PHYS THER-2013-Thomas-1563-74

doi: 10.2522/ptj.20120477Originally published online June 27, 2013

2013; 93:1563-1574.PHYS THER. Stanwell and Christopher R. LeviLucy C. Thomas, Darren A. Rivett, Grant Bateman, PeterCarotid Arterial Blood Flow and Cerebral InflowMechanical Neck Pain on Vertebral and Internal Effect of Selected Manual Therapy Interventions for

http://ptjournal.apta.org/content/93/11/1563found online at: The online version of this article, along with updated information and services, can be

Collections

Manual Therapy Injuries and Conditions: Neck

SystemAnatomy and Physiology: Cardiovascular/Pulmonary

in the following collection(s): This article, along with others on similar topics, appears

e-Letters

"Responses" in the online version of this article. "Submit a response" in the right-hand menu under

or click onhere To submit an e-Letter on this article, click

E-mail alerts to receive free e-mail alerts hereSign up

by guest on December 2, 2014http://ptjournal.apta.org/Downloaded from by guest on December 2, 2014http://ptjournal.apta.org/Downloaded from

http://ptjournal.apta.org/content/93/11/1563

http://ptjournal.apta.org/cgi/collection/anatomy_and_physiology_cardiovascular_pulmonary_system

http://ptjournal.apta.org/cgi/collection/anatomy_and_physiology_cardiovascular_pulmonary_system

http://ptjournal.apta.org/cgi/collection/injuries_and_conditions_neck

http://ptjournal.apta.org/cgi/collection/manual_therapy

http://ptjournal.apta.org/letters/submit/ptjournal;93/11/1563

http://ptjournal.apta.org/subscriptions/etoc.xhtml

http://ptjournal.apta.org/


Effect of Selected Manual TherapyInterventions for Mechanical NeckPain on Vertebral and InternalCarotid Arterial Blood Flow andCerebral InflowLucy C. Thomas, Darren A. Rivett, Grant Bateman, Peter Stanwell,Christopher R. Levi

Background. Manual therapy of the cervical spine has occasionally been associ-ated with serious adverse events involving compromise of the craniocervical arteries.Ultrasound studies have shown certain neck positions can alter craniocervical arterialblood flow velocities; however, findings are conflicting. Knowledge about the effectsof neck position on blood flow may assist clinicians in avoiding potentially hazardouspractices.

Objective. The purpose of this study was to examine the effects of selectedmanual therapeutic interventions on blood flow in the craniocervical arteries andblood supply to the brain using magnetic resonance angiography (MRA).

Design. This was an experimental, observational magnetic resonance imagingstudy.

Method. Twenty adult participants who were healthy and had a mean age of 33years were imaged using MRA in the following neck positions: neutral, rotation,rotation/distraction (similar to a Cyriax manipulation), C1–C2 rotation (similar to aMaitland or osteopathic manipulation), and distraction.

Results. The participants were imaged using 3T MRA. All participants had normalvascular anatomy. Average inflow to the brain in neutral was 6.98 mL/s and was notsignificantly changed by any of the test positions. There was no significant differencein flow in any of the 4 arteries in any position from neutral, despite large individualvariations.

Limitations. Only individuals who were asymptomatic were investigated, and ashort section of the arteries only were imaged.

Conclusions. Blood flow to the brain does not appear to be compromised bypositions commonly used in manual therapy. Positions using end-range neck rotationand distraction do not appear to be more hazardous to cerebral circulation than moresegmentally localized techniques.

L.C. Thomas, DipPhys, GradDip-AppSc(ManipPhty), MMedSc(Physiotherapy), Discipline ofPhysiotherapy, School of HealthSciences, Faculty of Health, TheUniversity of Newcastle, UniversityDrive, Callaghan 2308, NewSouth Wales, Australia. Address allcorrespondence to Ms Thomas at:[email protected].

D.A. Rivett, BAppSc(Phty), Grad-DipManipTher, MAppSc(Manip-Phty), PhD, Discipline of Physio-therapy, School of Health Sci-ences, Faculty of Health, TheUniversity of Newcastle.

G. Bateman, MBBS, FRANZCR,Department of Radiology, Facultyof Health, John Hunter Hospital,Newcastle, New South Wales,Australia.

P. Stanwell, PhD, School of HealthSciences, Faculty of Health, TheUniversity of Newcastle.

C.R. Levi, BMedSc, MBBS, FRACP,Department of Neurology, Facultyof Health, John Hunter Hospital.

[Thomas LC, Rivett DA, BatemanG, et al. Effect of selected manualtherapy interventions for mechan-ical neck pain on vertebral andinternal carotid arterial blood flowand cerebral inflow. Phys Ther.2013;93:1563–1574.]

© 2013 American Physical TherapyAssociation

Published Ahead of Print:June 27, 2013

Accepted: June 21, 2013Submitted: December 7, 2012

Research Report

Post a Rapid Response tothis article at:ptjournal.apta.org

November 2013 Volume 93 Number 11 Physical Therapy f 1563 by guest on December 2, 2014http://ptjournal.apta.org/Downloaded from


Manual therapy, includinghigh-velocity thrust (HVT)manipulation, is commonly

used for the management of neckpain and associated headache buthas in rare cases been associatedwith serious adverse neurovascularevents.1–3 These adverse events mostcommonly involve dissection of thevertebral or internal carotid arteries(ie, the craniocervical arteries),which supply blood to the brain.2,4–8

Dissection of these arteries maydirectly alter blood flow to the brainor trigger thrombus formation,potentially causing a stroke.9,10

Although concerns often have beenraised as to the safety of HVT manip-ulation, other manual techniquesusing sustained end-range positionsof the neck also are potentially impli-cated in causing changes in bloodflow in the craniocervical arter-ies.11,12 Where it has been possibleto identify particular manipulativeprocedures associated with cerebro-vascular complications such as dis-section, rotational techniques havebeen most frequently described.2,8

Dissections most commonly occurin the atlanto-axial portion of thevertebral artery and upper cervicalportion of the internal carotidartery.2,4,7 Most neurovascular inju-ries have been reported in individu-als between 30 and 40 years of agewho were healthy and had no othercardiovascular risk factors.6,13–15

Common cervical spine manual ther-apeutic procedures involve movingthe head and neck into variouspositions. It has been suggested thatone factor that may contribute toadverse neurovascular events follow-ing manual treatment of the neckmay be the positions of neck rotationclose to the end of the physiologi-cal range, which could temporarilycompromise blood flow.2 Theseflow changes may be an indication ofincreased biomechanical stress ofthe arterial wall.16 It is consideredthat the additional effect of an HVT

on a prestressed arterial wall maycause damage such as a dissection tooccur.

