Temporomandibular joint disorders and tinnitus: a systematic review and meta-analysis

Abstract

Background

Temporomandibular disorders (TMDs) are a common set of conditions affecting the jaw muscles and joint. Given the anatomical and neurological proximity, a conceptual link between TMDs and tinnitus has been hypothesized, but the strength and consistency of this association remain unclear.

Objectives

To systematically review and quantitatively synthesize evidence on the association between TMDs and tinnitus, specifically by pooling the prevalence of tinnitus in patients with TMDs and the prevalence of TMDs among patients with tinnitus, and by estimating corresponding odds ratios (ORs) in comparative studies.

Eligibility criteria

We included single- and double-arm observational studies (cross-sectional, case-control, and cohort) reporting the prevalence or odds of tinnitus in patients with TMDs or vice versa. Case reports, case series, reviews, and meta-analyses were excluded.

Information sources

PubMed, Web of Science, Cochrane Library, and Scopus databases were searched from inception to January 2026.

Risk of bias

The Newcastle-Ottawa Scale (NOS) was used to assess study quality and risk of bias for all included studies.

Synthesis methods

Pooled prevalence rates with 95% confidence intervals (CIs) were calculated for single-arm studies. Pooled ORs with 95% CIs were calculated for double-arm studies, using random-effects models.

Results

Thirty-four observational studies (27 cross-sectional, 5 case-control, and 3 cohort studies) involving 47,349 participants were included. The estimated pooled prevalence of tinnitus in patients with TMDs was 31.8% (95% CI, 24.2–40.5). The pooled prevalence of TMDs among patients with tinnitus was 63.4% (95% CI, 46.5–77.6). In comparative studies, TMD diagnosis was associated with higher odds of tinnitus compared with controls (OR = 3.924; 95% CI, 2.636–5.841). Similarly, the odds of TMD diagnosis were higher in patients with than in those without tinnitus (OR = 4.147; 95% CI, 1.877–9.164).

Strengths and limitations of evidence

The findings are limited by the observational nature of the included studies, which prevents establishing a temporal sequence or causal link. Furthermore, the high heterogeneity and reliance on self-reported tinnitus in many studies limit the certainty of the evidence. Approximately one-third of the cross-sectional studies were rated as fair or poor.

Interpretation

This meta-analysis indicates a consistent association between temporomandibular disorders and tinnitus. Pooled prevalence estimates and odds ratios indicate frequent co-occurrence of the two conditions in observational studies, although causality cannot be inferred.

Registration

PROSPERO (CRD42024616120).

Supplementary Information

The online version contains supplementary material available at 10.1186/s12903-026-07898-3.

Introduction

Tinnitus is the perception of a high-pitched sound in the ears or head in the absence of an external auditory stimulus [1]. Its estimated prevalence in adults worldwide is 4.1–37.2%, and it is severe in more than 2% of affected individuals [2]. Tinnitus is defined mainly as a subjective complaint, but it can be an objective symptom of conditions such as temporomandibular joint (TMJ) disease, aneurysm, tensor tympani muscle spasm, and other maxillofacial disorders [3]. Its pathophysiology is not fully understood and may be linked to numerous risk factors.

The TMJ is the articulation between the mandibular condyle and the glenoid fossa of the temporal bone. An articular disc divides this joint into two synovial cavities, enabling different types of complex movement [4]. Temporomandibular disorders (TMDs) are diverse clinical conditions that affect the masticatory (temporalis, masseter, and medial and lateral pterygoid) muscles, TMJ, and/or linked structures [5]. They include pain disorders such as myalgia, arthralgia, and temporomandibular headache, and TMJ-associated disorders such as arthritis and disc displacement. TMDs are highly prevalent, occurring in approximately 31% of adults [6]. Several biological and psychosocial risk factors have been identified [7], but the relationships between TMDs and otological symptoms remain a matter of debate.

Contemporary models of tinnitus pathophysiology suggest that this condition stems mainly from the central nervous system (CNS), although peripheral triggers may initiate the condition [1]. Peripheral causes such as cochlear or auditory nerve injury often initiate it, but its perception is more likely due to CNS alterations [1]. Findings regarding the association between tinnitus and TMDs are inconsistent. Tinnitus is more common in the TMD population than in the general population, but no causal relationship has been established [8]. Different mechanisms have been suggested to describe the association of these conditions. The TMJ and parts of the ear share common neural pathways, such as cranial nerves V and VII [9]. Furthermore, the middle ear muscles, jaw, and facial muscles have a common embryological origin [10]. Moreover, TMDs involve fatigue or spasm of the masticatory muscles, which are anatomically close to the muscles in the middle ear that regulate sound transmission [11]. Given these shared anatomical and embryological origins, ear problems such as vertigo, tinnitus, and otalgia have been hypothesized to be related to TMDs.

Tinnitus linked to somatic conditions such as TMDs is referred to as somatosensory tinnitus. In this condition, it has been hypothesized that sensory signals from the temporomandibular area may modulate activity in the dorsal cochlear nucleus (DCN) in the brainstem, thereby influencing tinnitus perception [12, 13]. The DCN plays a crucial role in the integration of information from the auditory and somatosensory systems and facilitates interactions between these systems, potentially contributing to the perception of tinnitus [14]. The reported prevalence of tinnitus in patients with TMDs across published observational studies ranges widely from 3.7% to 70% [15–20], and that of TMD diagnoses in patients with tinnitus ranges similarly from 9% to 98% [21, 22]. Previous meta-analyses have reported increased odds of co-occurrence, but estimates remain highly variable [23]. While several recent systematic reviews have examined aspects of this association [23–25], many have focused on selected TMD subtypes, specific diagnostic approaches, or comparative analyses alone. A quantitative synthesis that simultaneously pools prevalence estimates in both patient populations and comparative odds ratios across single- and double-arm observational studies remains limited. Therefore, we conducted a systematic review and meta-analysis with searches updated through January 2026 to estimate pooled prevalence and odds ratios for the association between tinnitus and TMDs.

