- Email: [email protected]
Relationship of skin temperature to sympathetic dysfunction in diabetic at-risk feet
Diabetes Research and Clinical Practice 73 (2006) 41–46 www.elsevier.com/locate/diabres
Relationship of skin temperature to sympathetic dysfunction in diabetic at-risk feet Pi-Chang Sun a,b, Hong-Da Lin c, Shyh-Hua Eric Jao a,1, Yan-Chiou Ku d, Rai-Chi Chan a, Cheng-Kung Cheng a,* a
Orthopedic Biomechanics Laboratory, Institute of Rehabilitation Science and Technology, National Yang Ming University, No. 155, Sec. 2, Li-Nung St, Taipei 11221, Taiwan b Division of Physical Medicine and Rehabilitation, Taoyuan Veteran Hospital, Taoyuan, Taiwan c Division of Endocrinology and Metabolism, Taipei Veteran General Hospital, Taipei, Taiwan d Department of Nursing, Taoyuan Veteran Hospital, Taoyuan, Taiwan Received 25 April 2005; received in revised form 28 October 2005; accepted 16 December 2005 Available online 17 February 2006
Abstract The relationship of plantar skin temperature to diabetic neuropathy was studied using clinical, nerve conduction and autonomic evaluations. The sympathetic skin response (SSR) was found present in both feet of 25 control subjects and 29 (out of 69) diabetic patients (SSR+ group). For those diabetic patients absent with the SSR in both feet, 18 patients (at-risk group) had preulcerative skin lesions (dry and fissured skin) and 22 did not (SSR group). The at-risk group showed significantly higher mean foot temperature (30.2 1.3 8C) than the SSR (27.9 1.7 8C), the SSR+ (27.1 2.0 8C) and the control group (26.8 1.8 8C). The SSR group had smaller temperature differences (7.2 1.7 8C versus 8.6 1.6 8C, p < 0.05) and smaller normalized temperature (referencing to the forehead temperature) (0.19 versus 0.24, p < 0.01) than the SSR+ group. Although the three diabetic groups had no significant differences in clinical and cardiovascular abnormalities, the at-risk group showed more nerve conduction abnormalities than the SSR and SSR+ groups (55% versus 23% and 14%, p < 0.02). This study indicated that the thermoregulatory sweating abnormality signified early sympathetic damage in diabetic feet. Assessing skin conditions and sudomotor activities should help identify small fiber neuropathy in diabetic patients with at-risk feet conditions. # 2006 Elsevier Ireland Ltd. All rights reserved. Keywords: Skin temperature; Diabetic foot; Sympathetic skin response
1. Introduction Foot diseases have been known as a frequent complication of diabetes mellitus. Severe foot complications such as ulcers and amputations result in higher mortality and medical costs for people with diabetes. Risk
* Corresponding author. Tel.: +886 2 28267020; fax: +886 2 28202519. E-mail address: [email protected] (C.-K. Cheng). 1 Current address: School of Engineering, University of Mississippi, USA.
identification is fundamental in the preventive healthcare management of the diabetic patients having potential foot problems [1]. The autonomic neuropathy in the lower limbs of diabetic patients leads to vasomotor disturbance, reduced sweating and abnormal skin conditions including anhydrosis, fissures and blisters [2]. Peripheral autonomic dysfunction was found to be associated with the development of foot lesions in diabetic subjects [3]. Traditional autonomic tests with evaluation mainly on cardiovascular reflex yield an assessment more of the central autonomic function. Clinical examinations (e.g. monofilament and tuning fork) and nerve conduction
0168-8227/$ – see front matter # 2006 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.diabres.2005.12.012
42
P.-C. Sun et al. / Diabetes Research and Clinical Practice 73 (2006) 41–46
studies do not adequately investigate the degree of peripheral autonomic involvement [4]. The sympathetic skin response (SSR) involving multiple levels of the nervous system can be used to assess the sudomotor function of the extremity by reflecting the integrity of unmyelinated sympathetic fibers [5]. It was reported that most diabetic patients with generalized autonomic symptoms revealed absent sympathetic skin response [6]. However, the decreased sudomotor activity has not yet been investigated using the SSR assessment in diabetic patients at risk of foot complications. Skin temperature has been used to detect acute tissue damage in neuropathic joint and ulcer [7,8]. Thermal images can be an adjuvant in the diagnosis of these major foot lesions. Nevertheless, little is known of the range of abnormal thermoregulation in the diabetic patients with minor skin lesions. The skin temperature was not well studied in diabetic feet with varying degrees of polyneuropathy. It might be of interest to determine whether thermal asymmetry in neuropathic feet is related to their clinical or electrophysiological abnormalities. The purpose of this study is to characterize the distribution of the diabetic foot skin temperature and seek correlation of the mean skin temperature with autonomic dysfunction as detected by sympathetic skin response as well as cardiovascular reflex tests. The correlation of the mean skin temperature with peripheral somatic neuropathy by using nerve conduction studies and clinical examinations was also studied.