It has been shown using ultrasoundimaging that certain neck move-ments, in particular cervical rotation,can alter blood flow velocities in thecraniocervical arteries in some indi-viduals.17,18 Numerous ultrasoundstudies over the last 2 decades haveexamined blood flow in differentpositions of the neck, but the resultshave been conflicting,17–22 withsome authors reporting that bloodflow was changed in contralateralrotation and others that it wasunchanged. Disagreement amongstudies may relate to a number ofmethodological considerations. Mostnotably, ultrasound is known to behighly operator dependent, particu-larly when sampling blood flowparameters in small-diameter tortu-ous vessels such as the vertebralartery.23,24 Studies also have com-monly measured parameters involv-ing flow velocity, which is subject towide fluctuation, particularly if sam-pled close to the vessel wall.23,24

Moreover, most studies typicallyexamined blood flow in a single ves-sel, usually the vertebral artery. How-ever, examination of blood flow inone vessel cannot provide a com-plete picture of blood flow to thebrain.

Angiography is considered to be thereference standard for imaging thecraniocervical arterial system25 andcurrently is most commonly per-formed using noninvasive methodssuch as magnetic resonance angiog-raphy (MRA) or computed tomo-graphic angiography (CTA). Mag-netic resonance angiography cangenerate a clear image of the vesselsusing “time of flight” MRA. Bloodflow quantification using phase con-trast magnetic resonance imaging(MRI) is a robust technique, withmeasured errors of �5%.26–29 Unlikeultrasound, MRA does not require

the operator to track the vessel ofinterest and thus avoids many of theproblems of measurement errorinherent in ultrasound. Magnetic res-onance angiography also providesmore detailed imaging and evalua-tion of blood flow characteristicsand cerebral perfusion than can beachieved using ultrasound.

It is important for manual therapypractitioners to be aware of theeffects their treatments may have onthe craniocervical blood vessels andbrain perfusion in order to avoid orminimize the use of any techniquesthat may have a greater effect onblood supply to the brain. It is pos-sible that rotation combined withlongitudinal distraction, a commoncomponent of manipulation tech-niques advocated by Cyriax andCyriax,30 may further increase themechanical stress on the arteries. Incontrast, it may be considered thatmanual techniques performed closerto the overall neutral position of theneck, perhaps involving more local-ized segmental rotation, such asthose described by Maitland31 andHing et al,32 will have less effect onthe arteries.

There have been few studies evalu-ating the effect on blood flow in thecraniocervical vessels during specificmanual therapy treatments, and all ofthose studies utilized ultra-sound.18,33,34 Some patient studiesexamined blood flow in neck rota-tion and extension to determine pre-surgical risk of stroke and showedthat vertebral artery blood flow wasaffected by contralateral rotation inabout 50% of patients, but thepatients investigated in these studieswere older individuals with symp-tomatic ischemic cerebrovasculardisease.35,36 We were unable to findany MRI studies of young individualswho were asymptomatic. Notably,no studies were identified that haveinvestigated blood flow during spe-

Manual Therapy Interventions for Mechanical Neck Pain

1564 f Physical Therapy Volume 93 Number 11 November 2013 by guest on December 2, 2014http://ptjournal.apta.org/Downloaded from


cific manual therapeutic techniquesusing MRI.

The purpose of this study, therefore,was to examine the effect of com-mon manual therapy procedures onblood flow in the craniocervicalarteries and blood supply to thebrain in participants who werehealthy using MRA. The objectivewas to determine whether therewere any differences in craniocervi-cal arterial flow or total blood supplyto the brain between the positionsinvolved in selected manual therapytechniques, which may help informrisk assessment by manual therapypractitioners. The effect of the thrustcomponent was not examined in thisstudy.

The specific research questionswere:

1. Do certain neck positions used incommon manual therapy proce-dures cause a greater differencein blood flow in the craniocervi-cal arteries compared with theneutral position than others?

2. Do certain neck positions used incommon manual therapy proce-dures cause greater difference intotal blood supply to the braincompared with the neutral posi-tion than others?

MethodDesignThe study was an experimentalMRI study examining blood flow inthe vertebral and internal carotidarteries in the neutral neck positionand comparing these measurementswith blood flow measurements in 7other neck positions used in com-mon manual therapy procedures.Informed consent was obtained fromall participants.

ParticipantsVolunteers who were between 18and 65 years of age and had no

reported mechanical neck pain orheadache were recruited into thestudy by advertisement and word ofmouth. Potential participants werescreened by interview and excludedif they reported any of the following(1–4 are standard contraindicationsfor manual therapy to the neck):(1) diagnosed inflammatory joint dis-ease; (2) any history of serious cervi-cal spine trauma, such as fractures;(3) any congenital disorder recog-nized as being associated with hyper-mobility or instability of the uppercervical spine; (4) diagnosed verte-brobasilar artery insufficiency (VBI);(5) claustrophobia or discomfort inconfined spaces (standard contrain-dication for MRI); or (6) any contra-indication identified by the localhealth authority MRI safety screen-ing questionnaire.

Cervical range-of-motion assessmentand testing for VBI as per AustralianPhysiotherapy Association guide-lines12 were undertaken prior to MRIexamination to ensure that partici-pants had no neck pain or limitationto neck movement and that no signsor symptoms of VBI were elicited.Participants with potential signs andsymptoms of VBI were excludedbecause of the risk of brain ischemiaimposed by the sustained neck posi-tions required for the study.

Demographic data, including ageand sex, were collected for all par-ticipants. Time of flight MRA wasperformed in the neutral neck posi-tion to give an image of the anatomyof the cerebral circulation. Thisimaging was reviewed post hoc forthe presence of any vascular anoma-lies such as hypoplasia or aplasia of avessel or anatomical variants such asthe vertebral artery ending in theposterior inferior cerebellar artery,suggestive of an incomplete Circle ofWillis. Dominance of one vertebralor one internal carotid artery wasidentified and was determined byvisual inspection of its relative size

compared with the contralateralvessel.

