Autonomic and Thermal Sensory Symptoms and Dysfunction After Stroke
Background and Purpose Symptoms interpreted as unilateral disturbances of autonomic function, such as coldness, dryness, sweating, and trophic changes, are well known but incompletely understood clinical problems after stroke. The present study provides data related to the incidence and mechanisms behind such symptoms.
Methods Temperature perception thresholds, skin temperatures, evaporation rates, and skin blood flow responses were measured bilaterally in 37 stroke patients aged 58±13 years (mean±SD) and in a control group of 15 patients aged 64±15 years with a single transient ischemic attack.
Results Of the 37 stroke patients, 43% reported a sensation of coldness in the contralesional side of the body. Basal skin blood flow and temperature were relatively lower in the contralesional side. There was an excess of evaporation in the contralesional side after brain stem lesions and in the ipsilesional side after hemispheric lesions. Vasomotor reflex asymmetries occurred in 34% of the patients and were due to weak vasodilator or vasoconstrictor reflexes in the ipsilesional side. These abnormalities correlated significantly to sensations of unilateral coldness, hypalgesia, and thermohypesthesia in the contralesional side and anatomically to lesions in spino-thalamo-cortical pathways.
Conclusions Focal central nervous system lesions due to stroke may result in symptoms and measurable evidence of unilateral disturbance of skin sympathetic function. Vasomotor asymmetries are probably due to lesions of vasomotor pathways descending uncrossed. Subjective coldness may be due to disturbed central processing.
Symptoms interpreted as a unilateral disturbance of autonomic function, such as coldness, dryness, sweating, and trophic changes, are well known clinical problems after stroke. The literature contains conflicting reports about the frequency and mechanisms behind these symptoms. For example, clinical hyperhidrosis of the paretic side in the acute phase of stroke was unusual in a large group of stroke patients,1 but hyperhidrosis measured with evaporimeter was quite common in other studies.2 3 4 Different temperature and blood flow asymmetries have also been described after brain lesions.5 6 7 8 9 10 11 In these studies, the patients were investigated with different methods at various time intervals after the stroke, which may explain the conflicting results concerning which side of the body displayed altered autonomic effector functions. In addition, there is no consensus as to which locations of central nervous system lesions result in autonomic dysfunction. The aim of the present study was to determine the incidence of symptoms of autonomic dysfunction after the acute phase of stroke in relation to the site of the lesion and objective measurements of autonomic function and to test whether the complaint of coldness is a symptom of autonomic dysfunction or a result of disturbed thermal perception. Patients without history or signs of former cardiovascular or neurological illness were examined during their first admittance for stroke.
Subjects and Methods
Thirty-seven patients (19 men, 18 women) with acute monofocal stroke (age, 58±13 years, mean±SD; range, 20 to 73 years) were included consecutively. Patients and control subjects with other central or peripheral nervous lesions, diabetes, or peripheral artery disease (indicated by history and physical examination) were excluded. Ten patients were taking antihypertensive medication at the time of investigation: diuretics, 4; ß-adrenergic blocking agents, 2; and calcium channel entry blockers, 4 with cerebral bleeding. The examination was made on day 6 to 113 (median, 14) after the acute incident in 35 patients (on day 6 to 25 in three fourths and on day 26 to 113 in one fourth of the patients) and after 1 year in 2 patients readmitted to the hospital because of pain in a paretic extremity. (These patients did not differ from the other stroke patients in any other aspect.) Twenty-six patients had hemispheric and 11 had brain stem lesions. The lesions were ischemic in 21 of the hemispheric strokes and in 9 of the brain stem strokes. Patients with unexpected radiological signs of preexisting lesions not indicated by the acute symptoms were never included in the study.
The control group consisted of 15 subjects, 5 women and 10 men aged 18 to 80 years (mean, 64±15 years), with a history of one transitory ischemic cerebral attack with a normal CT scan and no remaining symptoms of neurological or autonomic dysfunction. All were taking aspirin 75 mg/d, and 2 were taking ß-adrenergic blocking agents for hypertension.