Methods

Protocol and reporting

This systematic review and meta-analysis were conducted and reported following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 statement guidelines [26] The completed PRISMA 2020 checklist is provided in Supplementary File 1. The protocol for this systematic review was registered in the International Prospective Register of Systematic Reviews (PROSPERO) on 19 November 2024 (Registration number: CRD42024616120). Any protocol amendments (e.g., updated search date) were documented in the PROSPERO record, and deviations are transparently reported in the manuscript.

Search strategy and study selection

The search was re-run and updated until January 2026 to ensure inclusion of the most recent eligible studies. The search strategy was developed using a combination of controlled vocabulary (e.g., MeSH terms in PubMed) and free-text keywords to maximize sensitivity. PubMed, Cochrane Library, Web of Science, and Scopus databases were searched from inception. The full search strategy for each database, including all keywords and Boolean operators, is provided in Supplementary Tables 1 to ensure reproducibility. As an example, the PubMed search string was: ((“temporomandibular joint disorder“[MeSH] OR “temporomandibular joint dysfunction” OR “craniomandibular disorder” OR “temporomandibular disorder” OR TMD OR TMJ) AND (“tinnitus“[MeSH] OR “acufene” OR “ear noise” OR “ringing in the ears”)). No language restrictions were applied.

Two independent reviewers screened the titles and abstracts of the retrieved records to assess eligibility. Manual screening of the reference lists of included articles and relevant reviews, as well as a search of grey literature (e.g., conference abstracts, preprints) via Google Scholar, was also performed to identify any additional eligible studies. The full texts of reports found to be relevant were then reviewed. Disagreements on study eligibility were settled through consensus with a third reviewer.

Eligibility criteria

We defined the inclusion criteria based on the PICO framework: Population: Adult human subjects with either a diagnosis of TMD or a complaint of tinnitus; Intervention/Exposure: Diagnosis of TMD or presence of tinnitus; Comparator: Control groups without the exposure condition (non-TMD controls or non-tinnitus controls); Outcome: Prevalence of the co-occurring condition (tinnitus in TMD patients, or TMD in tinnitus patients) or the Odds Ratio (OR) comparing the two groups.

We included single-arm (prevalence) and double-arm (comparative) observational studies (cross-sectional, case-control, and cohort) reporting on the link between TMDs and tinnitus. Case reports, case series, narrative reviews, and meta-analyses were excluded. Reports lacking sufficient data on prevalence or effect estimate measures (e.g., ORs) were also excluded. Given the known heterogeneity in case definitions, studies were categorized based on the diagnostic tools used for subgroup analysis. For TMD, studies were grouped as: (1) Standardized criteria (DC/TMD or RDC/TMD), (2) Clinical examination only, or (3) Non-standardized/Self-report. For tinnitus, studies were grouped as: (1) Validated instruments (e.g., THI, TFI), or (2) Single-item question/Self-report. This explicit subgrouping plan was designed to control for the extreme case definition heterogeneity and assess its impact on pooled estimates.

Data extraction and quality assessment

Two researchers independently extracted data on study identifiers (author, year, country), study design, sample size, participant demographics, baseline condition (TMD or tinnitus), diagnostic criteria, prevalence estimates, effect sizes (ORs with 95% CIs), and quality scores. Study quality was assessed using the Newcastle–Ottawa Scale (NOS). For cross-sectional studies, the NOS adapted for cross-sectional designs by Herzog et al. [27] was applied. The scoring criteria and inter-rater agreement (assessed using Cohen’s kappa) are reported in Supplementary Table 2. The NOS categorizes study quality as “good,” “fair,” or “poor” across three domains: selection, comparability, and outcome assessment [28].

The certainty of evidence for the main outcomes was evaluated using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach. Assessment considered five domains: risk of bias, inconsistency (heterogeneity), indirectness, imprecision, and publication bias. Overall certainty was classified as high, moderate, low, or very low. Detailed GRADE evidence profiles are presented in the Results section and Supplementary Table S4.

Statistical analysis

For analysis, the studies were divided into two groups: those examining the odds of tinnitus in patients with TMDs and those examining the frequency of TMD diagnoses in patients with tinnitus. Each group was further subdivided into single- and double-arm studies, and the four groups were analyzed separately. For single-arm studies, pooled event rates were calculated using the logit transformation to stabilize the variance, and results were back-transformed to obtain pooled prevalence estimates with 95% confidence intervals (CIs). Pooled ORs with 95% CIs were obtained from the double-arm studies. A random-effects model was adopted for all analyses.

Heterogeneity was assessed using the I² statistic, where values of 30% to 60% may represent moderate heterogeneity, 50% to 90% substantial heterogeneity, and 75% to 100% considerable heterogeneity, as per Cochrane recommendations. In addition to 95% CIs, 95% prediction intervals (PIs) were calculated for pooled estimates to reflect the expected range of effects in future studies.

For the double-arm studies, subgroup analyses based on study design were conducted. Publication bias was assessed using Egger’s test and visual inspection of funnel plots. Given the limitations of Egger’s test, particularly with a small number of studies (< 10) and heterogeneity in observational prevalence data, the results were interpreted cautiously. Sensitivity analyses using the trim-and-fill method were performed to assess robustness of pooled estimates to potential publication bias. All statistical analyses were performed using Comprehensive Meta-Analysis (CMA), Version 3 (Biostat, Englewood, NJ, USA).

Results

Search results

The database search yielded a total of 593 records, from which 318 duplicates were removed. Titles and abstracts of the remaining 275 records were screened, and 60 reports were assessed for full-text eligibility. After full-text review, 34 studies met the inclusion criteria and were included in the meta-analysis [15–22, 29–52]. Figure 1 presents the PRISMA 2020 flow diagram summarizing the study selection process.

Fig. 1.

