Assessment of Pain (Distribution and Onset),Symptoms, SCL-90-R Inventory Responses, andthe Association With Infectious Events in PatientsWith Chronic Orofacial Pain

Neil R, McGfegor, MDScPenodontistCoilaborative Pain Research UnitFaculty of DentistryUniversity of SydneyWestmead HospitalWestmesd, New South WaiesAustralia

Henry L, Butt, PhDChief ScientistHunter Area Pathology Service

Mariann Zerbes, BScDivision of Microbiology and Infectious

Diseases

John Hunter HospitalNewcastle, New South WalesAustralia

Iven J, Klineberg, PhDProfessorCollaborative Pain Research UnitFaculty of DentistryUniversity of SydneyWestmead HospitalWestmead, New South WalesAustralia

R, Hugh Dunstan, DPhilSenior LecturerDepartment of Biological Sciences

Timothy K. Roberts, PhDAssociate ProfessorDepartment of Biological Sciences

University of NewcastleNewcastle, New South WalesAustralia

Correspondence to:Dr Iven KlinebergProfessonal UnitLevel 3Westmead HospitalDental Clinical SchooiWestmead, New South Waies 2145Australia

A visual analog pain scale and scalar responses to O pain/symp-tom indicator Symptom Checklist-90-Revised (SCL-90-R) ques-tions were used to assess symptom prevalence and pain severity in43 chronic orofacial muscle pain patients and 40 control subjects.The orofacial muscle pain group reported pain in an axial skeletaldistribution; neurocognitive, gastrogenitourinary, and muscu-loskeletal symptoms; infectious events at or preceding onset; simi-lar symptoms in sexual partners; and low prevalence of trauma.Sudden onset was reported by 30.2% of pain patients. Strong asso-ciations were found between chronic orofacial muscle pain and (1}onset-related infectiouslike events (67.4%); (2| a bigber prevalenceof history of respiratory and gastrogenitourinary infectious events;and (3) high prevalences of similar pain symptoms in long-termsexual partners. Tbe SCL-90-R somatization scores (> 62) had ahigher prevalence in tbe chronic pain group. No prevalence differ-ences or associations with pain/symptom indicators were foundfor depression or anxiety dimension scores. Tbese data suggestthat patients with recurrent systemic infectious events have ahigher prevalence of reporting of chronic orofacial muscle paincompared with control subjects, and tbese infectious events areassociated with the onset of chronic orofacial muscle pain in 67%of patients.J OROFACIAL PAIN 1996; 10:339-350,

key words; pain etiology, chronic disease, myofascial pain,fibromyalgia, cervical pain, temporomandihular pain

Chronic muscle pain conditions form a group of syndromesthat have many similar characteristics. These heterogeneotisyet overlapping groups of condition.s may intermittentlyaffect up to 70% of the population, with severe forms such asfibromyalgia occurring in between 5% to 10%,' They represent avery common clinical and therapeutic problem'-' with unknownetiology. Many etiologic bypotheses have heen proposed for theseconditions, including psychologic phenomena, neurotransmitter-related anomalies, reflex muscle hyperactivity, denervation anddeafferentation-related occurrences, immunologie- or cytokine-mediated events, and viral infections.^

Journal of Orofacial Pam 3 3 9

McGregor et ai

In a pilot study"* of 35 patients with chronic oro-facial muscle pain, an association was foundhetween muitiorgan symptoms and a history ofurinary tract infection. There was no increase inthe prevalence of reporting of viral infections suchas measles, mumps, or glandular fever, initial as-sessment of the urinary tract microbiology re-vealed that althotigh there was an increased preva-lence of history of urinary tract infection, thepatients with orofacial muscle pain did not havepyuria or hematuria and therefore did not havecurrent urinary tract infections. However, therewas an Increased prevalence of elevated midstreamurinary staphyiococcal counts {F < ,02) in thepatients with orofacial muscle pain (n = 9) com-pared to those in age-matched and sex-matchedcontrol subjects. Thus, the present study assesseschronic orofacial muscle pain patients with age-and sex-matched control subjects for pain charac-teristics, Hopkins Symptom Checklist-90-Revised(SCL-90-R) responses, microbiology, symptomprevalence, and onset events. This study presentsthe data on (1) pain distribution, (2) symptom pre-sentation, (31 SCL-90-R psychologic inventoryresponses, and (4) onset-related events and infec-tious histories in patients with chronic orofacialmuscle pain. The assessment of the microbiologyin this study cohort will he presented in subse-quent articles.

Materials and Methods

Selection of Patients and Control Subjects

Forty-six sequentially presenting patients, unrelatedto the pilot study cohort, were referred for assess-ment of chronic orofacial pain and were inter-viewed and chnically examined during a 7-monthperiod (January to July 1993), These patients weredesignated to he the chronic orofacial muscle pain(MP) group. Patients were selected as having mus-cle pain on the basis of a positive response on avisual analog pain scale |VAS) of average painintensity in the 2 weeks prior to consultation, aswell as the presence of palpable muscle pain in thereported pain areas.

The pain was present on greater than 50% ofdays during the 3 months immediately precedingconsultation. Pain was not associated with theteeth; temporomandihular joint (TMJ) clicking,arthritis, or crepitation; sinusitis; sahvaty glands; ornerve or vessel pathology.

Forty-one age- and sex-matched control subjectswere reeruited. Nine were relatives of the pain

patients and 32 were unrelated suhjects to ensuresimilar socioeconomic and ethnic backgrounds.Control suhjects were eligible for inclusion in thestudy if they had no response to the VAS of aver-age pain intensity in the 2 weeks prior to consulta-tion, did not give a history of chronic pain, andhad not required professional advice or treatmentfor chronic muscle pain in the previous 12 months.Acute pain associated with trauma during the pre-ceding 12 months was not an exclusionary crite-rion, Fach patient and control subject provided in-formed consent and was assessed by one clinician(NRM).

