Effects of Stretch-Based Progressive Relaxation Trainingon the Secretion of Salivary Immunoglobulin Ain Orofacial Pain Patients

Jeffrey J. Sherman, MSGraduate StudentDepartments of Psychoiogy and

Behaviorai Science, Orofaciai PainCenter

Coilege of Medicine

Charles R. Carlson, PhDAssociate ProfessorDepartment of Psychology andOrofacial Pain Center

James A. McCubbin. PfiDProfessorDepartment of Behavioral ScienceCoilege of Medicine

John F, Wilson, PhDAssociate ProfessorDepartment of Behaviorai ScienceCcliege of Medicine

University of KentuckyLexington, Kentucky

Correspondence to:Dr ChaHes R. CarisonDepartment of Psychology112 Kastle HaliUniversity of KentuckyLexington, Kentucky 40506-0044

There is a growing body of evidence that psychologic Stressors canaffect physical health and proneness to disease through depletionof the body's immune system. Relatively little research, however,has investigated the potential immunoenhancing effect of stress-relieving strategies such as progressive muscle relaxation. Thisstudy explored the relationship between immune functioning andrelaxation training with persons experiencing persistent facialpain, ¡n a single experimental session, 21 subjects either receivedrelaxation training or rested for an equivalent time period.Salivary immunoglobulin A, mood, pain, and tension levels weremeasured before and after relaxation and rest periods. Resultsindicated that a greater proportion of those receiving relaxationtraining had increases in secretion of salivary immunoglobulm A.Tbese findings suggest that immunoenhancement may be anotherpotential benefit of progressive relaxation training for personswith chronic pain conditions.J OROFACIAI, PAIN 1997;11:11.5-124,

key words: facial pain, relaxation, psycboneuroimmunology,salivary immunoglobulin A

Temporomandibular disorders are cbaracterized by pain anddysfunction of tbe masticatory and associated muscles. Ingeneral, cbronic pain syndromes are powerful Stressors,^

but facial pain is a particularly stressful experience because of tberegion in whicb tbe pain is localized. Pain in tbe facial area affectsbody image, subsistence, and tbe demonstration of pleasure in theform of smiling and laughter,^ Disruptions in the normal function-ing of tbe temporomandibular joint can have a significant mflu-ence on an individual's capacité' to adapt to her/his environments.

Facial pain conditions can have profound psychologic effectsalong witb important pbysiologic changes for tbe individual. Forexample, wben compared to control subjects, facial pain patientsare more depressed,"' distressed,' and demoralized,^ In tbe labora-tory setting, facial pain patients bave exaggerated heart rate andsystolic blood pressure responses to Stressors.'' Patients also reportmore pam-related illnesses, more life tbreatening pbysical prob-lems, more recent life events involving injury, and more nonpain-related pbysical illnesses. Taken togetber, tbese findings suggesttbat facial pain may be a Stressor of significant magnitude.

The immune system is sensitive to tbe psycbosocial experienceof stress." Components of the cellular immune system are affected

Journal of Orofacial Pain 115

Sherman et al

by the death of a spouse,^ recent separation from aspouse,^ medical school examinations,^" anddepression." A number of studies have demon-strated rhat secretory immunoglobuiin A (IgA) isalso sensitive to environmenral and laboratoryStressors. Secretory IgA is an important part of thebody's humotal immtine system. It is tbe predomi-nant immunoglobuiin found in body fluids, suchas rears, nasal secretions, saliva, milk, and gas-trointestinal secretions. Since it is most prevalentin surfaces tbat ate exposed to the external envi-ronment, it prevents penetration of microorgan-isms and acts as the front line of defense againstupper respirarory tract infections.'-"'''

As an important component of humoral immu-nity, the relationship between salivary IgA (S-IgA)and acute Stressors has heen investigated in severalstudies. Salivary IgA has been found ro decreaseafrer rhe stress produced hy college exams' "^^ anddaily hassles,"* Salivary IgA is also influenced hylevels of social support,'^ desirable and undesir-able life events,''' and positive mood states.''-''Furthermore, studies have shown that lower levelsof S-IgA are associated with highet incidence of ill-ness.'--^^'-'-^ These results suggest that S-IgA mayhe a useful measure of immune system functioningthat is sensitive to changes in the immediate psy-chologic environment.

Given the relationship between the experience ofstress, life events, and tbe immune system. Mar-bach et al- examined the cellular immune function-ing of facial patn patients. Results indicated thatlymphocyte activtty in response to concanavalin A(GonA) and pokcweed mitogen (PWM) wasdecreased in pain patients who had higher levels ofdistress and lower self-esteem, ln addition, thelevel of pam intensity reported by the patients wasinversely related to ConA tesponse. Although therewere no differences between pain patients and con-trol subjects in natural killer cell activity, T-lym-phocyte concentrations, and B cell counts, theauthors concluded that demoralization and level ofpain intensity ate associated wirh diminished invitro immune responses in facial pain patients.

