- Email: [email protected]
Sympathetic skin response in diabetic neuropathy: a prospective clinical and neurophysiological trial on 100 patients
JOURNAL OF THE
NEUROLOGICAL SCIENCES ELSEVIER
Journal of the Neurological Sciences 138 (1996) 120- 124
Sympathetic skin response in diabetic neuropathy: a prospective clinical and neurophysiological trial on 100 patients H.-J. Braune *, C. Horter Department
of Neurology,
Philipps
Uniwrsi~
Hospital,
Rudolf-Bultmann-Strasse
8, D-35033
Marburg,
Germany
Received 2 1 March 1995; revised 16 November 1995; accepted 2 December 1995
Abstract Sympathetic skin response(SSR) can be employed to assessaxonal damageof sympathetic sudomotor neurons in neuropathy. We examined a population of 100 patients with diabetic neuropathy, and 100 age- and sex-matchedhealthy volunteers. SSR was recorded from the skin of the hand and foot by meansof surfaceelectrodes.Stimulation was applied with 30 mA over the median nerve at the wrist contralateral to the derivation site. Irregular interstimulus intervals of more than one minute were applied. Peak-to-peakamplitude measurementshowed values of 2.91 f I .lO mV on the hand and 1.37 + 0.54 mV on the foot. Latencies lasted 1.44 f 0.10 s on the hand and 2.12 + 0.15 s on the foot. A total of 37% of the diabetics showed loss of SSR, a further 39% showed amplitude reductions or latency prolongations: Altogether, 76% of all diabetics did not show normal test results. The clinical stage of diabetic neuropathy exerted a significant influence: the further developed the clinical picture of the diseasewas, the more pathologic were the results of SSR testing. However, even at early stagesof neuropathy abnormal values of amplitude reduction or latency prolongation could be observedin over 50% of the patients. SSR examinations performed in diabetics with no or only slight clinical signs of neuropathy may be able to detect those subjectswith autonomic disturbancesat an initial stage. Kqwords:
Sympathetic skin response; Peripheral neuropathy; Diabetic autonomic neuropathy
1. Introduction Sympathetic skin response (SSR) is a transient change in the electrical potential of the skin which can easily be evoked by a variety of internal or external stimuli (Shahani et al., 1984). When elicited by electrical stimulation, the response is mediated via a reflex arc which includes large myelinated sensory fibers as afferent limb, central relays and efferent sympathetic pre- and postganglionic nerve fibers which activate eccrine sweat glands in the skin. Despite the complex nature of SSR, its synchronous activity at homologous skin sites suggests that it can be used to indicate the presence of sympathetic sudomotor activity, although it cannot quantitate the absolute level of that activity (Schondorf, 1993). A large number of influencing factors that may affect correct derivation of SSR potentials have been described (Baba et al., 1988; Schondorf and Gendron, 1990). Nevertheless, many investigators employ
this technique to assess autonomic function in polyneuropathies, especially in diabetic autonomic neuropathy (Shahani et al., 1984; Knezevic and Bajada, 1985; Niakan and Harati, 1988; Shahani et al., 1990). Since it has been maintained that SSR is present in all healthy subjects (Shahani et al., 1984; Knezevic and Bajada, 1985; Uncini et al., 1988; Elie and Guiheneuc, 19901, especially under the age of 60 (Drory and Korczyn, 1993), loss of potentials is regarded as a pathological sign. To investigate the correlation between SSR abnormalities and the clinical state of diabetic neuropathy, we re-examined the influencing factors on test results such as age, gender, site of derivation, intensity of electrical stimuli and habituation. Then we calculated normal values and examined a population of 100 patients with diabetic neuropathy of different stages.
2. Patients and methods * Corresponding author. Tel: +49 (6421) 285242. Fax: t49 288955.
(6421)
0022-510X/96/$15.00 0 1996 Elsevier Science B.V. All rights reserved PII SOO22-510X(96)00023-8
All healthy volunteers as well as all patients gave their informed consent according to the Declaration of Helsinki
H.-J. Braune,
C. Horter/Joumal
of the
as revised by the 35th World Medical Assembly, Venice, Italy, 1983. Apart from clinical examinations, SSR measurements were performed consecutively on 100 outpatients with diabetes mellitus (definition of WHO, 1985) at the Dept. of Neurology of the University Hospital Marburg, Germany. The study was carried out over a period of 18 months on 54 men and 46 women (age 18-82 years, mean 52.0 years). On the day of the examination the patients had their usual breakfeast and, if necessary, their morning insulin dose. None of the subjects was allowed to drink coffee or tea, smoke, or take any medication in the hours before and during the visit. Major physical activity/strain and hypo- or hyperglycemic periods which occurred up to three days before examination led to exclusion from the study. Patients with symptoms not due to diabetes or with case histories other than diabetes, e.g. previous injury, lumbar discogenic disease, carpal tunnel syndrome, inherited or alcoholic neuropathy or other neurological diseases as well as patients taking medication acting on the sympathetic system were excluded. Another exclusion criterion was the presence of acute focal neuropathy and acute painful neuropathic syndroms which are probably caused by diabetes. However, these are rare and presumably of varying pathogenesis. Several studies demonstrate that the clinical course of these syndromes is different from that of common chronic distal symmetric