Experimental ConditionsThe following sequence of neckpositions was used:

1. Neutral

2. Left rotation: the participant wasasked to turn his or her neck asfar as possible to the left.

3. Right rotation: same as left rota-tion but with a differentdirection.

4. Left rotation with distraction: astrong longitudinal stretch wasapplied to the neck by the inves-tigator (an experienced manipula-tive physical therapist); the headthen was fully rotated to the leftside and held in the rotated posi-tion while the distraction wasmaintained30 (Fig. 1A).

5. Right rotation with distraction:same as left rotation with dis-traction but with a differentdirection.

6. Left rotation localized to C1–C2:the C2 spinous process was stabi-lized in the neutral position bythe thumb and index finger of theinvestigator, and the participant’shead was rotated to the left untilend-range was perceived, localiz-ing rotation to the C1–C2 seg-ment31 (Fig. 1B).

7. Right rotation localized to C1–C2:same as left rotation localized toC1–C2 but with a differentdirection.

8. Distraction: a longitudinal stretchof the neck with the head in neu-tral then was applied by the inves-tigator, with one hand under theparticipant’s occiput and theother hand under the partici-pant’s chin.




9. Posttest neutral

The neck positions were selected toreplicate as closely as possible com-mon manual therapy procedureswithout application of the thrustcomponent. Neck rotation is a com-

mon component of a number ofmanual therapy procedures. Neckdistraction with rotation is a positiondescribed by Cyriax and Cyriax.30

This manual therapy procedureinvolves a longitudinal stretch (dis-traction) being applied to the neck

by the practitioner and the neckthen being taken to the limit ofrotation. The premanipulative posi-tion for this procedure was chosenbecause it might be expected toapply some degree of stress to thecraniocervical arteries, as high-lighted by recently published inter-national guidelines.37 Localizedrotation of the C1–C2 segment waschosen to simulate the premanip-ulative position for a proceduredescribed by Maitland31 in which theC2 vertebra is fixed by the operator’sthumb and index finger and the neckrotated by the other hand. This pre-manipulative position was chosenbecause it might be expected toapply less stress to the craniocervicalarteries. A second, posttest neutralposition measurement was taken toassess the variability of blood flowvolume.

Participants lay supine on the scan-ner bed with their head in a phasedarray head coil. This device is a rigidplastic box that encloses the headabove and on either side, with a barthat passes anterior to the chin.There is a space of approximately 4cm separating the box from the par-ticipant’s head on all sides (Fig. 2).Participants were asked to reportany restriction to rotation imposedby the box and were observed by theoperator. Participants were moni-tored closely throughout and imme-diately following the procedure forany symptoms or signs of discom-fort, claustrophobia, or vertebrobasi-lar insufficiency, in the case of whichthe examination would have beenterminated.

Measurement of Blood Flow inthe Craniocervical ArteriesBlood flow in each of the 4 cra-niocervical arteries was measuredwith MRI using a phase-contrastflow quantification sequence. Allparticipants were imaged on a 3-Tsuperconducting magnet (SiemensMagnetom Verio, Siemens AG, Erlan-

Figure 1.Participant position within the head coil for (A) rotation/distraction technique and (B)localized C1–C2 rotation technique.




gen, Germany). Participants werescanned with T1-weighted sagittaland axial images and 2-dimensionaltime of flight angiography. A retro-spective cardiac-gated phase-contrastflow quantification sequence wasused (repetition time�29 millisec-onds, echo time�7 milliseconds, flipangle�30°, slice thickness�6 mm,matrix�192 � 512, field of view�200, and number of excitations�1).This is a standard sequence availableon this imaging unit.38 A velocityencoding value of 100 cm/s wasused. The arterial plane of sectionwas selected to intersect the top ofthe atlas loop of the vertebral arteriesat the level of the C1 vertebra, withimaging extending to just below theatlas loop. The atlas loop segmentwas chosen because this is wheremost manipulative injuries havebeen reported to occur and, there-fore, where most changes in bloodflow might be expected to occur.The acquisition time was approxi-mately 2 minutes.

Blood flow measurements were ana-lyzed post hoc using the proprietarysoftware syngo Argus (Siemens AG).In order to analyze blood flow usingthe Argus system, a region of interestwas placed around each artery foreach of the neck positions. In orderto assess the effect of neck positionon blood flow, average blood flowvolume measured in milliliters persecond was used as the primary testvariable and was analyzed in neutraland each of the neck positions foreach artery. Average blood flow vol-ume in each artery then was com-pared between the neutral positionand each of the experimental neckpositions to determine whetherblood flow volume changed fromneutral. Velocity was not used forcomparison because it is subject towide variation depending on thearea of sampling, which is particu-larly a problem in small tortuousvessels.14,26,39

Measurement of Total BloodSupply to the BrainTotal blood supply to the brain wasdetermined from the sum of averageflow volume (mL/s) in both vertebraland both internal carotid arteries.The total blood supply for each ofthe neck positions then was com-pared with neutral. A difference inaverage total supply (increase ordecrease) of �10% compared withthe neutral position was consideredto be clinically important.40

Data AnalysisDescriptive statistics were used tosummarize the demographic data.Average blood flow in all 4 arteriesfor the neutral position and each ofthe experimental neck positions wasanalyzed with descriptive statisticsand tested for normal distribution. Alinear mixed-effects model was fit foreach artery (left, right) using a single-effect variable (neck position) toassess whether there were any differ-

ences between positions. The mod-els were fitted using SAS version 9.2(TS2M2) and SAS 2008 Proc Mixedtechnique (SAS Institute Inc, Cary,North Carolina) with restricted max-imum likelihood estimation and withthe Kenward-Roger adjustment fordownward bias in the variance-covariance matrix. Compound sym-metry was used. If the effect of aneck position was significant,follow-up testing of pairs of meanswas undertaken in 2 ways using Dun-nett’s adjustment to compare theneutral position with the other neckpositions and to examine all pairs ofmeans using a Bonferroni adjustmentto the significance level. For posthoc power analysis, the variabilitieswere determined using a random-effects mixed model for each artery,combining data for all 4 arteries, todetermine standard deviations dueonly to measurement and neck posi-tion sources of variation.

Figure 2.Participant positioned in scanner showing head and neck coil.




