W. H. Kirkaldy-Willis J. D. Cassidy

Spinal Manipulation in the Treatment ofLow-Back PainSUMMARYSpinal manipulation, one of the oldest formsof therapy for back pain, has mostly beenpracticed outside of the medical profession.Over the past decade, there has been anescalation of dlinical and basic science researchon manipulative therapy, which has shownthat there is a scientific basis for the treatmentof back pain by manipulation. Most familypractitioners have neither the time norinclination to master the art of manipulationand will wish to refer their patients to a skilledpractitioner of this therapy. Results of spinalmanipulation in 283 patients with low backpain are presented. The physician who makesuse of this resource will provide relief formany patients. (Can Fam Physician1985;31:535-540)

SOMMAIRELes manipulations vertebrales, qui sont l'une desformes les plus anciennes de traitement pour lalombalgie, ont, dans la majorite des cas, etel'apanage de professions autres que medicales. Aucours de la dernire decennie, la recherchefondamentale et clinique sur les manipulations s'estaccelerde et a pu demontrer qu'il existe une basescientifique pour justifier le traitement de lalombalgie par la manipulation. La plupart desmddecins de famille manquent soit de temps, soitd'int6rkt pour maitriser l'art des manipulations etpr6freront rdferer leurs patients aux praticienspossedant l'expdrience de cette forme de traitement.Cet article presente les resultats de manipulationsvertebrales chez 283 patients souffrant de lombalgie.Le medecin qui utilise cette forme de traitementcontribuera au soulagement de nombreux patients.

Key words: Spinal manipulation, low back,pain

Dr. Kirkaldy-Willis is a professoremeritus of orthopedics anddirector of the Low-Back PainClinic at the University Hospital,Saskatoon. Dr. Cassidy is achiropractor and a research fellowwith the Department ofOrthopedics at the UniversityHospital. Reprint requests to: Dr.W. H. Kirkaldy-Willis, Departmentof Orthopedics, UniversityHospital, University ofSaskatchewan, Saskatoon, SK.S7N OXO.

BACK PAIN is one of the com-monest presenting complaints in

office practice: almost 80% of thegeneral population will experiencelow back pain during adult life. Atany given time, 20-30% of adults suf-fer from low back pain. ", 2 In in-dustry, disorders of the lower backaccount for four hours per year perworker of lost productivity, and rank

CAN. FAM. PHYSICIAN Vol. 31: MARCH 1985

second only to upper respiratory in-fection as a cause of absenteeism.3Patients with low back pain representa major segment of the chronicallydisabled, comparable to the numberssuffering from heart disease, arthritisand rheumatism.3 Estimates for thecost of treatment and compensation inthe United States for those sufferingfrom back pain exceed $14 billion an-nually-notwithstanding the cost oflost productivity in the work place. 2Despite the high frequency and

enormous cost of low back pain, thecauses and effective therapeutic pro-grams remain highly problematic.Part of this problem is due to the na-ture of low back pain: it is a common,self-limiting disorder with a high rateof recurrence.4 Moreover, the manydifferent causes of back pain are notalways readily apparent. In fact, withthe exception of back pain and scia-tica resulting from entrapment of thespinal nerve root by degenerative

changes or by disc herniation, mostcauses of low back pain lack objectiveclinical signs and overt pathologicalchanges. Nevertheless, these obscurecauses are responsible for most of theback pain seen in clinical practice.3Less than 10% of low back pain isdue to herniation of the intervertebraldisc or entrapment of spinal nerves bydegenerative disc disease.",4'5 Ac-cordingly, the diagnosis of low backpain is a difficult matter. We attemptto group back pain patients into syn-drome categories on the basis of theirhistory, pain distribution, physical ex-amination and radiographic findings.6By this approach, we try to identifythe predominant pain-producing le-sion and direct our treatment accord-ingly. Although this method is clini-cally useful, it does not necessarilyreflect true pathogenesis.