2. Materials and methods Subjects were recruited from the members of a diabetes association, representing a diabetic population with fairly good medical conditions. Patients with history of foot ulcer, overt neurological symptoms or peripheral vascular diseases (ankle brachial pressure index < 0.9) were excluded from this
study. Informed consents were obtained from all the participants. There were 69 type 2 diabetic patients and 25 nondiabetic control subjects chosen for this study. The diabetic patients were classified depending on the presence of minor skin lesions and the result of the sympathetic skin response test in both feet. There were 29 diabetic patients who showed presence of the SSR and were classified as the SSR+ group. Among the remaining 40 diabetic patients with absent SSR, 18 patients having preulcerative foot lesions such as dry and fissured skin were considered as the at-risk group, and other 22 patients did not have the minor skin lesions (SSR group). Note that the SSR was found present over both feet in all the control subjects as expected. The criterion used to screen for minor skin lesions were based on the Seattle Wound Classification System graded 1.2–1.3, i.e., cracked or fissured skin involving only epidermis or dermis layers [9]. The clinical characteristics of all subjects were shown in Table 1. Skin temperatures of both feet were assessed using medical thermal imaging radiometer system (Spectrum 9000 MB; Biovision Technologies, Inc., Taipei, Taiwan). The operator who measured and stored the thermal data was blinded as to either the grouping or the condition of the feet of all the subjects. All measurements were performed in the morning to eliminate the influence of diurnal temperature variation on the subjects. All subjects were asked to remain in a seated position with bare feet for 15–20 min to achieve equilibration with the constant ambient temperature (21 1 8C). A cool test environment was required for reliable data acquisition during the thermographic measurement as reported by prior study [10]. The plantar thermal image was divided into six regions of interest as shown in Fig. 1, and the mean foot temperature was calculated from these six regions. The group temperature average was obtained from the average value of both feet of the subject. The forehead temperature was also measured as a reference in order to conduct a data process of temperature normalization by the forehead temperature as given below. DT TN ¼ ; TR where TN is the normalized temperature (no unit), DT = TR TMF the temperature difference (8C), TR the cor-
Table 1 Clinical characteristic
Age Men/women BMI (kg/m2) DM duration Treatment (I:O) FBG (mg/dl) HbA1c (%)
At risk (N = 18)
SSR (N = 22)
SSR+ (N = 29)
Control (N = 25)
56.6 7.9 11/7 25.4 3.7 18.3 7.1* 3:15 134 35 6.8 1.4
54.7 5.5 13/9 27.1 4.4 12.5 5.3 4:18 125 22 7.2 1.3
58 6.3 17/12 26.3 4.2 9.8 2.9 4:25 142 35 7.7 1.5
55.8 5.6 16/9 25.8 3.9 – – – –
Data are expressed as mean S.D. N: numbers of subjects; SSR: absent sympathetic skin response; SSR+: present sympathetic skin response; treatment (I:O): insulin:oral hypoglycemic agent + diet; FBG: fasting blood glucose. * p < 0.01, at-risk group compared with SSR and SSR+ groups.
P.-C. Sun et al. / Diabetes Research and Clinical Practice 73 (2006) 41–46
43
Fig. 1. Experimental set-up and graphic representation of designated areas in both feet: A and G, hallux; B and H, lesser toes; C and I, forefoot; D and J, arch; E and K, lateral sole; F and L, heel.