Collaborative Pain Research Unit Questionnaire

To facilitate standardized data collection, a com-prehensive questionnaire (Collaborative Pain Re-search Unit Questionnaire [CPRU], 1995) wascompleted by all study participants at initial pre-sentation. The questionnaire was compiled froman existing University of Sydney Orofacial PainQuestionnaire, together with items about featuresof other muscle pain conditions {generalized,regional, and localized pam syndromes) reportedin multiple articles from the literature (MEDLINE,1966 to 1991). The questionnaire contained fourbody diagrams (left and right head and neck, frontand back of whole body for pain distribution) and154 questions on medical and family history,onset, exacerbation, signs and symptoms, dura-tion, pain severity, previous treatment, and poten-tial etiologic conditions. The patienrs were askedto delineate on the body diagrams the usual areasof pain, irrespective of their perception of thepain's origin, and to qualify by notes what theyperceived to be the origin if they thought it wasunrelated to their orofacial pain.

Sytnptom Ch8cklist-90-Revised

The study participants completed an SCL-90-R';the prevalence and degree of responses to individ-ual questions, the index scores, and the dimensionscores were compared between the MP and controlgroups. The SCL-90-R raw dimension scores wereexpressed as a T score, which is a dimension indexscore standardized for changes in age and sex. TheSCL-90-R dimension T scores greater than 61were used to indicate psychologic morbidity perthe SCL-90-R handbook,-'' Psychologic syndrome Tscore profiles were allocated according to the SCL-90-R handbook. The ,SCL-90-R dimension ptofilepatterns were created using the K means clusteranalysis method. This allowed assessment of the

3 4 0 Voiume 10, Number 4, 1996

McGregor et al

various psychologic dimension profiles and the dis-tribution of these profile patterns between MP andcontrol groups.

Pain/Symptom Assessment

Pain was defined to he of muscle origin if therewas positive palpation of head, neck, and shouldermuscles''' in the sites reported to he painful.Patients reporting pain without associated muscletenderness were excluded. Patients were chosen tocomply with a myofascial pam diagnosis (a GroupI muscle disorder) of the temporomandihular dis-orders (TMD) research diagnostic criteria.^ Afterpatient selection, pain and somatic symptom sever-ity were retrospectively assessed using the scalarresponses to the VAS and 13 questions (Q) fromthe SCL-90-R (Ql, headaches; Q4, faintness anddizziness; Q12, chest pain; Q14, low in energy orrun down; Q27, lower-back pain; Q39, heart pal-pitations; Q40, nausea; Q42, muscle soreness;Q52, numbness or tingling; Q55, trouble concen-trating; Q56, weakness; Q58,, heavy feelings inlimhs; Q66, restless or disturbed sleep).'' An indexof symptom prevalence was calculated as the totalnumber of symptoms reported from the 54 symp-tom questions in the pain questionnaire. The SCL-90-R global lnde.xes were also used to assess inter-group differences. The SCL-90-R global symptomindex (GSI) is a measure of psychologic distresscomhining symptom number and severity.' TheSCL-90-R positive symptom distress index (PSDI)is a measure of psychologic severity corrected forsymptom numher. The positive symptom total(PST) is a measure of psychologic symptom preva-lence,̂ Suhjects were instructed to report any bodysymptoms, irrespective of the suhjects' impressionof their origin, and not to restrict reporting to theorofacial region. This was done to avoid restrictedreporting of whole body symptoms and to excludethe bias toward suhjects reporting only orofacialsymptoms that may be perceived to be of impor-tance in assessment of their orofacial condition.

Statistical Analysis

The clinical data were analyzed using f tests, thechi square test, the iVIann-Whitney U test, andmultiple regression analysis (significance P < .05).Cluster analysis by the K mean method was usedto differentiate the SCL-90-R dimension profilepatterns. Correction for multiplicity occurredwhere necessary. Sensitivity, defined as the preva-lence of true positives in the MP group, and speci-ficity, defined as the percentage of true negatives in

the control group, were determined. These datawere processed using Access (version 1.1,Microsoft, Redmond, WA), Excel (version 4,0,Microsoft), and Statatisca {version 4,5, Statsoft,Tulsa, OK).

Results

Patient Characteristics

Of the 46 interviewed MP patients who met thecriteria, three were excluded because of incompletequestionnaires. One of the 41 control subjects wasexcluded because of a positive response to the VASand a history of chronic pain within the previous12 months. The age and sex characteristics weresimilar for the MP and the control (C) groups (MPmean age and standard deviation [SD] 39,5 ± 11,6years, range 16 to 62 years, 79.1% female; Cmean age and SD 35.5 + 15,2 years, range 11 to72 years, 70% female). No statistically significantdifference was found in the number of MP or con-trol subjects in long-term marital or de facto rela-tionships (29 MP and 25 C subjects). No differ-ence was found m the ethnic hackgrounds of theparticipants. All were born in Austraha except two(both of Chinese origin). The majority of partici-pants were of Western European ethnic origin(Anglo-Celtic-Germanic, 74 of 83, 89,2%), seven(7 of 83, S.4%) were of Eastern European origin(Croatian/Slavic), and two were of Asian origin(Chinese, 2 of 83,2.4%),

Pain Distribution

Although all MP patients reported a history oforofacial muscle pain, on the day of the clinicalexamination, they also reported pain to be presentin other regions: (1} 38 (88,4%) had neck, shoul-der, and thoracic spine pain; (2) 35 (81.4%) hadface and head pain; (3) 30 (69,8%) had lower-limbpain; (4) 29 (67,4%) had lower-back pain; (5) 24(55.8%) had upper-limb pain; (6} 17 (39,5%) hadabdominal pain; and (7) 14 (32.6%) had anteriorchest wall and sternum pain. At initial examina-tion, three patients (7,0%) presented with painrestricted to the head, 12 (27,9%) had regionalpain distributed predominantly in a facioscapulo-humeral distribution, and 28 (65,1%) had pain inall four quadrants. Thus, at initial examination,7.0% were classified as having orofacial painalone, and 27,97o were classified as having myo-fascial pain syndrome; although the fibromyalgiapoints were not palpated, 65,1% had a pain distri-

Journal of Orofacisl Pain 341

McGregor et al

bution consistent with symptoms of fibromyalgia.Limb and anterior cbest wall and/or sternal painwere reported only by patients with pain in allfour quadrants. These patients characterized thefibromyalgidike patients.