The neurological evaluation was based on interviews and repeated clinical examinations. The patients were asked whether they had experienced a unilateral feeling of coldness or warmth or they had noticed asymmetry of perspiration, skin color, edema, “goose flesh,” or eczema. Determination of the lesion location was based on clinical evaluation and CT scanning of the brain.
The CT examination was performed with the scan plane parallel to the canthomeatal plane, ie, a gantry tilt about −10° from Reid’s baseline using 5- to 10-mm slice thickness. Examinations were made on day 1 or 2 in all cases and once more at least 4 weeks later in patients with hemispheric lesions (except for five patients who refused a second examination) to get a better delineation of the permanent damage and its location and extent in relation to a neuroanatomy atlas.12 The evaluation of the CT scans was made by two experienced neuroradiologists without knowledge of clinical data. With guidance from the atlas, 65 anatomic structures were defined in the scans, and the lesions were related to these structures and classified according to their vascular supply.
Temperature perception thresholds were recorded as the thermoneutral zone using the Marstock thermotest method (Somedic) in patients who experienced an asymmetrical temperature sensation.13 Skin temperature was measured with an electronic thermometer (Exacon) and skin evaporation with an evaporimeter (Servomed AB). Measurements were taken bilaterally on the plantar side of the big toes, the dorsum of the feet, pretibially, subcostally on the belly, at the elbows between the lateral epicondyle and the biceps tendon, in the palms, on the palmar side of the third finger, and on the forehead, always in the same order. Skin blood flow was monitored continuously and bilaterally on the plantar side of the big toes with laser-Doppler flowmeters (Periflux PF1a, Perimed). Respiratory movements were monitored by a strain gauge attached to the patient’s chest with a rubber band. Laser-Doppler and respiratory signals were recorded on an ink-jet recorder (Mingograph 800, Siemens-Elema) and an eight-channel FM tape recorder (Sangamo Sabre VI, Sangamo-Weston Schlumberger) and later transferred to a computer (PDP-11/70, sampling frequency of 25 Hz). Three types of stimuli were used to evoke vasomotor reflexes: a single deep breath, arousal (a sudden loud noise), and mental stress (subtraction of 6, 7, or 17 from 1000) for 60 seconds. The stimuli were given in random order, three times each, with at least 3 minutes between stimuli. The examinations were made in a quiet room by the same investigator (H.N.), after the subjects had rested in bed with clothes on and covered by two woolen blankets for 1 hour. Room temperature was 21°C to 23°C. Morning body temperature was normal in all patients.
Vasomotor Reflex Responses
For determination of the strength of the vasomotor response, the computer calculated for each stimulus the mean laser-Doppler signal from the three repetitions for a period of 30 seconds before and 150 seconds after the start of the stimulus. The mean laser-Doppler perfusion value during the 30 seconds before stimulation was defined as the basal perfusion. The response was defined as the deviation from the basal perfusion (see Fig 3⇓). For each subject, the difference between the responses on the left and right sides was calculated as (left response−right response)/([left basal perfusion+right basal perfusion]×).
Responses to a deep breath and arousal were determined for the first 15 seconds, whereas the response to mental stress was determined for the period of 15 to 75 seconds after the start. These analysis periods were chosen because the vasoconstrictor responses after arousal or a deep breath start after a few seconds and usually reach a maximum within 15 seconds. The responses to emotional stress start more slowly (usual duration over 60 seconds), and to avoid contamination with the arousal response, the first 15 seconds were excluded. To examine the symmetry of the time course of the responses, cross-correlation was performed for a period of 30 seconds (deep breath and arousal) or 60 seconds (mental stress) after the start of the stimulus. Cross-correlation indexes were calculated during successive displacements in each direction of corresponding laser-Doppler curves from the two sides. The correlation index value at zero time displacement was chosen as a measure of the degree of symmetry (see Fig 3⇓).