responding forehead temperature (8C) and TMF is the mean foot temperature (8C). This temperature normalization was adopted in this study for its reported good reliability [10]. SSR was investigated using a widely available electromyographic machine (Dantec, Keypoint, Skovlunde, Denmark). The method detected voltage changes from the skin surface as a result of the sudomotor fiber activities triggered by median nerve electrical stimulations at wrist with duration of 0.1 ms and intensity between 10 and 50 mA. The responses were filtered at a band pass of 0.5–2 kHz and recorded with active electrodes placed on both soles and reference electrodes on dorsal surfaces [11,12]. Amplitude and latency of each response was not assessed in this study. Absent response was defined when no voltage was found after 10 consecutive stimulations. Two standardized cardiovascular reflex tests were used to assess their autonomic function. The postural blood pressure change was measured from sitting to standing position. The heart rate variation during deep breathing was assessed by measuring the difference between maximum and minimum heart rate. All tests were performed and interpreted according to the described methods and threshold of normality [13]. All diabetic patients underwent a quantitative neurological examination, followed by a set of nerve conduction studies to evaluate peripheral somatic nerve function. The 128 Hz tuning fork and Semmes–Weinstein monofilament were applied to the dorsum of the great toe, and the patients were asked to respond if they felt the vibration or light touch. The sensation, muscle strength and tendon reflex was scored separately according to the described criteria [14]. The sural, peroneal motor, median sensory and motor, and ulnar sensory nerves
were evaluated [4]. The severity of somatic neuropathy was graded according to the Michigan Diabetic Neuropathy Score [14]. Each patient was given a composite score and graded based on the sum of scores on the clinical examination and the number of abnormal nerve conductions. Diabetic patients with neuropathy score > 6 points were considered clinically abnormal, and patients having two or more abnormal conduction studies were considered electrophysiologically abnormal. 2.1. Statistical analysis All data were expressed as means S.D. and analyzed by the SPSS software package (SPSS, Inc., Chicago, IL). ANOVA was used to determine overall temperature differences among the four groups. When significant differences were seen in the multiple groups’ comparison, the Scheffe post hoc test was used to identify which group differences accounted for the significant p-value. Differences in the frequency of CVR, clinical and nerve conduction abnormalities between groups were compared using Fisher’s exact test. The difference was considered statistically significant when the p-value was less than 0.05.
3. Results There were no significant differences in gender, age and BMI among the four groups (Table 1). No statistical difference was found between the three diabetic groups in the mean HbA1c. The patients with minor foot lesions, i.e., the at-risk group, had longer duration of
Table 2 Skin temperature measurements among four groups At risk (N = 18) Mean Mean Mean Mean
TMF TR DT TN
*
30.2 1.3 35.1 1.2 4.9 1.3* 0.14*
SSR (N = 22)
SSR+ (N = 29)
Control (N = 25)
p
27.9 1.7 35.3 1.1 7.2 1.7y 0.19yy
27.1 2.0 35.6 1.3 8.6 1.6 0.24
26.8 1.8 34.9 1.2 8.4 2.0 0.23
<0.001 >0.05 <0.0001 <0.0001
TMF: mean foot temperature; TR: reference (forehead) temperature; DT: TR – TMF; TN = DT/TR normalized temperature. * p < 0.0001, at risk compared with SSR and SSR+. y p < 0.05, SSR compared with SSR+. yy p < 0.01, SSR compared with SSR+.
44
P.-C. Sun et al. / Diabetes Research and Clinical Practice 73 (2006) 41–46
Table 3 showed the numbers of patients with cardiovascular, clinical and nerve conduction abnormalities of the three diabetic groups. There were no significant differences in the percentage of cardiovascular and clinical abnormalities among the three groups. The at-risk group had a significantly higher rate (10 out of 18, p < 0.02) of abnormal nerve conductions than the other diabetic groups. The SSR group seemed have slightly higher percentage of abnormal nerve conductions (5 of 22; 23%) than the SSR+ group (4 of 29; 14%), but their difference was not statistically significant ( p > 0.05). Fig. 2. Temperature distribution in six plantar areas among four groups.
4. Discussions
diabetes than the other two diabetic groups ( p < 0.01). The mean foot temperature was obtained in each group as shown in Table 2. The at-risk group with preulcerative lesion had significantly higher (30.2 1.7 8C) temperature than all other groups ( p < 0.0001). Our results showed that overall temperature was increased unanimously over the six plantar areas in the group with feet at risk (Fig. 2). In terms of the mean foot temperature, the SSR group was higher (27.9 2.0 8C) than the SSR+ group (27.1 2.4 8C), which in turns was higher than the control group (26.8 2.3 8C). However, the differences were not statistically significant among these 3 groups ( p > 0.05). The mean DT as defined by subtracting the foot temperature from the forehead temperature was shown in Table 2. The four groups had pretty close mean forehead temperatures (TR). The mean foot temperature was lower (4.9 1.5 8C) than the mean forehead temperature in the at-risk group. The mean DT in the SSR group was significantly smaller than that in the SSR+ group (7.3 1.8 8C versus 8.6 2.1 8C, p < 0.05) and noted that there was no significant difference when looking at the mean foot temperature alone as mentioned above. The normalized temperature TN by the forehead temperature also showed significant difference between the SSR group and the SSR+ group (0.19 versus 0.24, p < 0.001).