Since most low back pain is idio-pathic, the effects of many commonlyapplied therapies remain highly spec-

535

I

ulative. This is true for such populartreatments as spinal fusion, chemonu-cleolysis, facet injection and denerva-tion, transcutaneous electrical nervestimulation, acupuncture, exercise,traction, manipulation and so on. Onecould argue that spinal fusion relievespain by denervating spinal structuresrather than by stabilizing an unstablemotion segment. Perhaps extensiveoperative procedures relieve back painby lowering intraosseous venous hy-pertension or by enforcing bed restand a planned period of rehabilita-tion.7 Moreover, spinal fusion has notbeen subjected to rigorous clinicaltrial.8 This is also true for many othertherapies. Furthermore, the naturalhistory of low back pain ensures agood longterm result despite treat-ment.9 Therefore, it is wise to beginthe treatment of low back pain with asafe, cost-effective, conservative pro-gram of therapy and to proceed tomore expensive and invasive proce-dures in carefully selected patientsonly.

What is SpinalManipulation?

Spinal manipulation is essentiallyan assisted passive motion applied tothe spinal apophyseal and sacroiliacjoints. The terms mobilization andmanipulation require separate defini-tions.10 In Figure one, the motion of asynovial joint is shown in one plane.Beyond the end of the active range ofmotion (ROM) of any synovial joint,there is a small buffer zone of passivemobility. A joint can be only pas-sively assisted into this ROM. Thisconstitutes mobilization.At the end of the passive ROM, an

elastic barrier of resistance is encoun-tered. This barrier has a spring-likeend-feel which is the result of a nega-tive or subatmospheric intra-articularpressure. This negative pressure is astabilizing factor in the coaptation ofthe articular surfaces. If the separationof the articular surfaces is forcedbeyond this elastic barrier, the jointsurfaces suddenly move apart with acracking noise. This additional sepa-ration can only be achieved aftercracking the joint and has been la-belled the paraphysiological ROM.This constitutes manipulation.The cracking sound on entering the

paraphysiological ROM is the resultof the sudden liberation of synovialgases-a phenomenon known to

536

physicists as cavitation.11 The resul-tant synovial bubble can be demon-strated radiographically and is reab-sorbed over the following 30 minutes.During this period, the elastic barrierof resistance between the passive andparaphysiological zones is absent, andthere is an increase in the joint space.As the synovial gases are reabsorbed,the intra-articular pressure drops, thejoint space narrows, and the elasticbarrier of resistance is re-establishedbetween the passive and paraphysiolo-gical zones. During this refractoryperiod, the joint is somewhat unstableand cannot be remanipulated safely.

At the end of the paraphysiologicalROM, the limit of anatomical integ-rity is encountered. Movementbeyond this limit results in damage tothe capsular ligaments. During manip-ulation, a carefully graded anddirected thrust is applied across thejoint space at the end of the passiveROM. This force must be greatenough to overcome the elastic barrierof resistance, but not so great as toseparate the joint surfaces beyondtheir limit of anatomical integrity.This requires precise positioning ofthe joint at the end of the passiveROM and the proper degree of forceto overcome joint coaptation. Thisskill is not easily acquired; consider-able training and experience are nec-essary. In the hands of a skilled mani-pulator, the procedure is not painful.

Most lumbar spine manipulationsare done with the patient in the sideposture position (see Fig. 2). In thisposition, the knee and hip of theupper leg are flexed on the lower leg.This enables the upper thigh to beused as a lever. In the side posture,the lumbar lordosis is reduced, andthe spine is relatively straight.To begin the process of mobiliza-

tion and manipulation, the patient'supper body is twisted to introduce anelement of rotation and lateral flexioninto the lumbar spine (see Fig. 3). Inthis position, there is a counter-rotation of the upper torso on thepelvis, and the posterior facet jointsare at, or near, their limit of activeROM. During the next step, the mani-pulator must try to localize the pointof counter-rotation to the motion seg-ment to be manipulated, by varyingthe degree of flexion in the upperknee and hip. (see Fig. 4). This, inturn, varies the degree of tensionplaced on the lower lever and thepoint of counter-rotation between the

two levers. By increasing the tensionon the lower lever, the force of themanipulation can be localized tohigher levels of the lumbar spine.With experience, the manipulator canbe very specific in selecting the spinallevel to be manipulated.