Role of the Funding SourceThe study was funded by a grantfrom The University of Newcastle.

ResultsTwenty participants (10 male, 10female) with a mean age of 33.1years (SD�11.9) were recruited intothe study. All participants had nor-mal anatomy of their craniocervicalarterial circulation (Fig. 3). Threeparticipants (15%) had dominance ofone vertebral artery. No participantsexperienced any signs or symptomsof vertebrobasilar insufficiency inany of the neck positions. One par-

ticipant was excluded due to previ-ously unknown claustrophobiawhen scanning commenced.

Effect of Neck Position onBlood FlowMeasurements of average bloodflow volume (mL/s) for the internalcarotid and vertebral arteries in theneutral condition and each of theexperimental conditions are pre-sented in Table 1.

In order to compare differences inaverage volume flow between all 9positions to see if any of the neck

positions had greater effect on anyartery than another, we analyzedflow volume using a linear mixed-effects model with one categoricalvariable with 9 levels (neutral, leftrotation, right rotation, left rotation/distraction, right rotation/distrac-tion, C1–C2 left rotation, C1–C2right rotation, distraction, and a sec-ond posttest neutral position). Thesignificance level was set at ��.05,and following a significant effect,post hoc testing was carried out. Theresults showed no significant differ-ences in flow volume for either ofthe vertebral arteries (right vertebralartery, P�.28; left vertebral artery,P�.47) but a statistically significantdifference between positions forthe right internal carotid artery(P�.001) and left internal carotidartery (P�.01) (shown at bottom ofTab. 2).

Further analysis of the significantresults for the left internal carotidartery using the Dunnett test, inorder to adjust for multiple testsand comparing neutral with all theother positions, showed no signifi-cant differences, despite the signifi-cant overall test result. Similarly,analysis of the right internal carotidartery values using the Dunnett testshowed no significant differences.Further exploration of all pair-wisedifferences using a Bonferroni adjust-ment to control for family-wise errorrate (��.05/36�0.0014) showed nodifference for the left internal carotidartery, but 4 combinations were sig-nificantly different for the right inter-nal carotid artery. All combinationsinvolved left rotation. Mean averageflow volume for the right internalcarotid artery in left rotation wasunusually low and statistically differ-ent from the other positions. Remov-ing left rotation from the analysisresulted in a nonsignificant resultfor the right internal carotid artery.As the Dunnett test was negativefor all test positions compared withneutral, and because significant

Figure 3.Surface-rendered 3-dimensional multiplanar reformatted image of the carotid andvertebral arteries of a participant showing (A) normal anatomy and no dominance ofany vessel and (B) hypoplastic right vertebral artery and dominance of left vertebralartery.




changes were not similarly demon-strated for the left internal carotidartery in right rotation, it is likely thatthis finding for the right internalcarotid artery in left rotation is dueto a statistical anomaly.

Although mean values of averageflow volume were not significant forany position, there were certain indi-viduals with marked flow changes insome positions. Figure 4 shows par-allel plots of flow volume in all posi-

tions to demonstrate individualswith large changes.

Blood Supply to the BrainThe total blood inflow (mL/s) to thebrain for each position is shown in

Table 1.Mean (SD) Average Blood Flow Volume (mL/s) in the Craniocervical Arteries for Each Neck Position and the Mean Difference(mL/s, %) Between Each Neck Position and the Neutral Position With Linear Mixed Model Testing by Artery (P Value)a

Neck Position

Craniocervical Arteries Difference Between Neck Position and Neutral

RICA LICA RVA LVA

RICA LICA RVA LVA

NeckPositionMinus

Neutral

NeckPositionMinus

Neutral

NeckPositionMinus

Neutral

NeckPositionMinus

Neutral

Neutral 2.57 (0.86) 2.71 (1.03) 0.73 (0.62) 0.98 (0.64) N/A N/A N/A N/A

Left rotation 2.22 (0.76) 2.66 (0.97) 0.53 (0.72) 1.12 (0.71) �0.4 (�14%) �0.1 (�2%) �0.2 (�27%) 0.14 (14%)

Right rotation 2.61 (1.0) 2.56 (0.96) 0.80 (0.70) 0.87 (0.59) 0.04 (2%) �0.2 (�6%) 0.1 (10%) �0.1 (�11%)

Left rotation/distraction

2.79 (1.19) 3.01 (1.29) 0.65 (0.87) 0.87 (0.50) 0.2 (9%) 0.3 (11%) �0.1 (�11%) �0.1 (�11%)

Right rotation/distraction

2.98 (1.01) 2.61 (1.19) 0.73 (0.79) 0.86 (0.61) 0.4 (16%) �0.1 (�4%) 0.0 (0%) �0.1 (�12%)

C1–C2 leftrotation

2.84 (1.53) 3.01 (0.97) 0.66 (0.74) 1.03 (1.10) 0.3 (11%) 0.3 (11%) �0.1 (�10%) 0.1 (5%)

CI–C2 rightrotation

2.76 (0.87) 2.86 (0.96) 0.74 (0.69) 1.04 (0.71) 0.2 (7%) 0.2 (6%) 0.0 (1%) 0.1 (6%)

Distraction 2.83 (1.10) 2.68 (1.08) 0.55 (0.64) 0.88 (0.68) 0.3 (10%) �0.0 (�1%) �0.2 (�25%) �0.1 (�2%)

Posttest neutral 2.94 (1.0) 2.85 (0.91) 0.64 (0.55) 1.14 (1.0) 0.4 (14%) 0.1 (5%) �0.1 (�12%) �0.2 (�16%)

P valueb �.001 .01 .28 .47

a RICA�right internal carotid artery, LICA�left internal carotid artery, RVA�right vertebral artery, LVA�left vertebral artery, N/A�not applicable.b Linear mixed model P value for test of difference among the 9 position means.

