1765

The Cold Hemiplegic Arm

Peter Wanklyn, MRCP; David W. Ilsley, MB, ChB; David Greenstein, BSc, MB, ChB;

Ian F.G. Hampton, BSc, PhD; T.A. Roper, MRCP;Ralph C. Kester, MD, ChM, FRCS; Graham P. Mulley, DM

Background and Purpose Vasomotor changes occur in thearm after hemiplegic stroke. Previous studies have providedconflicting results, with most showing an increase in skintemperature of the hemiplegic arm. However, a number ofpatients complain of distressing coldness of the hemiplegicarm.

Methods Eleven patients with symptomatic coldness and 10patients with hemiplegia but no coldness were recruited. Theseverity of the symptom of coldness was compared by ques-tionnaire with other common symptoms after stroke. A ther-mographic camera was used to record the finger skin temper-ature response to cold stress. Blood flow to both hands wasalso measured simultaneously by means of two plethysmo-graphs. In all patients there were no symptoms in the unaf-fected arm, and this was used as a control.

Results The symptom of coldness rated highly comparedwith other symptoms. In the symptomatic group the finger

Several unilateral phenomena occur after paralyticstroke.1 We have noticed that some patientscomplain of coldness in the hemiplegic hand and

arm. This is often distressing and difficult to relieve withsimple local measures.

Vasomotor changes have been previously described inhemiplegic limbs.2-6 Whereas our patients complain ofcoldness, most reports have described an increase in skintemperature of the hemiplegic arm.2-4-6 One study thatused thermography found that the hemiplegic arm wascolder by 1°C to 5°C,7 and the only study that usedaxillary temperature measurements also showed re-duced local temperature.8 An interarm difference inblood pressure of 2 to 5 mm Hg has been shown inpatients with hemiplegia.9 With 24-hour blood pres-sure monitoring, an interarm difference of 4 mm Hgwas found in patients with acute hemiplegia, whichwas not related to muscle tone.10

The aims of this study were (1) to assess the degree ofdiscomfort caused by coldness of the hemiplegic upperlimb, (2) to objectively verify temperature differencesbetween the hemiplegic and normal arms, and (3) tocompare blood flow and skin temperature responses toimposed temperature changes in hemiplegic and unaf-fected hands.

Received January 24, 1994; final revision received May 30, 1994;accepted June 17, 1994.

From Seacroft Hospital (P.W., D.W.I., D.G., R.C.K.), the Phys-iology Department, University of Leeds (I.F.G.H.), and the De-partment of Medicine for the Elderly, St James' University Hospital(G.P.M.), Leeds; and Hope Hospital, Salford (T.A.R.), UK.

Correspondence to Professor G.P. Mulley, Department of Med-icine for the Elderly, St James' University Hospital, Beckett St,Leeds, LS9 7TF, UK.

© 1994 American Heart Association, Inc.

temperature on the hemiplegic side was lower at rest (mediandifference at rest, 0.65°C; P<.0001) and at all times after coldstress. In the asymptomatic group the fingers on the hemiple-gic side were colder at rest and after initial cooling (mediantemperature difference, 0.2°C) but at no other time. Handblood flow on the hemiplegic side was also decreased in thesymptomatic group by 35%. This was not seen in the asymp-tomatic group.

Conclusions Coldness of the hand may be a severe anddistressing symptom in some patients after hemiplegia. Symp-tomatic patients have lower finger skin temperatures at restand after standard cold stress. These symptomatic patientsalso had reduced blood flow to the hemiplegic hand. (Stroke.1994^5:1765-1770.)

Key Words • hemiplegia • thermography • vasomotorsystem

Subjects and MethodsStroke patients were recruited from the clinics and wards at

Seacroft and St James' hospitals and stroke clubs in Leeds. Noattempt was made to define a sample frame because we werestudying a clinical phenomenon rather than performing anepidemiological survey. Eligible patients were those withunilateral weakness with or without the symptom of coldnessin the affected arm. Patients with severe dysphasia or withspasticity resulting in an inability to open the hand fully wereexcluded.

All patients gave informed consent to participate in the trial,and approval for the trial was obtained from the local ethicscommittee.