The measurement of plantar skin temperature using infrared can discern small thermal difference beyond the palpable limit, particularly for the patients without gross inflammatory symptoms as in our experiment. Sympathetic dysfunction in the lower limbs leads to reduced sweating and dry skin, which is prone to crack and fissure. These dermatological conditions are highly related to thermoregulatory sweating abnormalities [1,15]. The diabetic patients in this study were categorized based on the degree of peripheral sudomotor dysfunctions. The concordance of results of the patients’ grouping and the skin temperature difference implies that small fiber neuropathy plays a preliminary role in the development of microcirculatory disturbance. The at-risk patients all showed absence of sympathetic skin response in both feet. The decreased sudomotor activity and concomitant thermoregulatory disturbance should be an early sign of sympathetic damage in diabetic feet [16]. Note that the peripheral small fiber neuropathy could not be accurately evaluated by the clinical and autonomic tests mentioned previously. The temperature and electrophysiological assessment conducted in this study proved to be useful for healthcare providers to identify the patients with sympathetic dysfunction in early stage to warrant their feet care. The sympathetic skin response is mediated by the postganglionic, unmyelinated small fibers via the
Table 3 The number and percentage of patients with abnormal studies
Cardiovascular reflex tests Postural blood pressure change Clinical score Nerve conduction *
At risk (N = 18)
SSR (N = 22)
SSR+ (N = 29)
3 1 4 10
5 0 4 5
4 1 3 4
p < 0.02, at risk compared with SSR and SSR+.
(16%) (5%) (22%) (55%)*
(23%) (0%) (18%) (23%)
(14%) (3%) (10%) (14%)
P.-C. Sun et al. / Diabetes Research and Clinical Practice 73 (2006) 41–46
sudomotor pathway, which is anatomically distinct from the vasomotor pathway [5]. Skin temperature is regulated by the sympathetic vasomotor and sudomotor functions. The neuropathic foot has elevated skin temperature primarily through increased arteriovenous shunt flow [17,18]. The absence of SSR indicated only sudomotor dysfunction to some degrees. However, The minimal temperature difference between the SSR and SSR+ group can be discerned by the normalized temperature, but not by the absolute plantar temperature. These findings show that the sympathetic skin response plays a certain role in the prediction of diabetic patients with thermoregulatory dysfunction. The SSR patients require further investigation to confirm if they have definite vasomotor problems. The peripheral neuropathy affects both somatic and sympathetic components in diabetic patients. Clinical and nerve conduction studies investigate only the somatic function of large myelinated fibers [4]. The atrisk group with longer diabetic duration is supposed to have more somatic nerves involved than the other diabetic groups. However, the clinical and nerve conduction examinations are not useful in discerning the two groups with and without SSR. The presence of SSR depends on intact innervations through the small unmyelinated fibers, which are frequently affected in diabetics [6]. Damage to the ummyelinated small fibers should be characteristic of neuropathic feet during its early stage [19]. The presence of peripheral small fiber neuropathy is more associated with temperature abnormality in the diabetic feet with subclinical symptoms than large fiber neuropathy. The SSR test presents itself as an adjunct method distinct from the traditional nerve conduction and autonomic studies in evaluating the peripheral nervous system. This study shows that the application of clinical examinations or nerve conduction study alone is not adequate in screening diabetic at-risk feet at early stage. Most previous studies emphasized the importance of identifying foot ulcerations in order to predict the risk of amputation [20,21], but they did not investigate on the roles of the preulcerative skin lesions in a cascade of worsening sequelae gradually leading to the ultimate disability of the lower extremity. The break of cutaneous barrier makes the skin more susceptible to undue mechanical injury [15]. The feet with elevated skin temperature should be indicative of early diabetic neuropathy as well reported in the literatures [1,18]. The potential arteriovenous shunting and compromised nutritive capillary flow could contribute to foot ulceration and amputation [17]. The minor skin lesions may be a pivotal precursor to subsequent foot problems,