Once the force of the manipulationhas been localized, the process of mo-bilization and then manipulation canbegin. With the patient rotated in theside posture, the counter-rotationforce on the spine can be increasedthrough the passive ROM up to theelastic barrier of resistance (see Fig.5). This constitutes spinal mobiliza-tion and can be repeated several timeswith increasing force. If enough forceis applied to overcome joint coapta-tion, a crack is produced, and theROM is increased into the paraphys-iological zone (see Fig. 6). An experi-enced manipulator can overcome theelastic barrier of resistance with acarefully applied, high-velocity,short-amplitude thrust. Less experi-enced clinicians should master the artof mobilization before attempting tomanipulate the spine.

What are the Effectsof Spinal Manipulation?

Research into the effects of spinalmanipulation has escalated over thepast decade,12-15 partly due to in-creased understanding of articular neu-rology and pain modulation.

Melzack and Wall16 first proposedthe gate theory of pain in 1965. Sincethat time, the basic principles of thistheory have withstood rigorous scien-tific scrutiny-even if the precisemechanisms and anatomical details arenot fully understood. In essence, theyproposed a spinal gating mechanismwithin the substantia gelatinosa(Rexed's lamina II) of the dorsal hornof the spinal cord. This gate controlsthe central transmission of sensory in-formation including pain, touch, tem-perature and proprioception. Theyhave shown that the central transmis-sion of pain can be blocked by in-creased proprioceptive input andfacilitated by a lack of proprioceptiveinput. This simple concept explainswhy rubbing an acute injury alleviatesthe pain and the importance of earlymobilization to control pain after mus-culosketetal injury.Wyke1 7 has shown that the articular

capsules of the spinal facet joints aredensely populated by mechanorecep-tors. These encapsulated nerve end-

CAN. FAM. PHYSICIAN Vol. 31: MARCH 1985

Fig.1Fig.4

Fig. 2

Fig. 3

SIDE POSTURE:Limit of Active R.O.M.

COUNTER-ROTATION of LEVERS

CAN. FAM. PHYSICIAN Vol. 31: MARCH 1985

JOINT MOBILIZATION and MANIPULATION

NACTIVE R.O.M.

El PASSIVE R.O.M.

*PARAPHYSIOLOGICAL9f :00 :0: }, ~~~~R.O.M.

'

I,..'- MOBIL (ZAtrSdN '' 96

'. MANIPUJLATION w

(Aftter Sandoz, 1976)

Segmental Localization of Force

(i) 14-5

(.) o 2-' H a E

tHIP and KNEE FLEXIO

SIDE POSTURE:Neutral Position

Fig. 5

Fig. 6

SIDE POSTURE:Limit of Passive R.O.M.

MOBILIZATION

SIDE POSTURE:Limit of Paraphysiological R.O.M.

../

MNIPULATION

Fig. I Fig. 4

537

ings relay proprioceptive informationon joint position and mobility throughlarge myelinated fibers to the substan-tia gelatinosa of the spinal cord. Theseimpulses then compete for centraltransmission with impulses from thesmaller unmyelinated pain fibers fromadjacent tissues. Hence, increased pro-prioceptive input in the form of spinalmobility tends to decrease the centraltransmission of pain from adjacent spi-nal structures by closing the gate. Anytherapy which induces motion into ar-ticular structures will help to inhibitpain transmission by this means. 18Wyke and others 19-21 have also

shown that articular mechanoreceptorstimulation has a reflexogenic effecton motor unit activity in the musclesoperating over the joint being stimu-lated. Stretching of apophyseal jointcapsules can therefore reflexly inhibitfacilitated motoneuron pools whichare responsible for the increasedmuscle excitability and spasms thatcommonly accompany low back pain.In more chronic cases, there is short-ening of periarticular connective tissueand intra-articular adhesions mayform.22' 23 We believe that in somecases, manipulation will stretch orbreak these adhesions. In fact, in mostcases of chronic low back pain, thereis an initial increase in symptoms afterthe first few manipulations. In almostall cases, however, this increase inpain is temporary and can be easilycontrolled by local application of ice.However, the gain in mobility must be

TABLE 1Results of Spinal Manipulation in 54Patients with Posterior JointSyndromeAverage duration of pain 5.6 yrs.Average length of follow-up 9.2 mo.Results:Grade 64%Grade 2 15%Grade 3 9%Grade 4 120/o

TABLE 2Results of Spinal Manipulation in 69Patients with Sacroiliac JointSyndromeAverage duration of pain 7.9 yrs.Average length of follow-up 10.3 mo.Results:Grade 1 71%Grade 2 22%Grade 3 3%Grade 4 4%

maintained during this period to pre-vent further adhesion formation.Through these mechanisms, spinalmanipulation can break the cycle ofpain, muscle spasm and immobilitywhich predominates in many cases oflow back pain.