Skin Evaporation and Skin Temperature
For each pair of measurement points, the difference in evaporation and skin temperature, respectively, was determined between the ipsilesional and contralesional sides. To obtain an overall value of the degree of symmetry, the mean difference of all eight measuring sites was calculated.
To evaluate whether the overall side asymmetry of skin temperature, evaporation, and blood flow, respectively, was significant in different groups of patients, Wilcoxon’s signed rank test of matched pairs was used. To classify the individual patient as being symmetrical or asymmetrical compared with the control subject, 95% confidence limits were used. Because six parameters were used in classification of the vasomotor response (strength and time course of the responses to mental stress, deep breath, and arousal, respectively), confidence limits of each of these were set to ±(100%−5/6%)=±99.2%. Spearman’s rank order correlation test was used to test correlation between blood flow response and skin temperature.
Sixteen of the 37 stroke patients (43%, seven brain stem infarcts, nine hemispheric infarcts) complained of an unpleasant (constant or temporary) feeling of coldness in extremities on the contralesional side, and three patients with hemispheric lesions had discomfort due to a feeling of warmth in a contralesional extremity. One brain stem infarct patient with subjective coldness in the contralesional side experienced a pleasant warmth in the ipsilesional side. No patient with right cortical hemispheric lesion had noticed coldness in the contralesional side.
Other autonomic symptoms reported included one patient with a brain stem infarct who noticed goose flesh on the contralesional side, which felt cold. Objectively, there was piloerection only on that side. One patient with a pontine lesion and one patient with a middle cerebral artery (MCA) infarct reported increased sweating in the contralesional hand during the first week. One patient with brain stem bleeding had initially profuse sweating in the ipsilesional side of the face. Two patients were troubled by a bluish discoloration in the contralesional extremity. Six patients had pain on the contralesional side, and three of them had a warm sensation; in addition, two had swelling in the paretic hand.
Location of the Lesion
Among the hemispheric stroke patients (n=26), there were five centrally located bleedings, seven lacunar infarcts, and 12 medial and two anterior cerebral artery infarcts. No correlation was found between the size of the lesion in the CT scan and asymmetries in autonomic function tests. Neither was there a difference with respect to autonomic symptoms or measurements between hemorrhagic or ischemic strokes.
Of the brain stem lesions (n=11), two were lateral pontine bleedings visible on CT (with corresponding symptoms) and nine were ischemic lesions with normal CT scans classified by unequivocal clinical findings as lateral medullary infarcts (n=6), lateral pontine infarcts (n=2), and combined lateral and medial medullary infarct (n=1).
Nineteen patients experienced subjective symptoms of temperature asymmetry and hypalgesia (reduced or absent pinprick sensation) on the contralesional side. In 11 of these patients, quantitative temperature perception tests were made; eight patients could not be reliably tested because of sensory neglect or aphasia. All patients tested had impaired temperature perception in the hand of the contralesional side; of the 10 patients whose feet were examined, eight had impaired temperature perception in the foot on the contralesional side (Fig 1⇓). The thermoneutral zone of the contralesional side was significantly wider than that of the ipsilesional side (P<.05).
The thermoneutral zone (mean±SD) of the ipsilesional side in the patients was 3.4±1.8°C in the thenar area and 9.1±3.1°C on the dorsum of the feet. Corresponding values in the control group were 1.9±0.8°C (P<.01) and 6.8±2.1°C (P<.05), respectively. All six patients with central pain had hypalgesia, and the three of them who could cooperate in the thermotest had impaired temperature perception in the contralesional side.
Skin temperature was lower in the contralesional side of the whole patient group (P<.002, n=36). The asymmetry was statistically significant also in the subgroup of patients with hemispheric stroke (P<.01; n=26) but not in the subgroup with brain stem stroke (Table 1⇓).
On an individual basis, seven of 35 patients (six hemispheric and one brain stem lesion) had an abnormal skin temperature asymmetry due to a lower skin temperature in the contralesional side (Fig 2⇓). Four of these patients had a subjective sensation of coldness. No correlation was found between skin temperature and the degree of temperature asymmetry.
