Pain Perception and Serum Beta-Endorphin in Trauma Patients

Pain Perception and Serum Beta-Endorphinin Trauma Patients

LAWSON BERNSTEIN, M.D., PAMELA D. GARZONE, PH.D.

THOMAS RUDY, PH.D., BRUCE KRAMER, M.D.

DWIGHT STIFF, PH.D., ANDREW PEITZMAN, M.D.

Acute traumatic injury engenders the production ofbeta-endorphin (BE) and otherendogenous opioids. Elevated BE concentration putatively correlates with pain percep­tion in trauma patients. The authors examined traumatic injury severity, pain percep­tion. and BE concentration in patients admitted to an urban trauma center. Briefrating instruments for pain and unpleasantness were administered. and blood wasdrawn for BE analysis in 48 trauma admissions and 33 age-. gender-. and race­matched control subjects for comparison. The authors found no correlation betweenseverity ofpain perception and BE. but a significant correlation was found betweenBE and patient body weight (P< 0.05), physician pain rating (P < 0.01), and InjurySeverity Score (P < 0.001). The results suggest that past findings associating traumapain perception and BE concentration are spurious.

(Psychosomatics 1995; 36:276-284)

T raumatic injury is a pervasive medical andpublic health problem that costs billions of

health care dollars each year.' Concomitantwith injury is the perception of pain, with andwithout tissue damage.2 A great variability inpain response is common in persons with simi­lar injuries, and it appears that both physiologicand psychologic factors play important roles in

Received March 12. 1993; revised May 7. 1993; ac­cepted June 23. 1993. From the Department of Psychiatry.Consultation Psychiatry Division. the Department of Phar­macy and Therapeutics. the Department of Anesthesiologyand Pain Institute. the Department ofCrilical Care Medicine.the Department of Surgery. Trauma Surgery Division. andGenetech. Inc.; all are at the University of Pillsburgh Medi­cal Center and Western Psychiatric Institute and Clinic(WPIC), Pillsburgh. PA. Address reprint requests to Dr.Bernstein. WPIC. 3811 O'Hara Street. Sle. 102. Pinsburgh.PA 15213.

Copyrighl © 1995 The Academy of PsychosomaticMedicine.

276

pain modulation.2-4 Unmodulated pain has beenshown to have a deleterious effect on patientsurgical outcome, including greater postopera­tive morbidity and increased mortality whencompared with more adequately treated controlsubjects.5-8

Psychiatric sequelae, including mood andanxiety disorders, have been noted in patientswith traumatic injury and severe pain.9-17 Re­search and clinical experience suggest that earlyaggressive pain treatment correlates with betteranalgesic efficacy's and later psychiatric out­come,IO·15-17 yet adequate pain assessment andtreatment remains an ongoing issue of debate inmanaging the trauma patient. 19.20

Acute traumatic injury engenders a varietyof physiologic changes involving the cardiovas­cular, endocrine, and inflammatory systems.6

Concomitant with the injury are the phenom­ena of pain and "stress-induced analgesia"21and neuroendocrine modulation of pain percep-

PSYCHOSOMATICS

tion. Tissue production of corticotropin-releas­ing hormone (CRH)-like substances after injuryappear to mediate stress-induced analgesiathrough increased hypothalamic production ofproopiomelanocortinin, the precursor proteinof adrenocorticotropin and beta-endorphin(BE).21.22 BE, along with other endogenousopioids, attenuates the neurophysiologic re­sponse to acute pain. 23-25 Cardiovascular col­lapse associated with hypotension parallels theacute rise in BE levels26 and can be attenuatedby opioid antagonists such as naloxone.26-31 BEreturns to baseline or subnormal levels after 5to 14 days, and BE has been correlated withincreased pain behaviors in the rat trauma-in­jury mode I.32

Transient elevation of BE also accompa­nies ethanol use,3] a frequent comorbid condi­tion with traumatic injury.34-36 In addition,obese people have greater baseline serum con­centrations of BE36-38 compared with normal­weight people.