At onset of the pain condition, head pain wasthe first site of pain reported hy 20 of the 43 MPpatients (46.5%], with 30 (69.8%) reporting afacioscapnlohumeral distribution. However, eight(IS.6%) also reported pain to he fnst noticed intbe lower back. Most patients reported localizedor regional pain at onset, whicb progressed tomore generalized pain involving all four quadrantsof 28 (65.1%) patients in this cobott.

Pain Assessment

As previously descrihed, there was a difference inthe VAS (mean ± standard error of the mean [SEM|MP = 2.72 ± 0.76; C ̂ 0) hetween the MP and con-trol groups. Forty-two of 43 197.7%| MP patientshad a VAS score of greater than 2. Forward step-wise discriminant function analysis using all painsymptom indicators and tbe SCL-90-R globalindexes, excluding the VAS, was used to assessintetgroup differences. Table 1 shows tbat symptomprevalence and SCL-90-R Q42 (muscle soreness),Q14 (low energy or run down—lethargy/fatigue),and Q40 (natisea) were important factors in differ-entiating hetween the MP and control groups. TheSCL-90-R glohal indexes were nor important inter-grotip determinants. Therefore, MP patients in thiscohort, selected on the basis of muscle pain, alsohad increased sytnptom prevalence, fatigue, andnausea compared with control subjects.

When the patients with four-quadrant pain(fibromyalgiclike group, n = 28) were compared

Table J Pain Assessment Questions FoundImportant in Determining the Difference BetweenMP Patients and Control Subjects

Pain/symptom indicators

Symptom prevaienceÇ42 Muscle soreness014 Low in energy orrun down

040 Nausea027 Lower back pain039 Palpitations05S Heaviness in limbs

Wilks'lainbda

.443462

429

.438

.423

.422

.417

f(l,75)7.09

10.59

4.466.183.463.232.31

l'-

< 009< .002

< .04< 02= 066= .076- 132

'Forward slepwise discriminan! function anaiysisP-:.00001, WiiiiB'iambda = .40192.

with che remainder of the MP patients (Other MPgroup, n = 15), forward stepwise discriminantfunction analysis revealed that Q40 (nausea, P <.003), Q14 (low energy or run down—lethargy/fa-tigue, P < .002), Q42 (muscle soreness), and symp-tom prevalence were also important factors in dif-ferentiating hetween tbe fihromyalgiclike patientsand the remainder of the MP patients (model:Wilks' lamhda = .387, F|4,38] = 15.03, F < .00001).Thus, nausea and fatigue were important in deter-mining the difference hetween the fibromyalgiclikepatients and the other MP patients.

Signs and Symptoms

Of the 54 symptoms reported in the CPRU ques-tionnaire, 35 symptoms and tbe sensitivity (Sen)and the specificity (Spec) of eacb are presented inTable 2. An increased prevalence of the symptomsin tbe MP patients was found (multiplicity correc-tion P < .005). The symptoms could be groupedinto musculoskeletal-, gastrointestinal-, genitouri-nary-, neurologic-, and infection-related and othergroups. There were highly significant prevalencedifferences in the musculoskeletal-, gasttogeni-tourinary-, and infection-related symptoms he-tween tbe MP and control groups. The fihromyal-giclike group also reported increased symptomprevalence (fibromyalgiclike - 21.2 ± 10.6 andOtber MP = 13.1 ± 8.5; F < .01), in particular nau-sea (Sen ^ 88%, Spec = 78%; P < .001) and dizzi-ness (Sen = 72%, Spec = 83%; P < .001) whencompared with the other MP patients (multiplicitycorrection P < .005).

Onset-Related Events

Thirty-one of tbe 43 (72.1%) MP patients associ-ateá onset of their symptoms with one or morecausally-related events including upper respiratoryttact/influenzalike infection, low-grade diarrhea,genitourinary tract infection, glandular fever, ortrauma (Table 3). Pain and symptom onset wasreported to be either sudden (13 of 43, 30.2%) orgradual (30 of 43, 69.8%). Patients witb suddenpain/symptom onset were more likely to report aninfectious event associated with onset (upper respi-ratory tract/influenzalike infections [sudden = 11of 13, gradual = 5 of 30; P < .001) or genitouri-nary tract infections [sudden = 3 of 13, gradual - 1of 30; F < .05j). Interestingly, pain onset followinga new sexual contact was reported by 3 of 43 (7%)MP patients, each of whom reported a genitouri-nary tract infection at onset, and one of whom alsoteported glandular fever at onset. Patients with

342 Volume 10, Number 4, 1996

McGregor et ai

Table 2 Sensitivity and Specificity of Questionnaire-Reported Symptomsin MP Patients (n = 43) and Control Subjects (n = 40]

Symptom

Wusouloskeletal symptomsHeadachesTension headachesTtviJ clickingMuscie weaknessMorning joint stiffnessMuscle fatigueMusole twitchesMusole crampsBruxismSciatica

Gastrointestinai symptomsNauseaAbdominal bloatingAbdominal painirntabie bowel syndromeLow-grade diarrheaGastnc refiuxGonsti patio n

Genitounnary symptomsUrinary frequencyGroin iymphodyniaRecurrent genitai infectionsDysuriaMenstrual pain (females)

infection-related symptomsFatigue or lethargyLow-grade feverGervicai iymphodyniaAxiai IymphodyniaRecurrent sore throats

