DESIGN OF THE STUDY The study sample consisted originally of 80 patients (18 men, 62 women) who were referred to the Oral and Maxillofacial Department in Oulu University Hospital, Oulu, Finland, for treatment of TMD-related facial pain. The patients were examined between March 2008 and August 2010. The inclusion criteria were the following: 1. clinically diagnosed TMD as defined by the Research Diagnostic Criteria for Temporomandibular Disorders (RDC/TMD) (1), 2. minimum of 20 years of age and 3. lack of long-term illnesses (i.e. rheumatoid arthritis) that may affect TMJs or the masticatory muscles.
Using computer generated random numbers, the patients were randomly assigned to two groups: splint group and control group. Patients in the control group (n=41) were given counseling and guidance for masticatory muscle exercises. In addition to these, the patients in the splint group (n=39) were also treated with a stabilization splint.
DATA COLLECTION The patients had four follow-ups: at about 1 month, 3 months, 6 months and one year after the baseline evaluation. The same dentist specialized in stomatognathic physiology (KS) conducted all of the follow-up examinations, being unaware of the group status of the patients. At baseline the psychosocial screening (depression symptoms, non-specific physical symptoms, pain-related disability) was performed using RDC/TMD Axis II instruments. At baseline and at all the follow-ups the intensity of the facial pain was assessed using Visual Analogue Scale (VAS). The patients' subjective estimates of the effects of the treatment were evaluated with a questionnaire, using a scale from 1 to 4 (1= Very good effect, 2= Treatment has helped to some extent, 3= No difference / Cannot tell and 4= Symptoms worsened). The patients' general health status was also inquired by a questionnaire (on a scale of 1-5, 1=excellent, 2=very good, 3=good, 4=moderate, 5=poor). The patients filled in the Oral Health Impact Profile-14 (OHIP-14) questionnaire before treatment and at 3 months, 6 months and 1 year.
TREATMENT PROCEDURES The stabilization splints were made of heat-cured acrylic by the same dental technician. The occlusion of the splint was defined in the centric relation occlusion using wax (Astynax, Associated Dental Products Ltd, UK). The patients were instructed to use the splint every night during the course of the study.
All the patients in both groups were instructed to perform a standardized program for masticatory muscle exercises as described by Carlsson and Magnusson (1992). At the beginning of the training program, active mouth openings, laterotrusion and protrusion movements were performed. The mandible was held in the maximal positions for a few seconds on each movement. Thereafter, these movements were made towards resistance (using the patient's own fingers). After jaw exercises, the patients were suggested to open the jaw wide, stretching it with fingers a few times for 10 to 20 seconds. These movements were repeated 7 to 10 times per training session, and the sessions were performed 2-3 times per day. The patients received written instructions, and the movements were also demonstrated by the dentist before the treatment, and reprised if necessary. The instructions for masticatory muscle exercises were given by the same dentist (KS) at the first visit. At every examination the patients were reminded to use the splint and/or to perform the exercises at a regular basis.
The stabilization splint treatments were performed by two other dentists who were instructed in the treatment method.
ATTRITION All of the patients in the total sample were defined as belonging to the "intention-to-treat" (ITT) population. Thus, the ITT also included those who switched groups or those who in whichever group received other treatment than initially planned based on the group criteria. In the analysis, the number of days each patient had stayed in the assigned groups was taken into account. The patients who had stayed in their assigned groups for the whole follow-up period were defined as forming the "per-protocol" (PP) population.
STATISTICAL ANALYSIS Statistical significances of the differences in patients' estimate of the effects of the treatment between the groups were analyzed using chi-square test. Differences in means of the change in VAS between the groups were tested using ANOVA. VAS change was calculated by subtracting VAS at one-year follow-up from baseline VAS. The fluctuation of facial pain during the study (VAS) was demonstrated with a box-plot diagram. At first, linear regression analysis was used to analyze factors associated with change in VAS after one-year follow-up in both the ITT and PP populations. Group status was included as the explanatory factor and baseline VAS, gender, age, length of treatment and general health status were included as confounders. Linear mixed-effect regression model was used to analyse factors associated with change in OHIP severity during the 1-year follow-up, taking into account treatment time, age, gender and group status. The effect of depression and non-specific physical symptoms on the intensity of facial pain, as measured with visual analogue scale (VAS) was estimated with linear mixed models.
