Study design and ethical aspectsThis is an observational cross-sectional study. The research was carried out at the Universidade CEUMA and the project was previously approved by the institution’s Research Ethics Committee (protocol number 5.674.373). The study followed the recommendations of the Declaration of Helsinki. Recruit your volunteers through social media and word-of-mouth. They underwent evaluation with the dentist.Previously, all volunteers with the potential to be included in the study were informed about the research objectives and the methods to be used. Once the volunteer agreed to participate in the investigation, the free and signed consent form was collected.ParticipantsAdults aged 18 to 55 with a confirmed diagnosis of myogenic TMD, indicated by a score of 50 or more on the Fonseca Anamnestic Index (FAI)16 and a pain level of 3 or above on the Numerical Pain Scale (NPS)17, were included. Individuals were excluded if they had clinical diagnoses of rheumatic, cardiovascular, metabolic, or respiratory conditions; wore full or partial dental prostheses; had systemic neuromuscular disorders; had a history of facial and/or temporomandibular joint trauma; experienced joint dislocation; had cancer; suffered from active inflammatory or infectious diseases; were diagnosed with fibromyalgia; were smokers; or had any other condition that would prevent them from undergoing the proposed evaluations.Numerical Pain Scale (NPS)The NPS17 is a simple and easy-to-use measure consisting of a sequence of numbers from 0 to 10. In this scale, the value 0 indicates “no pain” and the value 10 represents “the worst pain imaginable.” Thus, the volunteers rated their pain based on these parameters. Pain intensity was assessed with the individual at rest.Fonseca Anamnestic Index (FAI)The FAI was used to assess the presence and severity of TMD symptoms. This is a simple and easy-to-understand instrument, allowing patients to fill it out themselves, if necessary, without compromising the quality of the assessment. However, in this investigation, the FAI was carried out through an interview by a single previously trained examiner. Participants answered a total of 10 questions, with three response options: “yes” (score: 10), “no” (score: 0) and “sometimes” (score: 5). The final score of the instrument is determined by the sum of the scores of all items, allowing the following classifications: absence of signs and symptoms of TMD (0–15 points), mild TMD (20–45 points), moderate TMD (50–65 points) and severe TMD (70–100 points)18,19,20,21,22.Central Sensitization Inventory (CSI)The CSI which has been validated for the Brazilian population23, was utilized to evaluate central sensitization symptoms in participants with chronic pain. A single examiner conducted the questionnaire through an interview. It consisted of 25 questions about symptoms or health situations in the participant’s daily life, with five possible responses for each question: “never” (score: 0), “rarely” (score: 1), “sometimes” (score: 2), “often” (score: 3), and “always” (score: 4). The total score ranges from 0 to 100, with a previously established cutoff value of 30 points or more to address the possible presence of central sensitization24.R-R intervals recordings – heart rate variabilityStandardized procedures for collecting biological signals for subsequent analysis of heart rate variability have been previously published1.Data processingWe used the Kubios HRV standard analysis software (MATLAB, version 3.5, Kuopio, Finland) to process the HRV data11. To this end, the stretch with the greatest stationarity lasting 5 min, that is, short-term analysis, was selected for analysis. The choice of this section followed some previously established criteria: (1) absence of significant R–R interval outliers (i.e., R–R intervals markedly higher or lower than the overall R–R signal, as determined through visual inspection by the researcher); (2) equidistance of R–R intervals; and (3) Gaussian distribution of R–R intervals and heart rate distribution graphs12.All corrective filters from the Kubios software were used to evaluate the results11. All corrective filters available in the Kubios software were used to evaluate the results. Basically, the correction algorithm makes an RR interval value comparison with a local average interval. The local mean is obtained by filtering the median of the RR interval time series, and therefore, single outliers in the RR interval time series do not affect the local mean. If an RR interval differed from the local mean by more than a specified threshold value, the interval was identified as an artifact and marked for correction. The threshold value can be selected between: (1) None (no correction is performed); (2) Very low: 0.45 s (threshold in seconds); (3) Low: 0.35 s; (4) Medium: 0.25 s; (5) Strong: 0.15 s; and (6) Very strong: 0.05 s. To give an example, the “average” correction level will identify all RR intervals greater or less than 0.25 s compared to the local average. The correction is made by replacing the identified artifacts with interpolated values using a cubic spline interpolation.HRV was analyzed using overview parameters, linear statistical measures (time and frequency domain) and through non-linear statistical measures. In the overview, the following variables were used: parasympathetic nervous system (PNS) index, sympathetic nervous system (SNS) index and stress index. In the time domain, the following variables were analyzed: the average interval between R-waves (Mean RR) in milliseconds (ms); ms: milliseconds; standard deviation of all N-N normal intervals (SDNN) in ms; the average of heart rate (Mean HR) in beats per minute (bpm); square root of the mean squared differences of successive RR (RMSSD) in ms; number of interval differences of successive NN intervals greater than 50 ms divided by the total number of NN intervals (pNN50) in percentage (%); integral of the density of the RR interval histogram divided by its height (RR Tri); baseline width of the RR interval histogram (TINN) in ms. For frequency domain analysis, the following variables were evaluated: very-low-frequency band (0.0033–0.04 Hz) in milliseconds squared (ms2); LF: low frequency band (0.04–0.15) in ms2 and n.u.; HF: high frequency band (0.15–0.40) in ms2 and n.u.; total power in ms2 and LF/HF ratio8,25.For non-linear analysis, the following variables were considered: standard deviation perpendicular to the line of identity (SD1) in ms; standard deviation along the line of identity (SD2) in ms; ratio of SD2-to-SD1 (SD2/SD1); approximate entropy (ApEn); sample entropy (SampEn); detrended fluctuation analysis which describes short-term fluctuations (DFA α1); detrended fluctuation analysis, which describes long-term fluctuations (DFA α2)8,25.Statistical analysisData are presented as mean and standard deviation or absolute values and percentages of occurrence when appropriate. The Shapiro-Wilk test was used to verify the normality of the data. When the data met parametric assumptions, One-way ANOVA with Tukey-Kramer post-hoc was used to test the differences in HRV using the different Kubios Software artifact correction filters (none, very low, low, medium, strong and very strong).For data with non-parametric distribution, the Friedman test with Dunn’s post hoc was used. χ2 test was used to compare categorical variables. The effect size was calculated based on the Cohen d, according to the website: <https://www.psychometrica.de/effect_size.html>. It was considered the following interpretation of the d value: 0.2 (small), 0.5 (moderate) and > 0.8 (large effect size)26.All analyzes were performed using GraphPad Prism software (version 8.0.1 for Windows, GraphPad Software, San Diego, California USA). The probability of type 1 error occurrence was established at 5% for all tests (p < 0.05).
