Characteristics and impact of pain from root-filled teeth. A practice-based cross-sectional study comparing painful teeth with and without signs of inflammatory dental disease

2.1 Study design and participants

This is a cross-sectional study using data collected during the year 2015 in Örebro County, Sweden. All adult patients with at least one root-filled tooth who were scheduled for routine examination in April 2015 were eligible and invited to participate. The recruitment and examinations were performed within the 23 clinics of the Public Dental Service in Örebro County. The patients were thus examined by their regular dentist, who then provided the data to the study coordinator (JJS). For further details on design, see Jonsson Sjögren et al. [3] (2019).

2.2 Enrolment and data collection

If a patient agreed to participate, all their root-filled teeth were included in the data collection in 2015. The forms of informed consent are stored in a locked room of the Public Dental Service in Örebro County. Participants received no monetary compensation for study participation.

2.3 Inclusion in the present study

The individuals with one or more painful root-filled tooth identified in the data collection in 2015 were included in the analyses of this study. The definition of “painful” was the subjective experience reported by the patients.

The data collection included the following:

Clinical assessment of all included teeth to identify clear signs of infection/inflammation: presence of swelling, presence of sinus tract and greatest probing depth at the tooth.

The quality of the coronal restoration was rated as either good or poor (secondary caries, defective temporary restoration, permanent restoration with insufficient marginal integrity or no restoration). Tenderness to percussion and to apical palpation indicating hyperalgesia or allodynia were also noted.

• Radiographic examination of the root-filled tooth by intraoral periapical images with all apices clearly visible. It sufficed with one image if the first radiograph, exposed with an orthoradial projection, showed a clearly identifiable radiolucency around any apex. If the first image however appeared to show a normal periodontal ligament, i.e., healthy conditions, a second periapical radiograph was exposed, with a 10–15 degrees overaxial or a mesial or distal eccentric projection to improve detection of periapical radiolucency [24].

Clinical and radiographic examination findings were combined to form two groups based on whether the examination revealed signs of dental disease or not, see “Definition of groups” below.

• An interviewer-assisted questionnaire comprising questions regarding general health (diabetes, cardiovascular disease, gastrointestinal disorder, rheumatic disease, neurological disease, psychiatric disorder). Individuals were categorized as either “healthy” or “with one or more general health concerns”.

• Screening questions for TMD (3Q/TMD) [25]. The two questions specifically related to pain were used; the third question concerns locking of the jaw and was not considered relevant in this context.

Q1: Do you have pain in your temple, face, jaw or jaw joint once a week or more? (YES/NO)

Q2: Do you have pain once a week or more when you open your mouth or chew? (YES/NO)

• Occurrence of other chronic pain conditions, lasting at least 3 months [26] and persistent in character: “During the last three months, have you experienced pain in more than one place in the body four days per week or more often? (YES/NO)” [27].

The patients of this study all answered YES to the question if they had experienced any pain or discomfort (such as tenderness or swelling) related to the root-filled tooth during the last 3 months. They then completed a second questionnaire, one for each painful root-filled tooth, comprising questions regarding pain characteristics and consequences of pain:

(i) Pain intensity: current pain, average pain during the last month, and worst pain during the last month (on an 11-point numeric rating scale (NRS) ranging from 0 = no pain to 10 = pain as bad as could be).

(ii) Pain duration since onset: in months since the pain was first experienced.

(iii) Pain frequency: if the pain is continuous, recurrent or occasional.

(iv) Provoking factors: if the pain occurs spontaneously or when using, irritating or provoking the tooth.

(v) Pain persistence: if the pain from the tooth has lasted at least 8 hours a day and at least 15 days per month the last 3 months [28].

(vi) Pain quality: rated according to the short-form McGill Pain Questionnaire (SF-MPQ), in which 15 descriptive words of the pain are rated on a 4-point scale (0 = none, 1 = mild, 2 = moderate and 3 = severe). Eleven words describe the sensory part of the pain experience and four words describe the affective part [29, 30].