PRISMA 2020 flow diagram illustrating the identification, screening, eligibility assessment, and inclusion of studies in the systematic review and meta-analysis (searches updated through January 2026). The initial search across PubMed, Web of Science, Cochrane Library, Scopus, and other sources identified 593 records. After removal of 318 duplicates, 275 unique records were screened, and 215 were excluded based on title and abstract. Full-text assessment was performed for 60 reports, of which 26 were excluded for irrelevance (n = 21) or because they were case reports (n = 5). Ultimately, 34 studies met the eligibility criteria and were included in the qualitative synthesis and meta-analysis

Study characteristics

The 34 included studies were observational in design and involved a total of 47,349 subjects, approximately two-thirds of whom were female (Tables 1 and 2). The study design comprised 26 cross-sectional studies [11, 15–17, 19–22, 30–37, 40–43, 45, 46], case-control studies [18, 38, 48, 49], 3 cohort studies, [29, 51, 52] and 1 mixed two-phase study with cross-sectional and case-control components [50].

Table 1.

Summary of included studies

Study ID	Study design	Country	Institution	Recruitment period	Total Population (n)	Male n (%)	Assessment Tool		Conclusion
							Tinnitus	TMJ
Studies that reported the frequency of TMD in individuals with and without tinnitus
Bernhardt (I) 2004 [50]	Cross-sectional	Germany	(ENT) Clinic of the University of Greifswald	October 1997 - May 2001	4228	2069 (48.93)	Asking questions	Clinical examinations	A relationship between tinnitus and TMD was established.
Bernhardt (II) 2004 [50]	Case-control	Germany	(ENT) Clinic of the University of Greifswald	July 1997 - Oct 1998	1937	932 (48.11)	Asking questions	Clinical examinations	A relationship between tinnitus and TMD was established.
Buergers 2014 [32]	Cross-sectional	Germany	Department of Prosthetic Dentistry, Regensburg University Medical Center	May 2008 - April 2009	951	473 (49.73)	Asking question & THI	RDC/TMD	There is a significant correlation between tinnitus and TMD.
Didier 2023 [22]	Cross-sectional	Italy	Policlinic Hospital of Milan	NA	47	NA	A questionnaire	Axis I (DC/TMD)	A high prevalence of TMD in patients with tinnitus, as well as a not uncommon occurrence of tinnitus in patients presenting with TMD
Fernandes 2014 [20]	Cross-sectional	Brazil	(Universidade Estadual Paulista, Brazil).	NA	261	0	A question in RDC/TMD questionnaire	RDC/TMD	There is an association between painful TMD and self-reported tinnitus; however, no relationship of a causal nature could be established.
Kim 2018 [21]	Cross-sectional	Korea	The Korean national health and nutrition examination survey	2010–2011.	11 745	5778 (49.19)	Asking questions	Asking questions & Clinical examination	Subjects with tinnitus had more TMJ than subjects without tinnitus.
Lam 2001 [42]	Cross-sectional	Canada	Craniofacial Pain Unit of the Wasser Pain Management Center, Mount Sinai Hospital	May 1987 - June 1999.	192	NA	Asking questions	RDC/TMD	These findings indicate that TMD is significantly correlated to aural health, although no cause-and-effect relationship has yet been demonstrated.
Morais 2012 [44]	Cross-sectional	Brazil	Department of Speech and Hearing Therapy	NA	20	6 (30)	Asking question & THI	Asking questions	90% of patients withtinnitus had at least one sign or symptom of TMD and there was no correlation between the tinnitus and acuphenometry, THI and the TMD checklist.
Morgan 1992 [46]	Cross-sectional	USA	The American Tinnitus Association, Portland, Oregon.	NA	20	16 (80)	NA	Clinical examination, Mandibular scan and EMG	It appears that individuals who have tinnitus with no apparent otologic basis for this symptom should have a careful evaluation of the temporomandibular apparatus.
Ralli 2018 [29]	Cohort	Italy	Policlinico Umberto I, Sapienza University	January 2016 - June 2017	226	128 (56.63)	THI	DC/TMD	Patients with chronic tinnitus and normal hearing suggest that self-reported history for somatic dysfunction and modulation of tinnitus, when occurring simultaneously in the TMJ region, can be useful to preliminarily identify patients with TMJ disorders.
Rubinstein 1993 [47]	Cross-sectional	Sweden	Department of Audiology at the Sahlgrenska Hospital in Göteborg.	NA	102	62 (60.78)	NA	Clinical examinations & A questionnaire	There is a significant correlation between tinnitus and TMD.
Saldanha 2012 [34]	Cross-sectional	Brazil	Dental treatment at the Bauru School of Dentistry	NA	200	51 (25.5)	Asking a question & VAS	RDC/TMD, PPT & VAS	Results suggest that an association exists between TMD and subjective tinnitus.
Song 2018 [36]	Cross-sectional	Korea	5th Korea National Health and Nutritional Examination Survey	2010–2012	17 017	NA	Asking questions	Asking questions	TMDs and otolaryngologic disorders hold various correlations, suggesting the need for multitarget approaches to effectively address this phenomenon.
Studies that reported the frequency of tinnitus in Individuals with and without TMDs
Akhter 2013 [17]	Cross-sectional	Japan	Medical Center of Okayama University	Apr-08	1,930	981 (50.8)	A questionnaire	A questionnaire	TMD symptoms were significantly correlated to aural symptoms and headache.
Algieri 2016 [30]	Cross-sectional	Italy	Maxillo- Facial Science Department, “Sapienza” University of Rome	January - June 2013.	200	32 (16)	THI	Clinical examinations, MRI & Wilkes classification	This study shows that TMD-related tinnitus principally affects a younger population.
Al–Tuhafi 2005 [49]	Case-control	Iraq	Department of Oral and maxillofacial Surgery, University of Mosul	NA	200	77 (38.5)	Asking questions	Clinical examination	Tinnitus was found to be positively correlated with clicking and number of tender muscles.
Bernhardt 2004 [50]	Cross-sectional	Germany	(ENT) Clinic of the University of Greifswald	October 1997 - May 2001	4228	2069 (48.9)	Asking questions	Clinical examinations	A relationship between tinnitus and TMD was established.
Bernhardt 2011 [51]	Cohort	Germany	Center of Oral Health, University of Greifswald	October 1997 - May 2001	3134	1506 (48)	Asking questions	Clinical examinations	Pain on palpation of the TMJs predicts the development of tinnitus.
