higher AIP levels are associated with an increased likelihood of tinnitus, with a clear dose–response relationship observed across the four AIP quartiles. This relationship remained significant even after adjusting for various demographic and clinical covariates, highlighting the potential of AIP as a modifiable metabolic biomarker for tinnitus. However, it is important to note that, despite rigorous inclusion and exclusion criteria, there may still be potential selection bias in our study. For example, participants with missing data on key variables, such as AIP or tinnitus status, were excluded from the analysis, which may result in a non-representative sample. Although the analysis was adjusted for known confounders, the potential for selection bias in the excluded individuals—especially with respect to socioeconomic status, health conditions, or other unmeasured factors—cannot be ruled out, and this may impact the generalizability of our findings.
These findings are consistent with recent studies linking metabolic dysregulation to auditory dysfunction. Notably, our stratified analyses reveal that certain subgroups, including younger individuals (< 60 years), females, non-Hispanic Whites, and smokers, show a stronger association, further emphasizing the role of metabolic factors in tinnitus occurrence.
The association between AIP and Tinnitus was found to be partially mediated by hypertension, with hypertension explaining 17.53% of the observed relationship between AIP and tinnitus. The definition of hypertension in our analysis was aligned with NHANES criteria, based on self-reported physician diagnosis or antihypertensive medication use. This finding aligns with previous research suggesting that metabolic pathways involving lipid metabolism and blood pressure regulation may be interconnected and could influence auditory health [23]. Hypertension, a well-established risk factor for cardiovascular and metabolic diseases, may play a role in exacerbating the effects of metabolic dysregulation on tinnitus [24–26]. Previous studies have shown that metabolic dysregulation and tinnitus may be influenced by vascular changes, including those caused by atherosclerosis in the neck arteries, supporting the need for managing blood pressure in individuals with tinnitus [27].
While previous studies have focused on individual components of metabolic syndrome—such as obesity and dysglycaemia—our findings underscore the value of AIP as a comprehensive measure of atherogenic dyslipidemia in predicting tinnitus risk. AIP has been shown to be a superior predictor of cardiovascular risk compared to traditional lipid parameters, and our study extends this application to tinnitus. Moreover, we observed that participants with a normal BMI (< 25 kg/m2) were also at heightened risk, suggesting that metabolic dysregulation independent of obesity may contribute to tinnitus occurrence. This challenges the obesity-centric model that has traditionally dominated research on metabolic influences on health [28].
Although tinnitus is not a cardiovascular disease per se, emerging evidence suggests that vascular dysfunction and metabolic abnormalities contribute to its onset and progression. In this context, AIP offers unique advantages over traditional metabolic indices. Measures such as BMI, waist circumference, and glucose primarily reflect general adiposity or glycemic status, but they do not directly capture lipid-related atherogenic imbalance. In contrast, AIP simultaneously incorporates triglyceride and HDL-C levels, providing a more integrative indicator of dyslipidemia and vascular risk [10, 11, 18]. Importantly, HDL-C plays a neuroprotective and anti-inflammatory role, which may be relevant to auditory pathology. Compared with the TyG index and its derivatives, AIP additionally accounts for HDL-C, thus reflecting both atherogenic and protective lipid components. Prior studies have demonstrated that AIP outperforms TyG and conventional lipid parameters in predicting cardiometabolic risk stratification [29], which may explain its stronger association with tinnitus in our analysis. These comparative advantages support the application of AIP as a sensitive metabolic biomarker beyond cardiovascular disease, extending to conditions such as tinnitus that may share overlapping vascular–metabolic pathways.
Recent research supports the notion that lipid metabolism disturbances, including elevated triglycerides and decreased HDL cholesterol, may contribute to neural changes in the auditory system, which could influence tinnitus onset and progression [10]. Investigating the effects of atherogenic lipid markers, such as AIP, on the auditory pathways represents a promising direction for future research [11].
Despite the strengths of this study, including the use of the NHANES dataset and rigorous multivariable adjustment, several limitations must be acknowledged. First, the cross-sectional design of this study limits the ability to infer causality, and longitudinal studies are needed to confirm the temporal relationship between AIP and tinnitus onset. Additionally, the self-reported nature of tinnitus assessment may be subject to recall bias, and future studies should validate these findings using objective audiometric measures. The severity, type, and underlying causes of tinnitus were not analyzed in this study, which could affect the generalizability of the results. Furthermore, selection bias could be a factor, given that participants with missing data on essential variables were excluded from the study. These exclusions may have introduced a degree of non-random selection, potentially affecting the representativeness of the sample. Further investigations are warranted to explore how these factors interact with metabolic risk markers such as AIP.