Neuroiogic symptomsHyperesthesiaParesthesia

Other symptomsGardiac dysrythmiasGhest painAllergiesDiîziness or fainting spellsTinnitusEaraches

Sensitivity (%]

88.469.867.465.162.860.546.541.939 532 6

65.151 248 844 241.939.532 6

55 839.525.523 264 7

65 148.855.844.244.2

51.241.9

46.532 653.548.844.232 6

Specificity (%]

55.070.082.5

100.085.092.592.595.090.0

100.0

92.587.597.595.092.595.0

100.0

90.097.5

100.097.572.0

95.095.092.592.590.0

90.C97.5

95.095.087.595.090.092.5

< 001< 001< 001< .001< .001< .001< .001< .001< .005i.OOl

< 001< 001< .001< .001< .001< .001< 001

< .001< .001

McGregor et ai

graduai pain/symptom onset were more likely tobe unable to identify an onset-associated event(sudden = 0 of 13, gradual = 12 of 30; P < .007),

History of Infectious Events

Patients witb MP had more occurrences of historyof cbronic recurrent upper respiratory tract infec-tion (Sen = 39.5%. Spec = 95.0%; P < ,001), his-tory of genitourinary infection (Sen = 37.2%, Spec= 82.4%; P < .05), and history of appendectomies(Sen = 23,2%, Spec = 92.5%; P < ,05), No statisti-cally significant differences were noted for report-ing of glandular fever.

Pain in Long-Term Partners

In tbe MP group, a total of 29 patients were inlong-term relationships, and 17 of these (Sen =39.5%, Spec = 90,0%; P < ,02) reported tbat theirpartner had chronic mtiscle pain of varying sever-ity and distribution. When the participants in long-term relationships were compared between tbe MPand C groups, the difference was even more pro-nounced (Sen = 58.6%, Spec = 84,0%,; P < .002),Both of tbe control suhjects who reported thattheir partners had muscle pain also responded pos-itively to at least one of the SCL-90-R pain/symp-tom indicator questions.

Group Differences for the SCL-90-R

Table 4 shows tbe analysis of the responses to thevarious SCL-90-R global indexes and dimension

scores, Tbc major difference between cbe groupswas the higher global indexes and the somatizationdimension scores in the MP group, with higherscores also noted in the obsessive-compulsive anddepression dimensions. The somatization dimen-sion was the only group dimension in the SCL-liU-R-defincd psychologic morbity indexes range(greater than 62), and there was a bigber preva-lence of somatization T scores of greater tban 61in the MP patients (25 of 43 MP, 6 of 40 C; P <.0001). No other dimension T score of greaterthan 62 was found tu have an increased prevalencewhen MP and C groups were compared.

The fibromyalgiclike patients had higher SCL-90-R global indexes for the global symptom index(GSI) (P < ,005) and the positive symptom total(PST) (P < .02), but nut for tbe positive symptomdistress index (PSDI) (P < .054) when comparedwitb the otber MP patients (Mann-Whitney U test).Similarly, the fibromyalgiclike group had higherscores in the somatization dimension (P < ,0003),the depression dimension (P < .004), the obsessive-compulsive dimension (P < ,008), and the anxietydimension (P < ,02) when compared with the otherMP patients. Discriminant function analysis of thedifference between the fibromyalgiclike patientsand the other MP patients revealed tbat the somati-zation dimension (P < .03) was the prime assess-ment difference hetween tbe rwo groups (model:Wilks' lambda - .53, F[5,371 = 6,56, P < .0002),witb the hustility (P < .05) and tbe PST (P < ,005)being the second and tbird variables, respectively.Depression and anxiety were not important dis-criminant dimensions in this model.

Table 4 Mean i SEM SCL-90-R Global and Dimension T Scoresin MP Patients and Control Subjects

Giobsi indexesGiobal severity

Positive symptom totalPositive symptorn distress

Dimension scoresSomatiiation

Obsessive-compulsiveinterpersonai sensitivityDepression

AnxietyHostiiityPhobic anxiety

Paranoid ideationPsychoticism

Muscle painpatients

58,0 ± 1,3258,1 ± 1 1760 S I 0 97

64,2 ± 1 1558.1 ± 1,42

5!,9± 1,5958,0 ± 1.35

52.3 ± 1,5650.5 ± 1,2749.6 ± 1.22

48.3 ± 1 3754,8 ± 1,32

Controlsubjects

48,8 ± 1,7850,7+ 1,7752.3 ±1.15

50 2 ± 1 73

49 8 ± 1 6952,9 ± i,SO49,9 ± 1,7748,3 ± 1,5349 a ± 1,4347 a ± 1,07

49 1 ± 1,4750.8 ± 1.36

Statisticalsignificance*

< .02

McGregor et al

Global Index Associations With DimensionChanges and Pain/Symptom Indicators

Logistic regression analysis of the GSI against thedimension T scores revealed important associa-tions, in order of discriminant ability, with thedepression dimension (P < .000000001), the obses-sive-compulsive dimension {F < ,0000004), thesomatization dimension (F < ,00000006), and theinterpersonal sensitivity dimension (P < ,00008); anegative association was found for the phobic anx-iety- dimension (F < .03) (model: R- = ,960, f|7,75]= 255.87, F < .00001), The GSI was also associ-ated with SCL-90-R Q66 {restless or disturhedsleep, P < ,15), Q52 (numbness or tingling, P <.008), and Q12 (chest pain, P < ,007) (model: R- ^.413, F[4,781 = 13.71, F < .000001). The CSI wasnot associated with the index of symptom preva-tence or the VAS, and it was not an important fac-tor in determining the difference between the MPand controt groups.

Logistic regression analysis of the PSDI againstthe dimension T scores revealed a significant asso-ciation only with the somatization dimension {P <.0000002; model: R- = .529, f[9,73J = 9.10, P <.00001). The PSDI was also associated with SCL-90-R Ql (headaches, F < ,002), Q42 {muscle soreness,P < .002), and Q52 (ntimbness or tingling, P < .002)(model; R̂ = ,542, F[15,67] = 5,29, F < ,000001),The PSDI was not associated with the index ofsymptom prevalence or the VAS, and it was notimportant in determining differences between theMP and control groups.