At present, there is no evidence thatmanipulation replaces sublexated ver-tebrae. This theory was first put for-ward by the chiropractic professionmany years ago and has largely beenabandoned. However, changes in epi-durographic defects have been re-ported after manipulation,24 althougha similar study using myelographyshowed no changes in the defects, yetover 50% of the patients studied wereimproved by manipulation.25 More re-cently, manipulative therapy wasshown to be superior to shortwavediathermy and exercise in a rando-mized controlled clinical trial on pa-tients with prolapsed intervertebraldiscs.26 Our own studies and those ofothers suggest that success with mani-pulative therapy decreases with in-creasing neurological deficit.27' 28 Wewould therefore not recommend mani-pulative therapy in cases of prolapseddisc with marked neurological deficit.

How Successful isSpinal Manipulation?

Since 1952, there have been over 50clinical trials of spinal manipulationfor back pain.29 Of these studies, 13are randomized controlled clinicaltrials. Although some of these studiesshow faults in design and some degreeof variability in the results, certaintrends are emerging.

In the treatment of acute low backpain, most studies show that manipula-tion tends to shorten the episode ofpain,30 31 particularly over the shortterm. Longterm follow-up suggeststhat the initial advantage of manipula-tion over other therapies is lost withtime. This is also true for other treat-

TABLE 3Results of Spinal Manipulation in 48Patients with combined PosteriorJoint and Sacroiliac JointSyndromesAverage duration of pain 9.8 yrs.Average length of follow-up 13.9 mo.Results:Grade 1 67%Grade 2 21%Grade 3 6%Grade 4 6%

ment and is consistent with the recur-rent nature of low back pain.

Similar findings have been reportedfor the treatment of chronic low backpain by manipulation.32' 33 In mostcases, there is an initial improvementfollowed by a regression to the mean.These findings suggest that althoughspinal manipulation is successful in al-leviating low back pain, it does not af-fect the recurrent nature of the dis-order. Similarly, discectomy forlumbar disc herniation results in re-gression to the mean over a longertime.34

Several points should be made aboutclinical trials of manipulation. In mostcases, the method of manipulation isnot described; many of these trials uti-lized mobilization rather than manipu-lation. In a majority of the studies,very few treatments were given, andthe training and expertise of the mani-pulators are impossible to judge. Insome, there are obvious design errorsand experimental bias is likely. Inothers, the numbers are probably toosmall to show significance. Neverthe-less, there is ample evidence that spi-nal manipulation is a useful therapydeserving further study.

Which is the BestBack to Manipulate?Many attempts have been made to

identify patients who will best respondto manipulative therapy. Jayson et al.were unable to identify any prognosticmarkers other than a shorter history ofpain.35 Evans et al. found that patientsbenefiting from manipulation weremore likely to be older and to have hadsymptoms for a shorter period.32 Thesame group was unable to predicttreatment outcome on the basis of ra-diographic findings.36 In a retrospec-tive study, Morrison28 identified sev-eral parameters for success, includinga sudden onset of back or leg pain, re-duced spinal mobility, straight leg

TABLE 4Results of Spinal Manipulation in 31Patients with Posterior JointSyndrome and/or Sacroiliac JointSyndrome with Lumbar InstabilityAverage duration of pain 7.2 yrs.Average length of follow-up 8.0 mo.