Buergers 2014 [32]	Cross-sectional	Germany	Department of Prosthetic Dentistry, Regensburg University Medical Center	May 2008 - April 2009	951	473 (49.73)	Asking question & THI	RDC/TMD	There is a significant correlation between tinnitus and TMD.
Bush 1987 [37]	Cross-sectional	USA	TM J-Facial Pain Center, School of Dentistry, Virginia Commonwealth University.	NA	105	15 (14.28)	A questionnaire	NA	There is no statistical correlation between tinnitus and otalgia in patients with TMD.
Çebi 2023 [39]	Cross-sectional	Turkey	Department of Oral and Maxillofacial Surgery, Karabük University	January 2018 and July 2019	288	70 (24.2)	THI	(DC/TMD) & MRI	There is a relationship between aural symptoms, tinnitus, and temporomandibular disorders.
Ciancaglini 1994 [16]	Cross-sectional	Italy	San Raffaele, University of Milan	January 1985 - October 1991	797	263 (33)	Asking questions	Clinical examinations & CT	There is an association between TMD of arthrogenic origin and ear, nose, and throat symptoms, especially deafness.
de-Pedro-Herráez 2016 [52]	Cohort	Spain	Dental Clinic of the University San Pablo-CEU in Madrid.	2015, about 4 years after the initial evaluation	62	0	Asking questionnaire	RDC/TMD	It was found that the group affected by MTMD presented many more associated medical conditions than the control group.
Didier 2023 [22]	Cross-sectional	Italy	Policlinic Hospital of Milan	NA	50	NA	A questionnaire	A xis I (DC/TMD)	A high prevalence of TMD in patients with tinnitus, as well as a not uncommon occurrence of tinnitus in patients presenting with TMD
Effat 2016 [18]	Case-control	Egypt	Department of Otolaryngology, El-Sahel Teaching Hospital	over one year, starting on 10 July 2014.	214	64 (30)	Asking questions	History taking & The expanded DC/TMD	The main cause of otological symptoms and of audiometric findings in temporomandibular disorder patients is postulated to be an altered middle-ear to inner-ear pressure equilibrium.
Ferendiuk 2014 [15]	Cross-sectional	Poland	Department of Dental Prosthetics of the Jagiellonian Medical College, Cracow.	2008 - March 14, 2014	1208	256 (21.2)	Medical interview	Clinical examinations	In cases when otolaryngological origin of symptoms such as earache or sore throat, hearing impairment, tinnitus, or vertigo is excluded, functional disorders of the masticatory organ should be taken into consideration
Fernandes 2013 [38]	Case-control	Brazil	(Universidade Estadual Paulista, Brazil).	NA	224	40 (17.9)	A question in RDC/TMD questionnaire	RDC/TMD	Painful temporomandibular disorders, high levels of depression, and self-reported tinnitus are deeply associated. However, this association does not imply a causal relationship.
Fernandes 2014 [20]	Cross-sectional	Brazil	(Universidade Estadual Paulista, Brazil).	NA	261	0	A question in RDC/TMD questionnaire	RDC/TMD	There is an association between sleep bruxism, painful TMD and self-reported tinnitus; however, no relationship of a causal nature could be established.
Karacay 2023 [41]	Cross-sectional	Turkey	Kırsehir Ahi Evran University Faculty of Medicine	NA	129	25 (19.37)	A questionnaire	(DC/TMD)	TMD subtype was not associated with dizziness, tinnitus, or ear fullness.
Kim 2018 [21]	Cross-sectional	Korea	The Korean national health and nutrition examination survey	2010–2011.	11 745	5778 (49.19)	Asking questions	A questionnaire & Clinical examination	Subjects with TMD had more tinnitus than subjects without TMD.
Kusdra 2018 [40]	Cross-sectional	Brazil	Tuiuti University of Paraná	NA	485	NA	Asking questions	(RDC/TMD).	These data support the correlation between temporomandibular disorders and otological symptoms, even without being caused directly by the ear.
Machado 2010 [43]	Cross-sectional	Brazil	Department of Foundations of Speech Pathology at PUC-SP	NA	20	NA	A questionnaire	A questionnaire & Clinical examination	The results showed high percentage between TMD and otological symptoms. Other research with a larger sample are suggested with the aim of confirm the correlation between TMD and otological symptoms.
Maciejewska-Szaniec 2017 [33]	Cross-sectional	Poland	The Clinic of Oral Rehabilitation, Department of Prosthodontics, Poznan Univ	NA	246	99 (40.2)	Asking questions	RDC/TMD	The otological symptoms which most frequently accompany with TMD were fullness and otalgia.
Manfredini 2015 [35]	Cross-sectional	Italy.	The Temporomandibular Disorders Clinic, Section of Maxillofacial Surgery	The first six months of the year 2013.	238	192 (76.8)	Asking questions	An executive summary of the DC/TMD	Any correlation patterns were shown between tinnitus and TMD.
Mijiritsky 2020 [11]	Cross-sectional	Israel	Maurice and Gabriela Goldschleger Tel Aviv University Orofacial Pain Clinic.	2015–2017.	108	30 (27.8)	A question in DC/TMD questionnaire	DC/TMD	This study emphasizes the necessity of assessing both Axes I and II according to the DC/TMD in future studies and supports creating a standardized tinnitus screener tailored to TMD patients for future studies on tinnitus in TMD patients
Minervini 2024 [45]	Cross-sectional	Italy	Program of Orthodontics of the University of Campania Luigi Vanvitelli, Naples	NA	37	9 (24.32)	THI	DC/TMD.	The study provides deep insights into the prevalence of tinnitus in TMD patients, underscoring the need for comprehensive treatment strategies addressing both TMD and tinnitus.
Saldanha 2012 [34]	Cross-sectional	Brazil	São Paulo University	NA	200	51 (25.5)	Asking questions & VAS	RDC/TMD, PPT & VAS	Results suggest that an association exists between TMD and subjective tinnitus.
Song 2018 [36]	Cross-sectional	Korea	5th Korea National Health and Nutritional Examination Survey.	2010–2012	17 017	NA	Asking questions	Asking questions	These findings imply that TMDs, chronic diseases, and ophthalmologic and otolaryngologic disorders hold various correlation
Tuz 2003 [48]	Case-control	Turkey	Department of Oral and Maxillofacial Surgery at Ankara University	July 1997 - December 1998.	250	58 (23.2)	Asking questions	RDC/TMD	Patients in the TMD groups had high incidences of otologic complaints compared with the control subjects without TMD signs or symptoms.
Vasconcelos 2016 [19]	Cross-sectional	Brazil	Pain control of dental school, University of Pernambuco.	Sep 2011 - Aug 2012	100	12 (12)	Asking questions	The simplified Fonseca anamnestic questionnaire	There is a significant correlation between tinnitus and TMJ.