polyneuropathy which primarily affects the sensory nerve fibers (Behse et al., 1977; Dyck et al., 1992). The patients were assessed independently by two experienced neurologists, and, in accordance with their clinical stage of neuropathy, divided into five subgroups, as recommended by Dyck et al. (1985): 8 patients without any clinical signs of neuropathy (NO), 15 patients with only one clinical symptom of neuropathy (Nl), 24 patients with mild distal symmetric sensory polyneuropathy (N2a), 28 patients with distal symmetric sensorimotor polyneuropathy (N2b), 25 patients with severe symmetric sensorimotor polyneuropathy (N3). Autonomic symptoms included vesical incontinence (12 patients), constipation (9 patients), sexual dysfunction (6 of 54 males) and orthostatic dysregulation (23 patients). Clinical autonomic symptoms were mild in 14 N2b-patients (only one symptom) and moderate in 18 N3-patients (two symptoms). The duration of diabetes was estimated based on hospital files and information given by patients: 20 patients had been suffering from diabetes for less than five years, 31 patients for more than five but less than 10 years, and 49 patients for more than ten years. A total of 26 patients younger than 40 years had IDDM, 74 had NIDDM. In 72 patients, metabolic control was good with normal fasting blood sugar values, and normal or only slightly elevated HbAlc values. It was moderate in 28 patients. Tests were also performed in an control group of 51 men and 49 women (age 18-83 years, mean 50.3 years) to define the normal range of values. Subjects were seated comfortably in an armchair in a
Neurological
Sciences
138 (19961 120-124
121
quiet, dimly lit room at 22°C. A 4-channel EMG system (NIHON KOHDEN Neuropack 4) was used with a sensitivity of 0.5 mV/div, analysis time 10 s, and Cut-off frequency filters 0.1 Hz and 20 Hz. Latencies were measured manually to the first continuous deflection from the baseline with a cursor. Maximum amplitudes were measured from peak to peak. Stimulation cathode was placed over the median nerve on the wrist contralateral to the recording site. Ipsilateral stimulation induced muscle contraction artifacts on SSR derivation from the hand (Fuhrer, 1971). A stimulus strength of 30 mA with duration 0.2 ms was tolerable by all subjects. Lower values were associated with poor reproducibility and even absence of response. Skin impedance was reduced to < 5 k0 by gentle skin abrasion and application of conductive gel. The SSR was recorded from the hands and feet with surface electrodes (9 mm dia) on the volardorsal and plantardorsal surface. When recordings were made from other skin sites: e.g. fingertips or thenar-hypothenar, or when comparing results from the left and right hand or the left and right foot ot ten healthy volunteers, statistically significant differences in amplitudes and latenties were not detected. In addition, intraindividual differences of SSR amplitudes and latencies could not be shown by means of comparing the derivations from respectively. Minimum time between successive stimuli was 60 s. The greatest amplitude and shortest latency of five consecutive SSR measurements were recorded for further data processing, since averaging of consecutive potentials did not result in significantly different values (Schondorf and Gendron, 1990). Statistical evaluation was performed with the non-parametric U-test (Mann-Whitney). The analyzed control values determine the 95th percentile. The minimum criterion for assumed abnormal values in patients was a test result of > 95th percentile or absence of SSR.
3. Results Many investigators have noted that SSR tends to habituate itself to repetitive stimuli (Shahani et al., 1984; Knezevie and Bajada, 1985; Elie and Guiheneuc, 1990). We observed a decrease in amplitude of more than 50% between the first and fifth stimulus (stimuli applied regulary every 30 s) in ten healthy volunteers due to habituation. As a consequence, irregular interstimulus intervals of more than one minute were applied to ensure reproducibility of the SSR. The SSR is generally biphasic and consists of an initial negativity followed by a positive potential. Delayed negativity was observed in 27 of 100 healthy volunteers, which corresponds to other publications (Schondorf and Gendron, 1990). In one test series, more than one wave form was found in 43% of the subjects and more than two wave form was found in 13% of the subjects. These findings are
122
H.-J. Braune, C. Horter/Journal
of the Neurological Sciences 138 (19961 120-124
Table 1 Influence of clinical stage of diabetic neuropathy on sympathetic skin response Clinical stage
NO
Nl
N2a
N2b
N3
15 0 (0) 3 (20)
24 4 (16.61 8 (33.31
28 6 (21.4) 14 (501
25 5 (20) 9 (361
Patients with values outside of the 95th~percentile normal range in derivable SSR (o/o) Latency on hand 3 (37.5) 4 (26.6) 7 (291 Amplitude on hand 3 (37.51 6 (39.9) 5 (20.1) Latency on foot 3 (37.51 4 (26.6) 3 (12.5) Amplitude on foot 2 (13.3) 3 (12.5) 0 (0)
10 (35.71 9 (32.1) 2 (7.1) 0 (0)
12 (48) 8 (32) 7 (281 4 (161
Loss of SSR or at least two abnormal values ( % of patients in subgroup NO-N3) 7~ Of subgroup 50 70 62
76
88
Loss of sympathetic skin responses (%) Number of patients 8 SSR loss on hand l(12.5) SSR loss on foot 3 (37.5)
Clinical staging as recommended by Dyck et al. (1992): NO: diabetics with no clinical sign of neuropathy; Nl: diabetics with only one clinical sign of neuropathy; N2a: diabetics with mild distal symmetric sensory neuropathy; N2b: diabetics with distal symmetric sensorimotor neuropathy; N3: diabetics with severe symmetric sensorimotor neuropathy.