Table 2.Total Blood Inflow to the Brain Measured by Sum of Average Flow Volume (mL/s [95% Confidence Interval]) in Internal Carotidand Vertebral Arteries: Difference Between Neck Position and Neutral (mL/s, P Value, and Percentage Difference)

Neck PositionTotal Blood

Inflow

Difference Between Neck Position And Neutral

Neck PositionMinus Neutral P

PercentageDifference

Neutral 6.98 (5.83, 8.13) N/A N/A N/A

Left rotation 6.52 (5.43, 7.62) �0.5 .18 �6.6

Right rotation 6.84 (5.65, 8.03) �0.1 .68 �2.0

Left rotation/distraction 7.31 (5.96, 8.66) 0.3 .33 4.7

Right rotation/distraction 7.17 (5.78, 8.56) 0.3 .57 4.3

C1–C2 left rotation 7.53 (5.94, 9.12) 0.6 .10 7.9

CI–C2 right rotation 7.41 (6.29, 8.52)1 0.4 .21 6.2

Distraction 6.93 (5.61, 8.26)2 �0.1 .89 �0.7

Posttest neutral 7.58 (5.64, 9.52) 0.6 .38 8.6

a N/A�not applicable.




A

RVA

neut

RVA

neut

2

RVA

lr

RVA

rr

RVA

lrdis

t

RVA

rrdi

st

RVA

c1lr

RVA

c1rr

RVA

dist

LVA

lr

LVA

rr

LVA

lrdis

t

LVA

rrdi

st

LVA

c1lr

LVA

c1rr

LVA

dist

LVA

neut

LVA

neut

2

RIC

Ane

ut

RIC

Ane

ut2

RIC

Alr

RIC

Arr

RIC

Alrd

ist

RIC

Arr

dist

RIC

Ac1

lr

RIC

Ac1

rr

RIC

Adi

st

LIC

Alr

LIC

Arr

LIC

Alrd

ist

LIC

Arr

dist

LIC

Ac1

lr

LIC

Ac1

rr

LIC

Adi

st

LIC

Anu

et

LIC

Ane

ut2

B

0

1

2

3

4

5

6

7

8

–2

–1

0

1

2

3

4

Low-flow RVA and high-flow LVA

Stable flow profile

High-flow RICA and low-flow LICAStable low-flow profile

mL/

s m

L/s

Figure 4.Individual parallel plots of flow volume (mL/s) in (A) vertebral arteries and (B) internal carotid arteries for all participants, with specificplots highlighted to show individuals with large variation. For each, one individual with stable flow is shown for comparison. R�right,L�left, VA�vertebral artery, ICA�internal carotid artery, RVAneut�right vertebral artery neutral, RVAneut2�right vertebral arteryposttest neutral, RVArr�right vertebral artery right rotation, RVAlr�right vertebral artery left rotation, RVAlrdist�right vertebral arteryleft rotation/distraction, RVArr�right vertebral artery right rotation, RVAc1lr�right vertebral artery left rotation at C1–C2,RVAc1rr�right vertebral artery right rotation at C1–C2, RVAdist�right vertebral artery distraction, LVAlr�left verterbral artery leftrotation, LVArr�left verterbral artery right rotation, LVAlrdist�left vertebral artery left rotation distraction, LVArrdist�left vertebralartery right rotation distraction, LVAc1lr�left vertebral artery left rotation at C1–C2, LVAc1rr�left vertebral artery right rotation atC1–C2, LVAdist�left vertebral artery distraction, LVAneut�left vertebral artery neutral, LVAneut2�left vertebral artery posttestneutral, RICAneut�right internal carotid artery neutral, RICAneut2�right internal carotid artery posttest neutral, RICAlr�rightinternal carotid artery left rotation, RICArr�right internal carotid artery right rotation, RICAlrdist�right internal carotid artery leftrotation/distraction, RICArrdist�right internal carotid artery right rotation/distraction, RICAc1lr�right internal carotid artery leftrotation at C1–C2, RICc1rr�right internal carotid artery right rotation at C1–C2, RICAdist�right internal carotid artery distraction,LICAlr�left internal carotid artery left rotation, LICArr�left internal carotid artery right rotation, LICAlrdist�left internal carotid arteryleft rotation distraction, LICArrdist�left internal carotid artery right rotation distraction, LICAc1lr�left internal carotid artery leftrotation at C1–C2, LICAclrr�left internal carotid artery right rotation at C1–C2, LICAdist�left internal carotid artery distraction,LICAneut�left internal carotid artery neutral, LICAneut2�left internal carotid artery posttest neutral.




Table 2. The total inflow to the braindid not vary substantially from neu-tral in any test position. Flow gener-ally decreased slightly for both theend-range rotation and distractionpositions but increased in the otherpositions. The total flow volume forall positions was analysed with a lin-ear mixed effects model. No differ-ence was found among positions(P�.06). In addition, flow changeswere all less than 10%, which is con-sidered to be the normal variation forcerebral inflow.40

Post Hoc Power CalculationWe were unable to perform a powercalculation prior to the study due tothe lack of published studies usingMRI, but with the benefit of the datacollected, we performed a post hocpower analysis. The standard devia-tions for each artery due to the sumof measurement and positionsources of variation were 0.40 forthe vertebral artery and 0.54 for theinternal carotid artery. The powercalculation was based on a compari-son of 2 neck positions with an inde-pendent samples t test. For the ver-tebral artery, for a 50% change inaverage flow volume between neu-tral and any neck position, powerwas 0.97; for a 36.4% change fromthe neutral position, power was0.80. For the internal carotid artery,for a 50% change in flow volume,power was 1, and even for as low avalue as 16.4% change, power wasstill 0.80.

DiscussionThis comparative MRI study exam-ined blood flow in the craniocervicalarteries in different neck positionsand compared the measurements ofblood flow with that in the neutralposition to identify if any neck posi-tions were potentially more hazard-ous than others. The study showedthat blood flow was not negativelyaffected by any of the neck positionsused and that no position had signif-icantly greater effect on blood flow

than any other. Total cerebral inflowalso remained fairly constant in allpositions, suggesting that cerebralperfusion similarly was not nega-tively affected by any of the neckpositions. To date, there have beenno previous studies of young individ-uals who were healthy investigatingblood flow in the craniocervical ves-sels in different neck positions usingMRI. The results of this study suggestthat common neck positions used inmanual therapy practice do not, inand of themselves, appear to pose arisk to blood flow to the brain, andconcerns about the safety of com-monly used neck positions may beunfounded, at least on the groundsof their effects on blood flow.