BE serum concentrations have been used asa measure of acute pain and a measure of theefficacy of opiate analgesia in postoperative,39cancer,40.41 and pediatric bum patients.42.43 It hasbeen postulated that unmedicated acute injurypain proportionally correlates with increasedBE concentration. However, studies with hu­man and animal models of pain and opiate effi­cacy have presented varying and inconclusiveresults40.42-44 and have not controlled for otherphysiologic variables associated with BE pro­duction.

MATERIALS AND METHODS

This study was conducted in 1992 at the Uni­versity of Pittsburgh Trauma Center, wherethere are approximately 2,000 adult trauma in­jury evaluations a year. Forty-eight consecutivepatients admitted for trauma who met the inclu­sion criteria (systolic blood pressure [SBP]greater than 90 mmHg and Glasgow ComaScale [GCS] score greater than 8) were evalu­ated. The GCS is a standardized continuousrating scale to assess consciousness level, andit is an accepted measure of mental status ab-

VOLUME 36 • NUMBER 3 • MAY - JUNE 1995

Bernstein et al.

normality in the acute trauma setting.45 A GCSof 8 or higher generally indicates that a patientis able to follow commands and respond appro­priately to questions. An SBP of less than 90mmHg is accepted as clinical evidence of sig­nificant hypotension in the trauma patient,46 andfor the purpose of this study we used this stand­ard as the measure for hypotension. The lowestdocumented SBP was the standard used to de­termine significant hypotension. Patients givenmedication in the field or with significant pre­hospital hypotension were excluded from thestudy.

After initial evaluation by the trauma team,but before treatment, the patients who met theprotocol criteria were asked by the principalinvestigator/coinvestigator (LB/BK) to partici­pate in a brief (I-minute) pain assessment andto permit the investigator to withdraw a 5-ccaliquot of blood for BE measurement. Withincreasing evidence that two dimensions ofacute pain should be assessed,47 two Visual Ana­log Scales (VAS)-one for pain intensity andone for pain unpleasantness-were adminis­tered to the patient. A blinded physician VASco-rating was done by the investigator (LB/BK)before patient recruitment or review of clinicaldata, but after blood had been drawn for BEanalysis. The physician-raters relied on assess­ment of pain behaviors (e.g., grimacing) to de­rive VAS scores. The VAS is a well-validatedpain rating instrument used in a number of painstudies in trauma patients.48.49., The patientmarks a point on a 10-cm line that best reflectshis or her current level of pain. The left side ofthe line is marked "no pain" and the right "ex­treme pain." The score (0-10) is measured bythe distance in centimeters from the extreme leftof the line to the patient's mark.50

The 5-cc aliquot of blood drawn at admis­sion was placed in a vacutainer tube containingethylenediaminetetracetic acid and aprotininand immediately placed on ice. The time ofsample collection was recorded. Blood sampleswere decanted and the plasma frozen untilanalysis for BE. The elapsed time from admis­sion to assessment was approximately 10 min­utes in most cases. In most instances, when it

277

Pain and Serum Beta-Endorphin

did not interfere with clinical care (i.e., criticalillness, "stat" transfer to intensive care unit,etc.), informed consent was obtained before theinterview. Otherwise, informed consent was ob­tained after the patient was stabilized medically.This protocol was approved by the Universityof Pittsburgh Medical Center Institutional Re­view Board and is in accordance with its estab­lished guidelines on the treatment of humansubjects.

The medical record was screened for demo­graphic data such as age, gender, and weight. Inaddition, the Injury Severity Score (ISS)51 wasrecorded. The ISS is a point-system test thatyields a hierarchical rating for various injurytypes, thus allowing for comparison of injuryseverity between patients with different trau­mas. At admission, blood alcohol level and toxi­cology screen results were recorded and scoredcategorically (positive/negative) for the pri­mary analysis.

Thirty-three healthy volunteers served asthe control group. They were recruited througha local newspaper advertisement that asked forsubjects without an acute or chronic illness orcurrently on prescription medication. Patientswere also interviewed as to health status beforestudy entry. These healthy and unmedicatedvolunteers, matched to the trauma patients forage, gender, height, and race, also had a 5-ccaliquot of blood drawn, and the VAS was ad­ministered to them in the manner previouslydescribed. This ensured that both groups expe­rienced the minor trauma of phlebotomy priorto VAS rating. These blood samples were cen­trifuged, plasma decanted, and frozen untilassay.