Facial patn patients comptise a chronicallystressed population with high levels of depression,anxiety, helplessness, and life events. If these con-ditions are associated with an immunocompro-mised state, then patients cutrently in pam may beat increased risk for further diseases. For this rea-son, Schleifer et al-^ suggested rhat chronic facialpain patients provide a useful model for the studyof psychoimm uno logic processes. As Marhach andcolleagues^"^''' describe patients with temporo-mandibuiar disorders, rhey "are usually distressed

individuals wbo are beleaguered by recent physi-cal illnesses and injuries as well as by pam, whotend ro attribute tbeir fare ro external facrors, andwbo have fewer of some important type of socialsupport." This group of parients malees for anideal population in which to study immune pro-cesses,

A number of studies have demonstrated the use-fulness of various forms of progressive relaxationfor improving the condition of chronic painpatients. '*"^^ Many of these interventions havefocused on relaxation procedures that involvetensing and relaxing various muscle groups.However, tension-based relaxation procedures cansometimes increase pain and inhibit relaxation,^*For these reasons, an alternative muscle relaxationprocedure focusing on gentle muscle stretchingexercises was developed. Several experimentalstudies have demonsrtated the efficacy of astretch-based relaxation approach with musclepain patients.^^--^ Based on the results and effi-cacy of earlier controlled studies on this patientpopulation, the stretch-based relaxation approachwas cbosen for use in the present study,

A number of studies have demonstrated rhe use-fulness of relaxation training for enhancing eithercellular or humoral immune functioning in theelderly,^" in adult and pédiatrie cancer pa-tients,^''^^ in asymptomatic subjects in acutelystressful situations such as medical school exams, ^and in undergraduate students, "'• '' Thus, in addi-tion to relieving stress and pain, progressive relax-ation training may have other important benefitssuch as immunoenhancement. However, theauthors of the ptesent study were not able to findany publisbed studies that examined this relation-ship in facial pain patients.

This study investigated the potential ini-munoenhancing effects of a stretch-based progres-sive relaxation (SBPR) intervention, Ghanges in S-IgA secretion tares in patients undergoingrelaxation training were compared with changes inthose who rested quietly. It was hypothesized thatrhe relaxation intervention would increase S-IgAsecretion rates, improve mood, and decrease pain,as compared to resting quietly in a telaxed position.

Materials and Methods

Participants

Eighteen women and three men recruited from theOrofacial Pain Center at the University of Ken-tucky Dental School participated in the study.

116 Volume 11, Number 2, 1997

Sherman et al

Additionally, six patients refused ro participarebecause of time and travel constrainrs. Exclusioncriteria were (1) inflammation of rhe mouth orbody; (2> infections, tumors, or degenerative jointdisease; (3) history of alcohol or other substanceabuse; (4) recent dental surgery, invasive dentalwork, or open sores in the mouth; and (S) treat-ment with medications likely to have importantimmune or psychologic effects, such as antibiotics,corticosteroids, antidepressanrs, and hormonalagents. All participants had httle to no experiencewith relaxation training. Participants gaveinformed consent after the procedures and volun-tary nature of the experiment had been explained.This research was approved by the InstitutionalReview Board of the University of Kentucky fortbe protection of human subjects.

Dependent Measures

The secretion rates of S-lgA were measured using asingle radial immunodiffusion method. The assayhad a sensitivity range of 6.25 to 100.00 mg/dL.Because S-IgA concentration is influenced by rheflow rate of saliva, and relaxation stimulates sali-vary flow rate through parasympathetic acti-vation,^'' samples were timed and analyses wereconducted on both concentration level and secre-tion rate. Collected samples ranged in concentra-tion from 6.39 ro 93.58 mg/dL.

Pain severiry, emotion, and tension ratings weremeasured at rwo saliva collection periods. Painseverity was measured using the short form of tbeMcGill Pain Questionnaire (MPQ-SF). The MPQ-SF Is highly correlated wirh the original MPQ.^^ Itmeasures sensory and affective pain components aswell as intensity of pain using a present pain inten-sity' index and visual analog scale. Mood was mea-sured using the Emotion Assessment Scale (EAS),wbicb measures eight dimensions of emotionalresponding (sadness, happiness, fear, anger, guilt,anxiety, surprise, and disgusr). Tbe measure con-tains 24 visual analog scale items and bas a split-half reliabiliry of .94.^^ Muscle tension was mea-sured using tbe Tension Mannequin Scale (TMS).Tbis scale measures self-reports of muscle tensionm each of 19 muscle sites using a 100-mm visualanalog scale. ^

Procedures

Participants were randomly assigned to one oftwo groups. Those in group 1 received relaxationtraining, and those in group 2 rested quietly in areclined position.

Because the effects of a variery of variables, suchas exercise, use of medication and alcohol, food,and sleep, can all potentially affect S-IgA secretionrates, all participants refrained from exercise, medi-cation, and alcohol tise for 24 hours prior to partic-ipation. Participants wete also told to avoid brush-ing their teeth and eating or drinking anythingother than water for 2 hours prior to participatingin the srudy.

All parricipants washed rheir mouths with waterbefore beginning the experiment. Participantsrested quietly in a seated position on a recliner for5 minutes while acclimating ro the experimenralsetting and equipment. They then completed theMPQ-SF, EAS, and TMS, After completing theseinsrrumenrs, they were asked to swallow any salivathey had in rheir mouths, and they began expecto-rating into a 25-mL scintillation tube for 2 minutes.This was the basehne (time 1) sample. This samplewas then refrigerated before subsequent weighing,centrifuging, and freezing.

