Abstract
Objective To study clinical variables defining temporomandibular function in adults with juvenile idiopathic arthritis (JIA) and healthy controls.
Methods In this cross-sectional study, the temporomandibular joint (TMJ) screening protocol, mandibular range of motion (MROM), and anterior maximum voluntary bite force (AMVBF) were compared between adults with JIA and healthy controls. Unadjusted and adjusted models with corrections for sex and disease duration were constructed for active maximum interincisal mouth opening (AMIO) and AMVBF.
Results A total of 100 adults with JIA and 59 healthy adults were included in this study. In adults with JIA, 56% had clinically established TMJ involvement. AMIO was the MROM variable most reduced by TMJ involvement; AMIO was 8.8 mm (95% CI −11.40 to −6.12; P < 0.001) less in adults with JIA with TMJ involvement compared to JIA without TMJ involvement. No differences of AMIO were found between healthy adults and adults with JIA without TMJ involvement (−2.52, 95% CI −5.13 to 0.10; P = 0.06). Male sex was associated with a higher AMIO, and disease duration was associated with a decreased AMIO. Collinearity between the subtype prebiologic era and disease duration was found. AMVBF did not differ between adults with JIA and healthy adults.
Conclusion The high prevalence of clinically established TMJ involvement in adults with JIA indicates the need for awareness of TMJ problems in adults with JIA. TMJ involvement negatively influenced AMIO and should therefore be part of the TMJ screening in adults with JIA. AMVBF seems to have less utility for TMJ screening in adult populations.
The diagnosis of juvenile idiopathic arthritis (JIA) is based on chronic arthritis with an onset before the age of 16. In 30% to 50% of patients, the disease will remain active into adulthood.1-3 In a previous study, clinically established temporomandibular joint (TMJ) involvement was found in 30% of children with JIA.4 The prevalence of TMJ involvement varies widely (30-96%) because of the differences in TMJ involvement definition, diagnostic method, and characteristics of study participants.4-7 In 2019, an international consensus group defined TMJ involvement as abnormalities presumed to be the results of TMJ arthritis.7 In the current study we focused on clinical signs or symptoms of TMJ in patients with JIA, such as a reduced mouth opening, TMJ crepitation, and pain.4,7-9 These variables are widely studied in patients with JIA during childhood, but data on long-term clinically established TMJ involvement in patients with JIA who are now adults are scarce. Studies in adults with JIA determined TMJ involvement with radiological diagnostics, and therefore differ from our clinical approach.10-12 TMJ inflammation in childhood can lead to growth disturbances, TMJ damage, and eventually temporomandibular impairments in adulthood.10-12 TMJ damage can result in facial asymmetry and dentofacial deformity for which reconstructive jaw surgery can be needed. In the past, the cornerstone of JIA treatment was conventional synthetic disease-modifying antirheumatic drugs (csDMARDs) such as methotrexate. Around the turn of the 20th century, the introduction of biologic therapies strongly improved the treatment of JIA.13 It has been suggested that the TMJ benefits from biologic DMARDs (bDMARDs).14 In the current adult JIA population, patients had the opportunity for treatment with a biologic during childhood (diagnosed in the biologic era) whereas another group of patients with JIA has been treated in an era without the availability of biologics during childhood (prebiologic era). Focusing on these 2 groups may help to better understand TMJ involvement in adults and the supposed role of biologics. Therefore, the aim of this study is to compare the TMJ screening protocol score, mandibular range of motion (MROM) variables, and the anterior maximum voluntary bite force (AMVBF) in (1) adults with JIA and healthy controls, (2) adults with JIA with clinically established TMJ involvement and JIA without clinically established TMJ involvement, and (3) adults with JIA whose childhood was in the biologic era (diagnosis of JIA in or after 2002) and in adults with JIA whose childhood was in the prebiologic era (diagnosis of JIA earlier than 2002).