(vii) Pain intensity and pain-related disability: assessed using the compound measure Graded Chronic Pain Severity Scale (GCPS) [31, 32, 33]. The scale was modified to measure pain in a 1-month perspective [34]. The different measures of the GCPS were calculated; Characteristic Pain Intensity (CPI), disability points for number of days with interference, interference score and disability points for the interference score. These, in turn were used to calculate the compound measure Chronic Pain Grade, (0 = none, I = low intensity pain, with none-to-low pain-related disability, IIa = High intensity pain, without pain-related disability, IIb = High intensity pain, with low pain-related disability, III = Moderately limiting and IV = Severely limiting).

Characteristic pain intensity: On an 11-point Numeric Rating Scale (0–10 NRS) where zero represents “no pain” and 10 “pain as bad as could be”, participants rated the pain at the time of the examination, the worst pain experienced the past month, and the average pain experienced the past month. The mean of the three NRS scores was multiplied by ten to form the CPI, modified from Ohrbach (2010) [32].

Pain-related disability: The participants also stated the number of days the last month the pain was experienced, the number of days the pain had caused them to abstain from normal activities, and the degree of limitation that the pain caused for each of the three domains daily activities, recreational activities, and ability to work. Each was rated on the 11-point numeric rating scale (0–10 NRS), with 0 representing no limitation at all and 10 representing “unable to perform any activities”. The mean of the three NRS scores was multiplied by ten to form the disability score.

Disability points: The sum of the points of disability days and the disability score added together.

(viii) Present or recent intake of, e.g., analgesics, antibiotics, herbal remedies for the pain.

For further details on examination and data collection, see Jonsson Sjögren et al. [3] (2019).

2.4 Definitions of groups

Clinical and radiographic findings were combined when determining if a tooth was considered affected with dental inflammatory disease or not. Regarding the clinical findings, presence of swelling, a sinus tract and deep periodontal pocket/-s (≥6 mm) were considered as certain signs of an inflammatory condition affecting the surrounding tissues of the tooth. In contrast, we did not regard tenderness to percussion and to apical palpation as definite signs of inflammatory dental disease, since hyperalgesia and allodynia (gain in sensory function) often occur also due to pain of other origin [35, 36, 37].

Radiographic findings included presence/absence of a periapical radiolucency. Three specialists in Oral Radiology reviewed the radiographs independently of each other under optimal conditions without access to information about symptoms and clinical status. The observers noted if the apical condition was normal; if there was a widened periodontal ligament space ≤1 mm or if there was an apical radiolucency >1 mm wide. Given that generally, interobserver agreement on assessment of apical status is low for intraoral radiography [38], a majority principle in rating the apical status was followed, i.e., if two observers agreed on apical radiolucency the tooth was noted as having apical radiolucency. To obtain a strict definition of apical radiolucency and thereby avoid overestimating the presence of periapical disease, the ratings “normal apical condition” and “widened periodontal ligament space” were merged into one group, representing absence of radiolucency.

The two study groups were defined as follows:

Dental disease positive (DD+), a painful root-filled tooth with at least one positive finding of swelling, sinus tract, pocket depth ≥6 mm, or apical radiolucency.

Dental disease negative (DD−), a painful root-filled tooth without any of the findings swelling, sinus tract, pocket depth ≥6 mm, or apical radiolucency.

2.5 Statistical methods

A power analysis was performed. Data from 48 patients was available, 36 who had signs of dental disease and 12 without such signs. The basis for the analysis was pain intensity, and a difference of 3 units on the 0–10 NRS was assumed as clinically relevant [39]. A standard deviation (SD) of 1.8 was assumed from a study of similar patients [2]. The effect size was 1.67. With a 5% risk of type I error and a power of 80% at least six cases per pain group were needed to identify a statistically significant difference.