An initial assessment consisted of a questionnaire askingabout the presence and severity of symptoms in the arms. Ascale of 0 to 5 was used to quantify feelings of coldness as wellas heat, pain, pins and needles, and numbness in the arm. Thelimits of the scale were 0 (no symptoms) to 5 (very severe).This rating scale was performed to compare the subjectiveseverity of the symptom of coldness compared with some othercommon symptoms after stroke. Motor power was assessed ineach patient with the motoricity index. This involved powertesting of various movements. In the upper limb pinch grip,elbow flexion, and shoulder abduction were tested. In thelower limb ankle dorsiflexion, knee extension, and hip flexionwere tested. The motoricity index has been validated as ameasure of motor function after stroke.11 The maximum scoreattainable is 100, and when interpreting the score a highervalue represents greater motor power.

We used two methods to measure different aspects of thecirculation in the hands. A thermographic camera was used todetermine resting skin temperature in the fingers and theresponse to local cooling and subsequent rewarming. Venousocclusion plethysmography was used to measure the totalblood flow to the hands at a neutral temperature and aftercooling. We also performed plethysmography on 6 age-matched healthy control subjects who were taking no vasoac-

by guest on April 23, 2018

http://stroke.ahajournals.org/D

ownloaded from

1766 Stroke Vol 25, No 9 September 1994

TABLE 1. Characteristics of Stroke Patients With andWithout Cold Hemlpleglc Arms

TABLE 2. Mean Symptom Scores for Stroke PatientsWith and Without Cold Hemlpleglc Arms

Age,*y

Sex (M/F)

Hemiplegic side

Time since stroke,* mo

Motoricity score*

Smokers

Patients WithCold Arm

(n=11)

74.9(67-81)

10/1

5 left, 6 right

43.9(18-96)

75.6(57-100)

3

Patients WithoutCold Arm

(n=10)

74.2 (58-89)

5/5

2 left, 8 right

61.4(2-228)

83.6 (40-100)

2

*Values are mean, with range in parentheses.

tive medication at the time. This was to ascertain whetherthere is any asymmetry in the blood flow to normal hands inthis elderly group.

None of the patients in any of the groups had symptoms inthe unaffected hand, and this was used as a control in all cases.

We conducted a cold stress test using the method of Wilkinsonet al.12 On entering the laboratory, patients were allowed toacclimatize to indoor temperatures for 30 minutes and were thenseated in a constant temperature cabinet at 25 ± 1°C. Images weretaken at room temperature with a Starsight Thermographiccamera (Insight Vision Systems) with the images recorded di-rectly onto videotape. Latex gloves were placed on the hands,which were then immersed in water at 5°C for 1 minute. Tostandardize the cold stress, the patients were asked to keep thehands still and not allow them to touch the sides of the container.After removal of the gloves, images were recorded constantly

Symptom

No. Complaining ofSymptom

Patients PatientsWith Without

Cold Arm Cold Arm

Mean SymptomScore*

Patients PatientsWith Without

Cold Arm Cold Arm

Coldness

Pain

Heat

Numbness

Pins/needles

Weakness

Dependence

11

1

0

2

2

10

9

2.2

0.3

0

0.5

0.3

3.4

3.1

0

0.5

0.1

0

0

2

2.1

*Scored on a 5-point scale.

during the rewarming period of 10 minutes. The results wereanalyzed by a coworker who was unaware of the characteristics ofthe patients using the THERMOSOFT (EIC, Inc) computer soft-ware package. Values for the temperature of the tips and bases ofthe index, middle, and ring fingers were obtained for each timepoint in both hands.

Temperature results were obtained in the thermoneutralstate and at 0, 3, 5, and 10 minutes after cooling. The medianfinger temperatures at each time were calculated for bothhands.

We used water-filled plethysmographs to measure bloodflow in each hand. In healthy subjects there is a large normal

o

3 5 -

25

* Non-hemlpl»glc

•°" Hemiplegic

Pre 0 1 2 3 4 5 6

p<0.0005 p<0.0005 p<0.0005 p<0.02

7 8 9

Time (min)

10

p<0.002

Time (min)

Pre

0

3

5

10

Non-henuplegic hand (Temperature ° Q

Median

35.25

29.2

32.085

33.27

34.95

Std Error

0.34

0.65

0.42

0.35

0.36

StdDev

2.73

5.32

3.39

2.86

2.94

IQR

2.34

8.99

554

4.57

3.06

Hemiplegic hand (Temperature °C)