METHODS
In this cross-sectional study, participants were in care of the Department of Rheumatology & Clinical Immunology at the University Medical Center (UMC) Utrecht between November 2019 and September 2021. The inclusion criteria were adults (1) with JIA diagnosed according to the International League of Associations for Rheumatology criteria, and (2) aged between 18 years and 65 years. The exclusion criteria were as follows: (1) a history of mandibular trauma, (2) a history of maxillofacial surgery, and (3) an additional orofacial condition not related to JIA (eg, dental pain). All patients with JIA visiting the outpatient clinic in the past 12 months were invited to participate in the study. The invitation letter described the inclusion and exclusion criteria. None of the patients applying for our study were excluded based on the exclusion criteria. Healthy controls were recruited from a dental office and among colleagues. The inclusion criterion for healthy controls was age between 18 and 65 years. The exclusion criteria were equal to the JIA group. The study protocol was approved by the ethics committees of the UMC Utrecht (study ID: NL.METC-17-528/C). All participants received written information and provided their oral and signed informed consent. The following data from the electronic medical records were extracted: JIA subtype, date of diagnosis, disease duration, height, weight, sex, age, the presence of antinuclear antibody, rheumatoid factor (RF), HLA-B27, Disease Activity Score in 28 joints (DAS28), and current medication use such as bDMARDs and csDMARDs.15
Measurement instruments. In all participants the TMJ screening protocol, MROM, and AMVBF were assessed.
TMJ screening protocol. A total of 11 items regarding history, examination, and inspection produced the TMJ screening protocol.16 The history items addressed (1) problems in chewing, (2) eating slower than others, (3) difficulty in biting hard food, (4) pain while eating, and (5) a limited mouth opening. The clinical examination items of the TMJ screening addressed (6) limited mouth opening, the cutoff value for which was ≤ 40 mm without inclusion of the overbite,17 (7) crepitation on mandibular opening and closing, (8) pain on active maximum interincisal opening (AMIO), and (9) left or right mandibular midline deviation on opening wide. The inspection items of the TMJ screening protocol addressed (10) facial asymmetry and (11) retrognathia.
Each positive item of the TMJ screening protocol receives 1 point. All positive items produced the TMJ screening protocol score. The TMJ screening protocol has been validated by the Juvenile Arthritis Disease Activity Score in 27 joints (JADAS27).16 A TMJ screening protocol score of at least 2 was found to have the highest sensitivity and specificity to differentiate between a low disease activity with a JADAS27 score ≤ 2.2 and a higher disease activity score with a JADAS27 ≥ 6.4. Following our protocol, we operationally defined TMJ involvement as a TMJ screening protocol score ≥ 2 in this study.
MROM. The MROM, AMIO, passive maximum interincisal opening (PMIO), protrusion, and left and right laterotrusion were assessed. The overbite was not included in the AMIO and PMIO. The MROM was measured with a metal ruler to the nearest millimeter. During the AMIO and PMIO measurement, the participants were asked to open their mouth as wide as possible. During the PMIO measurement, the examiner applied a gentle stretch with the index finger and thumb on the incisal edges of the incisors to increase the mouth opening. Protrusion was assessed by requesting the participants to protrude the mandible as far anterior as possible. In the measurement of left and right laterotrusion, the dental midlines were used as reference points. In case of a midline shift with the teeth in occlusion, a correction was carried out for the size of this shift. The difference between left and right laterotrusion was labeled as “discrepancy between left and right laterotrusion.” Limitation in condylar sliding was assessed by palpation of the TMJ region during AMIO.