in accordance with the observations made by BABA et al. (1988). Stimulus duration was 0.2 ms. All patients and control subjects tolerated this value, even when multiple measurements were required (Knezevic and Bajada, 1985; Uncini et al., 1988). SSR measurement of peak-to-peak amplitude in 100 healthy volunteers showed values of 2.91 & 1.10 mV on the hand and 1.37 + 0.54 mV on the foot. Latencies lasted 1.44 IfI 0.10 s on the hand and 2.12 + 0.15 s on the foot. SSR latencies were significantly shorter on hands than on feet (p < 0.01) and SSR amplitudes were significantly lower on feet (p < 0.01). The standard deviation of amplitudes provided evidence of considerable interindividual variations. Latencies were quite stable interindividually. A statistically significant age dependency of normal values regarding SSR amplitudes and latencies on the hand and foot was not shown. Consequently, a single normal range for all ages was determined, and the cut-off values were defined as 95th percentiles: latency on hand < 1.58 s, latency on foot < 2.44 s, amplitude on hand < 1.20 mV, amplitude on foot < 0.50 mV. Statistical analysis showed significant differences between the group of diabetic and healthy subjects, even when patients with SSR loss and their counterparts of the control group were excluded. A strict age- and sex-matched control group was not regarded as necessary, because there is such large variability in the results that the influence of age could not be demonstrated. Mean and range of years of patients and controls are nevertheless compareable. In diabetic subjects, latencies were significantly longer (p < 0.01) and amplitudes significantly smaller (p < 0.01) than in the healthy control group. Most striking was the loss of potentials on hands (16%) and feet (37%) in some diabet-
its, since SSR was always detectable in the normal control group. All patients with absent SSR on hands have absent SSR on the feet, whereas 21 patients with absent SSR on feet had derivable SSR on hands, although in these cases values of latencies or amplitudes were abnormal. Two or more abnormal test results (prolongation of SSR and/or amplitude reduction on the hand and/or foot, but no loss of potentials) were found in 39% of the patients. A statistically significant influence of gender, age, metabolic control, duration or type of diabetes on the test results could not be established or was existent only to a minor degree. However, the clinical stage of diabetic neuropathy (NO-N3) exerted a significant influence on our study: the further developed the clinical picture of the neuropathy was, the more pathologic were the results of SSR testing (Table 1). A total of 84% of the patients with clinical signs of autonomic involvement had loss of SSR on the foot, including 14 patients with additional SSR loss on the hand.
4. Discussion The aim of this study was to investigate the prevalence of pathologic SSR tests in a population with diabetes mellitus. After establishing a standardized test procedure in accordance with recent publications (Niakan and Harati, 1988; Schondorf and Gendron, 1990), a normal range ‘of values for latencies and amplitudes of SSR on hands and feet was defined as lying within the 95% range of all data gathered from 100 healthy volunteers of all ages. Since no age dependency could be found, a single cut-off value for all ages was considered to be appropriate. These findings are contrary to the investigation of Drory and Korcyn (1993) who emphasize that SSR can be elicited in all healthy subjects up to the age of 60, whereas a loss of SSR
H.-J. Braune,
C. Horter/
Journal
of the Neurological
can be found in about 50% of the septuagenarians and octogenarians without clinical findings. The authors stimulated the median nerve ” slightly over motor threshold intensity” which is generally significantly lower than the value of 30 mA used in our study. At a stimulation intensity of less than 10 mA we found only poor potential reproducibility; a considerable interpersonal variation in SSR amplitudes was found even between 10 and 25 mA. This may also explain the mean (&SD) response amplitudes in the Drory/Korcyn study of 449 + 429 PV in the upper and 147 + 122 PV in the lower limbs which is remarkably lower than the results obtained in our study (2910 & 1100 PV on hands and 1370 f 540 PV on feet>. Furthermore, the authors did not perform nerve conduction studies in healthy elderly subjects with SSR loss in order to rule out subclinical neuropathies. Thus we would like to stress that, contrary to the findings of Drory and Korczyn, SSR can be recorded in all subjects. This includes healthy elderly subjects, provided that the stimulus is applied with sufficient intensity, as suggested by previous studies (Shahani et al., 1984; Knezevic and Bajada, 1985; Uncini et al., 1988; Elie and Guiheneuc, 1990). Since SSR includes a somatic afferent phase which may be impaired in polyneuropathies, and since SSR show a high degree of variability due to a number of influencing factors, many investigators consider only the complete loss of SSR as abnormal (Shahani et al., 1984; Van den Bergh and Kelly, 1986; Soliven et al., 1987). Uncini et al. (1988) suggest that absent SSR in autonomic neuropathy may be due to an involvement of the large-diameter sensory afferent fibers of the median nerve mediating the SSR. However, this can only be true for patients with loss of SSR on upper and lower limbs. A simultaneous presence of derivable SSR on hands and absence of SSR on feet cannot be explained by this suggestion. Here, disturbances on the efferent limb of the reflex arc are of major importance. In accordance with this opinion, a pathological loss of potentials on hands was found in 16% and on feet in 37% of the diabetics. Autonomic system deterioration in diabetic neuropathy affects a variety of functions, especially the loss of preganglionic neurons which are short and myelinated. Symptoms become apparent when more than 50% of the neurons are lost (Low et al., 1977). This loss may contribute to the decrease in amplitude, and, in the later course of the disease, to loss of the SSR. It does not explain, however, why the loss is more pronounced on the lower limbs. Here, the ‘dying-back-hypothesis’ is helpful. It suggests that the longer peripheral nerves are, the more easily injured (Dyck, 1987). The postganglionic autonomic fibers are unmyelinated; thus, only loss of axons can occur which also may result in SSR amplitude reduction. Latency prolongation may be due to changes occurring in preganglionic myelinated fibers, loss of thicker axons or sweat gland alterations. A total of 84% of the patients with clinical signs of
Sciences
138 (19961 120-124
123
autonomic involvement had loss of SSR on the foot, including 14 patients with further SSR loss on the hand, which tallies with the findings of Soliven et al. (1987). The presence of clinically apparent autonomic dysfunction only occurs late in the course of diabetic neuropathy and is associated with a bad prognosis (Ewing et al., 1980; O’Brian et al., 1991). In our study, all patients with signs of autonomic disturbances were stage N2b-N3, only 3 patients with SSR loss were stage NO, and 3 others were stage Nl. Thus, exclusive assessment of the loss of SSR does not permit new insights into advanced diabetic neuropathy which can instead be diagnosed correctly by means of comprehensive histories, appropriate clinical examination and conventional neurophysiological tests. However, we propose that prolongation of latencies and reduction of amplitudes, if carefully derived under standardized conditions, should also be taken into consideration. Two or more abnormal test results (prolongations of SSR and/or amplitude reduction on hand and/or foot, but no loss of potentials) could be found in 39% of the patients. Together with the 37% of the diabetics who showed a loss of SSR, three quarters (76%) of all diabetics showed abnormal test results. The clinical stage of diabetic neuropathy (NO-N3) exerted a significant influence on our study: the further developed the clinical picture, the more pathologic were the results of SSR testing (Table 1). However, prolongation of latencies and reduction of amplitudes were found in 14 of 23 patients in the early stage of neuropathy (8 patients without any clinical signs of neuropathy (NO), 15 patients with only one clinical symptom of neuropathy (N 1)). Thus, earlier detection of sympathetic involvement in diabetic neuropathy seems to be possible by evaluation of all information achieved by SSR sampling.