Results:Grade 1 26%Grade 2 19%Grade 3 29%Grade 4 26%

CAN. FAM. PHYSICIAN Vol. 31: MARCH 1985538

raising more than 60 degrees and fewneurological signs. Others have beenunable to differentiate responders fromnon-responders. 33' 37, 38We recently completed a prospect-

ive observational study of spinal ma-nipulation in 283 patients with chroniclow back and leg pain. The results ofthis study are summarized in Tables1-7. Details on our diagnostic criteriaand method of study are available else-where.27 Our patient population wastaken from a specialized universityback pain clinic reserved for patientswho have not responded to previousconservative or operative treatment.All of the patients in this study weretotally disabled (grade 4 disability) atthe onset of treatment. Therefore, ourresults might not be representative ofsimilar treatment given in primary careto patients who are not totally disa-bled.

In our study, the patient's responseto treatment was assessed by an inde-pendent observer and based on the pa-tient's impression of pain relief andloss of disability. The results weregraded as follows:Grade I. Symptom-free with no re-strictions for work or other activities.Grade 2. Mild intermittent pain withno restrictions for work or other activi-ties.Grade 3. Improved, but restricted inactivities by pain.Grade 4. Constant severe pain; dis-ability unaffected by treatment.

All patients entered the study at thegrade 4 level of pain and disability.They were given a two or three weekregimen of daily spinal manipulationsby an experienced chiropractor. Theresults of this treatment were assessedone month later and at three month in-tervals thereafter.No patients were made worse by

manipulation, yet many experiencedan increase in pain during the firstweek of treatment. Patients undergo-

TABLE 5Results of Spinal Manipulation in 60Patients with Nerve RootEntrapment SyndromeAverage duration of pain 7.2 yrs.Average length of follow-up 14.0 mo.Results:

Grade 1 25%Grade 2 25%Grade 3 17%Grade 4 330/o

ing manipulative treatment must there-fore be reassured that the initial dis-comfort is only temporary. In ourexperience, anything less than twoweeks of daily manipulation is inade-quate for chronic back pain patients.

Our results are summarized in Table8. Patients grouped under the referredpain syndromes include those withposterior joint and sacroiliac syn-dromes. These patients may presentwith low back and/or leg pain, buthave no signs of radiculopathy. Thedesignation of posterior joint or sacro-iliac syndrome does not necessarilyreflect pathogenesis, but rather thestructures that were manipulated.The patients grouped under nervecompression syndromes include thosewith nerve root entrapment syndromeand a small, select group of patientswith central spinal stenosis syndromewho were unfit for surgery. These pa-tients all had evidence of radiculo-pathy; most presented with leg pain.The level of manipulation in this groupwas determined by criteria outlinedelsewhere.39

If we consider a grade 1-2 responseto treatment as clinically significant, asignificantly better result was obtainedin the referred pain syndromes. Wealso found that patients with low backand/or proximal sciatica (pain not pastthe knee) responded significantly bet-ter than those with distal sciatic radia-tion of pain (p<0.001). In one groupof patients (Table 4), radiographic evi-dence of motion segment instability(as diagnosed from lateral views takenat the extremes of lumbar flexion andextension) was associated with a signi-ficantly poorer response to manipula-tion (p<0.01).40 Almost 25% of ourpatients had undergone previous surgi-cal treatment for their back pain, andalthough there was a trend towards apoorer response to treatment, it wasnot statistically significant. This was

TABLE 6Results of Spinal Manipulation inTen Patients with Nerve RootEntrapment Syndrome with LumbarInstabilityAverage duration of pain 11.5 yrs.Average length of follow-up 12.6 mo.Results:Grade 1 40%Grade 2 10%Grade 3 20%Grade 4 30%/

also true for the degree of degenerativedisc disease seen on the X-ray.

Who Should DeliverManipulative Therapy?

Several professional groups offermanipulative therapy; of these, chiro-practors are the largest. Most provin-cial health care plans insure their care.Recently, the physiotherapy profes-sion has become more interested inproviding this care. Many undergrad-uate physiotherapy colleges now offertraining in mobilization and manipula-tion. In the United States, some osteo-paths still provide manipulative care aspart of their allopathic practices. Somephysicians practice manipulative med-icine fulltime; most are members ofthe North American Academy of Man-ipulative Medicine which offers post-graduate training. It is not difficult toobtain manipulative treatment in mostNorth American and Europeancentres.