AMVBF. In a previous study, we found a 24 Newton (N) lower bite force in children with JIA compared to healthy children, and a 42 N lower bite force in case of JIA with TMJ involvement.4 The AMVBF was measured by placing a bite force transducer between the upper and lower central incisors.4 The bite force measurement consists of clenching as hard as possible for 10 seconds at maximum strength. Three attempts were documented. The highest bite force of the 3 attempts was defined as the AMVBF and is expressed in N. A good to excellent reliability of AMVBF measurements was found for children with JIA.18
Statistical analysis. In this study, we classified adults into groups: (1) JIA and healthy controls, (2) JIA with vs without clinically established TMJ involvement (TMJ protocol score ≥ 2),16 and (3) JIA diagnosed in the prebiologic era (biologics were not prescribed yet during childhood; defined as JIA diagnosed before 2002) and JIA diagnosed in the biologic era (adults with JIA who had the availability of biologics as a treatment option during childhood; defined as JIA diagnosed in 2002 or later). These groups are based on the date of authorization by the European Medicines Agency and Dutch Health Authorities on the use of biologics.19
Characteristics were presented as numbers and percentages for categorical variables and mean (SD) for normally distributed variables. Normality was assessed by plotting graphs (histograms and Q-Q plots). In our data, all variables were normally distributed. For the analyses of all clinical data, the unpaired t test was used for continuous data and the chi-square test was used for dichotomous or ordered categorical outcomes.
A clinically relevant effect of having JIA and clinically established TMJ involvement was mainly reflected in AMIO and AMVBF in children with JIA.4,20 In addition, AMIO is the most studied MROM variable in TMJ examination.8 Therefore, we further assessed AMIO and AMVBF in our predefined study groups by linear regression. We indicated disease duration and sex as potential confounders. In addition, we computed the previous biologic use duration as a percentage of disease duration. This constructed variable will take disease duration into account. First, unadjusted models were established to indicate variables of influence on AMIO and AMVBF. Second, we performed adjusted analyses with corrections for disease duration and sex. Validity of the models (ie, normality, homoscedasticity) was assessed with residual analysis.21 Results were reported as regression coefficients with 95% CI and P values. For dichotomous variables, regression coefficients represent the difference in mean AMIO or AMVBF. For continuous variables, the regression coefficient represents the increase in mean AMVBF or AMIO for each unit increase in the explanatory variable.
P < 0.05 was accepted as significant. Statistical analyses were performed using SPSS 25 (IBM SPSS Statistics for Windows).
RESULTS
In this cross-sectional study, 100 adults with JIA and 59 healthy adults were included. The mean age of adults with JIA was 28.8 (SD 9.2) years, and for healthy adults it was 32.4 (SD 15.1) years. Of the adults with JIA, 75 (75%) were female, and of the healthy adults, 32 (54%) were female (Table 1). Of the adults with JIA, 56 (56%) had a TMJ protocol score ≥ 2 and were therefore labeled with “TMJ involvement.” The adults with JIA and TMJ involvement had an older age (31.5 [SD 10.8] years) compared to the adults with JIA without clinically established TMJ involvement (25.3 [SD 5.1] years; P < 0.001). The age at onset was younger in the JIA with TMJ involvement group (6.5 [SD 5.1] years) compared to JIA without TMJ involvement (9.9 [SD 6.0] years; P = 0.003). The disease duration was longer in the TMJ involvement group (24.6 [SD 12.9] years) compared to the group without TMJ involvement (15.0 [SD 8.1] years; P < 0.001). The JIA subtype differed between adults with JIA with and without TMJ involvement (P = 0.03); polyarticular RF negative was more prevalent in the TMJ involvement group (38%) compared to the group without TMJ involvement (25%).
Demographics in adults with JIA and in healthy adults.
The group of adults with JIA was classified into 54 adults originating from the prebiologic era and 46 from the biologic era (Table 1). The age at onset was significantly lower (4.4 [SD 3.8] years) in the prebiologic era group than in the biologic era group (12.3 [SD 4.6] years; P < 0.001) as a result of our study design. The current biologic use did not differ between the prebiologic era group and the biologic era group (P = 0.78). As a result of the study design, patients with JIA from the prebiologic era were older (33.2 vs 23.5 years; P < 0.001) and had longer disease duration (28.5 vs 10.7 years; P < 0.001).