BETA-ENDORPHIN ASSAY PROCEDURE

Plasma concentrations of immunoreactive BE(, 25I_BE) were determined with a commerciallyavailable radioimmunoassay (RIA) kit (INC­STAR Corp., Stillwater, MN). Blood sampleswere collected as described. The samples wereimmediately centrifuged for 15 minutes at760 x g (40 C), and the serum was removed andstored at -850 C until assayed. The RIA used

278

the double-antibody technique. BE was ex­tracted from plasma by passage through a col­umn containing anti-BE-coated sepharoseparticles. Bound BE was subsequently elutedfrom the column and incubated with beta-en­dorphin antiserum, followed by '25I_BE. Phaseseparation was done with a preprecipitatedcomplex of second antibody and carrier. Allsamples and standards were subsequentlycounted in an 1272 CliniGamma gammacounter with a calculation method of %B/Bo(percent bound/free) vs. log concentration.

The antibody to BE has a cross reactivityof 100% to human BE, less than 5% cross-reac­tivity to B-lipotropin, and essentially no crossreactivity to other peptides or proteins such asdynorphin, enkephalin, adrenocorticotropichormone, luteinizing hormone, or follicle­stimulating hormone. The range of quantitationfor the assay is 5-80 pmol/L, with the limit ofdetection being approximately 3 pmol/L. Theintra-assay coefficient of variation (CV) is< 10% , except at the limit of quantitation,which has a CV of 13.7%. Concentration valuesfor quality control samples analyzed with eachset of study samples consistently fell within therange specified for each kit. In addition, thecalculated value for the control group was al­ways within 10% of its theoretical value.

In addition to performing the standard vali­dation described, several other factors andplasma constituents were evaluated for theireffects on the accurate determination of plasmaBE. Because trauma patients could potentiallyhave had their injury precipitated through theuse of ethanol or other drugs of abuse, it was ofinterest to evaluate the effects of these agentson the assay. In addition, these subjects mayalso have experienced posttraumatic metabolicdisturbances that produced changes in plasmacomponents such as albumin and alpha-I acidglycoprotein (AAG), as well as changes inplasma pH, which could potentially affect accu­rate analysis of BE. These factors also requiredevaluation of their possible effects on the assay.To I-rnl aliquots of control plasma was added100 IJ.g each of morphine, o-amphetamine, co­caine, and tetrahydrocannabinol. This concen-

PSYCHOSOMATICS

Bernstein et al.

TABLE 1. Patient descriptive statistics for predic­tors of posttrauma beta-endorphinconcentration

Comparison of Beta-EndorphinConcentration in Trauma

Patients and Control Subjects

The control subjects had significantlylower posttrauma PBEC than the patients (P <0.001, Table I). The mean PBEC for the con­trol subjects and patients was $ 5 pmol/L(SO = 1.25, range =$ 5-8.6 pmol/L) and16.39 pmollL (SO =16.83, range =5 $ -66.2pmollL), respectively. However, because the

36.68 (SO =15.37), respectively, served ascontrol subjects. The control group was notsignificantly different from the patient groupwith respect to age, gender, race, and height.The patients and control subjects differed as toweight: the patients mean weight was 170.26,Ib (SO = 37.43), compared with the controlgroup's weight of 147.33 Ib (SO = 23.48) (P <0.(01).

DescriptiveStatisticVariable

Statistical Methods·

tration is well above that reported to cause tox­icity and the level that would be expected tooccur after extensive abuse of these agents.Ethanol (2.5 fll) was added to a plasma sampleto give a concentration of 200 mg percent. Fortymg of albumin was added to plasma containing4.5 g/dL to give a final concentration of8.5 g/dL, whereas 8.2 mg of AAG was added toplasma containing 7.8 mg to give a final con­centration of 160 mg/dL. Plasma pH was ad­justed by adding 25 fll of either I mollLhydrochloric acid or I mollL sodium hydroxideto a I ml sample to give final pH values of6.5 and 9.4, respectively.