Data were entered in Excel® sheets (2401 Build 16.0.17231.20170, Microsoft Corporation, Redmond, WA, USA) and transferred to IBM SPSS version 27 (IBM, Armonk, NY, USA) for statistical analysis. Descriptive statistics was used for patient characteristics as well as pain characteristics. For comparisons between groups, Chi-square test and Fisher’s exact test was used for categorical variables and Mann-Whitney U test for continuous variables (skewed data, two groups). To make sure the results were robust to a potential clustering effect caused by some individuals contributing more than one tooth, Generalized Estimating Equations (GEEs) with the individual as the clustering effect were used. When comparing the two groups, the independent variables were tenderness to percussion, tenderness to apical palpation, positive answers(s) to Q1 or Q2, self-reported chronic pain and sex. For interobserver agreement, kappa statistics were used. κ ≤ 20 is considered to equal slight agreement, 0.21–0.40 fair, 0.41–0.60 moderate, 0.61–0.80 substantial and 0.81–1.0 almost perfect agreement [40]. Since missing data was excluded for individual variables, the valid number of included answers varied among variables. All significance tests were two-sided and conducted at the 0.05 level of significance.

3.1 Pain characteristics

Regarding frequency, intensity, duration, quality and provoking factors of pain or discomfort, the two groups were very similar. The findings are summarized below and details are found in Table 1. Pain intensity was significantly higher in DD− group; mean value 3.1 (range 0–7) compared to DD+ group with mean value 1.6 (range 0–6) (p = 0.030). The distribution of pain intensity ratings and ranges is found in Supplementary Figs. 1,2. For both groups, the most commonly reported pain or discomfort frequency was recurrent pain, selected by 48% of the DD+ group and 43.5% of the DD− group (p = 0.687). The pain or discomfort had lasted on average for 24.7 months for the DD+ group and 38.2 months for the DD− group (p = 0.476). The pain duration since onset can be expressed differently; 40.0% (n = 12) of the DD+ group and 52.2% (n = 12) of the DD− group reported a duration of >2 years (p = 0.38). In the DD+ group, the pain or discomfort was reported to start only on provocation by 62.5% of the patients (n = 15) and spontaneously by 37.5% (n = 9) while in the DD− group it started spontaneously in 56.5% of the patients (n = 13) and only on provocation in 43.5% (n = 10). However, no significant differences in distribution were found between the two groups, p = 0.191.

When the participants described the quality of their pain, the most common descriptors for the DD+ group was tender (n = 13, 48.1%), aching (n = 6, 21.4%), splitting (n = 5, 18.5%) and throbbing pain (n = 4, 14.8%), respectively. The most common descriptors for the DD− group was tender (n = 14, 60.9%), aching (n = 43.5%), heavy (n = 4, 17.4%) and throbbing pain (n = 4, 17.4%). No significant between-group differences were found in the choice of descriptors. Supplementary Table 1 shows the distribution of all verbal pain quality descriptors.

Some teeth in the DD+ group had multiple clinical and radiographic signs. The distribution is found in Table 2.

3.2 Clinical pain provocation tests

Table 1 shows clinical provocation test data. Tenderness to apical palpation was not reported by anyone in the DD− group, and was thus more common in the DD+ group (p = 0.007), while tenderness to percussion did not differ between the groups (p = 0.337).

3.3 Persistent tooth pain

Persistent tooth pain, defined as pain for at least 8 hours per day and at least 15 days per month the past 3 months was reported by seven participants with seven painful root-filled teeth in the DD+ group and by five participants with five painful root-filled teeth in the DD− group. The details are found in Table 3.

3.4 The impact of pain

The impact of pain on daily activities, expressed as Chronic Pain Grade and as intake of medications, showed that on average, the impact of pain was low with no differences between groups (Table 1).

3.5 Patient and clinical characteristics

Table 4 presents the demographics and patient-related and tooth-related characteristics. No significant differences were found between the two groups regarding sex (p = 0.961), age (p = 0.540) or general health, (p = 0.740). Furthermore, no differences could be seen between number of included teeth per participant (p = 0.617), jaw position (p = 0.431) or tooth type (p = 0.752).