TMJ screening protocol. All TMJ screening protocol score items were more common in adults with JIA than in healthy adults (P < 0.05; Table 2). Of all adults with JIA and clinically established TMJ involvement, the following TMJ screening protocol items were most prevalent: crepitation (50%), deviation > 2 mm during AMIO (48%), pain while eating (39%), asymmetry (39%), and retrognathia (39%; Table 2). In the uncorrected analysis for the prebiologic era group, TMJ involvement was more prevalent (70%) than in the biologic era group (39%; P = 0.002; Table 2). However, collinearity existed between the subtype prebiologic era vs biologic era and disease duration. This means that these variables are highly correlated with each other.
Clinical outcomes in adults with JIA and in healthy adults.
MROM. The mean uncorrected AMIO in adults with JIA was 47.5 (SD 8.3) mm, significantly lower than 55.3 (SD 6.0) mm in healthy adults (P < 0.001; Table 2). The mean PMIO, protrusion, and left and right laterotrusion were all significantly lower, and limited condylar sliding and the discrepancy between left and right laterotrusion were more prevalent in adults with JIA compared to healthy adults (Table 2). In adults with JIA and TMJ involvement, the mean uncorrected AMIO was 43.5 (SD 7.5) mm compared to 52.5 (SD 6.3) mm in adults with JIA without TMJ involvement (P < 0.001; Table 2). PMIO and condylar sliding were impaired in adults with JIA and TMJ involvement, compared to adults with JIA without TMJ involvement (P < 0.01). The other MROM variables were not significantly different in adults with JIA, with or without TMJ involvement, and not significantly different between the prebiologic era and biologic era group (Table 2).
The following variables were associated with a lower AMIO: JIA vs healthy adults, JIA with vs without TMJ involvement, JIA prebiologic era vs JIA biologic era, disease duration, age at onset, TMJ screening protocol score, limited condylar sliding, and medication use (P < 0.05; Table 3). Sex and height were associated with a higher AMIO (P < 0.05; Table 3). Adjusted analyses, with correction for sex, confirmed the lower values in adults with JIA for AMIO (−2.52, 95% CI −5.13 to 0.10; P < 0.001; Table 4) compared to healthy adults. However, when we include TMJ involvement in the adjusted analyses (Model 2, Table 4), there was no difference in AMIO between healthy adults and adults with JIA without TMJ involvement (−2.52, 95% CI −5.13 to 0.10; P = 0.06). A lower AMIO of 8.8 mm was found in JIA with TMJ involvement (95% CI −11.40 to −6.12; P < 0.001; Table 4) compared to JIA without TMJ involvement. No significant difference in AMIO was found between the prebiologic vs biologic era groups (P = 0.31; Table 4).
Unadjusted linear regression for AMIO and AMVBF in adults with JIA and healthy adults.
Adjusted linear regression for AMIO in adults with JIA and healthy adults (mm).
AMVBF. The uncorrected mean AMVBF was significantly lower in adults with JIA (160.8 [SD 62.5] N) compared to healthy adults (189.6 [SD 92.7] N; P = 0.04; Table 2). There were no differences between the mean AMVBF in adults with JIA with and without TMJ involvement (P = 0.19); likewise, there were no differences between the prebiologic era and the biologic era group (P = 0.95). The unadjusted analysis showed a 28.8 N lower AMVBF in adults with JIA compared to healthy adults (P = 0.02; Table 3). Sex, height, the TMJ screening protocol score, and limited condylar sliding were factors of influence on AMVBF (P < 0.01; Table 3). After correction for disease duration and sex, adults with JIA did not significantly differ in AMVBF from healthy adults (P = 0.24; Table 5).
Adjusted linear regression for AMVBF in adults with JIA and healthy adults (N).