None of these perturbations produced a sta­tistically significant effect on the measured con­centration of BE in a control subject's plasmasample. In addition, a 2-cycle freeze thaw of aplasma sample and allowing it to sit at roomtemperature for up to 2 hours was without ef­fect. The results of these validation studiesprove that the BE assay is sensitive and specificand that is also not easily influenced by severalphysiological factors that may be present intrauma patients.

38.7 ± 17.1

170.3 ± 37.4Chi-square tests and analysis of variancewere used to evaluate significant differencesbetween the control and patient samples. Linearregression analyses and Pearson correlationswere used to determine the association betweenposttrauma beta-endorphin concentration(PBEC) and selected predictor variables. Partialcorrelations were used to further clarify theassociation between patients' and physicians'pain ratings, injury severity, and PBEC. OnlyP values less than 0.05 were considered statis­tically significant.

RESULTS

Forty-eight patients, 28 men (mean age = 42.43years, SO =20.21), and 20 women (mean age =34.95, SO = 10.15) completed the study.Thirty-three subjects, 14 men and 19 women,with a mean age of 35.57 (SO =18.73) and

Age. years. mean ± SO

Weight. lbs. mean ± SOGender of patients, n

Men

Women

Type of trauma, %

Head injury

FallMotor vehicle accident

Assault

Amputation

Penetrating assault

Alcohol on breath. 0/0

Alcohol in blood

Loss of consciousness. 0/0Patient pain severity rating (0-10).

mean±SO

Patient pain unpleasantness rating <G-IO).mean ±SO

Physician pain unpleasantness rating (0-10)mean±SO

Injury Severity Scoremean±SO

28

20

8.3

8.3

68.38.3

4.2

2.1

10.416.2

28.3

4.2± 2.6

5.8 ± 3.1

4.6± 2.7

4.4 ± 4.2

VOLUME 36 • NUMBER 3 • MAY - JUNE 1995 279

Pain and Serum Beta-Endorphin

weight of the subjects was found to be signifi­cantly correlated with PBEC (r = 0.385,P <0.00 I), and because the control groupweighed significantly less than the patientgroup (P < 0.0 I), an analysis of covariance wasdone to test whether significant PBEC groupdifferences remained after statistically control­ling for body weight. This analysis indicated thepatient group still had a significantly higherPBEC than the control subjects after covaryingfor body weight (P < 0.01).

Predictors of PosttraumaBeta-Endorphin Concentration

Twelve variables were hypothesized as be­ing potential predictors of posttrauma PBEC.The variables and mean values are also pre­sented in Table I. Separate regression analyseswere computed to evaluate the contribution ofthese 12 variables to PBEC. The results of theregression analyses are presented in Table 2.

As shown in Table 2, the patient's weight,the physician's rating of the patient's pain se-

TABLE 2. Regression results testing predictorsof posttrauma beta·endorphine con·centration (PBEC)

StandardizedVariable Beta" rZb p

Age -0.114 0.013 NS

Weighl of palienls 0.318 0.101 <0.05

Trauma result of a -0.070 0.005 NShead injury

Alcohol on brealh -0.103 0.011 NS

Alcohol in blood -0.190 0.037 NS

Loss of consciousness 0.072 0.005 NS

Palienl pain severity 0.237 0.056 NSraling

Physician pain severily 0.394 0.155 <0.01raling

Physician pain unpleasanlness 0.197 0.0039 NSraling

Injury Severily Score 0.632 0.399 <0.001

Note: NS =nOI significan!."Measure is analogous 10 Pearson correlalions."The proporlion of variance in PBEC Ihal can be

explained by each of Ihe predictor variables(1.0 would indicale perfeci prediclion).

280

verity, and the ISS were significantly associatedwith PBEC. These three significant predictorsaccounted for 51.3% of the variance in PBEC(P < 0.001). The patients' pain severity or un­pleasantness ratings were not significant pre­dictors of PBEC.