Median

34 6

29.1

30.05

32.8

34.3

Std Error

0.30

0.77

0.56

0.38

0 39

StdDev

2.46

6.25

4.57

3.07

3.22

IQR

3.18

9.88

5.43

5.24

4.19

FIG 1. Line graph of the medianfinger temperature in each handduring cold stress test in patientscomplaining of cold in the hemi-plegic hand. Interquartile rangesquoted are from 25th to 75th per-centiles and represent 50% of thevalues (Wilcoxon test). IQR indi-cates interquartile range.

by guest on April 23, 2018

http://stroke.ahajournals.org/D

ownloaded from

Wankfyn et al The Cold Hemiplegic Ann 1767

FIG 2. Tnermograpnic images ot tne palmar aspect or tne nanasat thermoneutral temperature (25°C) in a patient with left hemiple-gia and symptomatic coldness (top) and In the same patent 3minutes after cold stress (bottom). The darker colors indicatelower temperature. Note the marked asymmetry in temperature,with the fingers of the hemiplegic hand being colder.

variation in hand blood flow. To avoid misinterpretations dueto this variation, we have assumed that, in any normal individ-ual, changes in blood flow occur simultaneously in both hands.Therefore, we used two plethysmographs arranged in parallelto measure flow in the two hands simultaneously.

The patients were seated comfortably and the hands placedin the water chambers, which were above heart level. Wristcuffs were then applied on both sides, and the water in bothchambers was circulated by a pump at a thermostaticallycontrolled temperature, initially 32°C. To obtain a reading ofblood flow the cuffs were inflated simultaneously to 60 mm Hg.This allows arterial inflow but prevents venous return, thuscausing the hand to swell. The change in volume of the handwith time causes water to be displaced, which is measured witha transducer and printed onto a chart recorder. At eachtemperature eight to 10 readings of blood flow were obtainedand the mean value calculated. This procedure was repeatedwith water at temperatures of 30°C and 28°C bathing thehands. Preliminary work had shown that below this tempera-ture the hand flow in many patients had fallen dramatically toless than 1 mL-100 mL"'-min"'. Finally, the volume of thehand that had been within the chamber was ascertained bymeasuring the volume of water displaced from a container withan already known volume. From the slope of the tracingsobtained and the known hand volume, we calculated bloodflow per 100 mL of tissue per minute.

Because the data were of a nonparametric nature, theWilcoxon matched-pairs signed rank test was used to test forsignificance between hemiplegic and unaffected hands. TheMann-Whitney rank sum test was used to test for significancebetween the patients with and without symptomatic coldness.

ResultsOf the 21 patients studied, 11 complained of a cold

hemiplegic arm, and 10 denied coldness in the limb.

Patients were matched for age but not for sex or side ofhemiplegia. In addition, the symptomatic group had alower mean motoricity score (Table 1).

Mean symptom scores were calculated for both patientgroups (Table 2). Coldness of the arm was a prominentsymptom in our study group (mean symptom score of 2.2versus 3.4 for weakness).

All 21 patients underwent thermography. In thegroup of patients complaining of a cold arm, the medianresting temperature of all points in the hemiplegic handwas 0.7°C cooler than the unaffected hand (P<.0001).In addition, the hand affected by stroke remainedsignificantly cooler than the unaffected hand at all timesafter cold provocation (Fig 1). Fig 2 shows a typicalresult in a patient with hemiplegia and symptomaticcoldness of the arm.

In the asymptomatic group the temperature of thehand affected by stroke at rest and immediately aftercooling was lower than the unaffected hand. The me-dian resting temperature difference was only 0.2°C.During rewarming, the hemiplegic hand was eitherwarmer or no different in temperature than the unaf-fected hand (Fig 3). A degree of asymmetry betweenthe two hands existed in the symptomatic and asymp-tomatic groups. Comparing this degree of asymmetry inthe two groups, we found a significantly larger differ-ence in temperature between the two hands in thesymptomatic group at all time points (Fig 4).

Twenty of the 21 patients underwent plethysmogra-phy, 10 with cold arms and 10 subjects without thissymptom (Table 3). In the symptomatic group bloodflow to the hemiplegic hand was significantly lower thanthe unaffected hand at all temperatures. At 32°C themedian blood flow difference was 1.3 mL • 100mL"1 • min"1; at 30°C and 28°C the median differenceswere 1.1 and 0.8 mL • 100 mL"1 • min"1, respectively. Inthose patients not complaining of a cold arm, there was nodifference in blood flow between the hands at any tem-perature. In the small group of healthy age-matchedsubjects (n=6) there was also no difference in blood flowbetween the two hands at any temperature (Table 4).