CEU Centro Escolar University, CT Computed tomography, DC/TMD Diagnostic Criteria for Temporomandibular Disorders, EMG Electromyography, ENT Ear, nose, and throat, MRI Magnetic resonance imaging, MTMD Myogenic temporomandibular disorder, NA Not applicable, PPT Pain pressure threshol, PUC-SP Pontifical Catholic University of São Paulo, RDC/TMD Research Diagnostic Criteria for Temporomandibular Joint Disorder, THI Tinnitus Handicap Inventory, TMD Temporomandibular disorder, TMJ Temporomandibular joint, USA United States of America, VAS Visual analog scale

Table 2.

Baseline characteristics of the included studies

Study ID	Comparison population (n)		Age (Years), mean ± SD (range)		Male sex, n (%)
	Case	Control	Case	Control	Case	Control
Studies that reported the frequency of TMD in individuals with and without tinnitus
Bernhardt (I) 2004 [50]	139	4089	20–79	-	72 (51.79)	1997 (48.83)
Bernhardt (II) 2004 [50]	30	1907	41 (18–71)#	49 (20–79)#	17 (56.66)	915 (48)
Buergers 2014 [32]	68	883	54.1 ± 17.1*	-	-	-
Didier 2023 [22]	47	-	-	-	-	-
Fernandes 2014 [20]	163	98	37.0 ± 12.7*	-	0	0
Kim 2018 [21]	2555	9190	49.1 ± 0.6*	44.5 ± 0.3*	1119 (43.8)	4659 (50.7)
Lam 2001 [42]	192	-	-	-	-	-
Morais 2012 [44]	20	-	31.2	-	6 (30)	-
Morgan 1992 [46]	20	-	55.3 (22–74)#	-	16 (80)	-
Ralli 2018 [29]	134	92	48.5 ± 14.1*	45.9 ± 13.6*	78 (58.2)	50 (54.3)
Rubinstein 1993 [47]	102	-	56 (15–84)#	-	62 (60.78)
Saldanha 2012 [34]	100	100	39.16	34.33	16 (16)	35 (35)
Song 2018 [36]	3866	13,151	-	-	-	-
Studies that reported the frequency of tinnitus in individuals with and without TMDs
Algieri 2016 [30]	200	-	44.1	-	32 (16)	-
Bernhardt 2004 [50]	745	3483	(20–79)	-	2069 (48.9)	-
Bernhardt 2011 [51]	191	2643	NA	-	1506 (48)	-
Buergers 2014 [32]	82	869	54.1 ± 17.1*	-	473 (49.73)	-
Ciancaglini 1994 [16]	797	-	37.4 ± 8.3*	-	263 (33)	-
de-Pedro-Herráez 2016 [52]	31	31	38.67	41.25	0	0
Effat 2016 [18]	104	110	35.3 ± 12.1*	30.9 ± 8.1*	20 (19)	44 (40)
Ferendiuk 2014 [15]	1209	-	(19–50)	-	256 (21.2)	-
Fernandes 2013 [38]	162	62	37.7 ± 12.6*	-	40 (17.9)	-
Fernandes 2014 [20]	216	45	37 (12.7)	-	0	0
Manfredini 2015 [35]	238	-	49.3 ± 13.5*	-	192 (76.8)	-
Saldanha 2012 [34]	140	60	36.7	-	51 (25.5)	-
Tuz 2003 [48]	200	50	29.6	37.2	35 (17.5)	23 (46)
Vasconcelos 2016 [19]	100	-	(13–70)	-	12 (12)	-

NA Not applicable, SD Standard deviation, TMD Temporomandibular disorder

mean ± SD^*

mean (range)^#

Tinnitus was the baseline condition in 13 studies (n = 7,821 cases), while TMD was the entry condition in 28 studies (n = 9,512 cases). The overlap reflects study designs that evaluated the occurrence of one condition among participants with the other condition in distinct study populations [22].

Tinnitus assessment methods were mainly subjective and included self-reported single-item or study-specific questions, validated instruments such as the Tinnitus Handicap Inventory [29, 30, 32, 39, 44, 45], visual analog scale [34], and general health survey items, [21, 36, 50] reflecting the heterogeneity of assessment approaches historically used in studies examining the relationship between tinnitus and craniomandibular disorders [47] Assessment of TMDs was more structured but very heterogeneous across studies, including clinical examination, Research Diagnostic Criteria for Temporomandibular Disorders (RDC/TMD), Diagnostic Criteria for Temporomandibular Disorders (DC/TMD), often complemented by imaging (e.g., computed tomography or magnetic resonance imaging) and additional functional tests (e.g., pressure pain threshold assessment, electromyography). A summary of diagnostic tools used across included studies is provided in Table 3.