Manipulation requires much prac-tice to acquire the necessary skills andcompetence. It is a fulltime vocation:few medical practitioners have thetime or inclination to master it. Mostdoctors, whether family physicians orsurgeons, will wish to refer their pa-tients to a practitioner of manipulativetherapy with whom they can cooper-ate, whose work they know and whomthey can trust. The professional back-ground of these practitioners may varyfrom case to case. The physician whomakes use of this resource will providerelief for many back pain patients. )

References1. Valkenburg HA, Haanen HCM. The epi-demiology of low-back pain. In: White AA,Gordon SL, eds. American Academy ofOrthopaedic Surgeons Symposium on Idio-pathic Low Back Pain. Toronto: CV MosbyCo., 1982:9-22.2. White AA. Introduction. In: White AA,Gordon SL, eds. American Academy of

TABLE 7Results of Spinal Manipulation in 11Patients with Central SpinalStenosis SyndromeAverage duration of pain 16.9 yrs.Average length of follow-up 7.0 mo.Results:Grade 1 18%Grade 2 18%Grade 3 18%Grade 4 46%8

CAN. FAM. PHYSICIAN Vol. 31: MARCH 1985 539

Orthopaedic Surgeons Symposium on Idio-pathic Low Back Pain. Toronto: CV MosbyCo., 1982:1-2.3. Kelsey JL. Idiopathic low-back pain:magnitude of the problem. In: White AA,Gordon SL, eds. American Academy ofOrthopaedic Surgeons Symposium on Idio-pathic Low Back Pain. Toronto: CV MosbyCo., 1982:5-8.4. Kelsey JL. Epidemiology of musculo-.skeletal disorders. New York: Oxford Uni-versity Press, 1982:145-67.5. Frymoyer JW, Newberg A, Pope MH,Wilder DG, Clements J, MacPherson B.Spine radiographs in patients with low-back pain. J Bone Joint Surg 1984; 66-A:1048-55.6. Kirkaldy-Willis WH, ed. Managing low-back pain. New York: Churchill Living-stone, 1983.7. Foley RK, Kirkaldy-Willis WH. Chronicvenous hypertension in the tail ofthe wistarrat. Spine 1979; 4:251-7.8. Farfan HF, Kirkaldy-Willis WH. Thepresent status of spinal fusion in the treat-ment of lumbar intervertebral joint dis-ease. Clin Orthop 1981; 158:198-214.9. Currey HLF, Greenwood RM, LloydGG, Murray RS. A prospective study oflow-back pain. Rheumatol Rehabil 1979;18:94-104.10. Sandoz R. Some physical mechanismsand effects of spinal adjustments. AnnSwiss Chirop Assoc 1976; 6:91-141.11. Unsworth A, Dowson D, Wright V.Cracking joints. Ann Rheum Dis 1971;30:348-58.12. Goldstein M. ed. The research status ofspinal manipulative therapy. National In-stitute ofNeurological and CommunicativeDisorders and Stroke. Monograph 15.Bethesda, MD.: U.S. Dept. ofHealth, Ed-ucation and Welfare, 1975:76-998.13. Buerger AA, Tobis JS, eds. Ap-proaches to the validation of manipulativetherapy. Springfield: Charles C Thomas,1977.14. Korr IM, ed. The neurobiologic mech-anisms in manipulative therapy. New York:Plenum Press, 1978.15. Haldeman S, ed. Modern developmentsin the principles and practice of chiroprac-tic. New York: Appelton-Century-Crofts,1980.16. Melzack R, Wall PD. Pain mech-anisms: a new theory. Science 1965;150:971-9.17. Wyke BD. Neurological aspects oflow-back pain. In: Jayson MIV, ed. The lumbarspine and back pain. London: Grune &Stratton, 1976:189-256.