5. Conclusion The results suggest that some changes leading to pathological SSR tests take place in a clinical stage in which no or only very slight clinical signs of diabetic neuropathy (NO-N 1) are apparent. Significant further deterioration of SSR test results can be expected in patients with clinically manifest neuropathy (stage N2a, N2b, N3) resulting in loss of potentials on feet and later on hands. Five-year-mortality of patients with autonomic disturbances is three times that of patients without autonomic involvement (Ewing et al., 1980; O’Brian et al., 1991; Sampson et al., 1990). But there is considerable controversy as to wether the worst in prognosis relates to vascular changes often occuring with diabetes or to autonomic failure itself. Nevertheless, SSR examination performed in diabetics with no or only slight clinical signs of neuropathy may help beside other tests determine those subjects with autonomic disturbances at an initial stage.
124
H.-J. Braune,
C. Horter/
Journal
of the Neurological
References Baba, M., Y. Watahiki, M. Matsunaga and K. Takebe (1988) Sympathetic skin response in healthy man. Electromyogr. Clin. Neurophysiol., 28: 277-283. Behse, F., F. Buchthal and F. Carlsen (1977) Nerve biopsy and conduction studies in diabetic neuropathy. J. Neurol. Neurosurg. Psychiatr., 40: 1072- 1082. Drory, V.E. and A.D. Korczyn (1993) Sympathetic skin response: age effect. Neurology, 43: 18 18- 1820. Dyck, P.J. (1987) Pathology. In: Dyck, P.J.. P.K. Thomas, A.K. Asbury, AI. Winegard and D. Porte (Eds.), Diabetic neuropathy, W. B. Saunders Company. Philadelphia, London, Toronto, Mexico City, Rio de Janeiro, Sydney, Tokyo, Hong Kong, pp. 223-236. Dyck, P.J., J.L. Karnes, J. Daube, P. O’Brian and F.J. Service (1985) Clinical and neuropathological criteria for the diagnosis ond staging of diabetic polyneuropathy. Brain, 108: 861-880. Dyck, P.J., J.L. Karnes, P.C. O’Brian, W.J. Litchy, P.A. Low and L.J. Melton III (1992) The Rochester diabetic neuropathy study: reassessment of tests and criteria for diagnosis and staged severity. Neurology, 42: 1164-1170. Elie, B. and P. Guiheneuc (1990) Sympathetic skin response: normal results in different experimental conditions. Electroencephalogr. Clin. Neurophysiol., 76: 258-267. Ewing, D.J., I.W. Campbell and B.F. Clarke (1980) The natural history of diabetics autonomic neuropathy. Quart. J. Med., 49: 95-99. Fuhrer, M.J. (1971) Effects of unilateral stimuli on the magnitude and latency of bilaterally recorded skin conductance responses. Psychophysiology, 8: 740-75 1. Knezevic, W. and S. Bajada (1985) Peripheral autonomic surface potential. A quantitative technique for recording sympathetic conduction in man. J. Neural. Sci., 67: 239-251. Low, P.A., H. Okazaki and P.J. Dyck (1977) Splanchnic preganglionic
Sciences
138 (19961 120-124
neurons in man. I. Morphometry of preganlionic cytons. Acta Neuropathol. (Berlin), 40: 55-61. Niakan, E. and Y. Harati (1988) Sympathetic skin response in diabetic peripheral neuropahty. Muscle Nerve, 11: 261-264. O’Brian, LA., J.P. McFadden and R.J.M. Corral1 (1991) The influence of autonomic neuropathy on mortality in insulin-dependent diabetes. Quart. J. Med., 79: 495-502. Sampson, M.J., S. Wilson, P. Karagiannis, M. Edmonds and P.J. Watkins (1990) Progression of diabetic autonomic neuropathy over a decade in insulin-dependent diabetics. Quart. J. Med., 75: 635-646. Schondorf, R. (1993) The role of the sympatheitc skin response in the assessment ot autonomic function. In: Low, P.A. (ed.), Clinical autonomic disorders, Little, Brown and Company, Boston, pp. 231241. Schondorf, R. and D. Gendron (1990) Properties of electrodermal activity recorded from non palmar/plantar skin sites. Neurology, 40: 128- 132. Shahani, B.T., T.J. Day, D. Cros, N. Khalil and C.S. Kneebone (1990) RR interval variation and the sympathetic skin response in the assessment of autonomic function in peripheral neuropathy. Arch Neurol., 47: 659-664. Shahani, B.T., J.J. Halperin, P. Boulu and J. Cohen (1984) Sympathetic skin response: a method of assessing unmyelinated axon dysfunction in peripheral neuropathies. J. Neural. Neurosurg. Psychiat., 47: 536542. Soliven, B., R. Maselli, J. Jaspan, A. Green, H. Graziano, M. Petersen and J.P. Spire (1987) Sympathetic skin response in diabetic neuropathy. Muscle Nerve, IO: 711-716. Uncini, A., S.L. Pullmann, R.E. Lovelace and D. Gambi (1988) The sympathetic skin response: normal values, elucidation of afferent components and application limits. J. Neurol. Sci., 87: 299-306. Van den Bergh, P. and J.J. Kelly (1986) The evoked electrodermal response in peripheral neuropathies. Muscle Nerve, 9: 656-657. WHO (I 985) Diabetes mellitus: Report of a study group, Technical report series, No. 727, World health organization, Geneva.