3.6 General health and chronic bodily or orofacial pain

There were no between-group differences for any of the variables. In the DD+ group 40.7% (n = 11) and in the DD− group 45.5% (n = 10) reported at least one general health concern (p = 0.740).

In the DD+ group 26.9% (n = 7) and 27.3% (n = 6) in the DD− group gave a positive answer to at least one of the two pain-related questions of the 3Q/TMD; Q1 or Q2 (p = 0.978).

Eleven of the participants (40.7%) in the DD+ group and nine of the participants (40.9%) in the DD− group experienced pain from ≥1 body site (p = 0.990).

3.7 Comparison between the groups—multivariate analyses

Univariate analyses do not take into consideration a possible clustering effect due to some participants contributing with more than one tooth. A multivariate analysis with GEE was therefore performed to confirm the robustness of the findings. This analysis found no statistically significant effects (ORs) (Supplementary Table 2, p = 0.33–0.86), indicating absence of a clustering effect. The variable “tenderness to apical palpation” was excluded from the analysis since no participant in DD− reported this.

3.8 Prevalence of odontogenic and non-odontogenic pain or discomfort associated with root-filled teeth

The original cross-sectional study involved 550 patients with 1256 root-filled teeth. The number of patients who reported a painful tooth was 53 (intention to include) rendering a pain prevalence of 9.6%. However, data was missing for four patients, thus 49 patients with 53 teeth were included in the analyses (Fig. 1); giving a prevalence of 8.9%. Twenty-two of the 550 originally included participants were assigned to the DD− group, which gives a 4.0% prevalence of pain or discomfort from root-filled teeth without signs of dental disease on individual level, and on tooth level a prevalence of 1.8% (23/1256 root-filled teeth). Twenty-seven of the participants were assigned to the DD+ group, which gives a 4.9% prevalence of pain or discomfort from teeth with signs of dental disease on individual level, and a prevalence of 2.4% on tooth level (30/1256 root-filled teeth).

3.9 Observer agreement of radiographic assessment

In the radiographic assessment all observers agreed on presence of apical radiolucency in 13 teeth (24.5%), two observers agreed on apical radiolucency in 8 teeth (15.1%) and only one observer noted apical radiolucency in 6 additional teeth (11.3%). All three observers agreed on absence of apical radiolucency in 26 teeth (49.1%), giving a total 3-observer agreement rate of 73.6%, and ≥2-observer agreement rate of 88.7%. κ values for the interobserver agreement ranged 0.525–0.755, which equals moderate to substantial agreement [40]. The details can be found in Supplementary Table 3.

4.1 Participants, group assignments and quality of data

The patients in this study were adult individuals who were scheduled for a regular check-up and did not seek care due to a problem, e.g., tooth pain, which would suggest that they may be of pain or discomfort from a root-filled tooth. In contrast, the majority of patients referred to a specialist clinic for a root-filled tooth had more pronounced symptoms such as pain and swelling [43].

Systematic misclassification of cases could be one explanation of the lack of identified differences. The DD+ and DD− groups were defined using a complex standard including both clinical and radiographic parameters, a strategy often reported to increase the diagnostic certainty, e.g., for TMD diagnoses [9, 44]. The criteria categorized the tooth as DD+ if any sign of dental disease (clinical and/or radiographic) was present. The sensitivity of the radiographic examination was increased by using more than one periapical intraoral radiograph per tooth [45]. Compared to periapical and panoramic images, Cone Beam Computed Tomography (CBCT) was reported to have higher sensitivity to identify apical bone destruction [46, 47], and yield a larger number of apical radiolucencies and other important features such as missed, c-shaped, or over-/underfilled root canals and vertical bone defects [10, 48, 49, 50]. It is possible that CBCT would have identified a larger proportion of the cases as DD+. However, since data were collected in general practice under standard routine check-up circumstances, this was not feasible. In addition, when CBCT imaging was compared to histopathologic examination, the positive predictive value of CBCT to identify AP was considerably lower for root-filled compared to non-root-filled teeth, in the range 0.48–0.64 [47]. Thus, it is conceivable that using CBCT would have overestimated the prevalence of AP; i.e., identified teeth without periapical disease as diseased.