Logistic regression analysis of the PST againstthe dimension T scores revealed strong associa-tions, in order of discriminant ability, with rhedepression dimension {F < .000002), interpersonalsensitivity dimension (P < .007), and the obsessive-compulsive dimension (P < .004) (model; R- =.887, f[6,7é] ^ 99.92, P < ,00001). The PST wasalso associated with SCL-90-R Q66 (restless ordisturbed sleep, P < ,06), Q52 (numbness or tin-gling, F < ,07), and Q12 (chest pain, P < .21)(modeh R̂ = .325, F[4,78] = 9.39, P < .000001).The PST was not associated with the index ofsymptom prevalence or the VAS,

Pain/Symptom indicators and SCL-90-RDittiensron Differences

Five pain/symptom indicators, the VAS, the indexof symptom prevalence, Q14 {low energy/lethargy/fatigue), Q40 {nausea), and Q42 {muscle soreness),were associated with differences between the MPand C groups. Analysis of the dimension score

changes with these pain/symptom indicators wasundertaken to assess the association between eachdimension and pain expression.

Logistic regression analysis of the VAS againstthe dimension T scores revealed that importantchanges in dimension scores occurred with increas-ing VAS scores (model: R'- = .337, f [9,73] = 4.12,F < .0003), A positive relationship was foundbetween the VAS and somatization scores {F <.0002), and a negative relationship was found withinterpersonal sensitivity (P < .02). The index ofsymptom prevalence was associared with strongdimensional changes (model: R̂ = ,614, F[9,73] =4,91, F < .00004), A positive relationship wasfound between the index of symptom prevalenceand somatization {P < ,008), obsessive-compulsive(P < ,05), and psychoticism (P < .05), and a nega-tive relationship was found with interpersonal sen-sitivit)' {P < ,04). The SCL-90-R Q42 {muscle sore-ness) was associated with pronounced dimensionalchanges {model: R' = .439, F|9,73J = 6,36, P <.00001) and had a positive relationship with thesomatization dimension (P < .000002). The SCL-90-R Q14 (low energy/run down) was also associ-ated with significant dimensional changes {model:R' = .372, /•T9,73J = 4.S1, P < .00005) and had anegative relationship with interpersonal sensitivir;'(P < ,005), No relationship was found between014 and rhe depression dimension, even though014 forms part of that dimension. The SCL-90-RO40 (nausea) was also associated with strongdimensional changes (model: R- = .392, P|9,73J =5,23, P < .00002) and had a positive relationshipwith somatization {P < .02) and a negative relation-ship with interpersonal sensitivity (P < .04), No as-sociation was found hetween any of the pain/symptom indicators and the depression and anxierydimension scores.

Three dimensions had increased responses in theMP patients: somatization; ohsessive-compulsive;and depression. Logistic regression analysis of thesomatization dimension against the pain/symptomindicators fotind positive associations with SCL-90-R Q52 (numbness or tingling; P < .009) and Q42(muscle soreness; P < .04) (model: R- = ,592,f[15,67] = 6.49, P < ,000001). The ohsessive-com-pulsive dimension was positively associated withSCL-90-R Q55 {ttouhle concentrating, P < .008),Q52 (numbness ot tingling; P < .009), and Q12(chest pain, P < .04). The depression dimension wasnot associated with any of the pain/symptom indica-tots. The anxiety dimension was positively associ-ated with SCL-90-R Q39 (heart palpitations, P <.05) and negatively associated with the VAS (P <,05) (model: R^ = .452, P[15,671 = 3.68, P < ,0002).

Journal of Orofaeial Pain 3 4 5

McGregor et al

SCL-90-R dimensions

Cluster No, (MP [n - 43], control [n - 40]; chi square ptobablility)

— O — Cluster 1 (6 MP, 13 control; not significant)....£].... Cluster 2 [16 MP, 3 control; P

ílcGregor et al

Table 5 Clinical Characteristics of the 29 Flevated Muscle Pain-RelatedSCL-90-R Questions

SCL-90-R, interpretation

Physical symptoms01 Headacbes04 Faintness or dizzinessQ12 Pains in heart or cbestQ27 Pains in lower back040 Nausea or upset stomach042 Muscle soreness048 Trouble getting your breatb049 Hotoroold spells052 Numbness or tingling in parts of your body053 A lump in your throat056 Weak feelings in bodyQ53 Heavy feelings in tlie arms or legs017 Trembling039 Heart pounding or racing

Sleep disturbance044 TroLible falling asleep064 Awakening in tbe early morning066 Sleep tbat is restless or disturbed

Neurocognitive everts09 Trouble remembering things038 Having lo do things very slowly to ensure correctness046 Difficulty making decisions051 Wind going blank055 Trouble concentrating

Possible disease consequences05 Loss of sexual interest or pleasure014 Feeling low in energy or slowed down071 Feeling everytbing is an effort087 The idea tbat something senous is wrong with your body

Possible psychologic responsesQ30 Feeling biue032 Feeling no interest in things050 Having to avoid certain things, places, or activities because they

frighten you

SomatizationSomatizationSomatizationSomatizationSomatizationSomatiïatiODSomatizationSomatizationSomatizationSomatizationSomatizationSomatizationAnxietyAnxiety

Additional itemsAdditional itemsAdditional items

Obsessive-compulsiveObsessive-compulsiveObsessive-compulsiveObsessive-compulsiveObsessive-compulsive

DepressionDepressionDepressionPsycboticism

DepressionDepression

Pbobic anxiety

from those of the control subjects. The SCL-90-Rinterpretation shows increases iti questions fromthe somatization, obsessive-compulsive., anddepression dimensions. Tbe somatization dimen-sion could be interpreted to indicate simple pbysi-cal symptom reporting, and the obsessive-compul-sive dimension responses could be interpreted toindicate neurocognitive changes. The sleep distur-bance gtoupmg was classified as an additionalitem by the SCL-90-R. The questions from thedepression dimension could be interpreted as reac-tive psychologic responses or symptoms, or conse-quences of the disease condition.