NEUROLOGICAL SCIENCES ELSEVIER
Journal of the Neurological Sciences 138 (1996) 120- 124
Sympathetic skin response in diabetic neuropathy: a prospective clinical and neurophysiological trial on 100 patients H.-J. Braune *, C. Horter Department
of Neurology,
Philipps
Uniwrsi~
Hospital,
Rudolf-Bultmann-Strasse
8, D-35033
Marburg,
Germany
Received 2 1 March 1995; revised 16 November 1995; accepted 2 December 1995
Abstract Sympathetic skin response(SSR) can be employed to assessaxonal damageof sympathetic sudomotor neurons in neuropathy. We examined a population of 100 patients with diabetic neuropathy, and 100 age- and sex-matchedhealthy volunteers. SSR was recorded from the skin of the hand and foot by meansof surfaceelectrodes.Stimulation was applied with 30 mA over the median nerve at the wrist contralateral to the derivation site. Irregular interstimulus intervals of more than one minute were applied. Peak-to-peakamplitude measurementshowed values of 2.91 f I .lO mV on the hand and 1.37 + 0.54 mV on the foot. Latencies lasted 1.44 f 0.10 s on the hand and 2.12 + 0.15 s on the foot. A total of 37% of the diabetics showed loss of SSR, a further 39% showed amplitude reductions or latency prolongations: Altogether, 76% of all diabetics did not show normal test results. The clinical stage of diabetic neuropathy exerted a significant influence: the further developed the clinical picture of the diseasewas, the more pathologic were the results of SSR testing. However, even at early stagesof neuropathy abnormal values of amplitude reduction or latency prolongation could be observedin over 50% of the patients. SSR examinations performed in diabetics with no or only slight clinical signs of neuropathy may be able to detect those subjectswith autonomic disturbancesat an initial stage. Kqwords:
Sympathetic skin response; Peripheral neuropathy; Diabetic autonomic neuropathy
1. Introduction Sympathetic skin response (SSR) is a transient change in the electrical potential of the skin which can easily be evoked by a variety of internal or external stimuli (Shahani et al., 1984). When elicited by electrical stimulation, the response is mediated via a reflex arc which includes large myelinated sensory fibers as afferent limb, central relays and efferent sympathetic pre- and postganglionic nerve fibers which activate eccrine sweat glands in the skin. Despite the complex nature of SSR, its synchronous activity at homologous skin sites suggests that it can be used to indicate the presence of sympathetic sudomotor activity, although it cannot quantitate the absolute level of that activity (Schondorf, 1993). A large number of influencing factors that may affect correct derivation of SSR potentials have been described (Baba et al., 1988; Schondorf and Gendron, 1990). Nevertheless, many investigators employ
this technique to assess autonomic function in polyneuropathies, especially in diabetic autonomic neuropathy (Shahani et al., 1984; Knezevic and Bajada, 1985; Niakan and Harati, 1988; Shahani et al., 1990). Since it has been maintained that SSR is present in all healthy subjects (Shahani et al., 1984; Knezevic and Bajada, 1985; Uncini et al., 1988; Elie and Guiheneuc, 19901, especially under the age of 60 (Drory and Korczyn, 1993), loss of potentials is regarded as a pathological sign. To investigate the correlation between SSR abnormalities and the clinical state of diabetic neuropathy, we re-examined the influencing factors on test results such as age, gender, site of derivation, intensity of electrical stimuli and habituation. Then we calculated normal values and examined a population of 100 patients with diabetic neuropathy of different stages.
2. Patients and methods * Corresponding author. Tel: +49 (6421) 285242. Fax: t49 288955.
(6421)
0022-510X/96/$15.00 0 1996 Elsevier Science B.V. All rights reserved PII SOO22-510X(96)00023-8
All healthy volunteers as well as all patients gave their informed consent according to the Declaration of Helsinki
H.-J. Braune,
C. Horter/Joumal
of the
as revised by the 35th World Medical Assembly, Venice, Italy, 1983. Apart from clinical examinations, SSR measurements were performed consecutively on 100 outpatients with diabetes mellitus (definition of WHO, 1985) at the Dept. of Neurology of the University Hospital Marburg, Germany. The study was carried out over a period of 18 months on 54 men and 46 women (age 18-82 years, mean 52.0 years). On the day of the examination the patients had their usual breakfeast and, if necessary, their morning insulin dose. None of the subjects was allowed to drink coffee or tea, smoke, or take any medication in the hours before and during the visit. Major physical activity/strain and hypo- or hyperglycemic periods which occurred up to three days before examination led to exclusion from the study. Patients with symptoms not due to diabetes or with case histories other than diabetes, e.g. previous injury, lumbar discogenic disease, carpal tunnel syndrome, inherited or alcoholic neuropathy or other neurological diseases as well as patients taking medication acting on the sympathetic system were excluded. Another exclusion criterion was the presence of acute focal neuropathy and acute painful neuropathic syndroms which are probably caused by diabetes. However, these are rare and presumably of varying pathogenesis. Several studies demonstrate that the clinical course of these syndromes is different from that of common chronic distal symmetric polyneuropathy which primarily affects the sensory nerve fibers (Behse et