In addition, since the time of RCT was not known in this study, it is unknown if a radiolucency observed represented an inflammatory state (AP) or healing in progress. If apical radiolucency is noted in combination with symptoms or abnormal clinical findings, it is more likely to represent active disease, such as apical periodontitis, marginal periodontitis [7], or vertical root fracture [51], especially in root-filled teeth which lack a physiological barrier against infection of the pulp space, generally have less remaining tooth structure and/or are heavily restored. In summary, our approach to group assignment was inclusive for DD+, increasing the chance of “true positives” but at the same time the risk of “false positives”. A more restrictive approach to inclusion in DD− group required absence of all clinical and/or radiographic signs of disease. Therefore, our approach is more likely to have misclassified cases as DD+ than the opposite.

Since the inclusion criteria was “pain during the last 3 months”, there is a theoretical risk of having included participants who had a painful tooth during the past three months but received RCT for it, after which the pain subsided. The risk of this is very low since the screening process took place some months before the data collection and only patients who already had root-filled teeth at the time of the pre-screening of the records were asked for participation.

The quality of clinical data may also affect the group assignment. Since this was a practice-based study, a possible limitation is that the dentists providing data were not calibrated in the clinical examination technique. To limit bias due to differences in procedures, the study coordinator (JJS) visited all clinics prior to the data collection to clarify how the clinical protocol should be interpreted.

To identify group differences, the sample size needs to be adequate. This study included 49 participants with 53 painful root-filled teeth, which surpassed the estimated sample size needed for the main outcome measure pain intensity. Several other studies have compared the characteristics of odontogenic and/or non-odontogenic pain (Atypical Odontalgia (AO), TMD) with similar sample sizes or smaller [52, 53, 54] whereas only few have reported on larger samples [55].

Data was missing for some teeth, n = 3–10; 5.7–18.9%. The variables with missing values were quality of the pain n = 3, present pain intensity n = 3, frequency n = 5, duration since onset n = 6, provoking factors n = 6, worst pain intensity n = 10, average pain intensity n = 10, days with pain past month n = 10, pain-related limitation n = 10, intake of analgesics, antibiotics and herbal remedies n = 10. No imputation was made for missing data in the statistical analyses.

4.2 Pain characteristics

On average, the pain characteristics were quite similar between the groups. The majority had mild pain, but the average intensity was higher in the DD− group. Similar relationships were reported in one study comparing odontogenic and non-odontogenic pain 6 months after RCT [10], and in another study comparing root-filled teeth with complete vs. incomplete healing 5–14 months post-RCT [5]. In all three studies, patients without clear signs of dental disease reported higher pain intensity than patients with dental inflammatory disease likely to be their pain cause. The potential implication is that with a higher pain intensity the patients are more prone to seek treatment [56] and the dentists could experience a pressure to initiate treatment [57], suggesting a risk of unnecessary endodontic retreatment being performed in this group.

Both groups had similar distribution of pain frequency; in both groups “recurrent pain” was reported by almost half of the participants. The duration since its onset was long for both groups, DD+ >2 years vs. DD− >3 years. Similarly, Daline and co-workers reported that in about one fifth of their patients, the pain continued and was still present 3.4 years post-RCT [58]. Recall bias is always a possible risk when working with data remembered in retrospect [59]. This means that, e.g., the reported duration of pain since the onset might not be exact. But overall, we estimate that the pain duration was likely to be long in both groups; roughly 4/10 in the DD+ group and half of the DD− group reported a pain duration since onset of >2 years.