Additional analyses were done to furtherclarify the apparent discrepancy between thepatients' and physicians' pain ratings. Results ofthese are presented in Table 3. As shown inTable 3, the patients' and physicians' pain se­verity and pain unpleasantness were signifi­cantly correlated. However, the physicians', butnot the patients' ,pain severity and pain unpleas­antness ratings were significantly associatedwith the ISS. Controlling for the significantassociation between the ISS and pain severityscore, the partial correlation between the phy­sicians' rating of pain severity and PBEC wasnot statistically significant (r =0.12, P =0.39).

DISCUSSION

The production of BE in the acute trauma set­ting is a complex physiologic event. In addition,other psychologic and pharmacologic variablesalso affect the BE response to injury. Previoushuman studies examining the BE-acute painrelationship have not taken this multivariatepathophysiologic approach. In this study, anassociation between patient pain intensityand/or pain unpleasantness perceptions and ele­vated BE could not be demonstrated.

As noted, local tissue response to injuryproduces CRH-like factors that augment physi­ologic secretion of BE from the hypothalamusand the adrenals. The positive correlation be­tween BE and body weight may be attributableto increased local tissue response to injury(more soft tissue to injure), augmented hypotha­lamic response to CRH, or increased adrenalproduction of BE in heavier persons. Becausethe hypothalamic-pituitary-adrenal axis may re­act differently in the obese,36 and because BEconcentrations are generally higher in the over­weight as compared with normal weight per­sons,37.38 these factors would seem a likelyexplanation. Likewise, obesity itself could be a

PSYCHOSOMATICS

Bernstein et al.

TABLE 3. Pearson correlations among pain ratings and Injury Severity Scores

Measure PPS PUP MDPS MDUP ISS

Patient pain severity (PPS) 1.00

Patient unpleasantness (PUP) 0.27 1.00Physician pain severity (MDPS) 0.54- 0.19 1.00

Physician unpleasantness (MDUP) 0.38- 0.46- 0.54- 1.00Injury Severity Score (ISS) 0.27 0.02 0.46- 0.41- 1.00

-P<O.OI.

risk factor for greater tissue damage caused byblunt trauma, because there is a relationshipbetween force of impact and weight. However,neither of these conclusions can be substanti­ated by our study.

When weight was removed as a potentialconfound, the positive correlation between ISSand BE remained, underscoring the robust rela­tionship between the two. This relationship isconsistent with the role of tissue trauma as theprimary injury-associated variable responsiblefor elevated BE in the trauma patients vs. con­trol subjects. Contrary to other studies, we didnot find a positive association between alcoholand elevated BE concentration, although thiseffect may be undetectable when compared tothe overwhelming stimulus of tissue injury.

Almost 50% of the variance in BE remainsunexplained in this sample, highlighting thecomplexity of BE production. The influence ofalcohol and other factors that affect BE (e.g.,head trauma, loss ofconsciousness) are difficultto separate from the cascade of physiologicevents associated with the neurohormonal re­sponse to acute injury.52 No correlation wasfound between the presence of opiates (11.4%of the sample), cocaine (18%), or marijuana( 17%) in serum toxicologies and elevated pa­tient BE. In addition, the effect of chronic vs.acute ethanol and illicit substance use is notquantifiable from this study. Although six casesexperienced transient loss of consciousness atthe scene, none had significantly elevatedPBEC levels. The presence of premorbid sig­nificant gastrointestinal (10.4%) or cardiac dis­ease (14.6%) in patients did not correlate withchanges in BE, nor did the presence of current

VOLUME 36 -NUMBER 3 -MAY - JUNE 1995

prescription drug use (35.3%). Other medicaldiagnoses or illicit substance effects could notbe assessed because of the low presence ofthese in the sample. These other conditionsmay be related to perturbations in BE physiol­ogy or premorbid pain experience, which couldaffect acute injury BE response and/or painperception.