Blood flow in the craniocervicalarteries varied between neck posi-tions but was not significantlychanged by any neck positions usedin common manual therapy proce-dures compared with the neutralposition. In addition, no position hadany significantly greater effect onblood flow than any other, includingthose using potentially more stress-ful positions of end-range rotation orrotation and distraction. Previousstudies examining blood flowchanges during neck rotation havegenerally examined flow in specificarteries only, notably the vertebralarteries,10,17,18 which does not allowconsideration of the cerebral circula-tion as a whole. Concerns have beenraised that flow reduction in rotationmay represent a risk factor for neu-rovascular complication subsequentto cervical manual therapy.41 More-over, it has been suggested that cli-nicians should measure flow in thevertebral arteries as part of prema-nipulative screening of the cervicalspine.12 The current study contrastswith some previous studies wherereduction in flow velocity was dem-onstrated in the vertebral arteriesassociated with contralateral rota-tion.10,17,18 However, these were allultrasound studies that did not mea-

sure actual flow volume. Using MRA,Weintraub and Khoury35 showedthat blood flow may be reduced inthe vertebral artery on contralateralrotation, but their study was of olderpatients with advanced cerebro-vascular disease who do not repre-sent the population of interest forthe current study aim, and even inthis group only about 50% showedchanges in flow. Interestingly,although not statistically significant,the current study also suggests thatwhen flow is decreased in one vesselby a neck position, it appears to becompensated by increased flow inanother. For example, in left rota-tion, there was marked reduction inflow in the right internal carotid andvertebral arteries but with a markedincrease in flow in the left vertebralartery. Measurement of flow in thevertebral arteries for premanipula-tive screening purposes, therefore,would not seem to be particularlyuseful.

Studies also have demonstratedgreater reduction in blood flow withincreasing range of rotation16,18 andsuggested that end-range contra-lateral rotation positions might havea greater effect on blood flow thanmore localized segmental rotation.The results of this study suggestthat common manual therapy proce-dures including end-range neck rota-tion and rotation/distraction do notappear to be any more hazardous interms of blood flow than more local-ized positions. It is important toconsider, however, that although ahealthy artery should have the capac-ity to easily withstand such mechan-ical stress, the effects on arteries thatare in a weakened state either fromunderlying arteriopathy or a tempo-rary friable state due to infection orthe potential presence of proinflam-matory factors in the circulation areas yet unknown in the human model.Future research into the effects ofneck position on blood flow, there-fore, probably is not warranted.




However, other factors such as thestate of the arterial wall and theeffect of the manipulative thrust maybe more important determinants ofrisk from manual therapy applied tothe neck, and future studies couldinvestigate these factors.

The present study showed that totalblood inflow to the brain was notsignificantly changed from the neu-tral position by any of the test posi-tions, and none of the positionsshowed a compromise to total cere-bral blood supply. In the manipula-tion therapeutic positions, flow vol-ume was increased somewhat. Thisfinding was further supported by thefact that no signs or symptoms ofcerebral ischemia were evidentdespite marked reduction of flow inthe vertebral arteries of some indi-viduals. A previous ultrasound studyalso demonstrated marked reductionor even cessation of vertebral arteryflow on neck rotation in some indi-viduals, yet no signs of vertebrobasi-lar insufficiency being elicited,42 sug-gesting cerebral blood supply wasnot compromised. This conclusion isnot surprising given the homeostaticfunction of the Circle of Willis inmaintaining a constant blood supplyto the brain. The position of C1–C2rotation to the left showed a largechange in flow volume from neutral,which may help account for theoverall P value being close to signif-icance (.06); however, this changewas actually an increase in flow(7.9%) and does not suggest a risk tobrain perfusion. As shown in arecent study, small fluctuations(�10%) in flow volume to the brainnormally occur according to an indi-vidual’s arousal state and increase ifthe individual is anxious or doing amental task, which is not consideredto be detrimental to brain function.40

Clinically, the results of this studysuggest that even if blood flow inone vessel is markedly altered by aneck position or an individual haslow or absent flow in one vessel, it

may still be adequately compensatedfor via the other vessels, so concernsabout the safety of particular neckposition in terms of brain ischemiamay generally be unwarranted. How-ever, if the Circle of Willis is notintact (eg, missing a posterior com-municating artery), reduction in totalblood flow to the brain may be pos-sible, particularly if there is no othersource of collateral circulation. Inthe absence of radiological imaging,clinicians can generally rely only onproduction of signs or symptoms ofbrain ischemia to identify individualsat potential risk of an adverse neuro-vascular event.

Strengths and WeaknessesThe main strengths of this study arethe use of MRI, which provides high-quality images of the craniocervicalarteries, and the ability to track theseimages in different neck positions,which has not been undertaken pre-viously. The advantage of using MRIis that flow can be measured simul-taneously in all 4 vessels, whichallows comparison of the overalleffect of the different neck positionson blood flow, as well as the estima-tion of total cerebral inflow. Thestudy used neck positions designedto replicate those used in typicalmanual therapy practice, so theresults can be related to the clinicalsituation. In addition, the neck posi-tions selected allowed comparisonbetween potentially more stressfulend-range rotation and rotation/dis-traction positions with more local-ized segmental rotation positions inorder to see whether any were morehazardous than another in terms ofeffects on blood flow.

There were some limitations of thestudy, including the scanner setup,with the participant’s head enclosedwithin a head and neck coil. In orderto achieve the manipulation posi-tions of the neck as described, it wasnecessary to modify some hand posi-tions: (1) to have the top hand

around the chin and the underneathhand under the occiput for therotation-distraction position and (2)to rotate the head from the vertex ofthe skull rather than the chin for theC1–C2 rotation position. However,only one hand was in the coil for themanipulation positions and was posi-tioned under the head and neck, sodid not limit the amount of rotationavailable. Although it was more chal-lenging to perform the technique, itwas possible to replicate the tech-nique position as described. Due toMRI safety restrictions, it was notpossible to formally evaluate rangeof rotation inside the coil. In addi-tion, in order to minimize the timespent in each position, only a shortsegment of the arteries was exam-ined at the level of C1; however, it ispossible that during the rotationpositions the arteries moved slightlycephalad, causing a less straight sec-tion of artery to be imaged. Thisapproach may have made it moredifficult in some cases to accuratelyidentify the center of the arteriallumen in cross-section to undertakeflow analysis. Flow analysis, there-fore, may have been performednearer the arterial wall in someinstances where flow might be morevariable. Future studies could imagea longer section of artery to mini-mize this problem when associatedacquisition times are reduced byimprovements in technology.