In conclusion, elevated AIP levels are associated with an increased likelihood of tinnitus, with hypertension partially mediating this relationship. This study highlights AIP as a promising metabolic biomarker for tinnitus and calls for the integration of lipid ratio metrics into routine clinical practice for tinnitus risk stratification. Given the modifiable nature of AIP, early interventions targeting atherogenic dyslipidemia and hypertension may help reduce tinnitus risk, particularly in vulnerable populations. Future longitudinal studies and mechanistic research are needed to further explore the pathways linking metabolic dysregulation to tinnitus and to develop effective prevention strategies.
Supplementary Information
Acknowledgements
We thank the National Health and Nutrition Examination Surveys for providing the data.
Authors’ contributions
Conceptualization, W.M.; data curation,W.M. and T.W.; formal analysis, W.M.; methodology, C.W.; software, W.M.; supervision, C.W.; writing-original draft, D.Y.; writing—review and editing, Y.L. and W.M.; visualization, W.M.; All authors have read and agreed to the published version of the manuscript.
read and agreed to the published version of the manuscript.
Funding
No funding.
Data availability
The survey data are publicly available on the internet for data users and researchers throughout the world (www.cdc.gov/nchs/nhanes/).
Declarations
Ethics approval and consent to participate
The studies involving human participants were reviewed and approved by NCHS Ethics Review Board. The participants provided their written informed consent to participate in this study.
Consent for publication
All participants in the NHANES study provided consent for publication.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Henton A, Tzounopoulos T. What’s the buzz? The neuroscience and the treatment of tinnitus. Physiol Rev. 2021;101(4):1609–32. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Perumareddi P. Tinnitus. Prim Care. 2025;52(1):61–70. [DOI] [PubMed] [Google Scholar]
- 3.Fuller T, Cima R, Langguth B, Mazurek B, Vlaeyen JW, Hoare DJ. Cognitive behavioural therapy for tinnitus. Cochrane Database Syst Rev. 2020;1(1):Cd012614. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Mazurek B, Böcking B, Dobel C, Rose M, Brüggemann P. Tinnitus and influencing comorbidities. Laryngorhinootologie. 2023;102(S 01):S50–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Molnár A, Mavrogeni P, Tamás L, Maihoub S. Correlation between tinnitus handicap and depression and anxiety scores. Ear Nose Throat J. 2025;104(8):Np565-np571. [DOI] [PubMed] [Google Scholar]
- 6.Holmes S, Padgham ND. “Ringing in the ears”: narrative review of tinnitus and its impact. Biol Res Nurs. 2011;13(1):97–108. [DOI] [PubMed] [Google Scholar]
- 7.Wang J, Xiang JH, Peng XY, Liu M, Sun LJ, Zhang M, et al. Characteristic alterations of gut microbiota and serum metabolites in patients with chronic tinnitus: a multi-omics analysis. Microbiol Spectr. 2025;13(1):e0187824. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Firdaus M, Mathew MK, Wright J: Health promotion in older adults: the role of lifestyle in the metabolic syndrome. Geriatrics 2006, 61(2):18–22, 24–15. [PubMed]
- 9.Onaolapo AY, Ojo FO, Adeleye OO, Falade J, Onaolapo OJ. Diabetes mellitus and energy dysmetabolism in Alzheimer’s disease: understanding the relationships and potential therapeutic targets. Curr Diabetes Rev. 2023;19(8):e020123212333. [DOI] [PubMed] [Google Scholar]
- 10.Dobiásová M, Frohlich J. The new atherogenic plasma index reflects the triglyceride and HDL-cholesterol ratio, the lipoprotein particle size and the cholesterol esterification rate: changes during lipanor therapy. Vnitr Lek. 2000;46(3):152–6. [PubMed] [Google Scholar]