The finding of a lack of association be­tween patient pain/unpleasantness and BE con­trasts with the studies cited before. Past studieshave suggested that elevated BE is a marker forpain39 and associated with pain perceptions.4J

One hypothesis suggests that elevated BE isassociated with undertreated acute or chronicpain42 and that adequate analgesia is reflectedin normalization of BE levels. Another hypo­thesis suggests that BE itself has autonomousanalgesic properties40 that modulate the efferentlimb of cutaneous and possibly visceral painperception. Thus, BE levels, either elevated ordepressed, should be positively or inversely re­lated to pain severity perception. Both hypothe­ses require assumptions as to cause and effectand impute mechanistic connections wherenone may exist.

This study assumes that peripheral meas­urement of BE reflects central nervous system(CNS) levels of the hormone. It is possible thatthere is a dissociated response between CNSand peripheral production of BE and that differ­ent physiological compartments reflect this in­equality. Acute CNS production of BE maysignificantly lag behind that of the adrenalgland. Moreover, the ependymal barrier be­tween the brain and the systemic circulation isdiscontinuous. One prominent locus of dis-

281

Pain and Serum Beta-Endorphin

continuity is the hypothalamic-pituitary stalk.This suggests that peripheral measurements ofBE may reflect some percentage of centrallyproduced hormone, although the total amountis not known. A dissociated response is difficultto document, however, and requires simultane­ous cerebrospinal fluid and serum assays. Stud­ies simultaneously sampling peripheral andcentral BE concentrations are inconsistent indocumenting any correlation between thetwO. 39.43.52-54

An interesting and potentially importantfinding from this study is the significant corre­lation between physician-rated pain/unpleas­antness assessment and ISS. These findingssuggest that the physicians' ratings of pain se­verity were determined, in part, by their assess­ment of the degree of trauma present, followedby an inference about how much pain the pa­tient "should be" experiencing. based on behav­ioral observations (e.g., facial grimacing). Thisreplicates and extends previous findings fromthe nontrauma literature indicating that physi­cians have great difficulty in estimating pa­tients' pain from the physical examination.55

.56

In addition, these findings underscore the com­plex relationship between injury and pain,which may encompass psychologic and situ­ational responses idiosyncratic to the patient.

References

I. Mackenzie EJ. Smith GS. Black LW. et al: Acute hospi­tal costs of trauma in the United States: implications forregionalized systems of care. J Trauma 1990; 30: 1096­1101

2. Melzack R. Wall PD. Ty TC: Acute pain in an emergencyclinic: latency of onset and descriptor patterns. Pain1982; 14:33-35

3. Peck C: Psychological factors in acute pain manage­ment. in Acute Pain Management. edited by Cousins MJ.Phillips GO. Edinburgh. Churchill-Livingstone. 1986

4. Wall P: On the relationship of anxiety to pain. Pain 1979;6:253-264

5. Craig DB: Postoperative recovery of pulmonary func­tion. Anesth Analg 1981; 60:46-51

6. Kehlet H: The stress response to analgesia and surgery.Clinical Anesthesiology 1984; 2:315-318

7. Sjogren S. Wright B: Circulatory changes during con­tinuous epidural blockage. Acta Anaesthesiol Scand1972; 46:5-10

282

A hypothesis inferred from this study is thatphysicians underrate psychic experiences ascompared to overt tissue trauma/physical injuryin their assessment of acute trauma injury pain.

CONCLUSION

Assessment and management of acute injurypain remains a difficult clinical issue. Our studyhighlights the problems inherent in correlatingpain assessment to a single BE assay and eluci­dates other confounding physiologic mecha­nisms that affect BE production. It alsodemonstrates that physician pain assessmentmay overlap only slightly with the patient'sexperience and may rely excessively on injuryassessment alone.

This work was supported by Western Psy­chiatric Seed Grant No. 2-9087/.

This research was presented at the Ameri­can Psychiatric Association Annual Meeting,May /993, San Francisco, CA.

The authors thank Erin Gordish for dataentry and the housestafffrom the Departmentsof Surgery, Emergency, and Critical CareMedicinefor their help in identifying study can­didates.