After expectorating, group 1 participants wereguided through a 20- to 30-minute stretch-basedrelaxation procedure according ro instructions byCarlson and Collins.'* All procedures were identi-cal for participants in group 2, except they weretold to recline in the chair, find a comfortable posi-tion for themselves and rest for approximately 25minutes.

Ar time 2, participants expectorated for a secondrime following the same guidelines as before. Fol-lowing this 2-minute time period, they completedthe EAS, MPQ-SF, and the TMS scale. The samemale experimenter was presenr rhroughout bothtbe relaxation and resting conditions. While the re-laxation training required that tbe experimenterread instructions throughout the procedure, he wasseated out of the patient's view and there was notwo-way interaction between tbe experimenter andthe participant during the procedure. In the restingcondition, the experimenter sar quietly in the roomour of tbe patient's view, and rhere was no interac-tion between the experimenrer and rhe participant.

Immunoassay

After collection, saliva samples were weighed andcentrifuged at 2,000 rotations per minute 1656g)for at least 15 minutes. The supernatant was pipet-ted from the sample and stored in a manual defrostfreezer for no mote than 1 month. Before assaying,samples were defrosted and then inverted severaltimes to mix rhe contents.

Five microlirers of each sample were micropipet-ted Into individual wells in a single radial immuno-

Joumal of Orofscial Pain 117

Sherman et al

diffusion plate (Kent Labs, Redmotid, WA) coti-taitiJng secretory IgA antibodies and standards. Sixreference samples of known dilutions (100.00,50,00, 25.00, 12.50, and 6.25 tng/dL) weretnicropipetted into six wells to serve as referenceconcentrations. Plates were then incubated upsidedown at room temperature (23°C ± 2''C] unrii end-point diffusion (approximately 48 bours). End-point diffusion is marked by tbe appearance of aprecipitin ring around each well. Tbe diatneter ofeatb ting was then measured to tbe nearest balfmillimeter with a gridded magnifying plate reader.Each ring was measured by two readers to ensurereliability.

The collection technique involves centrifugingentire saliva samples, aliquoting the supernatants,discarding tbe sediment, then freezing tbe super-natant until samples could be defrosted and de-posited into the wells for reading. Little is knownabout tbe effects of freezing S-IgA over extendedperiods of time. To clarify this question, four sam-ples from subjects and two reliabihty samples ofknown concentrations were defrosted and assayed 1month and 2 months after freezing. Tbe intraclasscorrelation'"* for these concentrations was R = ,40.The mean level of S-IgA for samples frozen for 1month was 39,9 mg/dL (standard deviation [SD]10,5). Tbe mean level of S-IgA for samples frozenfor 2 montbs was 34,5 mg/dL (SD 9.1). Tbe differ-ence between these means was not statistically sig-nificant (i[5j = 1.36, P > .05), wbicb suggests tbatsamples did not significantly change over cbemontb-long freezing period.

Statistical Analysis

All statistical analyses were carried out using tbeStatistical Analysis System,'' and t tests were per-formed on all baseline measures, such as age, emo-tional state, anxiety, pain levels, and baseline S-IgA levels, to confirm initial equivalence betweengroups, Cbi square analyses were performed onthe number of participants in eacb group who badincreases versus decreases in S-IgA. A nonparamet-ric test was cbosen to reflect proportions of partic-ipants wirb increases in S-lgA.

Results

Reliability and Validity of the S-IgA Measure

Initial analyses were conducted to determine thereliability and validity of tbe assay procedures.Two observers read 74 precipitin ring diameters

during rbe course of tbe study, Tbesc readingsincluded participant samples {n = 40| .nid stan-dards from colostrum (n = 34). Tbe mti .n-lass cor-relation for interrarer reliability'"' of thcsr readingswas R = ,998, Tbis measure of interratet reliabilityindicated tbat ring diameter reading was bigblyconsistent from one rater to tbe next.

To test tbe intraplate reliability of the measure,seven standards of known dilutions were depositedinto two different wells on cacb of seven platesused for tbe study. Similarly, nine saliva sampleswere deposited into two wells on eacb plate. Theintraclass correlation between the ring diameterreadings is a measure of agreement of predicted S-IgA concentration for the same plate. Reliability oftbese readings was R - .996, To calculate a mea-sure for interplate reliability, nine samples weredeposited into wells on different plates. Tbe intra-class correlation between these readings was R =.948. Tbis level of intraplate and interplate agree-ment between values suggested minimal variabilitywitbin and between plates and negligible measure-ment error for tbe tecbnique itself.