al., 1977; Dyck et al., 1992). The patients were assessed independently by two experienced neurologists, and, in accordance with their clinical stage of neuropathy, divided into five subgroups, as recommended by Dyck et al. (1985): 8 patients without any clinical signs of neuropathy (NO), 15 patients with only one clinical symptom of neuropathy (Nl), 24 patients with mild distal symmetric sensory polyneuropathy (N2a), 28 patients with distal symmetric sensorimotor polyneuropathy (N2b), 25 patients with severe symmetric sensorimotor polyneuropathy (N3). Autonomic symptoms included vesical incontinence (12 patients), constipation (9 patients), sexual dysfunction (6 of 54 males) and orthostatic dysregulation (23 patients). Clinical autonomic symptoms were mild in 14 N2b-patients (only one symptom) and moderate in 18 N3-patients (two symptoms). The duration of diabetes was estimated based on hospital files and information given by patients: 20 patients had been suffering from diabetes for less than five years, 31 patients for more than five but less than 10 years, and 49 patients for more than ten years. A total of 26 patients younger than 40 years had IDDM, 74 had NIDDM. In 72 patients, metabolic control was good with normal fasting blood sugar values, and normal or only slightly elevated HbAlc values. It was moderate in 28 patients. Tests were also performed in an control group of 51 men and 49 women (age 18-83 years, mean 50.3 years) to define the normal range of values. Subjects were seated comfortably in an armchair in a
Neurological
Sciences
138 (19961 120-124
121
quiet, dimly lit room at 22°C. A 4-channel EMG system (NIHON KOHDEN Neuropack 4) was used with a sensitivity of 0.5 mV/div, analysis time 10 s, and Cut-off frequency filters 0.1 Hz and 20 Hz. Latencies were measured manually to the first continuous deflection from the baseline with a cursor. Maximum amplitudes were measured from peak to peak. Stimulation cathode was placed over the median nerve on the wrist contralateral to the recording site. Ipsilateral stimulation induced muscle contraction artifacts on SSR derivation from the hand (Fuhrer, 1971). A stimulus strength of 30 mA with duration 0.2 ms was tolerable by all subjects. Lower values were associated with poor reproducibility and even absence of response. Skin impedance was reduced to < 5 k0 by gentle skin abrasion and application of conductive gel. The SSR was recorded from the hands and feet with surface electrodes (9 mm dia) on the volardorsal and plantardorsal surface. When recordings were made from other skin sites: e.g. fingertips or thenar-hypothenar, or when comparing results from the left and right hand or the left and right foot ot ten healthy volunteers, statistically significant differences in amplitudes and latenties were not detected. In addition, intraindividual differences of SSR amplitudes and latencies could not be shown by means of comparing the derivations from respectively. Minimum time between successive stimuli was 60 s. The greatest amplitude and shortest latency of five consecutive SSR measurements were recorded for further data processing, since averaging of consecutive potentials did not result in significantly different values (Schondorf and Gendron, 1990). Statistical evaluation was performed with the non-parametric U-test (Mann-Whitney). The analyzed control values determine the 95th percentile. The minimum criterion for assumed abnormal values in patients was a test result of > 95th percentile or absence of SSR.
3. Results Many investigators have noted that SSR tends to habituate itself to repetitive stimuli (Shahani et al., 1984; Knezevie and Bajada, 1985; Elie and Guiheneuc, 1990). We observed a decrease in amplitude of more than 50% between the first and fifth stimulus (stimuli applied regulary every 30 s) in ten healthy volunteers due to habituation. As a consequence, irregular interstimulus intervals of more than one minute were applied to ensure reproducibility of the SSR. The SSR is generally biphasic and consists of an initial negativity followed by a positive potential. Delayed negativity was observed in 27 of 100 healthy volunteers, which corresponds to other publications (Schondorf and Gendron, 1990). In one test series, more than one wave form was found in 43% of the subjects and more than two wave form was found in 13% of the subjects. These findings are
122
H.-J. Braune, C. Horter/Journal
of the Neurological Sciences 138 (19961 120-124
Table 1 Influence of clinical stage of diabetic neuropathy on sympathetic skin response Clinical stage
NO
Nl
N2a
N2b
N3
15 0 (0) 3 (20)
24 4 (16.61 8 (33.31
28 6 (21.4) 14 (501
25 5 (20) 9 (361
Patients with values outside of the 95th~percentile normal range in derivable SSR (o/o) Latency on hand 3 (37.5) 4 (26.6) 7 (291 Amplitude on hand 3 (37.51 6 (39.9) 5 (20.1) Latency on foot 3 (37.51 4 (26.6) 3 (12.5) Amplitude on foot 2 (13.3) 3 (12.5) 0 (0)
10 (35.71 9 (32.1) 2 (7.1) 0 (0)
12 (48) 8 (32) 7 (281 4 (161
Loss of SSR or at least two abnormal values ( % of patients in subgroup NO-N3) 7~ Of subgroup 50 70 62
76
88
Loss of sympathetic skin responses (%) Number of patients 8 SSR loss on hand l(12.5) SSR loss on foot 3 (37.5)
Clinical staging as recommended by Dyck et al. (1992): NO: diabetics with no clinical sign of neuropathy; Nl: diabetics with only one clinical sign of neuropathy; N2a: diabetics with mild distal symmetric sensory neuropathy; N2b: diabetics with distal symmetric sensorimotor neuropathy; N3: diabetics with severe symmetric sensorimotor neuropathy.