8. Yeager MP. Glass DO. Neff RK. et al: Epidural anesthe­sia and analgesia in high risk surgical patients. Anesthe­siology 1987; 66:729-736

9. Adler A: Neuropsychiatric complications in victims ofBoston's Cocoanut Grove Disaster. JAMA 1943;123:1098-1101

10. Andreasen N. Norris A. Hartford C: Incidence of long­term psychiatric complications in severely burnedadults. Ann Surg 1971; 174:785-793

II. Andreasen N. Noyes R. Hartford C. et al: Managementof emotional reactions in seriously burned adults. N EnglJ Med 1972; 286:65-69

12. Blank K. Perry S: Relationship of psychological processduring delirium to outcome. Am J Psychiatry 1984;141:843-847

13. Cella D. Perry S. Kulchycky S. et al: Stress and copingin relatives of bum patients: a longitudinal study. HospCommunity Psychiatry 1988; 39: 159-166

14. Hamburg D. Hamburg B. DeGoza S: Adaptive problems

PSYCHOSOMATICS

and mechanisms in severely burned patients. Am J Psy­chiatry 1953; 16: 1-20

15. Horowitz M. Wilner N. Alvarez W: Adaptive problemsand mechanisms in severely burned patients. Am J Psy­chiatry 1953; 16:1-20

16. Perry S. Cella D. Faltenberg. et a1: Pain perception inpatients with stress disorders. Journal of Pain SymptomManagement 1987; 2:22-23

17. Stoddard F. Norman D. Murphy J. et a1: Psychiatricoutcome of burned children and adolescents. J Am AcadChild Adolesc Psychiatry 1989; 28:589-595

18. Cousins MJ. Phillips GO (eds): Acute Pain Management.Edinburgh. Churchill-Livingston. 1986

19. Perry S. Heidrich G: Management of pain during de­bridement: a survey of US bum units. Pain 1982; 13:267­280

20. Perry S. Inturrisi C: Analgesia and morphine dispositionin bum patients. Journal of Bum Care and Rehabilitation1983; 26:276--279

21. Guillemin R. Vargo T. Rossier J. el a1: 13-endorphin andadrenoconicOlrophin are secreted concomitantly by thepituitary gland. Science 1977; 197:1367-1374

22. Mains RE. Elipper BA. Ling N: Common precursor toendorphin and conicotrophin. Proc Natl Acad Sci USA1977; 74:3014-3018

23. Basbaum A. Fields HL: Endogenous pain control mecha­nisms: review and hypothesis. Ann Neural 1976; 4:451­462

24. Chang 10. Cuatrecases P: Multiple opiate receptors:enkephalins and morphine bind to receptors of differentspecificity. J Bioi Chem 1979; 254:2610-2618

25. Hill RG. Hughes J: Opioid analgesics: CNS sites andmechanisms of action. Circ Shock 1989; 75:411-414

26. Daly T. Beamer KC. Vargish T. et a1: Correlation ofplasma 13-endorphin levels with mean anerial pressureand cardiac output in hypovolemic shock. Cril Care Med1987; 15:72~725

27. Faden A. Holaday JW: Opiate antagonists: a role in thetreatment of hypovolemic shock. Science 1979;199:317-321

28. Gao GC. Lin BC. Sun K. et al: The role of 13-endorphinin experimental bumshock. Advances in the Biosciences1989; 75:631-{j38

29. Gurll N. Reynolds J. Vargish T. et a1: Primate hemmhor­agic shock responds to naloxone. Circ Shock 1982;9:189-192

30. Shatney CH. Cohen RM. Cohen MR. et a1: Endogenousopioid activity in clinical hemorrhagic shock. SurgGynecolObstet 1985; 160:547-551

31. Tuggle D. Honon J. Kiser S: Beta-endorphin in hem­morhagic shock. Federation Proceedings 1983; 42:981­987

32. Przewlocka B. Wladyslaw L. Halina M. et al: Bum injuryin the hindlimb activates endogenous opioid symptomsin the rat. in New Leads in Opioid Research. edited byRee V. Noordwijkerhout. Scandinavia. ExcerptaMedica. 1990

33. Balon-Perin S. Kolanowski J. Berbinschi A. et a1: The

VOLUME 36-NUMBER 3 -MAY -JUNE 1995

Bernstein et al.