Previous researcb suggests tbat S-IgA concentra-tion is negatively associated witb salivary flowrate. Therefore, studies using tbis method mustcorrect for volume of saliva produced,''^ Tbe rela-tionsbips between concentration and weigbt ofsaliva sample at times 1 and 2, were r(18) = -.40,F < ,1 , and )'[18) = -.48, P < .05, respectively.These findings cunfirtn tbe importance of control-ling for salivary flow rate by deriving S-IgA secre-tion rates as a function of salivary flow. Secretionrates more accurately reflect S-IgA availability insaliva because they account for tbe relationshipbetween S-IgA concentration and saliva volume,Tbe correction for secretion rates was accom-plished by multiplying S-IgA concentration in mil-ligrams per tnillihter by ibe volume produced petminute in millilirers per minute. All subsequentanalyses are performed on corrected S-IgA secre-tion rates reported as milligrams per minute.

Descriptive Population Statistics

Twenty-one participants (3 men, 18 women) com-pleted the study. One participant was unable toexpectorate at rime 2, so her S-IgA data were notincluded in tbe analyses. Consistent witb tbe litera-ture," sex differences in S-IgA values were not sta-tistically significant {t[lQ] = ,78, P > .05). Mean S-IgA secretion rates for men and women were 0.32mg/min and 0.24 mg/min, respectively. Initial val-ues for S-IgA, and baseline values for age, duration

118 Volume 11, Numbet2, 1997

Sherman et al

Table 1 Means and Standard Deviations of Etnotional State, Pain, and MuscleTension Levels at Baseline for Relaxation ¡n = 10) and Rest (n = 10) Groups

Variable

S-lgA img/min)Age

Pain duration (months)EAS (mm)

FearDisgustAngerAnxietySadnessHappinessSurpriseGuilt

MPQ.SF (mm)SensoryAffectPresent pain intensity indexVisuai analog scaie

Muscletension (mm)

Relaxation

Mean

.21030

8

5.16.48.6

20.010.116.6

9.38.4

4.80.91 4

26

164

SD

.1608.48.4

4.88.9

10.817.8

9.913.910.214.2

4.61.61.Î

19.6

12.3

Rest

Mean

.2123613

5.7

8.57,4

27 01011979.4

3 9

6 90 81 3

35

12 7

SD

.14214.410.9

4.65.77 5

22 15 8

18412.43.5

7.72.10.6

22 6

90

F

0 0 11.46i.06

0,090,460,090 630 000 190 00t 02

0 570,010,100,91

0 65

of pam, emotional state, and pain levels are shovi'nin Table 1. No statistically significant differenceswere found between groups ior any of the mea-sured variables at baseline. Mean duration of pamwas 11 months, with a range of 7 days to 36months. Twelve participants had primary diag-noses of masticatory or cervical myofascial pain.Other primary diagnoses included disc displace-ment with reduction (four participants), chronictension tj'pe headache (two participants), capsulitis[two participants], and osteoarthritis (one partici-pant).

Salivary IgA Changes

Figure 1 shows the mean S-lgA values at both col-lection periods. To determine whether relaxationeffectively increases S-lgA, individual changes in S-IgA fur both groups were examined. Figures 2 and3 show changes in S-IgA for all participants atboth collection periods. Figure 2 reveals that S-IgAsecretion rate mcreased for 9 of the 10 subjectswho received relaxation training. Figure 3 revealsthat 6 of the 10 participants who rested for anequivalent time period had increases in S-IgA.Pearson's chi-square statistic was used to test thehypothesis that these proportions were equal.Results indicate (x^[l,0.05] ^ 6.8, P < .05) thatthese proportions are tiot equal, and it wasunlikely that this pattern of frequencies wouldoccur by chance.

— Relaxation

Fig 1 Salivary IgA secretion rates for relaxation (n = 10)and rest (n - 10) groups.

Journal of Orofacial Pain 119

Sherman et al

I 0,8

Fig 2 Individual changes in S-IgA for relaxation group Fig 3 Individual changes in S-IgA for rest group (n .10).

Salivary Flow Rate

Auronomic nervous system acrivity can cbangesahvary flow rate because increased sympatheticnervous system activity tends to decrease salivaproduction. In turn, volume of saliva can affectconcentration levels of S-IgA. Saliva flow for thoseparticipants receiving relaxation training betweenrimes 1 and 2 (group 1) increased a mean of 0.15mL/minute, from 0.63 mL/minute to 0,78inL/minute. For those who rested, saliva flowincreased 0.08 mL/minute, from 0,63 mL/minuteto 0,71 mL/minute. Adjusting for initial valueswitb analysis of covariance (ANCOVA), tbe differ-ence between groups at time 2 was not statisticallysignificant (P[l,16] = 0,63, P > .05). This suggeststhat the intervention did not increase saliva flowsubstantially more than the rest period.

Self-Ratings of Emotion, Muscle Tension, andPain Measijres

To determine wbetber the relaxation training wasmore effective tban rest for improving self-ratingsof muscle tension, pain, and emotion, ANCOVAwas conducted on time 2 data after adjusting forbaseline differences. No statistically significantgroup differences were apparent for the pain ormuscle tension measures. Analyses on the EASscores indicated tbat initial values for sadness sig-nificantly predicted levels at time 2 (/'[1,16] -7,68, P < ,0.5), The effect of group membersbip onsadness, adjusting for initial levels, was F{\,16) =7.08, P < .05. Tbe interaction between tbe covari-ate and the group effect was not statistically signif-icant (f[l,16] = 3,09, P > .05), Tbese results indi-

cated tbat tbe SBPR group bad lower levels of self-reported sadness tban tbose wbo rested. No othermeasured emotional differences were apparent.