DiscussionThis study was prompted by the observation that

some patients have an unpleasant feeling of coldness inthe hemiplegic arm after a stroke. The symptom devel-oped within days in some patients but more commonlytook weeks or months to become prominent. The sever-ity of the problem varied considerably, but most of ourpatients were constantly aware of the symptom andobtained only limited relief from local measures such asgloves or hot water bottles. While our subjective ratingsystem is crude, the comparison with other problemsafter stroke showed that coldness in the arm wassurprisingly prominent in our study group (mean symp-tom score of 2.2 versus 3.4 for weakness). It is of interestthat the group with coldness complained of more symp-toms overall.

We investigated skin temperature and blood flow, butanother possible mechanism producing the symptommight be abnormal temperature perception due to thestroke. Damage to the parietal lobe might cause alteredinterpretations of limb temperatures. In general, pa-tients with left-sided paralysis are more likely to haveperceptual problems than those with right-sided weak-

by guest on April 23, 2018

http://stroke.ahajournals.org/D

ownloaded from

1768 Stroke Vol 25, No 9 September 1994

rees

C)

1CD

P.T

empe

ra

3 4 -

32 -i

3 0 -

28

2 6 -

24-

22-

\

V-1

Pre 0

(

1

1 1

2 3

All

/

4

points

/Jk

1 i

5 6

p=ns

"^ Non-hemipleglc

-°- Hemiplegic

i i i

7 8 9 10

Time (mln)

Time (nun)

Pre

0

3

5

10

Non-hemiplegic hand (Temperature °C)

Median

31 3

22 6

25.7

28.58

29.4

Std Error

046

0.67

071

064

0 56

StdDev

3 47

509

5.37

4 8 4

4 26

IQR

3 57

690

8 52

7 47

5 23

Hemiplegic hand (Temperature ° Q

Median

31 5

22.6

25 7

29 3

30 8

Std Error

044

0 69

0.72

0.63

0 49

StdDev

3.37

5 27

5 4 6

4.79

3 80

IQR

3 61

7 15

9 20

8 37

5.82

Fra 3. Une graph of the medianfinger temperature in each handduring cold stress test in patientsnot complaining of cold In thehemiplegic hand. Interquartileranges quoted are from 25th to75th percentiles and represent50% of values (Wilcoxon test).IQR indicates interquartile range.

0.6-

0.2

• Symptomatic

I I Asymptomatic

5 $ -0.2

E <u2 Q

- 0 . 4 -Pre

p<0.02 p<0.05 p<0.005 p<0 02

7 8 9

Time (mln)p<0.0001

Time (rrun)

Pre

0

3

5

10

Temperature difference between the two hands( ° Q Median (IQR)

Symptomatic patients

0 45 (1 3)

0.4 (2.4)

0.3 (3.4)

0.3(1.1)

0 2(1.4)

Asymptomatic patienti

0.1(0.3)

0 3 (0 9)

-0 3 (0 9)

-0 1(1 1)

-0.3 (0.9)

FIG 4. Bar graph shows median temperaturedifference between hands during cold stresstest. Difference represents temperature of un-affected hand minus temperature of hemiplegichand. If the hemiplegic hand is colder, thedifference will be positive. There is a significantdifference between the symptomatic andasymptomatic patients at all time points (Mann-Whitney U test). IQR indicates interquartilerange.

by guest on April 23, 2018

http://stroke.ahajournals.org/D

ownloaded from

Wankfyn et al The Cold Hemiplegic Ann 1769

TABLE 3. Differences In Median Hand Blood Flow as Measured by Plethysmography

Blood Flow, mL • 100 mL"' • mlrr1

Patients With Cold Arm Patients Without Cold Arm

Temperatureof Water, °C

28

30

32

HemiplegicArm

1.5(1.9)

2.1 (3.6)

2.4 (3.9)

UnaffectedArm

2.3 (3.3)

3.2 (5.7)

3.7 (5.7)

P

<.O5

<.O5

<.05

HemiplegicArm

1.8(1.7)

2.9 (2.0)

2.5 (2.4)

UnaffectedArm

1.6 (1.0)

2.9 (4.4)

3.4 (3.7)