effects of glucose ingestion and fasting on plasma im­munoreactive beta-endorphin. adrenoconicotropic hor­mone and conisol in obese subjects. J Endocrinol Invest1991; 14:919-925

34. Eckardt MJ. Kaelber CT: Health hazards associated withalcohol consumption. JAMA 1981; 246:648-{j66

35. Howland J. Hingson R: Alcohol as a risk factor forinjuries or death due to fues and bums: review of theliterature. Public Health Rep 1987; 102:475-483

36. Lindenbaum GA. Carroll SF. Daskall, et al: Patterns ofalcohol and drug abuse in an urban trauma center: theincreasing role of cocaine. J Trauma 1989; 29: 1654­1658

37. Facchinetti F. Giovannini C. Petraglia F. et a1: Plasma13-endorphin resistance 10 dexamethasone suppression inobese patients. J Endocrinol Invest 1988; II: 119-123

38. Giovannini C. Ciucci E. Cassetta M. et a1: Unresponsive­ness of the endorphinergic syslem to its physiologicalfeedback in obesity. Appetite 1991; 16:39-43

39. Cohen M. Pickar D. Dubois M. et a1: Surgical stress andendorphins. Lancet 1981; 1:213-215

40. Roquefeuil B. Bressot N. Colson P. et a1: Plasma assaysof beta-endorphin and conisol in patients with cancerpain under morphine treatment. in New Leads in OpioidResearch, edited by Ree V. Noordwihkerhout. Scandina­via. Excerpta Medica. 1990

41. Foley KM. Kourides IA. Inturissi CEo et al: Beta-endor­phin: analgesic and hormonal effects in humans. ProcNatl Acad Sci USA 1979; 76:5366--5381

42. Lloyd DA. Teich S. Rose M: Serum endorphin levels ininjured children. Surg Gynecol Obstet 1991; 172:449­452.

43. Syzfelbein S. Osgood P. Carr 0: The assessment of painand plasma !3-endorphin immunoreactivity in bltrnedchildren. Pain 1985; 22: 17~182

44. Osgood PF. Murphy JL. Carr DB. et a1: Increases inplasma 13-endorphin and tail flick latency in the ratfollowing bum injury. Life Sci 1987; 40:547-554

45. Teasdale G. Jennet B: Assessment of coma and impairedconsciousness: a practical scale. Arch Neurol 1974;2:81-83

46. Eastman AB. West JG: Field triage. in Trauma. editedby Moore L. Mastox J. Felliciano E. New York. Apple­ton. 1991

47. Price DO. Harkins SW: The affective-motivational di­mension of pain. American Psychological Society Jour­nal 1992; 1:229-239

48. Menegazzi J: A randomized control trial of the use ofmusic during laceration repair. Ann Emerg Med 1991;20:348-350

49. Wallenstein SL: Oinical evaluation of mild analgesics:the management of clinical pain. Br J Clin Pharmacol1980; 10:3195-3275

50. Huskisson E: Measurement of pain. Lancet 1974;2:1127-1131

51. Baker S, O'Neill B. Haddon W. et a1: The Injury SeverityScore: a method for describing patients with multipleinjuries and evaluating emergency care. J Trauma 1974;

283

Pain and Serum Beta-Endorphin

3:187-19652. Barton RN. Stoner HB. Walson SM: Relalionship among

plama conisol. adrenoconicolrophin. and severily ofinjury in recently injured palients. J Trauma 1987;4:384-392

53. Puig MM. Laorden ML. Miralles FS. et al: Endorphinlevels in cerebrospinal nuid of patienls with postopera­tive and chronic pain. J Anesthesiol 1982; 57: 1-4

284

54. Carr DB. Ballantyne JC. Osgood PF. el al: Pituitary­adrenal stress response in the absence of brain-pituitaryconnections. Anesth Analg 1989; 69: 197-20 I

55. Pettingill BF: Physicians estimates of disability vs. pa­tients repon of pain. Psychosomalics 1979; 20:827-830

56. Wanman SA, Morlock LL, Malilz FE. et al: Impact ofdivergent evaluations by physicians and patients of pa­tient's complaints. Public Health Rep 1983; 98: 141-144

PSYCHOSOMATICS