Discussion

The patients with chronic orofacial pain in thisstudy were compared with a control group of sub-jects who did not respond to a VAS of average painintensity in the previous 2 weeks and did not reporta history of or treatment for chronic pain in the last12 months. Pain and symptom presentation wasassessed against the scalar responses to 13 SCL-90-R questions that were predominantly from tbesomatization dimension and that represented arange of symptoms likely to be higher in painpatients than in control subjects. A symptom preva-

Journal of Orofacisl Pain 347

McGregor et al

lence score derived from the qiiestiontiaire was alsoused. In studies of this type, the reporting ot painand treatment sought for pain are influenced bymany factors that have been shown to be poor indi-cators of the presence of physical symptoms.''"*The SCL-90-R questions were used in the presentstudy, even though they were not specific for oro-facial pain symptom expression, heeause they(1) gave a scalar response; (2) indicated pain/symptom expression ¡n rhe previous 7 days only,representing current symptom expression; (3) werecotnponents of a larger qtiestionnaire and wereunlikely to register to the study participant as painor symptom response indicators; and (4) may moreaccurately reflect overall symptom expression. Inan attempt to more completely understand the eti-ology of these conditions and not simply eachpatient's major complaint uf pain, a hroader assess-ment of the study participants was made. In sup-port of this approach, symptom prevalence wasfound to be a much stronger indicator of increasingcondition severity and intergroup differences thanwas the assessment of muscle soreness. In supportof these observations, other studies tising palpablemuscle tenderness found no correlation betweenpalpable tenderness and pain.' Fatigue or lethargy(SCL-90-R QI4) and nausea (SCL-90-R Q40) werealso found to he important intergroup determinantsthat were also more prevalent tn patients withfibromyalgialike pam than m those with myofascialor localized pain. The assessment of these symp-toms may allow better quantification of the pamcondition than would pain scores alone.

In the present study, patients with orofacialmuscle pain reported a wide pain distribution thatwas prominent in the axial skeleton, but alsoextended to the limbs and anterior chest wait inpatients with more severe symptoms. The distri-bution of patients in this study cohort wirh fibro-myalgiclike, myofascia! pain, and localized painpresentation appeared skewed toward the fibro-myalgiclike patients. This may reflect che methodof patient selection, the difference in approach inobtaining patient reporting of whole body symp-toms, or simply a random event. However, everyeffort was made to request that patients reportsymptoms distant from the face, symptoms thatrhey may not normally report in a dental setting.Irrespective of these considerations, rhis cohort ofpatients with chronic orofacial muscle pain had ahigher prevalence of musculoskeletal, gastrointesti-nal, genitourinary, and neurologic symptoms con-sistent with a systemic condition; symptoms andhistories suggestive of nasopharyngeal, gastroin-testinal, and genitourinary infectious events, par-

ticularly in patients reporting a sudden onset; low-grade fever and Iymphodynia; reporting of pain mtheir sexual partners; and infectious events occur-ring at the time of onset. The TMD research diag-nostic criteria'* were devised for a dental settingand do not differentiate between patients on thebasis of mosr of these associated symptoms. Thepresent study showed that patients with orofacia!muscle pain (Group I, myofascial pain**) reportedmultiple symptoms distant to the orofacial area,suggesting that the condition has a systemic basisand that the orofacial muscle pain is unlikely torepresent a localized functional phenomenon. In-terpretations of these data include ('I) patients pre-senting with orofacial muscle pain have a hetero-geneous group of etiolugic medical conditions, (2)orofacial muscle pain patients have increasedprevalence of infectious events as a result of theirrelated medical condition(s), or (3) a medical con-dition with orofacial muscle pain as a presentingsymptom is the result of a transmissible agent[s).In support of these possibilities, chronic musclepain conditions have been associated with entero-virus" and Epstem-Barr virus^^ infections; chronicbacterial infections, such as Lyme disease'^; andparasitic infections, such as toxoplasmosis.''' In ad-dition, injection of inflammatory mediating cyto-kines such as inrerleukin-2'^ and interferon'^ in-duces myalgia and many signs and symptomsreported by patients with chronic pam, supportingthe possibility of an immune-mediated condition.This model might explain the development of mus-cle pain symptoms via multiple etiologic pathways,

A psychophysiologic model has been proposedfor chronic pain, and several studies have shown anassociation with increased psychologic dimensionchanges, particularly depression scores, in musclepain patients.^-'*'•'''"-' In che present scudy, patiencshad increased T scores in the somatization, obses-sive-compulsive, and depression dimensions. TheMP group had a higher prevalence of somatizationdimension T scores greater than 62 but did not havea higher prevalence of obsessive-compulsive ordepression dimensions scores greater than 62. TheSCL-90-R GSI, PSDI, and PST were not found to beimportant incergroup determinants, and none wasassociated with the factors found by logistic regres-sion to be important intergroup determinants, eg,VAS, symptom prevalence index, or SCL-90-R Q14(fatigue/letharg}'), Q40 (nausea), and Q42 (musclesoreness). The GSI (rhe SCI-90-R psychologic dis-tress index) and the PST (the SCL-90-R symptomprevalence index) were predominantly associatedwith Increasing depression scores; the PSDI (theSCL-90-R symptom severity measure) was strongly

348 Volume 10, Number 4, 1996

McGregor et al

associated with the somatization dimension. Thepatients witb higber GSI and PST scores had strongassociations witb tbe depression dimension andwere found to bave symptom associations differentfrom tbose most commonly expressed by the MPpatients. Cluster analysis revealed a subgroup ofpatients with increased multiple SCL-90-R dimen-sion T scores greater than 62 and with a higherprevalence in the MP group. Possible interpretationsinclude (1) patients in tbis subgroup bave a psycho-logic morbidity, (2) patients in this subgroup repre-sent an augmenting group,^ or (3) a combination ofthe other two, Tbe distinction between these threepossibilities can only be determined by the appropri-ate clinical investigations. However, these data donot provide strong evidence for a direct associationbetween increasing affective disturbance and musclepain or physical symptom prevalence in the studycohort, but a subgroup that may have a psychologicmorbidity was found.