Clinically, hyperhidrosis was noted only in one patient (with a left-sided brain stem hemorrhage and sweating in the right side of the face). In patients with hemispheric lesions, the skin evaporation was lower, ie, there was a relative hypohidrosis in the contralesional side compared with the ipsilesional side (mean side difference, 2.0 g H2O/m2 skin per hour, P<.02) (Table 1⇑); in contrast, there was a relative hyperhidrosis of about the same magnitude (2.2 g H2O/m2 skin per hour, P<.01) in the contralesional side of patients with brain stem lesions (Table 1⇑). The asymmetry was similar but more pronounced when measured only in hands and feet instead of the whole body. The two patients who noticed an initial increased sweating in the paretic hand had no side asymmetry (clinically or objectively) at the time of investigation. On an individual basis, eight of 35 patients (23%) had an abnormal asymmetry of sweating. There was no correlation between side differences in skin temperature and evaporation or between evaporation and subjective sensation of temperature asymmetry (Fig 2⇑). Neither was any correlation found between the degree of motor impairment and evaporation asymmetry.
Skin Blood Flow
Resting Blood Flow
Measurements of skin blood flow were made in 15 control subjects and 32 patients. The mean skin blood flow level at rest was lower in the contralesional than in the ipsilesional side (P=.05). The degree of asymmetry was similar but not significant in the subgroups with hemispheric and brain stem stroke. No side difference was present in control subjects.
Vasomotor Responses to Stimuli
The response pattern varied both between maneuvers and between subjects. In control subjects, both a deep breath and arousal usually led to short-lasting flow reductions (with a minimum flow within 10 seconds) followed by transient flow increases for 10 to 20 seconds. Occasionally, however, the response was either a monophasic reduction (n=1) or increase of flow (n=3). The response to mental stress, on the other hand, was most commonly a sustained increase of blood flow (sometimes preceded by a short vasoconstriction as after arousal), which almost always outlasted the 60-second stress period (Figs 3⇓ and 4⇓). In three subjects, the stress response was a monophasic vasoconstriction. Mean skin temperature in the contralesional toes was 29.7°C (range, 23.5°C to 34.5°C). No correlation was found between skin temperature and blood flow response.
In 11 of 32 stroke patients (34%), the vasomotor responses were asymmetrical. In five patients, there were asymmetries in both the strength of the flow response and its time course; in two cases, only the time course was abnormal, and in four only the strength. All asymmetries in the strength of vasomotor responses, whether they were vasodilatations evoked by mental stress (n=3) or vasoconstrictions after a deep breath (n=4) or arousal (n=2), had weaker or absent responses for the ipsilesional side.
Relation to Clinical Symptoms
Of the 11 patients with asymmetrical responses, six had no or only a discrete motor paresis, and three had moderate and two severe hemiparesis. Ten had reduced pain and temperature sensibility, suggesting disturbance of the spinothalamic pathways. The proportion of asymmetrical vasomotor reflexes in patients complaining of coldness in the contralesional extremity was significantly higher (P<.001) than in patients without such symptoms (Table 2⇓). No patient complaining of warm sensation in a contralesional extremity had asymmetrical vasomotor reflexes.
The mean vasomotor response for the patients complaining of coldness was asymmetrical with respect to vasodilatation, which was significantly weaker in the ipsilesional side (P<.05).
Relation to the Site of the Lesion
Six of the 11 patients with asymmetrical vasomotor response had hemispheric lesions (three lacunar and three MCA infarcts), and five had lateral brain stem lesions with medial medullary extension in one case. No significant asymmetries were seen except for the subgroup of patients with MCA strokes (n=10), in whom the vasoconstriction response to a deep breath was asymmetrical and weaker in the ipsilesional side (P<.05).