P

<.79

<.13

<.O7

Values in parentheses are interquartile range. Significant differences between the hemiplegic and unaffectedhands were only found in the patients complaining of cold ami (Wiicoxon test).

ness. In our small series, 6 of 11 symptomatic patientshad right-sided weakness, suggesting that altered per-ception is unlikely to be the major factor in the genesisof this phenomenon. However, sensory impairment orinattention may play a role in deciding who developssymptoms. One patient in the asymptomatic group hada markedly colder hemiplegic hand and reduced handblood flow. This patient had sensory deficits on exami-nation and a right parietal infarct on cranial computedtomographic scan. Patients with other neurological le-sions such as poliomyelitis develop cold limbs, and somework on peripheral circulatory changes has been report-ed.1316 Little evidence has been found for sympatheti-cally mediated vasospastic responses,16 although initialsuccess was claimed for sympathectomy in cold postpo-lio limbs.17 Instead, it is suggested that venous conges-tion of the limb results in uncontrolled heat loss.16

Venous congestion is suggested because of the violethue of extremities in these patients and vascular micros-copy showing dilated venules.14 Our patients did notshow any red or purple color change in the affectedlimb, and there was no evidence of edema on clinicalexamination.

Infrared thermography provides an accurate measureof skin temperature. Furthermore, it is noninvasive andwell tolerated by patients and has been used extensivelyin the evaluation of vasospastic conditions.18 There maybe a temperature difference of up to 0.25°C between thehands in healthy people.19 The thermographic resultsshowed that members of our group with symptomaticcoldness do indeed have lower resting finger tempera-tures in the hemiplegic hand, and the variation betweenthe two sides is well in excess of that seen in healthysubjects. Those hemiplegic stroke patients who did notcomplain of a cold arm had a much smaller difference intemperature between the hands at rest. This difference

TABLE 4. Differences In Median Hand Blood Row asMeasured by Plethysmography In Age-Matched ControlSubjects (n=6)

Water Temperature, °C

28

30

32

Blood Row,mL-IOOmL-1 • mln'1

Left Hand

4.4 (3.4)

6.3 (2.9)

7.4 (8.0)

Right Hand

4.25 (4.5)

6.45 (3.4)

6.65 (8.2)

P

.68

.89

.68

Values in parentheses are interquartile range. No significantdifferences in flow were detected between the two hands.

in temperature was within the normal range describedin healthy people. However, the finger temperatureswere lower in both hands in the asymptomatic group. Itis not clear why these patients did not complain ofcoldness. We did not test thermal sensory thresholds inour patients, but this has been studied formally in strokepatients.20 The sensory threshold to external skin cool-ing in the hemiplegic arm was elevated in 6 of 17patients. It was also found that sensory thresholds fortemperature were elevated in 3 of 10 patients with noclinically detectable deficit in temperature sensation,and ipsilateral alterations in temperature perceptionwere also noted. As already mentioned, one asymptom-atic patient was known to have a sensory deficit, whichmay explain the lack of cold sensation, since the armwas colder by several degrees.

Previous temperature studies in the hemiplegic limbhave been conflicting. Most of these studies showed anincreased temperature of the hemiplegic arm.2-46 Bucy5

suggested that there are two stages after hemiplegia,with initial warmth of the limb in the acute phase andcoldness in the chronic phase. More recent work foundthat the skin temperature in the axilla of the hemiplegicarm was lower than that in the unaffected arm.8 All ofthese reports have investigated unselected hemiplegicpatients with no perception of abnormal temperature inthe arm. By selecting a group of patients with symptom-atic coldness of the hemiplegic arm, we have shown alower skin temperature and markedly lower blood flowin the hemiplegic hand compared with the unaffectedhand. In some of the asymptomatic patients the affectedarm was indeed warmer than the contralateral arm, insome cases by more than 1°C, suggesting that hemiple-gic patients have a range of temperature changes in theaffected arm.