To further explore tbese findings, the relation-sbip between S-!gA and all related variables wasexamined. Tbe S-IgA values were averaged andcorrelated with age; the change in S-IgA because ofrelaxation; and initial levels of emotion, pain, andtension. Salivary IgA was negatively correlatedwitb age {r[18] = -.43, P < .06), initial levels offear (r[18] = -.43, P < .06), and anger (r[18] =-.39, P<.OS).

Discussion

The results supported the hypothesis tbat a stretcb-based progressive relaxation intervention mayincrease salivary IgA secretion rates in facial painpatients. A greater proportion of tbose receivingrelaxation training bad increases in S-IgA whencompared to tbose who rested quietly. In addition,tbose receiving the relaxation training reportedlow cr ratings of sadne.ss than did resting controls.These findings demonstrate two potential benefitsof relaxation training for the management ofcbronic pain, namely tbe enhancement of humoralimmune function and reductions in self-reports ofsadness.

Because S-IgA helps to prevent patbogen im-plantation in mucous membranes,**^ increases in S-IgA could be associated witb reduced susceptibilityto upper respiratory tract infections. 'Ihese find-ings may be of particular importance for tbe facialpain patient. Marbacb and colleagues-* demon-strated tbat facial pain patients have more pain-

120 Volume 11, Number 2, 1997

Sherman et al

and nonpain-related illnesses than do asymp-tomatic subjects. By enhancing the secretion rateof S-IgA, SBPR may not only reduce their risk forupper respiratory tract infections, but also forother diseases of the otal cavity. Previous findingshave indicated that lowered S-IgA levels are associ-ated with an increase in rhe incidence of dentalcaries''"' and periodontal disease.'*'' The use of theSBPR approach may influence physical health onseveral important dimensions of health and psy-chosocial functioning.

The rapid effect of the relaxation was apparentby the greater proportion of increases in S-IgAsecretion for the relaxation group after a brief, 25-minute intervention, but how long can this effecthe mainrained? McClelland et al'^ found that S-IgA concentration was higher for students immedi-ately after a stressful examination than on morerelaxed days. These authors also found an equiva-lent, significant drop in S-IgA within 2 houts of theStressor. Similarly, Dillon et al-" found increases inS-IgA concentration after subjects viewed a hu-morous video, but increased levels declined withinminutes after viewing. In agreement with ourresults. Green et al''' found tmmediate increases inS-IgA with a 20-niinutc relaxation intervention in agroup of students. They also found that theseiticreases in S-IgA were maintained during a 3-week period of daily relaxation practice so thateven before relaxation on the twenty-second dayof practice, initial levels of S-IgA were greater thantwice that of initial levels before relaxation on thefirst day. Longitudinal studies over extended peri-ods of time would be necessary to clarify the main-tenance of S-IgA changes resulting from relaxationamong chtonic facial pam patients.

It is interesting to note that the obsetved ln-ctease in S-IgA during relaxation occurred with aconcomitant decrease in level of sadness. Althoughstudies have found that S-IgA is related to ade-quacy of social suppott'^ and loneliness,'''' themechanisms underlying these relationships and theeffects of relaxation are poorly understood. It ispossible that other emotional variables such assadness or depression may mediate the relationshipbetween immunologie changes and relaxation.Similarly, it is possible rhat otber variables notadequately accounted for, other than the relax-ation intervention, may have influenced S-IgAsecretion. For example, the impact of sitting in aprivate room with a male therapist may have haddifferential effects on participants based on theirpersonality characteristics and psychosocia! his-tory. While for some, the experience of interactingwith an experimenter may have heen pleasant.

challenging, and socially suppottive, others mayhave perceived the experience as an anxiety-pro-voking test. The expetience of either teceivingrelaxation training or resting quietly, and expecto-rating into a tuhe, may have very different emo-rional connotations for different subjects. Thishighlights the need to futther research the mecha-nisms underlying these telationships and is of par-ticular relevance to those experiencing chronicpam who may have a higher incidence of depres-sion and the perception of less than adequate levelsof social support.-'

Many factors affect the secrerion rate and com-position of saliva,'"' Stressful situations teduce sah-vary fiow through sympathetic activation,'^ whilerelaxation increases flow rate.'^''" All else beingequal, increased flow of sahva would decrease S-igA concentration.'^ As expected in both groups,S-IgA concentration was inversely related to sali-vary flow, which suggests the need for adjustingfor salivary flow and using secretion rates rarherrhan concenttarion values. However, although theintervention increased salivary flow a minimal0.15 mL/minute, concentration rates still increasedfor those subjects receiving the intervention.Furthermore, since the intervention did not signifi-cantly change flow täte in the telaxation group,our results cannot be accounted for by changes insalivary flow alone.