The validity of tbe SCL-90-R interpretation forpain patients bas been previously questioned''' andwas examined in the present study from twoaspects: the affective psychologic assessment of theSCL-90-R and as a physical disease with possiblepsychologic consequences. Table 5 presents thesetwo possible interpretations of the SCL-90-R re-sponses and shows tbat the SCL-90-R can be inter-preted in two ways. It was not witbin tbe scope ufthis article to attempt to differentiate betvi'een tbesetwo possibilities, but these data suggest that the MPgroup could be divided into patients with a predom-inantly physical illness, and another smaller sub-group (about 15%) with potential psychologic mor-bidity similar to tbat described in the TMD researchdiagnostic criteria,^ The SCL-90-R responses dosuggest tbat patients witb cbronic orofacial mnsclepain have a higher prevalence of altered neurocogni-tive responses particularly associated with memoryand concentration. These data are interestingbecause alterations of N-methyl-D-aspartate(NMDA) receptors in the brain and the spinal cordare associated witb not only modulation of painresponses -̂ but also memory functions.̂ ^ The possi-bility of cbanges in brain metabolism and functionare also supported by the observation of alterationof the hypothalamic-pituitary-adrenal axis in pa-tients with fibromyalgia syndrome,^* depression,and chronic fatigue syndrome.-^

Conclusions

The present study indicates that for patients withchronic orofacial muscle pain, (Í) muscle pain symp-

toms are not restricted to the orofacial areas; (2)multiorgan involvement and diverse symptomexpression suggest tbat a systemic condition maybe occurring in at least a subgroup of patients; (3)gastrogenitourinary symptoms are prominent; (4)onset is associated witb infectious events in a sub-group of patients; I5) there is an increased preva-lence of bistory of infectious events and recurrentinfections; (6) partners of muscle pain patientsbave an increased prevalence of muscle pain; (7)tbere are increased SCL-90-R inventory scores inthe somatization, obsessive-compulsive, anddepression dimensions; (8) tbe SCL-90-R responsessuggest that neurocognitive cbanges are occurringin the pain patients; (9) tbe SCL-90-R inventoryresponses can bavc multiple interpretation; and(10) multiple etiologies are likely. Future researchshould focus on assessment of the many possibleetiologic agents or events associated witb onset ofcbronic orofaciai muscle pain.

Acknowledgments

The authors wish tu acknowledge the support of the GideLang Foundation fur this research.

References

1. de Girolamo G. Epidemiology and social cosis of lowhack pain and fibromyaigia. Chn J Pain 1991;7(suppl):S1-S7,

2. Magni G, Canton G, Gallimheni L, Pain symptoms in anaduluscem kalian population, EurJ Pediatr 199Ü;149:592-593,

3, Vaeroy H, Merskey H (eds), Ptugress in Fibromyalgia andMyufascial Pain, Amsterdam, The Netherlands: Elsevicr,1993,

4. McGregor NR, Butt HL, Klineberg l j , Craniocervicalmyalgia: An association with genitoutinary infection[abstract 39]. J Dent Res 1991;72-670.

5, Derogatis LR, SCL-90-Ri Administration, Scoring andProccdu:-es Manual-II for the Revised Version, Towson,.MD: Clinical Psychometrics Research, 1975,

6. Helkimo M. Correlation of the electromyographic silentperiod duration and che Helkimo Dysfunction Index, ActaOdont Scand 1974;37:51-56.

7, Wolfe F, Smyrlie HA, Yunus MB, Bennett RM, Bom-bardier C, Goldberg DL, et al. The American College ofRheumatology 1990 criteria for the dassificarion offibromyalgia. Report of the multicenter criteria commit-tee. Arthritis Rheum 1990;33:160-172,

8, LeResche L, Von Korff MR. Research diagnostic criteria.Axis I: Clinical TMD conditions. Axis U: Pain-related dis-ability and psychological status, In; Dworkin SF, LeRescheL leds}. Research Diagnostic Criteria for Temporomandib-ular Disorders: Review, Criteria, Examinations and Speci-fications, Critique, J Craniomandib Disord Facial OralPain 19 92;é 1.327-334,

Journal of Orofacial Pain 349

McGregor et al

9, Pullinger AG, SeUgman DA, Solberg WK, Temporo-mandibular disorders. Part 1: Functiunal status, dcntomor-pbnliigical features, and sex differences in s nonpatientpopulation, J Prosthet Dent 1988;59:228-235,

10, KIcinknecht PA, Alexatider LD, Maliony ER, Dworkiti SF,Correspondence between subjective report of temporo-mandibular disorder symptoms and clinical findings. | AmDent Assoc ISSO^lOiIiy-îol,

11, Archard LC, Bclian PO, Bell EJ, Doyle D, Bowles NE,Postviral fatigue syndrome; Persistence of enteriivirusRNA in muscle and elevated creiitinme kinase. J R SocMed 198S;S 1:326-329.

12, Straus SE, Tosato G, Armstrong G, Persistmg illness .indfatigue in adults with evidence of Epstein-Bair virus infec-tion. Ann Intern Med I 985;102:7-16.

13, Dinerman H, Steere AC. Lyme disease associated withfibromyalgia, Ann Intern Med 1992;117:281-.285,

14, Feldman HA, Toxoplasmosis: An overview. Bull N YAcad .Med 1974;50il 10-115,

15, "Wallace DJ, Margolin K, Acute onset fihromyalgia as acomplication of interleukin-2 therapy |ahstract 54| .Arthritis Rheum ]988;31(suppl):S24.

