Prevalence of temporomandibular disorders diagnosis in patients treated with Herbst appliance: a systematic review and meta-analysis

We looked through Scopus, Web of Science, and PubMed for publications published between the beginning and May 1, 2023. The search method used to look up documents in search engines is shown in Table 1.

Additionally, we manually looked through published topical and systematic reviews on related subjects.

The Cochrane Handbook for Systematic Reviews of Interventions and the Preferred Reporting Items for Systematic Reviews (PRISMA) guidelines 2020 served as our guides when conducting this systematic review. The protocol for the systematic review has been registered as CRD42022354011 on the International Prospective Register of Systematic Reviews (PROSPERO).

We applied the following Population, Intervention, Comparator, and Outcomes (PICO) model to assess the document eligibility:

P) Participants consisted of Class II division 1 patient.

I) The Intervention consisted of orthodontic treatment with the Herbst appliance.

C) The Comparison consisted of a clinical analysis for temporomandibular disorders before and after treatment with Herbst’s appliance in the same patients.

O) The Outcome was the occurrence of temporomandibular disorders in orthodontic patients treated with Herbst applications.

Only studies that provided prevalence data before and after treatment were included. We established the following exclusion criteria: 1) diagnosis of rheumatic diseases or chronic inflammatory disorders (e.g., rheumatoid arthritis, juvenile arthritis, idiopathic arthritis, psoriatic arthritis); 2) diagnosis of fibromyalgia; 3) congenital abnormalities or neoplastic conditions in the TMJ region; 4) studies that included subjects who underwent arthrocentesis or intra-articular infiltration; 5) cross-over study design; 6) language other than English (we considered only articles written in English); 7) unavailability of full-text (e.g., posters and conference abstracts); 8) studies involving animals; 9) review articles (topical or systematic); 10) case reports/series; 11) evaluation of TMD prior Herbst therapy.

The data extracted were collected in a digital database.

The following data were extracted by two reviewers (MDB and BV). 1) First author; 2) Year of publication; 3) Nationality; 4) Number of study participants; 5) Age of study participants; 6) Diagnostic criteria/tools used for the diagnosis of TMD; 7) Prevalence in patients before treatment; 8) Prevalence in patients post-treatment.

The risk of bias in the included studies was evaluated by two reviewers (MC and GM) using the Cochrane risk-of-bias instrument for randomized trials (RoB 2), Version 2. An established method for assessing the caliber of randomized trials is the Cochrane RoB 2 instrument. Random sequence generation, allocation concealment, participant and staff blinding, outcome assessment blinding, inadequate outcome data, and selective reporting are the six potential bias domains that are taken into account. Any differences of opinion were discussed with a third reviewer (MDB) until a consensus was obtained.

The software Review Manager version 5.2.8 (Cochrane Collaboration, Copenhagen, Denmark; 2014) was used to perform the pooled analysis. We measured the Odds Ratio (OR) between the two groups (prevalence of TMD in pre-Herbst and post-Herbst). The Higgins Index (I²) and the chi-square test were implemented to assess Heterogeneity among studies. We classified heterogeneity as follows: low heterogeneity (< 30%), medium heterogeneity (30–60%), and high heterogeneity (> 60%).

Three studies were included in the systematic review and were considered for the metanalysis, as illustrated in the PRISMA 2020 flowchart in Fig. 1. Sixty articles were selected after the search. One report was excluded before the screening because it was not in English. Twelve records were duplicates and, therefore, were excluded. The remaining 47 articles were selected for the title and abstract screening to evaluate whether they met the PICO criteria. Among these, ten studies were excluded because they were reviews, clinical cases, book chapters etc. The remaining 37 articles were selected, and full-text screening was performed. Therefore, 34 pieces were eliminated, as 28 did not meet PICO, and six were off-topic. Therefore three articles were considered by an electronic search. In addition, a manual search of the bibliography and major sites was performed, and an additional article was included. In the end, four papers were included in the study. The included studies have been published between 1982 and 2019. The studies are longitudinal study. All these studies considered the presence of TMD before and after Herbst therapy. They all assessed symptoms with the Helkimo index and Manual Functional analysis. Helkimo was credited as being a pioneer in the development of an index to gauge the level of pain and severity in TMD sufferers. Registration of subjective symptoms applying for the Helkimo Index required a questionnaire-based survey. Questionnaire comprised two parts: Anamnestic component which includes answers to questions in “yes” or “no”.Clinical analysis evaluate the pain and the mouth opening. The two data are combined and based on responses and symptoms a grade from 0 no symptoms to 2 disabling symptoms. Three additional categories were added to Helkimo’s index: occlusal dysfunction, clinical, and anamnesis.Manual Functional Examination assessed patients’ pain and difficulty in opening by categorizing them from 0 to 2. The data extracted from each study, as reported in the paragraph “data extraction,” were written in Table 2.

The included subjects in this review were 147 (71 females and 76 males, mean age 13,3). For each patient, the presence of TMD was assessed at least immediately before treatment with Herbst and immediately after the end of therapy; in the case of studies 1 and 4, the last follow-up is up to 12 months; in the case of study 3 it is greater than or equal to 15 years. A population of 186 patients in whom the onset of symptoms was assessed either by the Helkimo score or by Manual functional analysis was taken for the meta-analysis.