Some important considerations for futute stud-ies were revealed during a dehriefing of suhjectsshortly after participation was complete. Severalrelaxation subjects commented that learning theprocedures for the first time and concentrating ontbe stretches made it difficult to "feel telaxed," Itis not uncommon in tbe clinical setting to see diffi-culties in achieving a truly relaxed and comfort-able state in a patient's firsr arrempr at telaxation.The stretches require rhe patienr to follow simpleinstructions and are genetally easy to follow, butthey tequire some concenttation and practicebefore patients feel ptoficient in their performance.Thus, future studies should emphasize additionalsessions and practice training until sub|ects feeladept at the procedures. Regardless, even with thedifficulties associated with learnmg a new tech-nique, subjects receiving relaxation training for thefirst time still had Increases in S-IgA secretion rate.

In addition, some subjects in the control groupalso commented about the botedom that set induring the rest period. Although it is possihle thatdifferences between groups may have resultedfrom the tedium of resting and not from benefitsof the training, tbis explanation is not supportedby the evidence from other relaxation studies. For

Journal of Orofacial Pain 121

Sherman et al

example, Green and Green" compared relaxationwith massage, touching, visualization, and lyingquietly. They found that relaxation increasedsecretion of S-IgA more than the alternative con-trol procedures. Thus, the evidence would suggestthat some aspect of the relaxation training influ-enced S-IgA. It is possible that this aspect mayhave been the socially supportive role of the thera-pist, and such hypotheses merit further explo-ration in future studies.

The general procedure for assay of S-IgA usedin the present study (single radial immunodiffu-sion) has been used in much of the psychoneu-roimmunology literature. Tbe plates used in ourstudy, however, were made specifically for secre-tory IgA detection (Kent Labs) and used secretoryigA antibody and standards rather than antibodyand standards from serum. Previous stud-|ç5is,2i,34,4s using serum IgA standards requiredthat obtained concentration values be correctedfor the differences in sedimenration coefficientsberween serum and salivary IgA. Using secretory-specific antibodies and standards required no con-version from serum to secretory standards. Highinterrater reliability demonstrates the ease andaccuracy with which the precipitin rings can hemeasured. Moreover, high interplate andinrraplate reliabilities also showed that the mea-sure is consistent and accurate.

It is important to note these results do not sug-gest a relationship between relaxation and theproduction of serum anribodies because such aprocess takes days as opposed to minutes. We didnot measure a change in tbe number of antibodiesactually produced. Rather, we observed changesin antibodies released and available in saliva.Results suggest tbat relaxation resulted in moreincreases in S-IgA secretion than did sitting qui-etly. The magnitude of this change and the chnicalsignificance of such change require further study.

Another important consideration when review-ing these results is the association between lowpower and small sample size. The low power asso-ciated with a smail sample generally results m abias toward null findings. Given the small samplesize in rhis study, the likelihood of finding thehypothesized effects was low. Despite the lowpower, however, statistically significant effectsand relationships did emerge in the predicteddirection.

Schwartz"" ' ) stated tbat "measuring what iseasy to measure in studying rhe effects of psy-chosocial factors [on immunity) may not be clini-cally relevant." In detecting an effect between S-IgA and relaxation, we must be cautious to avoid

overinterpreting the restilts. The present findingsdo nor allow for broad statemenrs about improve-ments in health or well-being as a rtsiilt of prac-ticing relaxation. However, one can cc tbe acutebenefits of a srrerch-based relaxation approach forimproving mood, and increasing S-IgA secretionrate in chronic orofacial pain patients. Theseresults demonstrate tbat relaxation may at leastoffer a brief respite from the chronic stress thatlikely contributes to the psychologic and physicaldistress that many of these pain patients report.

Acknowledgments

This sludy wab supported by a University of KentuckyFouiidariori Grant awarded to Dr Charles R, Carlson.

References

1. Okeson JP. Bell's ürofacial Pains, cd S. Chicago:Quintessence, 1995.

2. Marbach JJ, Schleifer SJ, Keller SE. Facial pain, distress,and immune functioning. Brain Behav Immunity 19904-243-254.

3. Carlson CR, Bertrand PM, Okeson JP. The role of behav-ioral medicine in rhe treatment of chronic temporoman-dibular disorders: Neuromuscular re-education in rhe den-tal setting. In: Kail R, Shtark M, Sokhad?,e T (eds). TheEncyclopedia of Applied Psychophysiologj' and BehavioralMedicine. Bensalen, PA: Future Health Press, 1997 (inpress),

4. Marbach JJ, Lund P. Depression, anhedoma, and anxiet)'in tempormaiidihular joint and other facial pain syn-dromes. Pain 1981;11:73-84.

5. Marbach JJ, Lennon MC, Dohrenwend BP. Candidate riskfactors for temporomandibular pain and dysfunction syn-drome: Psychosocial, health behavior, physical illness andinjury. Pain ]988;34:139-151.

6. Carlson CR, Okeson JP, Falace DA, Nitz AJ, Curran S,Anderson D. Comparison of psychologic and physiologicfunctioning hetween patients with masticatory muscle painand matched controls. J Orofacial Pain 1993;7:Ii-22,

7. Ader R. Psychoneuroimmunology. New York: AcademicPress, 1981.

8. Schleifer SJ, Keller SE, Camerino M, Thornton JL, Stein M.Suppression of lymphocyte stimulation following bereave-ment. J Am Med Assoc 19S4;250J 74-378.