In this group of previously healthy subjects with monofocal stroke, 43% had a sensation of coldness in the contralesional side; this was the only common hemisymptom suggestive of disturbance of autonomic function. The prevalence of this symptom has not been reported previously, probably because it is usually overshadowed by other functional disturbances. This would agree with our experience that most patients did not mention the symptom spontaneously; in only one case was coldness the main complaint from the beginning and continued to be the overshadowing chronic problem after discharge from the hospital. Because of aphasia or neglect in some patients, we may even have underestimated the prevalence. Clinically overt signs of autonomic dysfunction were rare, but laboratory tests confirmed that unilateral vasomotor and/or sudomotor dysfunction was present in approximately 45% of the investigated patients. The sensation of coldness in the contralesional side correlated significantly to an asymmetry of vasomotor reflexes and a reduced cutaneous sensibility to pinprick and temperature stimuli. In addition, basal skin blood flow, skin temperature, and basal evaporation were significantly lower in the contralesional than in the ipsilesional side of the body in the whole group, but on an individual basis there was no correlation to the symptom of unilateral coldness.
In theory, the varying time between the acute stroke and the investigation may have affected the results. However, since examinations were made more than 6 days after the stroke, irritative (hyperexcitability) phenomena probably do not confound the picture. Furthermore, the results found in patients investigated before and after day 25 did not differ; therefore, it seems likely that the time factor is of minor importance. Transient ischemic attack patients were preferred to healthy subjects for control purposes to include persons with similar vascular status and risk factors as stroke patients. This should have minimized misinterpretation of changes in temperature and autonomic functions due to general manifestations of atherosclerosis as the result of clinically manifest central nervous system damage.
Basal Blood Flow
Because of variations in epidermal and/or vascular microstructure, resting cutaneous blood flow measured with the laser-Doppler technique may differ markedly between sites located only a few millimeters apart.14 Such random variations should not, however, affect group comparisons, and we therefore conclude that our finding of a lower basal blood flow in the contralesional side represents a stroke-induced effect. The underlying mechanism is unclear. Theoretically, increased cutaneous vasoconstrictor nerve traffic in the contralesional side or decreased traffic in the opposite side (or a corresponding contrary asymmetry of vasodilator nerve traffic) would be possible explanations. The literature contains conflicting reports on temperature and skin perfusion asymmetries in cerebral lesions.5 6 7 8 9 10 11 Our finding of more pronounced vasoconstriction in the contralesional side agrees with the most recent of these reports,11 in which stroke patients complaining of coldness in the paretic hand were found to have a reduced skin temperature at rest and after cold stimuli, as well as a reduced blood flow in the symptomatic hand.
Vasomotor reflex responses were asymmetrical in 34% of the patients, and the asymmetries occurred both with brain stem and hemispheric lesions. On the basis of the CT scan or the clinical findings, all but one of the patients with hemispheric strokes who had asymmetrical responses had lesions in the region of the posterior branch of capsula interna and the posterolateral part of the thalamus. The brain stem stroke patients with vasomotor reflex asymmetry all had symptoms suggesting disorder of spinothalamic pathways.
The close anatomic relationship between central skin vasomotor pathways and afferent temperature and pain pathways may explain the correlation between asymmetries of vasomotor reflexes and temperature perception: the lesion may have affected both structures. Alternatively the vasomotor asymmetry may be a reflex consequence of a lesion of afferent thermal pathways. In a physiological study, Cooper and Kerslake15 found that short-lasting heating of the legs induced reflex vasodilatation in the hands; in another report,6 it was concluded that a normal vasodilatation response required intact afferent and efferent fibers in structures above the brain stem level. Interruption of this reflex pathway may contribute to our finding of a correlation between disturbed sensibility and asymmetry of vasomotor responses.
Vasomotor responses that were asymmetrical in the strength of the response were always weaker in the ipsilesional side regardless of whether a vasoconstriction or a vasodilatation occurred. A possible explanation would be that excitatory nerve fibers involved in vasodilator and vasoconstrictor responses descend uncrossed. This would agree with results from previous studies suggesting that the majority of sympathetic hypothalamic projections to the spinal cord are ipsilateral.16 17 An alternative (or additive) mechanism could be that pathways with an inhibitory effect on vasomotor reflexes descend crossed. A loss of inhibitory influence would then lead to amplification of reflexes in the contralesional side.