Table 3.

Summary of diagnostic tools used to identify TMDs and tinnitus across included studies

Diagnostic Tool Category	TMD Studies (n=28)	Tinnitus Studies (n=13)
Standardized Criteria
DC/TMD or RDC/TMD	12	5
Clinical Examination Only	8	3
Non-Standardized/Self-report	8	5
Tinnitus Assessment
Validated Tinnitus Instruments	10	8
Single-item Question/Self-report	18	5

Meta-analysis results

Prevalence of tinnitus in individuals with TMDs

Based on 28 single-arm studies, the pooled prevalence of tinnitus among patients diagnosed with TMDs was 0.318 (95% CI, 0.242–0.405; Fig. 2). Substantial between-study heterogeneity was observed (I² = 98.2%, 95% CI: 97.9–98.5%), indicating wide variability in prevalence estimates. Funnel plot findings were interpreted cautiously given the observational study designs and heterogeneity of included studies (Supplementary Fig. 1). Prediction intervals were wide, reflecting substantial between-study variability.

Fig. 2.

Forest plot summarizing the prevalence of tinnitus in patients with temporomandibular disorders (TMDs) across 28 included studies. The analysis yielded a pooled prevalence of 0.318 (95% CI, 0.242–0.405). The size of each square is proportional to the study’s weight in the meta-analysis, and the diamond represents the overall pooled estimate and its 95% confidence interval. Heterogeneity was substantial (I² = 98.2%, 95% CI: 97.9–98.5%)

In contrast to the prevalence analyses, data from 12 comparative studies showed that individuals with TMDs had significantly higher odds of tinnitus compared with controls (OR = 3.924; 95% CI, 2.636–5.841; Fig. 3). Stratified analyses by study design yielded ORs of 4.69 for case-control studies, 3.9 for cohort studies, and 3.5 for cross-sectional studies. Substantial heterogeneity was also observed for this analysis (I² = 90.4%, 95% CI: 85.4–93.7%). Egger’s test suggested potential small-study effects; therefore, publication bias findings were interpreted with caution (Supplementary Fig. 2).

Fig. 3.

Forest plot of the meta-analysis assessing the association between temporomandibular disorders (TMDs) and tinnitus, stratified by study design. The plot presents odds ratios (ORs) with 95% confidence intervals (CIs) from individual case–control, cohort, and cross-sectional studies comparing patients with TMDs to controls. Each square represents the OR from an individual study, and the diamond represents the overall pooled OR derived from a random-effects model (OR = 3.924; 95% CI, 2.636–5.841). Substantial heterogeneity was observed (I² = 90.4%, 95% CI: 85.4–93.7%)

TMD diagnoses in patients with tinnitus

From single-arm studies, the pooled prevalence of TMD diagnoses among patients with tinnitus was 0.634 (95% CI, 0.465–0.776; Fig. 4). Considerable between-study heterogeneity was observed (I² = 78.5%, 95% CI: 6.4–95.0%), reflecting wide variability in reported prevalence across studies. Funnel plot asymmetry suggested potential publication bias; however, these findings were interpreted cautiously in light of the substantial heterogeneity and observational nature of the included studies (Supplementary Fig. 3).

Fig. 4.

Forest plot summarizing the prevalence of temporomandibular disorders (TMDs) in patients with tinnitus across 13 included studies. The analysis yielded a pooled prevalence of 0.634 (95% CI, 0.465–0.776). The diamond represents the overall pooled estimate derived from a random-effects model. Considerable heterogeneity was observed (I² = 78.5%, 95% CI: 6.4–95.0%)

Based on eight comparative studies, individuals with tinnitus had significantly higher odds of a TMD diagnosis compared with those without tinnitus (OR = 4.147; 95% CI, 1.877–9.164; Fig. 5). Considerable heterogeneity was observed (I² = 95.8%, 95% CI: 88.1–98.5%). Egger’s test indicated possible small-study effects, and publication bias assessments were therefore interpreted cautiously (Supplementary Fig. 4).

Fig. 5.

Forest plot of the meta-analysis assessing the association between tinnitus and temporomandibular disorders (TMDs), stratified by study design The plot presents odds ratios (ORs) with 95% confidence intervals (CIs) from individual case–control, cohort, and cross-sectional studies comparing patients with tinnitus to controls without tinnitus. Each square represents the OR from an individual study, and the diamond represents the overall pooled OR derived from a random-effects model (OR = 4.147; 95% CI, 1.877–9.164). Considerable heterogeneity was observed (I² = 95.8%, 95% CI: 88.1–98.5%)

Study quality

The quality of all cohort and case-control studies was rated as good according to the Newcastle–Ottawa Scale (Supplementary Table S2). In contrast, the quality of one-third of the cross-sectional studies was rated as fair or poor, primarily due to limitations in exposure and outcome assessment and insufficient control for confounding factors (Supplementary Table S3). Overall certainty of evidence ranged from low to very low across all primary outcomes according to the GRADE assessment (Supplementary Table S4).

Discussion

This meta-analysis included data from 34 studies, in which tinnitus was the baseline condition in 13 studies and TMDs were the baseline condition in 28 studies. Collectively, these findings indicate an association between TMDs and tinnitus; however, the certainty of this evidence is limited by the observational study designs and diagnostic heterogeneity. Similar investigations have reported increased odds of tinnitus among individuals with TMDs, as well as increased prevalence of TMD diagnoses among patients with tinnitus; however, the predominantly cross-sectional nature of the evidence limits inference regarding temporal direction or causality [24, 25].