9. Kiecolt-Glaser JK, Fisher LD, Ogrocki P, Stout JC,Speicher CE, Glaser R, Marital quality, marital disruption,and immune functioning. Psychosom Med I987;49i13-34.

10, Kiecolt-Glaser JK, Garner W, Speicher CE. Penn G, GlaserR, Psychosocial modifiers of im mu no competence in medi-cal students. Psychosom Med 19S4;46:7-14.

11, Schleifer SJ, Keller SE, Bond RN, Cohen J, Stein M. Majordepressive disorder and immunity. Arch Gen Psychiarry

12. Jemmot JB, McClelland DC. Secretory IgA as a measure ofresistance to infectious disease: Comments on Stone, Can,Valdimarsdottir and Neale. Behav Med 1989-12'63-71.

122 Volume 11, Number 2, 1997

Slierman et al

13, Rossen RD, Butler WT, Waldman RH, The prorein innasal secretions, JAMA 1970;211:1157-116 I,

14, Yodfat Y, Sylvian H. A prospective study of acute upperrespiratory infections among children in a kibbutz, JInfect Dis I977;3é:26-3O.

15, Jemniort JB, Borysetiko JZ, Borysenko M, McClelland D,Chapman R, Meyer D, Benson M, Academic stress,power motivation, and decrease in secretion rate of sali-vary secretory imunnoglubuÜn A, Lancet19S3;25:1400-1402,

16, Jemmott JB, Magloire K. Academic stress, social support,and secretory imruunoglobulin A, J Pcrs Sue Psycliol1988;55:803-810.

17, McClelland DC, Ross G, Patel V, The effect of an aca-demic e.\aminatiün on saliv:iry norepinephtine andimmunoglobulin levels, J Human Srress 1985;11(21:Í2-

18, Martin RA, Dubbin JP, Sense of hunior, hassles andimmunciglobulin A: Evidence for a stress moderatingeffecr of humor, Int J Psychiatry Med 1988;18:93-105,

19, Evans P, Bristow M, Hucklebridge F, Clow A, Walters N,The relationship between secretory immunity, mood andhfe events, Br J Clin Psychology 1993;32:227-236,

20, Dillon KM, Minchoff B, Baker KH, Positive emotionalstates and enhancement of the immune system, Inr JPsychiatry Med 19S5;15:13-18,

21, McClelland DC, Alexander C, Marks E, The need forpower, stress, immune function and illness among maleprisoners, J Abnorm Psychology 1982;91:él-70,

11. McClelland DC, Floor E, Davidson R, Saron C, Srressedpower motivation, sympathetic activation, immune func-tion and illness. J Human Srress 1!I8O;6(2¡:11-19,

23, Schleifer SJ, Marbach J, Keller SE, Psychoneuroimmu-nology: Potential relevance to chronic facial pain, AnesdiProg 1990;34:93-98,

24, Crockett DJ, Foremen ME, Alden L, Blasberg B. A com-parison of treatment modes in the management ofmyofascial pain dysfunction syndrome, Biofeedback SelfRegul 1986;11:279-291,

25, Rider MS, Achterberg J, Lawlis F, Govern A, Toledo R,Butler JR, Effect of immune system imagery on secretoryIgA. Biofeedback Self Regul 1990;15:317-333,

26, Turk CD, Meichenbaum D, Genest N, Pain andBebaviural Medicine: A Cognitive Behavioral Perspective,NewYork:Guilford, 1983,

27, Carlson CR, Okesun JP, Falace DA, Nitz AJ, AndersonD, Stretch-based relaxation and rhe reduction of EMGactivity among masticatory muscle pain patientb. JCraniomandib Disord Facial Oral Pain 1991;5:205-212,

28, Carlson CR, VenTrella MA, Sturgis ET, Relaxation train-ing through muscle stretching procedures: A pilot case. JBehav Ther Exp Psychiatry 1987;18:121-126,

29, Kay JA, Carlson CR, The role of stretch based relaxationin rhe rreatmenl of chronic neck tension, Behav Ther1992;23:423-431,

30, Kiecok-Glasifr JK, Glaser R, Willinger D, Stone J, MessickG, Shippard A, et al, PsychosocisI enhancement ofimmunocumpeteiicc in a geriatric population. HealthPsychology 1985;4:25-41,

31, Fawïy FI, Immune effects of a short term intervention forcancer patients. Advances 1990; 10:32-33,

32, Olness K, Culbert T, Udcn D, Self regulation of salivaryimmunoglobulin A by children. Pediatrics 1989;83:6Ê-71,

33, Kiecok-Glaser JK, Glaser R, Psychological influences onimmunity. Psycbosomatics 1986^27:621-624,

iA. Green ML, Green RG, Santoro W, Daily relaxation modi-fies serum and salivary immunoglobulins and psychophys-iologic symptom severity, Biofeedback Self Regul 1988;13:187-199,

35, Murse DR, Schacterle GR, Esposite JV, Chod SD, FürstML, DiPonziano J. Zaydenberg M, Stress, meditation andsaliva: A study of separare salivary gland secretion inendodontic patients, J Oral Med 1983;38:15O-]59,