The blood flow to the hemiplegic hand was signifi-cantly reduced compared with the unaffected arm in thesymptomatic group but not in the asymptomatic group.We measured whole hand blood flow, which has ahigher proportion of flow for nutrition compared withfinger flow, which is largely related to thermoregulatoryfunction. This explains why there is no direct correlationbetween our blood flow measurements and finger tipskin temperature. The mechanism causing the reduc-tion in blood flow to the hemiplegic hand is unknown.One might postulate that simple disuse could accountfor the change, and this may be true for the forearmwith its large muscle groups. However, the hand com-prises proportionately more skin and less muscle, sothat sympathetically mediated vasomotor tone is prob-

by guest on April 23, 2018

http://stroke.ahajournals.org/D

ownloaded from

1770 Stroke Vol 25, No 9 September 1994

ably more important in determining blood flow thanlocal mediators produced by active muscle. Reflex sym-pathetic dystrophy is another complication of strokethat produces symptoms in the hand, but our patientsdid not exhibit other recognized features of the condi-tion. Only one symptomatic patient described any painin the arm. Pain is not always present in the extremity inreflex sympathetic dystrophy, although it is reported tooccur in the majority of cases.21 In addition, none of ourpatients reported worsening of the symptom after use,and none had evidence of swelling of the hand. Skintemperature, particularly in the fingers, largely reflectsthe state of tone in the skin vasculature. Large numbersof arteriovenous anastomoses are present in the fingers.Constriction in the precapillary sphincters due to vari-ous influences causes blood to be diverted away fromthe skin capillaries via these anastomoses. Constrictinginfluences include a2 receptor-mediated sympatheticstimulation, circulating norepinephrine, and possibly5 -hydroxytryptamine, although the former mechanismis by far the most important.22

Is it possible that vascular reactivity is altered in thehemiplegic arm after stroke? After deep inspiration orcough, vasoconstriction occurs in the limbs, which isspinally mediated.23-24 This vasoconstriction is greater intetraplegic subjects, indicating a descending inhibitoryinfluence on the area of the cord responsible for theresponse.23 Abnormal persistent vasoconstriction due toa spinal reflex would explain the reduced blood flow tothe hemiplegic hand shown in this study. Anotherpossibility is that these patients have an enhancedsensitivity to endogenous vasoconstrictors only on thehemiplegic side. An animal model of arterial denerva-tion demonstrated hypersensitivity of vascular smoothmuscle to constrictor influences.26 Lesions in the inner-vation of the fingers commonly result in cold intolerancedue to a disturbance in vasoregulation.27 There is alsoevidence that diabetic patients with peripheral neurop-athy have such enhanced sensitivity to vasoconstric-tors.28 Although this is a lower motor neuron lesion,sympathetic fibers are affected, and it may be that theautonomic control of vasomotor tone is altered inhemiplegic patients via a lesion in the sympatheticsystem.

Coldness in the hemiplegic hand is a troublesome andunpleasant symptom in some patients after stroke. Wehave objectively shown marked vasomotor abnormalitiesin this small group of symptomatic patients. Clinicianscaring for stroke patients should be aware of the symptom.Explaining that the symptom is not uncommon may givesome reassurance, and local warming can sometimes af-ford relief. The next step is to determine what might bedone to improve this symptom, and future studies mightinvestigate the efficacy of vasodilators.

AcknowledgmentsWe would like to thank the St James' Bed Push Club for

financial support, without which the project would not havesucceeded. We would also like to thank the staff in the

Physiology Department of the Medical School, particularlyBarbara Royle and Sue Shaw for their invaluable assistance.

References1. MuUey G. Practical Management of Stroke. London, England:

Chapman and Hall; 1988.2. Chevallier PE. De la paralysie des nerfi vaso moteurs dans I'hemi-

plegie. Paris, France: E. Martinet; 1867.3. Kennard MA. Vasomotor disturbances resulting from cortical

lesions. Arch Seuwl Psychiatry. 1935^33:537-545.4. Ellis L, Weiss S. Vasomotor disturbance and oedema associated

with cerebral hemiplegia. Arch Neurol Psychiatry. 1936^36:363-372.5. Bucy P. Vasomotor changes associated with paralysis of cerebral

origin. Arch Neurvl Psychiatry. 1935;33:30-52.6. Herbault A-G, Cole J, Sedgewick E. A cerebral hemisphere

influence on cutaneous vasomotor reflexes in humans. / NeurolNeuwsurg Psychiatry. 1990,53:118-120.

7. Thurston N, Kent B, Jewell M, Blood H. Thermographic eval-uation of the painful shoulder in the hemiplegic patient. Pftys Ther.1988;66:1376-1381.

8. Mulley G. Axillary temperature differences in hemiplegia.Postgrad MedJ. 198046:248-249.

9. Dewar R, Sykes D, Mulkerrin E, Nicklason F, Thomas D, SeymoreR. The effect of hemiplegia on blood pressure measurement in theelderly. Postgrad Med 3. 1992;68:888-891.