There were also no differences between groups regarding what triggered the isolated pain episodes, although the patients in the DD+ group tended to report that it started more often on provocation rather than appearing spontaneously while the DD− group reported the opposite. In the clinical assessment, pain on palpation of the apical area was reported by about a quarter of the individuals in the DD+ group, and not by any participant in the DD− group. The latter conflicts with another study in which 38% of patients with non-odontogenic persistent pain after RCT reported pain on apical palpation [10].

Tenderness to percussion was equally frequent in both groups. Both parameters represent a state of allodynia, usually attributed to central sensitization, either induced by locally released inflammatory mediators sensitizing peripheral nociceptors [60] as in AP [61], or the sensitization of second-order neurons by ongoing and intense afferent input leading to increased processing of the signals in the higher brain centers [62, 63]. Patients with persistent tooth pain are known to exhibit tenderness to percussion [64], experienced by a majority of patients with tooth pain of odontogenic as well as non-odontogenic origin [10]. It has previously been suggested that spontaneous pain is a clinical feature of neuropathic pain [65] while pain on provocation would indicate inflammatory tooth pain associated with a specific tooth [62].

4.3 Prevalence of painful root-filled teeth

The participants in this study were drawn from a larger sample of patients with root-filled teeth presenting for their annual check-up. In the larger sample of 550 patients with 1256 root-filled teeth, the prevalence of odontogenic tooth pain or discomfort (represented by the DD+ group) was 4.9% on an individual level and 2.5% on tooth level. Our post-analysis of published data from 10 retrospective studies and one prospective study [4, 8] estimated the prevalence of odontogenic pain ≥6 months after root canal treatment roughly to 2–3% (on tooth level), which is consistent with our findings.

Correspondingly, the prevalence of pain or discomfort from teeth without signs of inflammatory dental disease was found to be 4.0% on the individual level, in agreement with a meta-analysis of 10 studies reporting a pooled prevalence of non-odontogenic tooth pain 1–6 years after RCT to be 3.4% [8], and a more recent retrospective practice-based study reporting 3.1% prevalence of non-odontogenic tooth pain 3–5 years post-RCT [4].

4.4 Persistent tooth pain

Persistent pain has been reported to affect root-filled teeth, and as mentioned above, several pain origins are possible. In the current study, the specific pain origin for each case in the DD− group was not possible to establish. Nevertheless, it was of interest to examine the prevalence of persistent pain on the population level. Five participants with five root-filled teeth (a prevalence of 0.9% on the individual level and 0.4% on tooth level) fulfilled the temporal criteria for persistent dento-alveolar pain disorder (PDAP), “pain more than 8 hours a day for at least 15 days per month the last three months” [28]. In contrast, another study establishing pain diagnoses >6 months after RCT reported that 2/19 participants (10.5%) of participants with persistent pain were diagnosed with PDAP [10]. The large difference is likely a reflection of participant selection and study design; the present cross-sectional study examined an adult population visiting their dentist for a check-up, while the Nixdorf et al. [10] (2015) study was a prospective study of RCT. It must be emphasized that PDAP is a diagnosis of exclusion, appropriate only when all other possible pain origins (e.g., TMD, sinusitis, neuropathic pain, headache) have been excluded with reasonable certainty, which was not done in this study.

Our findings underscore that persistent pain is not always a reason to seek care. The intensity may be mild and might therefore be neglected, ignored or missed by patients and dentists, possibly due to a lack of understanding of its cause. Future research could focus on identifying reasons for not addressing pain in a root-filled tooth. Discomfort from root-filled teeth may not always be identified as pain. Eleven of the participants answered YES to the question “Have you experienced pain or discomfort (swelling, tenderness on chewing or any other problem) from any of your root-filled teeth at any time during the last 3 months?” but still rated their average pain intensity as zero. Similarly, 29 out of 264 participants (11%) in a prospective study rated their pain intensity as zero but also described the quality of their symptoms according to the McGill Pain questionnaire [5]. One possible interpretation of this is that they experience dysesthesia rather than pain. The definition of dysesthesia according to IASP is “an unpleasant abnormal sensation, whether spontaneous or evoked” [66]. It is often associated with damage to the peripheral sensory nerves, including the trigeminal nerve [21], and reported iatrogenic reasons include endodontic treatment and local anesthesia [37]. In the present study no differences were seen between the two groups concerning the SF-MPQ, which concurs with other studies exploring the diagnostic value of pain quality to distinguish between pain conditions [52, 67]. Based on descriptive words from patients, a fourteen-item screener was developed with the purpose to differentiate AO and PDAP patients from other oral pain conditions. The reported sensitivity was 0.77 and specificity was 0.69, suggesting that the screener could be used with caution together with standard clinical examination [68].