18. Terrett ACJ, Vernon H. Manipulationand pain tolerance: a controlled study ofthe effect of spinal tnanipulItion on para-spinal cutaneous pain tolerance levels. AmJPhys Med 1984; 63:217-25.19. Buerger AA. Experimental neuro-muscular models of spinal manual tech-niques. Manual Med 1983; 1:10-17.20. Wyke BD. Articular neurology andmanipulative therapy. In: Idczak RM, ed.Aspects of manipulative therapy. Carlton:Lincoln Institute of Health Sciences,1980:67-71.21. Korr IM. Proprioceptors and somaticdysfunction. J Am Osteopath Assoc 1975;74:638-50.22. Akeson WH, Amiel D, Woo S. Immo-bility effects on synovial joints. The patho-mechanics of joint contracture. Biorheo-logy 1980; 17:95-110.23. Kirkaldy-Willis WH, Heithoff KB,Tchang S, Bowen CVA, Cassidy JD, Shan-non R. Lumbar spondylosis and stenosis:correlation of pathological anatomy withhigh-resolution computed tomographicscanning. In: Donovan Post MJ, ed. Com-puted tomography of the spine. Baltimore:Williams and Wilkins, 1984:546-69.24. Mathews JA, Yates DAH. Reduction oflumbar disc prolapse by manipulation. BrMed J 1969; 3:696-7.25. Chrisman DD, Mittnackt A, Snook GA.A study of the results following rotatorymanipulation in the lumbar intervertebraldisc syndrome. J Bone Joint Surg 1964;46-A:517-24.26. Nwuga VCB. Relative therapeutic effi-cacy of vertebral manipulation and con-ventional treatment in back pain manage-ment. Am J Phys Med 1982; 61:273-8.27. Cassidy JD, Kirkaldy-Willis WH,McGregor M. Spinal manipulation for thetreatment of chronic low-back and legpain: an observational study. In: BuergerAA, ed. Empirical approaches to the vali-dation of manipulative medicine. Spring-field: Charles C. Thonmas, in press.

28. Morrison MCT. The best back to ma-nipulate? Ann R Coll Surg Engl 1984;66:52-3.29. Brunarski DJ. Clinical trials of spinalmanipulation: a critical appraisal and re-view of the literature. J Manipulative Phy-sio Ther 1984; 7:243-9.30. Rasmussen GG. Manipulation in treat-ment oflow backpain-a randomized clin-ical trial. Manuelle Medizin 1979; 17:8-10.31. Farrell JP, Twomey LT. Acute low-back pain. Comparison oftwo conservativetreatment approaches. Med J Aust '1982;1:1604.32. Evans DP, Burke MS, Lloyd KN, Rob-erts EE, Roberts GM. Lumbar spinal ma-nipulation on trial. Part --clinical assess-ment. Rheumatol Rehabil 1978; 17:46-53.33. Coxhead CE, Inskip H, Meade TW,North WRS, Troup JDG. Multicentre trialof physiotherapy in the management ofsciatic symptoms. Lancet 1981; May16:1065-8.34. Weber H. Lumbar disc herniation. Acontrolled prospective study with ten yearsofobservation. Spine 1983; 8:131-40.35. Jayson MIV, Sim-Williams H, YoungS, Baddeley H, Collins E. Mobilizationand manipulation for low-back pain. Spine1981; 6:409-16.36. Roberts GM, Roberts EE, Lloyd KN,Burke MS, Evans DP. Lumbar spinalmanipulation on trial. Part ll-radiologi-cal assessment. Rheumatol Rehabil 1978;17:54-9.37. Doran DML, Newell DJ. Manipulationin treatment of low-back pain: a multi-centre study. Br Med J 1975; 2:161-4.38. Buerger AA. A controlled trial of rota-tional manipulation in low-back pain.Manuelle Medizin 1980; 2:17-26.39. Cassidy JD, Potter GE. Motion exam-ination ofthe lumbar spine. J ManipulativePhysiol Ther 1979; 2:151-8.40. Kirkaldy-Willis WH, Farfan HF. Insta-bility of the lumbar spine. Clin Orthop1982; 165:110-23.

TABLE 8Results of Spinal Manipulation In 283 Patients With Referred PainSyndromes or Nerve Compression Syndromes

Improved Not ImprovedSyndromes (Grade 1 & 2) (Grade 3 & 4)Referred Pain 163(81%) 39 (190/%)Nerve compression 39(48%) 42 (52%)

X2= 29.7,df = 1,p<00O1

540 CAN. FAM. PHYSICIAN Vol. 31: MARCH 1985