.îli, Melzack R, The short form of the McGill Pain Question-naire, Pain I987;3O;191-197,

37, Carlson CR, Collins FL, Porzelius J, Stewarr JE, Nitz AJ,Lind C, The assessment of emotional reactivity, J Psycho-pathol Behav Assessment I989;11J13-325,

3S, Webster JS, Ahles TA, Thompson JK, Raczynski JM, Theassessment of objective tension levels among several mus-cle groups: The tension mannequin, J Behav Ther ExpPsycbiatry 1984;15;323-323,

39, Carlson CR, Collins FL, Stretch Relaxation: A TrainingManual, Wheaton, IL: 1989,

40, Fleiss JL, The Design and Analysis of ClinicalExperiments, New York: Wiley & Sons, 1980,

41, SAS User's Guide: Statistics, version 5, Cary, NC: SASInsriwce, 1985,

42, Stone AA, Cox DS, Valdimarsdottir H, Nealc JM.Sccretory-lgA as a measure of lmmunocompetence, JHuman Stress 19S7;13:136-140,

43, Toma5i TB, The Immune System of Secretions, EnglewoodCliffs, NJ:Prencice Hall, 1976,

44, Lehner T, Cardwell JE, Clarry ED, Immunoglobulins insaliva and serum in dental caries. Lancet 1 9 6 7 Í 1 Í 1 2 9 4 -1297,

45, Genco RJ, Evans RT, Ellison SA, Dental researcb inmicrobiology with emphasis on periodoncal disease, J AmDentAssoc 1960;78:1016-1036,

46, Mandel ID. Wocman S, The salivary secretion in healthand disease. Oral Sci Rev 1976;4:2i^7,

47, Feldman H, Sahvary buffer capacity, pH and stress. AmSoc Psychosom Dem Med 1974;21:25-30,

48, Green ML, Green RG, Relaxation increases salivaryimmunoglobulin A, Psycbol Rep 1987;6]:623-629.

49, Schwartz CE, Introduction: Old methodological chal-lenges and new mind body links in psycho ne uro immunol-ogy. Advances 1994;10:4-7,

Journal of Otofacial Pain 123

Sherman et al

Resumen

La Relación Entre el Relajamiento y los Niveles deSaliva-lmmunoglobulira A en Personas con DoloresFaciales Persistentes

Existe una creciente evidencia que indica que ei estréssicológico puede afectar ia saiud y ia tendencia a la enfer-medad a través dei desgaste dei sis terna inmune,Reiativamerte pocos estudios han investigado ta capacidad deaiivio y fortalecimiento del sistema inmjne de las estrategiaspara aliviar ei estrés, como la técnica dei reiajamienlo progre.sivo musciiiar. Este estudio explora la relación entre ei fiín.cionamierto del sistema inmune y el reiajamiento en personascon dolores faciales persistentes. En una sola sesión expéri-mentai, 25 pacientes recibieron entrenamiento de reiajación, opor ei mismo periodo de tiempo soio descansaron. Se midieronios niveles de Saiiva-lnmiinogiobiiiina A (S-igA), disposición,dolor, y tension, antes y después dei relajamiento o de los peri-odos de solo descanso. Los resultados indicaron que una granproporción de aquelios que recibieron ei entrenamiento parareiajamiento tuvieron incrementos en ia secreción de S-igA.Estos resuitadOE sugieren que el fortalecimiento dei sistemainmune pjede ser otro beneficio de la técnica de relajamientoprogresivo para las personas qje padecen de doior crónico.

Zusammenfassung

Ausvuirkungen des auf Dehnung beruhendí^ri progres-siven Entspsnnurgstrainings auf die Elc-kretion desSpeichel-Immunoglobulin A bei Patienten mit orofazialenSchmerzen

Es existieren immer mehr Beweise dafür, dass psychologischeStressoren die körperliche Gesundheit und dieErkrankungsneigung durch eine Schwächung desKórperimmunsysterris beeinflussen können. Dagegen ist relativwenig erforscht über die mögliche rmmunoerhöhende Wirkungvon stressabbauenden Strategien wie die zunehmendeMuskelentspannung, Diese Studie untersucht die Beziehungzwischen der Imniunarbeit und Entspannungstraining beiPersonen, weiche unter andauerndem Gesichtsschmerz ieider.in einer einzigen Versuchssitïjng erhieiten 21 Leute entwederEntspannungstraining oder verblieben für die entsprechendeZeitpefiode, Speichei-Immunogiobulin A. Stimmung. Schmerzenund Spannungsziistand wurden vor und nach der Entpannungund Ruheperiode gemessen. Die Resultate weisen darauf hin,dess ein grösserer Anteii der Personen, weicheEntspannungstraining erhieiten, eine Erhöhung der Sekretionvon Speichel-immunoglobulin A aufweisen. Diese Befundelegen nahe, dass die Immunoerhöhung ein weiterer möglicherVorteil des progressiven Entspannungstrainings bei Personenmit cilronischen Schmerzzusténden sein kann.

124 Volume 11, Number 2, 1997