Future ResearchOne of the key components of neckmanipulation is the HVT. It was notpossible during this study to look atthe effects of the manipulative thruston blood flow as the participant’shead would have had to be removedfrom the head coil for the procedureand then returned to it, making itdifficult to compare flow measuresbecause the baseline neutral positionwould be slightly different. Futurestudies could examine the effects ofthe manipulative thrust on bloodflow in the craniocervical arteries.




The study examined relatively youngparticipants who were healthy andin whom it was assumed arterieswere healthy and without anypathology. The effect of neck posi-tion on blood flow in arteries withsome form of arteriopathy may bedifferent and could be investigatedin the future. Future studies alsocould look at other neck positionsused in manual therapy, such asextension or lateral flexion. The posthoc power calculation may be usefulto determine appropriate samplesize for future flow studies.

ConclusionAlthough concerns have been raisedabout the safety of manual therapyapplied to the neck, in particular theupper cervical spine segments, noneof the positions tested in this studydemonstrated any significant changein blood flow volume from the neu-tral position. Moreover, no positionincluding end-range rotation, uppercervical rotation, or strong distrac-tion had any greater effect on bloodflow than any other. Total blood sup-ply to the brain was not adverselyaffected by any positions, and inmost positions relating to commonmanual therapy procedures such asrotation/distraction and C1–C2 rota-tion, supply was increased some-what. Reduction in flow in one ves-sel appeared to be compensated forby an increase in another. This find-ing suggests that the neck positionsthemselves are not inherently haz-ardous in terms of compromise toblood flow in the craniocervicalarteries, and it is more likely, there-fore, that other factors such as thestate of the arteries and the effect ofthe manipulative thrust may be moreimportant. Future imaging studiesfocusing on blood flow in normal orindividual craniocervical arteriesmay not be particularly useful.

All authors provided concept/idea/researchdesign. Ms Thomas, Dr Rivett, Dr Stanwell,and Professor Levi provided writing. Ms

Thomas, Dr Bateman, and Dr Stanwell pro-vided data collection. Ms Thomas, Dr Rivett,Dr Bateman, and Professor Levi provideddata analysis and project management. MsThomas provided study participants. DrBateman provided facilities/equipment. DrRivett, Dr Bateman, and Professor Levi pro-vided institutional liaisons. Dr Stanwell andProfessor Levi provided consultation (includ-ing review of manuscript before submission).

The study protocol was approved by TheUniversity of Newcastle Human ResearchEthics Committee.

The study was funded by a grant from TheUniversity of Newcastle.

DOI: 10.2522/ptj.20120477

References1 Thiel HW, Bolton JE, Docherty S, Portlock

JC. Safety of chiropractic manipulation ofthe cervical spine: a prospective nationalsurvey. Spine (Phila Pa 1986). 2007;32:2375–2378.

2 Haldeman S, Kohlbeck FJ, McGregor M.Risk factors and precipitating neck move-ments causing vertebrobasilar artery dis-section after cervical trauma and spinalmanipulation. Spine (Phila Pa 1986).1999;24:785–794.

3 Cassidy D, Boyle E, Cote P, et al. Risk ofvertebrobasilar stroke and chiropracticcare: results of a population-based case-control and case-crossover study. Spine(Phila Pa 1986). 2008;33(4 suppl):S176–S183.

4 Frisoni G, Anzola G. Vertebrobasilar isch-aemia after neck motion. Stroke. 1991;22:1452–1460.

5 Assendelft WJJ, Bouter LM, Knipschild PG.Complications of spinal manipulation: acomprehensive review of the literature.J Fam Pract. 1996;42:475–480.

6 Caso V, Paciaroni M, Bogousslavsky J.Environmental factors and cervical arterydissection. In: Baumgartner RW, Bogouss-lavsky J, Caso V, Paciaroni M, eds. Hand-book on Cerebral Artery Dissection. Vol20. Basel, Switzerland: Karger; 2005:44–53.

7 Hurwitz E, Aker P, Adams AH, et al. Manip-ulation and mobilization of the cervicalspine. A systematic review of the litera-ture. Spine (Phila Pa 1986). 1996;21:1746–1760.

8 Paciaroni M, Bogousslavsky J. Cerebrovas-cular complications of cervical manipula-tion. Eur Neurol. 2009;61:112–118.

9 Ohsaka M, Takgami M, Koyanagi I, et al.Cerebral ischaemia originating from rota-tional vertebral artery occlusion caused byC5/6 spondylotic changes: a case report.No Shinkei Geka. 2009;37:797–802.

10 Mitchell J, Kramschuster K. Real-timeultrasound measurements of changes insub-occipital vertebral artery diameter andblood flow associated with cervical spinerotation. Physiother Res Int. 2008;13:241–254.

11 Magarey M, Rebbeck T, Coughlan B, et al.Pre-manipulative testing of the cervicalspine review, revision and new clinicalguidelines. Man Ther. 2004;9:95–108.

12 Rivett D, Shirley D, Magarey M, RefshaugeKM. APA Clinical Guidelines for Assess-ing Vertebrobasilar Insufficiency in theManagement of Cervical Spine Disorders,2006. Melbourne, Australia: AustralianPhysiotherapy Association, Musculoskele-tal Physiotherapy Australia; 2006.

13 Arnold M, Pannier B, Chabriat H, et al. Vas-cular risk factors and morphometric datain cervical artery dissection: a case-controlstudy. J Neurol Neurosurg Psychiatry.2009;80:232–234.

14 Debette S, Leys D. Cervical-artery dissec-tions: predisposing factors, diagnosis andoutcome. Lancet Neurology. 2009;8:668–678.

15 Thomas L, Rivett D, Levi C. Risk factorsand clinical features of craniocervical arte-rial dissection. Man Ther. 2011;16:351–356.

16 Refshauge KM. Rotation: a valid prema-nipulative dizziness test? Does it predictsafe manipulation? J Manipulative PhysiolTher. 1994;17:15–19.

17 Mitchell JA. Changes in vertebral arteryblood flow following normal rotation ofthe cervical spine. J Manipulative PhysiolTher. 2003;26:347–351.

18 Zaina C, Grant R, Johnson C, et al. Theeffect of cervical rotation on blood flow inthe contralateral artery. Man Ther. 2003;8:103–109.