in accordance with the observations made by BABA et al. (1988). Stimulus duration was 0.2 ms. All patients and control subjects tolerated this value, even when multiple measurements were required (Knezevic and Bajada, 1985; Uncini et al., 1988). SSR measurement of peak-to-peak amplitude in 100 healthy volunteers showed values of 2.91 & 1.10 mV on the hand and 1.37 + 0.54 mV on the foot. Latencies lasted 1.44 IfI 0.10 s on the hand and 2.12 + 0.15 s on the foot. SSR latencies were significantly shorter on hands than on feet (p < 0.01) and SSR amplitudes were significantly lower on feet (p < 0.01). The standard deviation of amplitudes provided evidence of considerable interindividual variations. Latencies were quite stable interindividually. A statistically significant age dependency of normal values regarding SSR amplitudes and latencies on the hand and foot was not shown. Consequently, a single normal range for all ages was determined, and the cut-off values were defined as 95th percentiles: latency on hand < 1.58 s, latency on foot < 2.44 s, amplitude on hand < 1.20 mV, amplitude on foot < 0.50 mV. Statistical analysis showed significant differences between the group of diabetic and healthy subjects, even when patients with SSR loss and their counterparts of the control group were excluded. A strict age- and sex-matched control group was not regarded as necessary, because there is such large variability in the results that the influence of age could not be demonstrated. Mean and range of years of patients and controls are nevertheless compareable. In diabetic subjects, latencies were significantly longer (p < 0.01) and amplitudes significantly smaller (p < 0.01) than in the healthy control group. Most striking was the loss of potentials on hands (16%) and feet (37%) in some diabet-
its, since SSR was always detectable in the normal control group. All patients with absent SSR on hands have absent SSR on the feet, whereas 21 patients with absent SSR on feet had derivable SSR on hands, although in these cases values of latencies or amplitudes were abnormal. Two or more abnormal test results (prolongation of SSR and/or amplitude reduction on the hand and/or foot, but no loss of potentials) were found in 39% of the patients. A statistically significant influence of gender, age, metabolic control, duration or type of diabetes on the test results could not be established or was existent only to a minor degree. However, the clinical stage of diabetic neuropathy (NO-N3) exerted a significant influence on our study: the further developed the clinical picture of the neuropathy was, the more pathologic were the results of SSR testing (Table 1). A total of 84% of the patients with clinical signs of autonomic involvement had loss of SSR on the foot, including 14 patients with additional SSR loss on the hand.
4. Discussion The aim of this study was to investigate the prevalence of pathologic SSR tests in a population with diabetes mellitus. After establishing a standardized test procedure in accordance with recent publications (Niakan and Harati, 1988; Schondorf and Gendron, 1990), a normal range ‘of values for latencies and amplitudes of SSR on hands and feet was defined as lying within the 95% range of all data gathered from 100 healthy volunteers of all ages. Since no age dependency could be found, a single cut-off value for all ages was considered to be appropriate. These findings are contrary to the investigation of Drory and Korcyn (1993) who emphasize that SSR can be elicited in all healthy subjects up to the age of 60, whereas a loss of SSR
H.-J. Braune,
C. Horter/
Journal
of the Neurological
can be found in about 50% of the septuagenarians and octogenarians without clinical findings. The authors stimulated the median nerve ” slightly over motor threshold intensity” which is generally significantly lower than the value of 30 mA used in our study. At a stimulation intensity of less than 10 mA we found only poor potential reproducibility; a considerable interpersonal variation in SSR amplitudes was found even between 10 and 25 mA. This may also explain the mean (&SD) response amplitudes in the Drory/Korcyn study of 449 + 429 PV in the upper and 147 + 122 PV in the lower limbs which is remarkably lower than the results obtained in our study (2910 & 1100 PV on hands and 1370 f 540 PV on feet>. Furthermore, the authors did not perform nerve conduction studies in healthy elderly subjects with SSR loss in order to rule out subclinical neuropathies. Thus we would like to stress that, contrary to the findings of Drory and Korczyn, SSR can be recorded in all subjects. This includes healthy elderly subjects, provided that the stimulus is applied with sufficient intensity, as suggested by previous studies (Shahani et al., 1984; Knezevic and Bajada, 1985; Uncini et al., 1988; Elie and Guiheneuc, 1990). Since SSR includes a somatic afferent phase which may be impaired in polyneuropathies, and since SSR show a high degree of variability due to a number of influencing factors, many investigators consider only the complete loss of SSR as abnormal (Shahani et al., 1984; Van den Bergh and Kelly, 1986; Soliven et al., 1987). Uncini et al. (1988) suggest that absent SSR in autonomic neuropathy may be due to an involvement of the large-diameter sensory afferent fibers of the median nerve mediating the SSR. However, this can only be true for patients with loss of SSR on upper and lower limbs. A simultaneous presence of derivable SSR on hands and absence of SSR on feet cannot be explained by this suggestion. Here, disturbances on the efferent limb of the reflex arc are of major importance. In accordance with this opinion, a pathological loss of potentials on hands was found in 16% and on feet in 37% of the diabetics. Autonomic system deterioration in diabetic neuropathy affects a variety of functions, especially the loss of preganglionic neurons which are short and myelinated. Symptoms become apparent when more than 50% of the neurons are lost (Low et al., 1977). This loss may contribute to the decrease in amplitude, and, in the later course of the disease, to loss of the SSR. It does not explain, however, why the loss is more pronounced on the lower limbs. Here, the ‘dying-back-hypothesis’ is helpful. It suggests that the longer peripheral nerves are, the more easily injured (Dyck, 1987). The postganglionic autonomic fibers are unmyelinated; thus, only loss of axons can occur which also may result in SSR amplitude reduction. Latency prolongation may be due to changes occurring in preganglionic myelinated fibers, loss of thicker axons or sweat gland alterations. A total of 84% of the patients with clinical signs of
Sciences
138 (19961 120-124
123
autonomic involvement had loss of SSR on the foot, including 14 patients with further SSR loss on the hand, which tallies with the findings of Soliven et al. (1987). The presence of clinically apparent autonomic dysfunction only occurs late in the course of diabetic neuropathy and is associated with a bad prognosis (Ewing et al., 1980; O’Brian et al., 1991). In our study, all patients with signs of autonomic disturbances were stage N2b-N3, only 3 patients with SSR loss were stage NO, and 3 others were stage Nl. Thus, exclusive assessment of the loss of SSR does not permit new insights into advanced diabetic neuropathy which can instead be diagnosed correctly by means of comprehensive histories, appropriate clinical examination and conventional neurophysiological tests. However, we propose that prolongation of latencies and reduction of amplitudes, if carefully derived under standardized conditions, should also be taken into consideration. Two or more abnormal test results (prolongations of SSR and/or amplitude reduction on hand and/or foot, but no loss of potentials) could be found in 39% of the patients. Together with the 37% of the diabetics who showed a loss of SSR, three quarters (76%) of all diabetics showed abnormal test results. The clinical stage of diabetic neuropathy (NO-N3) exerted a significant influence on our study: the further developed the clinical picture, the more pathologic were the results of SSR testing (Table 1). However, prolongation of latencies and reduction of amplitudes were found in 14 of 23 patients in the early stage of neuropathy (8 patients without any clinical signs of neuropathy (NO), 15 patients with only one clinical symptom of neuropathy (N 1)). Thus, earlier detection of sympathetic involvement in diabetic neuropathy seems to be possible by evaluation of all information achieved by SSR sampling.