Demographics and clinical data of the involved subjects are presented in Table 2.

In the study by Pancherz and Anehus-Pancherz, 20 male patients were enrolled and successfully treated for six months. The evaluation of the temporomandibular joint included a questionnaire and a clinical examination. The anamnestic part had questions regarding sounds and pain from the TMJ, pain in the masticatory muscles, and biting and chewing difficulties. The answer could be yes or no. The clinical evaluation included the range of mandibular movements (maximum opening capacity, maximum lateral movements and maximum protrusion) according to Ingervall (1970), the presence or absence of joint noises during mandibular movements and tenderness on palpation using the method of Carlsson and Helkimo (1972) of TMJ and masticatory muscle. The questionnaire results indicated that joint sounds (clicking) were noticed by two patients before treatment, but the clicking had disappeared; furthermore, no patients reported pain in the masticatory muscles or TMJ during treatment. On clinical examination, the maximum and maximum protrusive openings remained unchanged before and after treatment; only the lateral movements showed a slight decrease (1.9 mm on average) during treatment, returning to the initial values after treatment. No joint noise was detected twelve months after the end of treatment (although two patients were positive for this sign before treatment). Tenderness on palpation at the TMJ was detected in four patients before treatment, only in two twelve months after treatment. Tenderness on palpation of one or more masticatory muscles was seen in five patients before and in two twelve months after treatment (mainly the posterior belly of the digastric muscle followed by the lateral pterygoid muscle) [28].

In the study conducted by Aidar et al., patients were clinically examined to determine the Helkimo CDI at three-time intervals: immediately before Herbst treatment (T1), at the end of Herbst.

treatment (T3) and the end of phase 2 orthodontic treatment (T4). There were no changes in 26 (86.6%), 16 (76.2%) and 16 (76.2%) patients at T1-T3 (p = 1.000), T3-T4 (p = 1.000) and T1-T4 (p = 1.000), respectively. There were changes in 4 (13.3%), 5 (23.8%) and 5 (23.8%) patients at T1-T3, T3-T4 and T1-T4, respectively [29].

In Ruf and Bock’s work, 33 patients were analysed at three-time points: before treatment (T0), after treatment (T1) and at an average follow-up of 19.2 years (T2). The patients were evaluated based on the Helkimo index and with the research diagnostic criteria for temporomandibular disorders (RDC/TMD) and diagnostic criteria for temporomandibular disorders (DC/TMD). At each of these three points, most patients (82–88%) never presented anamnestic symptoms of TMD: an improvement trend between T0 and T1 and a recurrence trend between T1 and T2 was noted, however, without statistical significance. Clinical dysfunction (Helkimo index D) was never shown between 55 and 73% of the patients at the three-time points. An improvement between T0 and T1 and a recurrence between T1 and T2 can also be seen here, but without statistical significance [30].

In the study by Ruf et al., 62 patients (35 females and 27 males) were analyzed, with a mean age at the start of treatment of 14.4 years. The patients were clinically examined according to Manual Functional Analysis (MFA) at three time points: before Herbst treatment (T1), immediately after Herbst removal (T2), and one year after Herbst removal (T3). At T1, capsulitis of the low stratum was detected in 21 patients (5 symptomatic and 16 sub-symptomatic); at T3, only five patients. Clinical signs of disc displacement were detected at T1 in 9 patients (11 joints): in two joints, a partial disc displacement with reduction (PDDwR); in five joints, a total disc displacement with reduction (TDDwR); and in four joints, a full disc displacement without reduction (TDDnoR). At T2, only one joint showed clinical signs of disc displacement, and at T3, two affected joints (1 patient) with a TDDwR were identified. Crepitus was detected in two patients (number 2 and 39) at both joints at T1, in one joint (patient number 2) at T2 and in three joints (patients number 3 and 39) at T3 [31].

The overall effect, reported in the forest plot (Fig. 2), showed that there was no difference in TMD prevalence between pre-Herbst and post-Herbst therapy (OR 0.74; 95% CI: 0.33–1.68), suggesting that Herbst is not a factor that can cause or at least exacerbate the symptomatology of TMD.

The risk of bias in the included studies was reported in Fig. 3. Regarding the randomization process and allocation concealment, two studies ensured a high risk of bias. Only one study excluded a performance; all the studies confirmed an increased risk of detection bias (self-reported outcome), and 2 of the included studies present low detection bias (objective measures) (Fig. 3). Two studies ensure a low risk regarding attrition and reporting bias.

The meta-analysis has several limitations and strengths. The study analyzes the effects of Herbst therapy on the onset or exacerbation of TMD symptoms. The powers are undoubtedly the homogeneity of the population considering a population with an average age of 12 years. Critical issues of the studies considered, however, are that they do not think other cofactors in the development of TMD and, therefore, bias may be present in the studies. Therefore, more studies will be needed, perhaps with the presence of multiple linear regressions that can evaluate possible confounding factors in the ‘incidence of TMD in patients treated with Herbst.