19 Arnetoli G, Amadori A, Stephani P, Nuz-zaci G. Sonography of vertebral arteries inDe Kleyn’s position in subjects and inpatients with vertebrobasilar transientischemic attacks. Angiology. 1989;40:716–720.

20 Haynes M, Cala LA, Melsom A, et al. Ver-tebral arteries and cervical rotation: mod-elling and magnetic resonance angiogra-phy studies. J Manipulative Physiol Ther.2002;25:370–383.

21 Weingart JR, Bischoff HP. Doppler sonog-raphy of the vertebral artery with regard tohead positions appropriate to manualmedicine. Manuelle Medizin. 1992;30:62–65.

22 Thiel H, Wallace K, Donat J, Yong-Hing K.Effect of various head and neck positionson vertebral artery blood flow. Clin Bio-mech. 1994;9:105–110.

23 Freed KS, Brown LB, Carroll MD. Theextracranial cerebral vessels. In: RumackCM, Wilson SR, Charboneau JW, eds.Diagnostic Ultrasound. Vol 1. 2nd ed. StLouis, MO: Mosby Year Book Inc; 1998:885–919.

24 Zwiebel WJ. Introduction to VascularUltrasonography. 4th ed. Philadelphia,PA: Harcourt Health Sciences; 2000.

25 Rodallec MH, Marteau V, Gerber S, et al.Craniocervical arterial dissection: spec-trum of imaging findings and differentialdiagnosis. Radiographics. 2008;28:1711–1728.




26 Evans A, Iwai F, Grist T. MR imaging ofblood flow with a phase subtraction tech-nique: in vitro and in vivo validation.Invest Radiol. 1993;28:109–115.

27 Laffon E, Valli N, Latrabe V, et al. A valida-tion of a flow quantification by MR phasemapping software. Eur J Radiol. 1998;27:166–172.

28 Paciaroni M, Caso V, Agnelli G. Magneticresonance imaging, magnetic resonanceand catheter angiography for diagnosis ofcervical artery dissection. In: BaumgartnerRW, Bogousslavsky J, Caso V, Paciaroni M,eds. Handbook of Cerebral Artery Dissec-tion: Frontiers in Neurology and Neuro-science. Vol 20. Basel, Switzerland: Karger;2005:102–118.

29 Power A, Maier S, Chung T, Geva T. Phase-velocity cine magnetic resonance imagingmeasurement of pulsatile blood flow inchildren and young adults: in vitro and invivo validation. Pediatr Cardiol. 2000;21:104–110.

30 Cyriax J, Cyriax P. Cyriax’s IllustratedManual of Orthopaedic Medicine. 2nd ed.London, United Kingdom: ButterworthHeinemann; 2000.

31 Maitland GD. Vertebral Manipulation.5th ed. London, United Kingdom: Butter-worth; 1986.

32 Hing W, Reid D, Monaghan M. Manipula-tion of the cervical spine. Man Ther. 2003;8:2–9.

33 Bowler N, Shamley D, Davies R. The effectof a simulated manipulation position oninternal carotid and vertebral artery bloodflow in healthy individuals. Man Ther.2011;16:8–93.

34 Licht PB, Christensen HW, Svendsen P,Hoilund-Carlsen PF. Vertebral artery flowand cervical manipulation: an experimen-tal study. J Manipulative Physiol Ther.1999;22:431–435.

35 Weintraub MI, Khoury A. Critical neckposition as an independent risk factor forposterior circulation stroke: a magneticresonance angiographic analysis. J Neuro-imaging. 1995;5:16–22.

36 Weintraub MI, Khoury A. Cerebral hemo-dynamic changes induced by simulatedtracheal intubation: a possible role in peri-operative stroke? Magnetic resonanceangiography and flow analysis in 160 cas-es. Stroke. 1998;29:1644–1649.

37 Rushton A, Rivett D, Carlesso L, et al.International Framework for Examina-tion of the Cervical Region for Potentialof Cervical Arterial Dysfunction Prior toOrthopaedic Manual Therapy Interven-tion. Auckland, New Zealand: Interna-tional Federation of Orthopaedic Manipu-lative Physical Therapists; 2012.

38 Bateman GA. The pathophysiology of idio-pathic normal pressure hydrocephalus:cerebral ischaemia or altered venoushemodynamics? Am J Neuroradiol. 2008;29:198–203.

39 Tay KY, King-Im JM, Trivedi RA, et al.Imaging the vertebral artery. Eur Radiol.2005;15:1329–1343.

40 Kreiger SN, Streicher MN, Trampel R,Turner R. Cerebral blood volume changesduring brain activation. J Cereb BloodFlow Metab. 2012;32:1618–1631.

41 Mitchell JA, Keene D, Dyson C, et al. Iscervical spine rotation, as used in the stan-dard vertebrobasilar insufficiency test,associated with a measureable change inintracranial vertebral artery blood flow?Man Ther. 2004;9:220–227.

42 Rivett D, Sharples K, Milburn P. Effect ofpre-manipulative tests on vertebral arteryand internal carotid blood flow: a pilotstudy. J Manipulative Physiol Ther. 1999;22:368–375.




doi: 10.2522/ptj.20120477Originally published online June 27, 2013

2013; 93:1563-1574.PHYS THER. Stanwell and Christopher R. LeviLucy C. Thomas, Darren A. Rivett, Grant Bateman, PeterCarotid Arterial Blood Flow and Cerebral InflowMechanical Neck Pain on Vertebral and Internal Effect of Selected Manual Therapy Interventions for

References

http://ptjournal.apta.org/content/93/11/1563#BIBLfor free at: This article cites 37 articles, 5 of which you can access

Information Subscription http://ptjournal.apta.org/subscriptions/

Permissions and Reprints http://ptjournal.apta.org/site/misc/terms.xhtml

Information for Authors http://ptjournal.apta.org/site/misc/ifora.xhtml

by guest on December 2, 2014http://ptjournal.apta.org/Downloaded from

http://ptjournal.apta.org/content/93/11/1563#BIBL

http://ptjournal.apta.org/subscriptions/

http://ptjournal.apta.org/site/misc/terms.xhtml

http://ptjournal.apta.org/site/misc/ifora.xhtml


Documents

PHYS THER-2013-Thomas-1563-74