Clinical hyperhidrosis was present in only one of our patients. This agrees with the results of Labar et al,1 who observed that about 1% of a large group of stroke patients sweated more on the paretic than the nonparetic side during days 1 to 3. The same direction of asymmetry during the first week after the stroke was found in patients with hemispheric stroke studied by Korpelainen et al.3 Because cortical electrical stimulation also results in contralateral perspiration,18 these asymmetries may in theory be irritative phenomena, not present in our patients who had ipsilesional hyperhidrosis when examined on day 6 or later. However, this explanation does not agree with the results from the longitudinal study of Korpelainen et al, who found that, although the magnitude of the asymmetry in their study had decreased 1 month after the stroke, the direction of asymmetry was still unchanged (and opposite to ours). Our results are also different in that we found no correlation to the degree of motor paresis in patients with hemispheric stroke. In contrast to Korpelainen et al, we did not warm the patients, and we also kept them in bed for 1 hour before taking measurements. This may have affected the magnitude, but it is not probable that it would have affected the direction of the asymmetry. The group of patients with brain stem lesions who had contralesional hyperhidrosis in our study had no or sparse motor impairment, suggesting that factors other than motor deficit are of importance for the hyperhidrosis. In a study of sympathetic skin responses, Korpelainen et al4 found contralesional increased sweat responses after brain stem lesions only and not after hemispheric lesions, which is in accordance with our results. Previous studies give evidence for dual central excitatory sudomotor pathways: a crossed cortico-limbico-spinal tract18 and an uncrossed hypothalamo-reticulo-spinal pathway.18 19 Hypothetically, therefore, the asymmetry of evaporation after hemispheric lesions in our study may be explained by lesions primarily affecting the crossing cortico-limbico-spinal pathway, and the asymmetry after brain stem lesion may be a result of lesions primarily affecting the hypothalamo-reticulo-spinal thermoregulatory pathway.
Skin temperature was about 0.4°C lower in the contralesional side in patients with hemispheric lesions. The mean side difference in the smaller group of patients with brain stem infarcts was similar but did not reach statistical significance. The temperature asymmetry did not correlate to the sensation of coldness, but that does not exclude a correlation under other circumstances. Our patients were resting in bed covered by blankets for 1 hour before measurements. This procedure was adopted to achieve standardized measurement conditions, but it may have reduced asymmetries. In support of this possibility, one patient (without evidence of neuropathy or leg artery stenosis) had skin temperature of 9°C to 10°C lower in the paretic leg on several occasions when out of bed. In contrast, the difference was only 1°C to 2°C when the patient was immobilized under blankets in bed.
In theory, the lower skin temperature in the contralesional than in the ipsilesional side may be due to either increased evaporation or reduced blood flow in the contralesional side. Since our patients with hemispheric lesions had lower evaporation and cutaneous blood flow in the contralesional side, the lower temperature in that side was probably due to the reduced blood flow. This would also agree with Thiele and van Senden,20 who found that cutaneous vasoconstriction caused a lowering of skin temperature, which resulted in a decreased insensible perspiration.
Determination of temperature perception thresholds requires a high degree of cooperation of the subject. Many stroke patients have difficulty with the test because of impaired concentration, slow reactions to stimuli, inattention to the paretic side, or sensory dysphasia with difficulty in understanding instructions. Although patients were excluded from the temperature perception test whenever such symptoms were detected, widened thermoneutral zones in the ipsilesional side were found. Since patients with pure central hemispheric lesions also had widened thermoneutral zones in the ipsilesional side, compliance of the patients is probably not the cause of this result. An alternative explanation would be that some spinothalamic pathways have bilateral thalamic projections.21 The high prevalence of reduced temperature perception in patients with coldness may be an overestimation, since not all were tested. However, the pinprick sensation test, which does not demand the same degree of cooperation from the patient and therefore was studied in all patients, was also reduced in all patients with coldness, which supports the thermotest result. Our findings of impaired pinprick sensation and temperature perception in the patients with pain in the paretic side agree with the results of Boivie et al,22 who studied patients with poststroke pain. However, our results also show that these findings are not specific but may occur in patients without poststroke pain.