Tinnitus assessments in the included studies depended primarily on self-reports, consistent with the methods used in other studies to evaluate tinnitus severity [21, 36, 50, 53]. TMDs were diagnosed using more structured approaches, including clinical examination and, in some studies, standardized diagnostic criteria and imaging assessments. This methodological difference is a source of significant variability among studies, emphasizing the importance of the use of standardized assessments in research to enhance the comparability of findings. Future studies should employ validated tinnitus tools (e.g., THI, TFI) and standardized TMD diagnostic criteria (e.g., DC/TMD), with uniform data collection protocols, to improve cross-study comparability.

Prevalence of tinnitus in patients with TMDs

‏This meta-analysis yielded an overall pooled prevalence of tinnitus among patients with TMDs of 31.8%, suggesting an association between the two conditions. This finding is consistent with previous reports [23, 54]. For example, Kusdra et al. [40] found that patients with TMDs were more likely to report auditory symptoms such as tinnitus, and a review conducted by Bader and Lavigne [55] suggested that chronic pain conditions, including TMDs, are associated with a higher prevalence of tinnitus. However, estimates of the prevalence of tinnitus among individuals with TMDs varied significantly among studies included in this meta-analysis (5–10% to 70%), compatible with other reports [24, 56]. This variation may be attributed to differences in study design, sample size, participants’ demographic characteristics, and the diagnostic criteria used for TMDs and tinnitus [2, 57]. For instance, studies conducted with smaller, more homogeneous samples may yield lower prevalence estimates, whereas those conducted with larger, more diverse samples may capture a broader range of experiences, leading to higher estimates [53]. Additionally, the assessment of tinnitus using self-reported questionnaires in some studies and clinical evaluation in others can lead to differences in reported prevalence. Hypothesized mechanisms underlying the association between TMDs and tinnitus may also vary among individuals, with psychological factors such as stress and anxiety and physical factors such as jaw dysfunction and nerve involvement playing divergent roles in different populations [23, 58]. These contributing elements highlight the complexity of both conditions and the need for standardized research methodologies to better understand their relationship. Overall pooled prevalence estimates were calculated in this study to provide a more comprehensive overview of the association between TMDs and tinnitus despite these reporting variations [59].

This analysis indicated that individuals with TMDs are nearly four times more likely to experience tinnitus than are those without TMD. This observed association may be partially explained by the shared anatomical and neurological connections between the stomatognathic and auditory systems, although such mechanisms remain hypothetical. Importantly, association does not imply causation, and the findings are constrained by the observational nature of the included studies. TMD-related factors such as muscle tension, psychological stress, and occlusal alterations have been hypothesized to influence auditory processing and tinnitus perception in some individuals [60]. Evaluation of potential treatment effects requires dedicated longitudinal and interventional studies [61].

The high level of between-study heterogeneity observed in studies assessing tinnitus occurrence among patients with TMDs indicates substantial variability in effect estimates across included populations [62]. This observation is consistent with some previous findings [21, 62]. However, evidence of publication bias in studies comparing patients with TMDs to controls was detected. Studies conducted with smaller (and often lower quality) samples tend to show more extreme effects than those conducted with larger samples, and statistically significant or favorable results are more likely to be published than null findings, meaning that pooled estimates may not accurately reflect the true effect size in the broader literature [63, 64]. This may result in overestimation of pooled effect sizes. Our findings highlight concerns about publication bias in pooled data and the need for caution when interpreting results derived from them.

Diagnosis of TMDs in patients with tinnitus

The pooled prevalence of TMDs among patients with tinnitus in this meta-analysis was 63.4%, with a corresponding 95% confidence interval of 46.5% to 77.6%. This wide confidence interval, together with substantial variability across studies, indicates considerable uncertainty in the precision of the pooled estimate [60, 65]. Thus, these findings should be interpreted with caution, and additional well-designed studies using standardized criteria are needed to explore this association further.

This meta-analysis demonstrated that TMDs were more frequently diagnosed among individuals with tinnitus than among those without tinnitus, based on comparative analyses. The high level of between-study heterogeneity observed in these analyses (I² = 95.8%) indicates substantial variability in effect estimates across the included studies [64]. However, evidence of potential publication bias was detected, indicating that studies reporting higher prevalence of TMDs among patients with tinnitus may be more likely to be published, which could contribute to overestimation of the pooled effect. These results are similar to those of other studies [66, 67]. For example, a meta-analysis conducted by Fabrizia et al. [54] also revealed an increased prevalence of TMDs among patients with tinnitus. However, other researchers such as Oliveira et al. [68] did not observe a significant difference in this prevalence between tinnitus and non-tinnitus groups, highlighting the heterogeneity of the available evidence and the complexity of this relationship.

Variability in prevalence and association estimates

Discrepancies in reported TMD and tinnitus prevalence rates and associations between the conditions can be attributed to several factors, including differences in the diagnostic criteria and assessment instruments used, study population characteristics, and the presence of comorbidities. Notably, some studies focused on the associations of specific TMD subtypes with tinnitus. For example, Akhter et al. [17] found that tinnitus was more prevalent among patients who experienced difficulties with mouth opening and those with TMJ clicking and pain. Karacay and Korkmaz [41] found no significant difference in the tinnitus prevalence based on the TMD subtype, reporting notably high rates across various TMD categories.

Further complicating the understanding of the interplay between TMDs and tinnitus, the recent meta-analysis of Omidvar and Jafari [23] showed that referral bias may contribute to the variability of findings, potentially resulting in the overrepresentation of patients with both conditions in clinical settings. They reported ORs for the prevalence of tinnitus in patients with TMDs ranging from 1.78 to 7.80, and elevated odds of TMD diagnosis in patients with tinnitus [23]. They noted that the strength of the association between the conditions appeared to vary based on the population studied, being greater, for instance, among patients receiving dental care than in the general population [23]. The present meta-analysis, conducted with a larger dataset, yielded higher ORs for the prevalence of tinnitus among patients with TMDs relative to controls. Additionally, the OR for TMD diagnosis among patients with tinnitus varied across study types, further indicating frequent co-occurrence of tinnitus and TMD. These findings, similar to those of other studies [40, 62], highlight the relevance of considering TMDs as a potential comorbidity in patients with tinnitus in research and clinical contexts.