DISCUSSION
TMJ involvement, defined as a TMJ protocol score ≥ 2, is highly prevalent in adults with JIA (56%). Damage-related items such as crepitation, deviation during AMIO, and retrognathia are the most scored items on the TMJ screening protocol in adults with JIA. Of all MROM variables, AMIO is the variable with the most distinct reduction in adults with JIA and TMJ involvement. We found that adults with JIA have an 8.8-mm lower AMIO compared to adults with JIA without TMJ involvement. No differences in AMIO were found between healthy adults and adults with JIA without clinically established TMJ involvement. AMVBF seems not to differ in adults with JIA and healthy adults. Since the variable disease duration and the subgroup prebiologic vs biologic era are highly correlated because of our study design, we cannot distinguish the effects of these 2 variables on TMJ involvement, AMIO, and AMVBF.
Most TMJ studies in adults with JIA used cone beam computed tomography (CBCT), panoramic radiographs, or magnetic resonance imaging (MRI) as diagnostic tools to indicate TMJ abnormalities, affected TMJ, growth disturbances, TMJ involvement, and condylar deformities.10-12,22-24 An international consensus-based recommendation for standardizing terminology defined TMJ involvement as “clinical situations in which no contrast-enhanced MRI verification of active TMJ inflammation has occurred but where signs, symptoms, and/or radiological findings suggest the presence of actual or former TMJ arthritis.”7 In our study we used a clinical approach to define TMJ involvement.16 Consequently, this may underestimate the underlying pathology of TMJ involvement such as TMJ arthritis and TMJ damage, as we did not use imaging techniques such as the gold standard MRI.25 TMJ abnormalities were found in 55% of a study population of adults with JIA confirmed by CBCT.12 In a study in which MRI was used, 70% of the adults with JIA had TMJ involvement and thereby 57% had TMJ growth disturbances.10,22 TMJ involvement confirmed by panoramic radiographs was found in 67% of adult women with a history of JIA.23 These prevalence rates are in line with our results, in which clinically established TMJ involvement was found in 56% of adults with JIA.
The most prevalent TMJ screening protocol items in this study were deviation during AMIO (31%), crepitation (30%), retrognathia (25%), problems in chewing (23%), pain while eating (22%), limited mouth opening (22%), and asymmetry (22%). The prevalence of deviation during AMIO is reported in 11% to 39% of adults with JIA.11,12 Crepitation was present in 25% of adults with JIA, and in 41% of women with a history of JIA.23,26 Retrognathia was found in 27% of adults with JIA.10 Facial asymmetry was found in 76% of adults with JIA using cephalometric measurement, which may explain the lower prevalence of asymmetry (22%) in our study.12 A limited mouth opening, defined differently in each study, was found in 29% to 55% in adults with JIA.11,12,22,23,26 A correlation between mouth opening and disease duration, TMJ involvement, and molar bite force is described.23 In a study to assess AMVBF in children with JIA, lower AMVBF was found for children with JIA compared to healthy controls.4 However in our study, AMVBF was not impaired in adults with JIA.
Screening and monitoring of the TMJ is advised by a consensus-based recommendation for patients with JIA based on literature of children with JIA.8 Our study confirmed that TMJ involvement influenced AMIO negatively and should therefore be part of the TMJ screening in adults with JIA. AMIO is a quick, easy, and highly reliable measurement for every day clinical practice to be used by the rheumatologist. AMVBF did not differ between adults with and without TMJ involvement; therefore, AMVBF seems to have less utility for TMJ screening in adult populations. Screening of the TMJ may help the rheumatologist to determine if a referral to a multidisciplinary team consisting of a maxillofacial surgeon, orthodontist, orofacial pain specialist, and/or physiotherapist for further diagnosis and management is needed.