16, Farkkila M, I.ivanainen M, Roine R, Neimi ML, GantellK, Neurotoxic and other side-effects of high-dose inter-feron in ainyotrophic lateral sclerosis. Acta Neurol Scand

17. Buckelew SP, DcGood DE. Schwartz DP, Kerler RM, Cog-nitive and somatic item response pattern of pain patients,psychiatric patients, atid hospital empkij'ces, J Glin Psychol19S6;42:852-860.

18. Clark S, Camphell SM, Forehand ME, Clinical character-istics of fihrositis, II. A 'hlinded' controlled study usingstandard psychological tests. Arthritis Rheum 1985;28:132-137,

19. Lee YO, Lee SW, A study of the emotional characteristicsof temporomandihular disorder patients using SCL-90-R.J Graniomandib Disord Facia! Oral Pain 19H9;3;25-34.

20. Peiz M, Merskey H, A description of the psychologicaleffects of chronic painful lesions. Pain 1982^14:293-301,

21. Von Korff M, Dworkin SF, LeResclie L, Kruger A, An epi-demiologic cotnparison of pain complaints. Pain I988;32:173-1 S3.

22. Nasstrom J, Karlsson U, Post G, Antinotiteptive action,?of different dasses of escitatory ammo acid receptor an-tagonists in mice. Eut J Pharmacol 1992;212:21-29.

2,3. Venable N, Kelly PH, Effects of NMDA receptor antago-nists on passive avoidance learning and retrieval in ratsand mice. Psychopharmacology 1990;100;2I5-22I,

24. Demitrack MA, Dale JK, Straus SE, Laue L, Listwak SJ,Kruesi MJ, et al. Evidence for impaired activation of theliypothalamic-pituitary.adrenal axis in patients withchcotiic fatigue syndrome. J Clin Endocrinol Metal) 1991;73:1224-1234,

25, Cleare AJ, Bt-arn J, Allain T, McGregor A, Wessely S,Murray RM, O'Koane V. Contrasting neuroendocnne re-sponses in depression and chronic fatigue syndrotne. J Af-fective Disord 199S;35:283-289.

Resuttien

Evaluación del Dolor (Distribución y Ataque), Síntomas.Respuestas al Inventario SCL-90-H, y la Asociación conEventos Infecciosos en Pacientes con Dolor OrofacialCrónico

Zusammenfassutig

Einschätzung von Schmerzen (Verteilung und Anfang).Symptome, SCL-90.R und der Zusammenhang mit infek-tiven Ereignissen in Patienten niit chronischen oro-fazialen Schmerzen

Se Jtiiizó una escaia de doior análoga y las respuestas escaion-adas 3 13 respuestas indicadoras de doior y sintomas correspon.dientes a ia Lista de Venñcación de Sintomas Revisada-90 ISCL-90-R). para evaluar la prevaiencia de los sintomas y la severidaddel dolor en 43 pacientes con doior muscuiar orofacial crónico y40 pacientes de controi. Ei grupo con dolor muscuiar orofacialpresentó doior caracterizado por una distribución esqueiéticaaxiai; lo mismo que sintomas neurocognoscitivos, gastrogenitoun-nanos, y muscuioesqueiéticos. eventos infecciosos antes o en elmomento del ataque; compañeros sexuaies con sintomas simi-lares; y una baja prevaiencia de trauma durante ei ataque. Elataque repentino fue reportado por ei 30,2% de ios paoientescon doior. Se encontraron asociaciones fuertes entre ei doiormuscular orofacial crónico y (¡> los eventos de carácter infec-cioso relacionados al ataque (67,4%); (2) una prevaiancia aita deantecedentes respiratorios y eventos infecciosos gastrogenituri.narioŝ y (3) prevaiencias aitas de sintomas doiorosos simiiaresen ios compañeros sexuaies con quienes se había mantenido unareiación a iargo piazo. Los vaiores de somatiíación (> 62) de iaSCL-90-R presentaron una prevaiencia aita en ei grupo con doiorcrónico; no se encontró ninguna diferencia en cuanto a ia preva-iencia de ios valores de \s depresión o ia ansiedad No se encon-tró ninguna asociación entre ios valores de ia depresión y los indi-cadores de doior y síntomas. Estos datos indican que iospaoientes con infecciones sistémicas recurrentes tienen unamayor prevaiencia en cuanto al hecho de reportar dolor muscularorofaoial crónico en comparación con los sujetos de control, yestos eventos infecciosos están asociados con los ataques dedolor muscular orofacial crónico er 67% de los pacientes.

Um die Prävalenz von Symptomen und die Intensität derSchmerzen einzuschätzen, wurden bei 43 chronischen oro-fazialen Wuskeischmerzpatienten und bei 40 Kontroilpersoneneine Visuai Analog Scale und der SCL-90-R Fragebogenbenutzt. Die orofaziaie Muskelschmerzgruppe berichtet jber:Schmerzen in einer axial siieiettalen Verteiiung; neuroiiognitive.gastrourogenitaie und musiiuloskelettale Symptome; infeiftiveEreignisse zu oder vor Beginn der Schmerzen; ähniiche Symp-tome beim Partner, und eine itieine Pravalenz von Traumata beiBeginn. Ein plötziicher Beginn der Schmerzen fand man in30.2% der Patienten, Außerdem wurde ein stari 62). Eswurde kein Zusammenhang zwischen Depression undSchmerzen gefunden. Diese Studie zeigt, dass Patienten mitwiederhoiten systemisohen Infeiftionseriirankungen häufigerunter chronischen orofaziaien Muskeischmerzen leider.Außerdem sind diese infektiuen Erkrankungen in 67% der Fäiieder ausiösende Faktor für chronische orofaziaie Muskei-schmerzen.