Comparison with recent systematic reviews

Our findings align broadly with recent systematic reviews, such as those by De La Torre Canales et al. and Frosolini et al. (CRANIO 2025), which reported an association between TMD and tinnitus. The present review adds incremental value by offering an updated and comprehensive quantitative synthesis, including a larger number of studies and a recent search update. Importantly, this study provides pooled prevalence estimates for both directions of the association alongside comparative odds ratios, with prevalence analyses conducted using logit transformation to improve statistical handling. While the overall conclusions regarding the association remain consistent, our detailed subgroup and sensitivity analyses, particularly those addressing diagnostic heterogeneity, offer a more granular understanding of the variability in reported effects, which was a key limitation in prior reviews. This approach contributes to a clearer characterization of co-occurrence patterns and highlights the influence of methodological differences across studies.

Study quality

The methodological quality of the studies included in this meta-analysis was generally good, but differed according to the study design. All cohort and case-control studies were of good quality, highlighting the comparatively higher methodological rigor of these designs, which allow better control for confounding and more reliable assessment of associations. Consistently, previous research has emphasized that cohort and case-control studies provide dependable data for epidemiological investigations [69, 70]. In contrast, about one-third of the cross-sectional studies included in this meta-analysis were of fair or poor quality, raising concerns about the reliability of their findings. Cross-sectional designs entail challenges such as recall bias, selection bias, and the inability to establish temporal direction or causality. In similar research, significant portions of cross-sectional studies in the same field have also received lower quality ratings [23, 53].

Thus, despite the statistically significant associations observed in the present analysis, the findings should be interpreted with caution due to the prevalence of cross-sectional studies in the sample. Additionally, variability in study quality and the potential for publication bias could have influenced association strength. Future research efforts should focus on the application of more robust methodologies; in particular, prospective longitudinal studies are needed to clarify the temporal relationship between TMDs and tinnitus and the mechanisms underlying these conditions.

Limitations

This systematic review and meta-analysis has several limitations that warrant cautious interpretation of the findings. First, the predominantly observational and cross-sectional nature of the included studies limits the ability to infer temporal relationships or establish causality between TMD and tinnitus. Such designs are susceptible to reverse causality and residual confounding, which may influence the observed associations.

Second, diagnostic heterogeneity represents an important source of bias. Criteria used to diagnose TMD varied widely, ranging from standardized tools such as DC/TMD and RDC/TMD to non-standardized self-report measures or clinical examinations without clearly defined protocols. Similarly, tinnitus assessment methods ranged from validated instruments to single-item questions. This variability increases the risk of misclassification bias and may have contributed to inconsistent effect estimates across studies.

Third, the substantial statistical heterogeneity (I² values consistently exceeding 75%) observed across most analyses indicates considerable between-study variability. This heterogeneity likely reflects differences in study populations, methodological approaches, and diagnostic definitions. Although random-effects models and subgroup analyses were applied, the presence of high heterogeneity suggests that pooled estimates should be interpreted cautiously.

Fourth, evidence of potential publication bias, indicated by Egger’s test in several analyses, suggests that studies reporting significant or favorable findings may be more likely to be published. This selective reporting could result in an overestimation of true effect sizes, despite the use of sensitivity analyses such as trim-and-fill.

Finally, the lack of robust longitudinal studies prevents definitive conclusions regarding the temporal or “bidirectional” nature of the association. At present, the relationship between TMD and tinnitus is more appropriately described as co-occurrence or association rather than causal linkage. Future research should prioritize prospective longitudinal designs using standardized diagnostic criteria for both conditions to improve causal inference and reduce methodological bias.

Conclusion

Evidence suggests an association between temporomandibular disorders and tinnitus, but certainty is limited by diagnostic heterogeneity, predominantly cross-sectional study designs, and potential publication bias. The pooled prevalence of tinnitus among individuals with TMDs was 31.8%, and that of TMDs among patients with tinnitus was 63.4%; however, the considerable variability across studies indicates a complex and heterogeneous relationship. These findings therefore warrant cautious interpretation. Future longitudinal studies using standardized diagnostic criteria (e.g., DC/TMD) and validated tinnitus assessment instruments are essential to clarify temporal relationships and strengthen the evidence base.

Abbreviations

TMD

Temporomandibular disorder

DC/TMD

Diagnostic Criteria for Temporomandibular Disorders

RDC/TMD

Research Diagnostic Criteria for Temporomandibular Disorders

OR

Odds ratio

CI

Confidence interval

PRISMA

Preferred Reporting Items for Systematic Reviews and Meta-Analyses

Authors’ contributions

**Hebah A. Alghamdi: ** Conceptualization, Methodology, Software, Validation, Formal analysis, Investigation, Resources, Data Curation, Writing – Original Draft, Writing – Review & Editing, Visualization, Supervision, Project administration.**Alanoud A. Alwably: ** Methodology, Software, Validation, Formal analysis, Investigation, Resources, Data Curation, Writing – Review & Editing, Visualization.**Faisal I. Alsaad: ** Methodology, Software, Validation, Formal analysis, Investigation, Resources, Data Curation, Writing – Review & Editing, Visualization.**Reem J. Alghamdi: ** Methodology, Software, Validation, Formal analysis, Investigation, Resources, Data Curation, Writing – Review & Editing, Visualization.**Asmaa Faden: ** Conceptualization, Methodology, Validation, Writing – Review & Editing, Supervision. All authors have read and approved the final version of the manuscript.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Data availability

All data analysed during this study are derived from published articles and their supplementary materials. Extracted data and analytic outputs are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.