- 11.Shen SW, Lu Y, Li F, Yang CJ, Feng YB, Li HW, et al. Atherogenic index of plasma is an effective index for estimating abdominal obesity. Lipids Health Dis. 2018;17(1):11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Maihoub S, Mavrogeni P, Molnár V, Molnár A. Tinnitus and its comorbidities: a comprehensive analysis of their relationships. J Clin Med. 2025. 10.3390/jcm14041285. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Capucho AM, Chegão A, Martins FO, Vicente Miranda H, Conde SV. Dysmetabolism and neurodegeneration: trick or treat? Nutrients. 2022. 10.3390/nu14071425. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Wang K, Cui Y, Wu T. Association between tinnitus and suicidal ideation in postmenopausal women: an observational study. Am J Geriatr Psychiatry. 2025;33(4):405–12. [DOI] [PubMed] [Google Scholar]
- 15.Chasca W, Nerada S, Zenone M, Barbic S. TikTok and #occupationaltherapy: cross-sectional study. JMIR Form Res. 2023;7:e45554. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Koru-Sengul T, Clark JD, 3rd, Ocasio MA, Wanner A, Fleming LE, Lee DJ: Utilization of the national health and nutrition examination (NHANES) survey for symptoms, tests, and diagnosis of chronic respiratory diseases and assessment of Second hand Smoke Exposure. Epidemiology (Sunnyvale). 2011;1(2):104. [DOI] [PMC free article] [PubMed]
- 17.Li X, Hu X, You C. Interactive effects of composite dietary antioxidant index with body mass index for the risk of stroke among U.S. adults: insight from NHANES 2001–2018. Front Nutr. 2024;11:1378479. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Zhu XW, Deng FY, Lei SF. Meta-analysis of atherogenic index of plasma and other lipid parameters in relation to risk of type 2 diabetes mellitus. Prim Care Diabetes. 2015;9(1):60–7. [DOI] [PubMed] [Google Scholar]
- 19.Du KJ, Mo BY, Hou T, Chen L, Su DR, Yin SH. Association of serum Klotho with tinnitus prevalence, duration and severity: a cross-sectional study in middle-aged and older adults. PLoS ONE. 2025;20(7):e0327228. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Lai H, Yin J, Yang H. Association of oxidative balance score with hearing loss and tinnitus: NHANES 1999–2018. Front Nutr. 2024;11:1421605. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Zhang H, Zhang G, Fu J. Exploring the L-shaped relationship between Atherogenic Index of Plasma and depression: results from NHANES 2005–2018. J Affect Disord. 2024;359:133–9. [DOI] [PubMed] [Google Scholar]
- 22.Casagrande SS, Lee C, Stoeckel LE, Menke A, Cowie CC. Cognitive function among older adults with diabetes and prediabetes, NHANES 2011–2014. Diabetes Res Clin Pract. 2021;178:108939. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Molnár A, Molnár V, Mavrogeni P, Maihoub S. Fasting glucose, haemoglobin A1C (HbA1c), blood lipid, and triglyceride-glucose index parameters in relation to subjective tinnitus. Biomedicines. 2025. 10.3390/biomedicines13040824. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Spinar J, Vítovec J. Metabolic syndrome and cardiovascular disease. Vnitr Lek. 2009;55(7–8):653–8. [PubMed] [Google Scholar]
- 25.Gupta S, Gupta BM. Metabolic syndrome: diabetes and cardiovascular disease. Indian Heart J. 2006;58(2):149–52. [PubMed] [Google Scholar]
- 26.Wen ZZ, Yang YH, Liu D, Shi CX. NAD(+) metabolism in cardiovascular diseases. Sheng Li Xue Bao. 2025;77(2):345–60. [DOI] [PubMed] [Google Scholar]
- 27.Molnár A, Molnár V, Mavrogeni P, Maihoub S. The influence of carotid and vertebral Doppler ultrasonography and brain MRI abnormalities on hearing levels, tinnitus intensities and frequencies. Audiol Res. 2025;15(2):29. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Vahedian Z, Fakhraie G, Bovet J, Mozaffarieh M. Nutritional recommendations for individuals with Flammer syndrome. Epma j. 2017;8(2):187–95. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Won KB, Kim YS, Lee BK, Heo R, Han D, Lee JH, et al. The relationship of insulin resistance estimated by triglyceride glucose index and coronary plaque characteristics. Medicine (Baltimore). 2018;97(21):e10726. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The survey data are publicly available on the internet for data users and researchers throughout the world (www.cdc.gov/nchs/nhanes/).