Sensation of Coldness on the Contralesional Side
Because skin temperature was lower on the contralesional side, the sensation of coldness on this side may reflect an adequate perception of the temperature asymmetry. In view of the markedly impaired temperature perception on the contralesional side, this explanation seems unlikely. However, it cannot be totally disregarded, since we measured only the width of the thermoneutral zone; therefore, a selective impairment of warmth perception (with preserved cold perception) is difficult to exclude. The coldness probably is not due to inactivity of the extremity, since more than 50% of patients with that symptom had no or only a slight hemiparesis. A more likely explanation for the sensation of coldness may be that it is caused by a disturbed central processing. This would be compatible with the correlation between hypalgesia, thermohypesthesia, and coldness in the contralesional side and would be an explanation that is analogous to that used to explain the occurrence of poststroke pain.22
Symptoms suggestive of autonomic dysfunction on one side of the body after stroke are common, and objective measurements demonstrate the occurrence of sweat, temperature, and skin perfusion asymmetries. The symptom of coldness on the contralesional side is related to skin vasomotor reflex asymmetry and a lesion of spino-thalamo-cortical pathways and may be due to a disturbed central processing.
This study was supported by the foundation “1987 års stiftelse för Strokeforskning” and Swedish Medical Research Council grant 3546.
- Received January 13, 1995.
- Revision received April 24, 1995.
- Accepted April 27, 1995.
- Copyright © 1995 by American Heart Association
Labar DR, Mohr JP, Nichols FT, Fenwick TN, Tatemichi TK. Unilateral hyperhidrosis after cerebral infarction. Neurology. 1988;38:1679-1682.
Korpelainen JT, Sotaniemi A, Myllylä VV. Hyperhidrosis as a reflection of autonomic failure in patients with acute hemispheral brain infarction. Stroke. 1992;23:1271-1275.
Korpelainen JT, Sotaniemi KA, Myllylä VV. Asymmetric sweating in stroke: a prospective quantitative study of patients with hemispheral brain infarction. Neurology. 1993;43:1211-1214.
Korpelainen JT, Tolonen U, Sotaniemi KA, Myllylä VV. Suppressed sympathetic skin response in brain infarction. Stroke. 1993;24:1389-1392.
Olsen A. Hemiplegi og hudtemperatur. Hospitalstidende. 1933;76:1097-1103.
Cooper KE, Ferres HM, Guttman L. Vasomotor response in the foot to raising body temperature in the paraplegic patient. J Physiol (Lond). 1957;136:547-555.
Ellis LB, Weiss S. Vasomotor disturbance and oedema associated with cerebral hemiplegia. Arch Neurol Psychiatry. 1936;35:362-372.
Wanklyn P, Isley DW, Greenstein D, Hampton IFG, Roper TA, Kester RC, Mulley GP. The cold hemiplegic arm. Stroke. 1994;25:1765-1770.
Kretschman H-J, Weinrich W. Clinical Neuroimaging and Clinical Neuroanatomy. New York, NY: Thieme Medical Publishers; 1992.
Fruhstorfer HG, Lindblom U, Schmidt WG. Method for quantitative estimation of thermal thresholds in patients. J Neurol Neurosurg Psychiatry. 1976;39:1071-1075.
Cooper KE, Kerslake DM. Abolition of nervous reflex vasodilatation by sympathectomy of the heated area. J Physiol (Lond). 1953;119:18-29.
Yassui Y, Breeder CD, Saper CB, Cechetto DF. Autonomic responses and efferent pathways from insular cortex in the rat. J Comp Neurol. 1991;303:335-374.
Albe-Fessard D, Berkeley KJ, Kruger L, Ralston HJ, Willis WD Jr. Diencephalic mechanisms of pain sensation. Brain Res Rev. 1985;9:217-296.