4.5 The impact of pain

On average, the impact on life was very low. Despite long pain duration since onset in both groups, the participants reported little effect on daily activities (GCPS grades I–II with either low or high pain intensity) and very few took analgesics for the pain. A study evaluating pain from root-filled teeth 6 months after RCT reported similar impact but a larger proportion (42.6%) of participants took pain medications [2] which may reflect a shorter pain duration since onset and thus possibly still acute pain after RCT. A follow-up study of patients with AO reported more severe impact on life from persistent pain; about 1 in 10 patients originally recruited from orofacial pain clinics were in grade III or IV, and 3/4 in grades I–II after seven years [54]. Again, the differences are likely explained by differences in participant selection.

4.6 Patient characteristics: General health, TMD and other chronic pain

In patients with chronic pain compared to patients without chronic pain, increased odds ratios have been reported for general health issues such as diabetes, hypertension, ischemic heart disease, heart failure, acute myocardial infarction, parkinsonism and mood/anxiety disorders in patients [69]. This may conceivably be associated with systemic factors, but the direction of a possible causal relationship remains unclear. This study found no difference between the two pain groups regarding presence of general health problems. A retrospective cross-sectional study from specialist care similarly reported no group differences regarding presence of “other health comorbidities” between five different orofacial pain diagnoses [70], suggesting that general health concerns are similar between pain conditions.

TMD pain is a frequent finding in the adult population in epidemiological studies, (5–12%) [71]. In the current practice-based study, a complete examination of the masticatory system such as the DC/TMD [9] was not performed, so no TMD pain diagnoses, e.g., myofascial pain or arthralgia were made. However, the 3Q/TMD screening instrument was validated in its entirety and has shown good validity with sensitivity and specificity >0.8 [25]. We did not analyze the third question unrelated to pain, and therefore identification of all TMDs is less certain, but positive responses to the pain-related questions 1 and 2 are still likely to reflect the occurence of painful TMD specifically. The estimated prevalence was about one in four participants, with no difference between the two pain groups. Other studies have reported an association between pain from root-filled teeth and TMD [1, 5, 10]. In a study of chronic tooth pain and TMD, half of the patients with painful teeth also had myofascial TMD [52]. A study on presence of TMD among endodontic patients reported a 54% prevalence of painful TMDs [41]. The specific pathophysiologic mechanisms of myofascial pain are unknown, but a theory is that sustained contraction of the muscles may be a factor, and that persistent and annoying pain e.g., toothache can produce a local muscle response, triggering muscle pain [72, 73].

Chronic Overlapping Pain Conditions (COPCs) are pain conditions that frequently occur together; some examples are fibromyalgia, irritable bowel syndrome, TMD, lower back pain and migraine [74] and possibly also PDAP [75]. Among the COPCs, pain amplification and an abnormal sensitivity to pain is a common feature [74]. This begs the question: do patients with pain from their root-filled tooth also have other bodily pains, and does it differ between inflammatory dental pain origin and non-odontogenic pain origin? In this study, no association could be seen between long-lasting pains in the rest of the body and tooth pain origin; the prevalence was similar in both groups. Similarly, a study comparing patient questionnaire data 3 years post-RCT performed by mainly general dentists in a practice-based study found no differences [58].