5. Conclusion The results suggest that some changes leading to pathological SSR tests take place in a clinical stage in which no or only very slight clinical signs of diabetic neuropathy (NO-N 1) are apparent. Significant further deterioration of SSR test results can be expected in patients with clinically manifest neuropathy (stage N2a, N2b, N3) resulting in loss of potentials on feet and later on hands. Five-year-mortality of patients with autonomic disturbances is three times that of patients without autonomic involvement (Ewing et al., 1980; O’Brian et al., 1991; Sampson et al., 1990). But there is considerable controversy as to wether the worst in prognosis relates to vascular changes often occuring with diabetes or to autonomic failure itself. Nevertheless, SSR examination performed in diabetics with no or only slight clinical signs of neuropathy may help beside other tests determine those subjects with autonomic disturbances at an initial stage.
124
H.-J. Braune,
C. Horter/
Journal
of the Neurological
References Baba, M., Y. Watahiki, M. Matsunaga and K. Takebe (1988) Sympathetic skin response in healthy man. Electromyogr. Clin. Neurophysiol., 28: 277-283. Behse, F., F. Buchthal and F. Carlsen (1977) Nerve biopsy and conduction studies in diabetic neuropathy. J. Neurol. Neurosurg. Psychiatr., 40: 1072- 1082. Drory, V.E. and A.D. Korczyn (1993) Sympathetic skin response: age effect. Neurology, 43: 18 18- 1820. Dyck, P.J. (1987) Pathology. In: Dyck, P.J.. P.K. Thomas, A.K. Asbury, AI. Winegard and D. Porte (Eds.), Diabetic neuropathy, W. B. Saunders Company. Philadelphia, London, Toronto, Mexico City, Rio de Janeiro, Sydney, Tokyo, Hong Kong, pp. 223-236. Dyck, P.J., J.L. Karnes, J. Daube, P. O’Brian and F.J. Service (1985) Clinical and neuropathological criteria for the diagnosis ond staging of diabetic polyneuropathy. Brain, 108: 861-880. Dyck, P.J., J.L. Karnes, P.C. O’Brian, W.J. Litchy, P.A. Low and L.J. Melton III (1992) The Rochester diabetic neuropathy study: reassessment of tests and criteria for diagnosis and staged severity. Neurology, 42: 1164-1170. Elie, B. and P. Guiheneuc (1990) Sympathetic skin response: normal results in different experimental conditions. Electroencephalogr. Clin. Neurophysiol., 76: 258-267. Ewing, D.J., I.W. Campbell and B.F. Clarke (1980) The natural history of diabetics autonomic neuropathy. Quart. J. Med., 49: 95-99. Fuhrer, M.J. (1971) Effects of unilateral stimuli on the magnitude and latency of bilaterally recorded skin conductance responses. Psychophysiology, 8: 740-75 1. Knezevic, W. and S. Bajada (1985) Peripheral autonomic surface potential. A quantitative technique for recording sympathetic conduction in man. J. Neural. Sci., 67: 239-251. Low, P.A., H. Okazaki and P.J. Dyck (1977) Splanchnic preganglionic
Sciences
138 (19961 120-124
neurons in man. I. Morphometry of preganlionic cytons. Acta Neuropathol. (Berlin), 40: 55-61. Niakan, E. and Y. Harati (1988) Sympathetic skin response in diabetic peripheral neuropahty. Muscle Nerve, 11: 261-264. O’Brian, LA., J.P. McFadden and R.J.M. Corral1 (1991) The influence of autonomic neuropathy on mortality in insulin-dependent diabetes. Quart. J. Med., 79: 495-502. Sampson, M.J., S. Wilson, P. Karagiannis, M. Edmonds and P.J. Watkins (1990) Progression of diabetic autonomic neuropathy over a decade in insulin-dependent diabetics. Quart. J. Med., 75: 635-646. Schondorf, R. (1993) The role of the sympatheitc skin response in the assessment ot autonomic function. In: Low, P.A. (ed.), Clinical autonomic disorders, Little, Brown and Company, Boston, pp. 231241. Schondorf, R. and D. Gendron (1990) Properties of electrodermal activity recorded from non palmar/plantar skin sites. Neurology, 40: 128- 132. Shahani, B.T., T.J. Day, D. Cros, N. Khalil and C.S. Kneebone (1990) RR interval variation and the sympathetic skin response in the assessment of autonomic function in peripheral neuropathy. Arch Neurol., 47: 659-664. Shahani, B.T., J.J. Halperin, P. Boulu and J. Cohen (1984) Sympathetic skin response: a method of assessing unmyelinated axon dysfunction in peripheral neuropathies. J. Neural. Neurosurg. Psychiat., 47: 536542. Soliven, B., R. Maselli, J. Jaspan, A. Green, H. Graziano, M. Petersen and J.P. Spire (1987) Sympathetic skin response in diabetic neuropathy. Muscle Nerve, IO: 711-716. Uncini, A., S.L. Pullmann, R.E. Lovelace and D. Gambi (1988) The sympathetic skin response: normal values, elucidation of afferent components and application limits. J. Neurol. Sci., 87: 299-306. Van den Bergh, P. and J.J. Kelly (1986) The evoked electrodermal response in peripheral neuropathies. Muscle Nerve, 9: 656-657. WHO (I 985) Diabetes mellitus: Report of a study group, Technical report series, No. 727, World health organization, Geneva.