The strengths of this study are the comprehensive TMJ screening, the comparison with a control group, and assessment of the oral function variables MROM and AMVBF. A limitation of this study is the potential selection bias. Of all adults with JIA, a selected group of patients was visiting the university hospital because 30% to 50% of the children with JIA will have active JIA in adulthood.1-3 We noticed a variation in the subtypes between the adults with JIA in this study and children with JIA studied before.20 On the other hand, adults with JIA visiting the university hospital are likely the patients with higher disease activity, with the need for screening of all joints, including the TMJ. In future studies it will be interesting to follow-up a wider group of adults with JIA—patients with JIA treated in nonacademic hospitals and adults with JIA who have a clinical remission of the disease. This may solve the potential selection bias in our study. Because of our cross-sectional study design, age at onset in the biologic era group (12.3 [SD 4.6] years; P < 0.001) is higher than in the prebiologic era group (4.4 [SD 3.8] years). As a result, the disease duration is lower in the biologic era group compared to the prebiologic era group as well. Patients assigned to the biologic era group have to be diagnosed with JIA in 2002 or later. Therefore, at the initiation of our study (November 2019), adults with JIA from the biologic era group have a higher chance to be diagnosed at an older age. Children with a lower age at onset in the biologic era group are not yet adults at the time of our study, and are therefore not included.
Another limitation of this study is the collinearity between disease duration and the subcategory prebiologic era vs biologic era caused by our study design. The prebiologic era group consisted of adults with JIA with an older age (33.2 years) and a longer disease duration (28.5 years), compared to the biologic era group (23.5 years old and 10.7 years of disease duration). A longer disease duration is associated with a higher prevalence of TMJ involvement.27-29 Also, biologics were revolutionary in treatment of JIA, and may positively affect the TMJ as well.13,14 However, we could not distinguish the effect of disease duration from the subdivision of prebiologic era vs biologic era on TMJ involvement, AMIO, and AMVBF. In an additional analysis, we calculated the history of biologic medication use in months as a percentage of disease duration to deal with these correlated variables. However, we did not find a significant effect on AMIO and AMVBF (Table 3). The history of biologic use may not be appropriate to analyze the effect of biologics on the TMJ. This may be because of the step-up treatment schemes that are used by most rheumatologists (ie, biologics are recommended in case nonsteroidal antiinflammatory drug and DMARD treatments fail) for the subtype polyarthritis and for patients with features of poor prognosis.30 Therefore, the group of adults with a few months of biologic use may consist of adults with mild disease activity and of adults from the prebiologic era group.
Another limitation of this study is the absence of a validated disease activity score in patients with JIA.15,31 As an alternative we used the DAS28, although, in 28 of the 100 patients included in this study, the DAS28 was missing. In most cases, the TMJ measurements of our study were combined with a visit of the patients with their rheumatologist. In 14 patient visits, patients did not visit their physician on the day of the measurement. Disease activity was scored every visit, but DAS28 was not always calculated because of a missing value (usually the visual analog scale). Therefore, this score likely only partially displayed the disease activity in our study population (72%).
In 56% of our adult JIA study group, we found TMJ involvement. As TMJ involvement is a clinical diagnosis, we cannot differentiate between the underlying pathology of TMJ damage and TMJ arthritis. Our clinical approach is a first step to distinguish patients with potential TMJ pathology. A MRI-based TMJ study can help to further differentiate between TMJ damage and TMJ arthritis. In future studies, it may be interesting to correlate the TMJ screening protocol items with MRI results. This information will answer the question of whether the clinical signs of adults with TMJ damage and TMJ arthritis differ. This may help to narrow the indication for MRI diagnostics, since a suspicion of TMJ inflammation is an indication for MRI, whereas TMJ damage can be diagnosed by panoramic radiographs or CBCT. A next step and the ultimate goal is to determine the best treatment options for adult patients with JIA with TMJ involvement, as this strategy is currently missing.
The high prevalence of clinically established TMJ involvement in adults with JIA indicates the need for awareness of TMJ problems in both children and adults with JIA. TMJ involvement influenced AMIO negatively and should therefore be part of the TMJ screening in adults with JIA. AMVBF did not differ between adults with and without TMJ involvement; therefore, AMVBF seems to have less utility for TMJ screening in adult populations.
Footnotes
The authors declare no conflicts of interest relevant to this article.
- Accepted for publication June 12, 2023.
- Copyright © 2023 by the Journal of Rheumatology