10. Panayiotou B, Harper G, Fotherby M, Potter J, Castleden C.Interarm blood pressure differences in acute hemiplegia. / AmGeriatrSoc. 1993;41:422-423.

11. Demeurise G, Demol O, Robaye E. Motor evaluation in vascularhemiplegia. Eur Ncurol 1980;19:382-389.

12. Wilkinson D, Vowden P, Kester R. Use of thermography in theassessment of Raynaud's syndrome and vibration white fingerdisease. ConfProc Biol Eng Soc. 1988;4:20-25.

13. Abramson D, Flachs K, Frieburg J, Mirsky I. Blood flow inextremities affected by anterior poliomyelitis. Arch Intern Med.1943;71:391-396.

14. Kozak P. Circulatory changes of the paretic extremities after acuteanterior poliomyelitis. Arch Phys Med RehabiL 1968;49:77-81.

15. Trott A. Peripheral circulatory changes in patients with poliomy-elitis. Clin Orthop. 1968;61:213-222.

16. Bruno R, Johnson J, Berman W. Vasomotor abnormalities aspost-polio sequelae: functional and clinical implications. Ortho-pedics. 1985;8:865-869.

17. Stenport K. Treatment of cold feet and legs after poliomyelitis.AngMogy. 1951;2:345-349.

18. Darton K, Black C. Pyroelectric vidicon thermography and coldchallenge quantify the severity of Raynaud's phenomenon. Br JRheumatoL 1991^0:190-195.

19. Uematsu S. Symmetry of skin temperature, comparing one side ofthe body to the other. Thermology. 1985;l:4-7.

20. Adams R, Burke D. Deficits of thermal sensation in patients withunilateral cerebral lesions. Electroencephalogr Clin NeurophysioL1989;73:443-452.

21. Veldman P, Reynen H, Arntz I, Goris R. Signs and symptoms ofreflex sympathetic dystrophy: prospective study of 829 patients.Lancet 1993^42:1012-1016.

22. Cofrman J. Raynaud's Phenomenon. New York, NY: Oxford Uni-versity Press, Inc.; 1989.

23. Bolton B, Carmichael E, Stump G. Vasoconstriction followingdeep inspiration. J PhysioL 1936;86:83-94.

24. Gilliatt R, Guttmann L, Whitteridge D. Inspiratory vasocon-striction in patients after spinal injuries. / PhysioL 1948;107:67-75.

25. Cole J, Mani R, Sedgwick E. Cutaneous vasomotor reflexes fol-lowing spinal cord injury in man. J PhysioL 1985^369:134. Abstract.

26. Bevan R, Clemenson A, Joyce E, Bevan J. Sympathetic dener-vation of resistance arteries increases contraction and decreasesrelaxation to flow. Am J PhysioL 1993;264:H490-H494.

27. Nylander G, Nylander E, Lassvik C. Cold sensitivity after replan-tation in relation to arterial circulation and vasoregulation. J HandSurgBr. 1987;12-B:78-81.

28. Masson E, Bodmer C, Benbow S, Savage M, Williams G. Vascularreactivity and peripheral nerve function in diabetes. Diabet Med.1992;9(suppl 1):33. Abstract.

by guest on April 23, 2018

http://stroke.ahajournals.org/D

ownloaded from

P Wanklyn, D W Ilsley, D Greenstein, I F Hampton, T A Roper, R C Kester and G P MulleyThe cold hemiplegic arm.

Print ISSN: 0039-2499. Online ISSN: 1524-4628 Copyright © 1994 American Heart Association, Inc. All rights reserved.

is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Stroke doi: 10.1161/01.STR.25.9.1765

1994;25:1765-1770Stroke. 

http://stroke.ahajournals.org/content/25/9/1765World Wide Web at:

The online version of this article, along with updated information and services, is located on the

  http://stroke.ahajournals.org//subscriptions/

is online at: Stroke Information about subscribing to Subscriptions: 

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

  document. Permissions and Rights Question and Answer available in the

Permissions in the middle column of the Web page under Services. Further information about this process isOnce the online version of the published article for which permission is being requested is located, click Request

can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office.Stroke Requests for permissions to reproduce figures, tables, or portions of articles originally published inPermissions:

by guest on April 23, 2018

http://stroke.ahajournals.org/D

ownloaded from