45
though their causal relationship has yet to be clearly established. Follow-up of this patient population is needed to determine whether the preulcerative lesion is a strong indicator of the development of severe diabetic foot complications later. In conclusion, the thermoregulatory disturbance and sweating dysfunction should be a sign of early sympathetic damage in diabetic feet. Assessing skin conditions and sudomotor activities shall help healthcare providers to identify the patients with small fiber neuropathy in early stage more accurately, in order to alert them of their feet conditions. References [1] P.J. Watkins, The diabetic foot, BMJ 326 (2003) 977–979. [2] P.J. Watkins, P.K. Thomas, Diabetes Mellitus and the nervous system, J. Neurol. Neurosurg. Psychiatry 65 (1998) 620–663. [3] J.E. Gilmore, J.A. Ellen, J.R. Hayes, Autonomic function in diabetic neuropathic patients with foot ulceration, Diabetes Care 16 (1993) 61–67. [4] J. Kimura, Principles and variations of nerve conduction studies, in: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice, Oxford University Press, New York, 2001, pp. 91– 129. [5] B.T. Shahani, J.J. Halperin, P. Boulu, J. Cohen, Sympathetic skin response—a method of assessing unmyelinated axon dysfunction in peripheral neuropathies, J. Neurol. Neurosurg. Psychiatry 47 (1984) 536–542. [6] B. Soliven, R. Maselli, J. Jaspan, Sympatheic skin response in diabetic neuropathy, Muscle Nerve 10 (1987) 711–716. [7] M. McGill, L. Molyneaux, T. Bolton, K. Ioannou, R. Uren, D.K. Yue, Response of Charcot’s arthropathy to contact casting: assessment by quantitative techniques, Diabetologia 43 (2000) 481–484. [8] E.J. Boyko, J.H. Ahroni, V.L. Stensel, Skin temperature in the neuropathic diabetic foot, J. Diabetes Complications 15 (2001) 260–264. [9] D.K. Litzelman, D.J. Marriott, F. Vinicor, Independent physiological predictors of foot lesions in patients with NIDDM, Diabetes Care 20 (1997) 1273–1278. [10] P.C. Sun, S.H. Jao, C.K. Cheng, Assessing foot temperature using infrared thermography, Foot Ankle Int. 26 (2005) 847–853. [11] T. Zgur, D.B. Vodusek, M. Krzan, M. Vrtovec, M. Denislic, B. Sibanc, Autonomic system dysfunction in moderate diabetic polyneuropathy assessed by sympathetic skin response and Valsalva index, Electromyogr. Clin. Neurophysiol. 33 (1993) 433–439. [12] C. Ionescu-Tirgoviste, S. Pruna, I. Mincu, Peripheral sympathetic neuropathy evaluated by recording the evoked electrodermal response using an impedance reactometer, Diabetes Res. Clin. Pract. 9 (1990) 201–209. [13] H.S. Chen, C.M. Hwu, B.I. Kuo, Abnormal cardiovascular reflex tests are predictors of mortality in type 2 diabetes mellitus, Diabet. Med. 18 (2001) 268–273. [14] E.L. Feldman, M.J. Stevens, P.K. Thomas, M.B. Brown, N. Canal, D.A. Greene, A practical two steps quantitative clinical electrophysiological assessment for the diagnosis and staging diabetic neuropathy, Diabetes Care 11 (1994) 1281–1289.
46
P.-C. Sun et al. / Diabetes Research and Clinical Practice 73 (2006) 41–46
[15] K. Springett, R.J. White, Skin changes in the at risk foot and their treatment, Br. J. Community Nurs. 12 (2002) 25–32. [16] R.D. Hoeldtke, K.D. Bryner, G.G. Horvath, R.W. Phares, L.F. Broy, G.R. Hobbs, Redistribution of sudomotor response is an early sign of sympathetic dysfunction in type 1 diabetes, Diabetes 50 (2001) 436–443. [17] L. Uccioli, L. Mancini, A. Giordano, A. Solini, P. Magnani, A. Manto, et al., Lower limb arterio-venous shunts, autonomic neuropathy and diabetic foot, Diabetes Res. Clin. Pract. 16 (1992) 123–130.
[18] M.D. Flynn, J.E. Tooke, Diabetic neuropathy and the microcirculation, Diabet. Med. 12 (1995) 298–301. [19] C.J. Tack, P.J. van GurP, C. Holmes, D.S. Goldstein, Local sympathetic denervation in painful diabetic neuropathy, Diabetes 51 (2002) 3545–3553. [20] E.J. Peters, L.A. Lavery, Effectiveness of the diabetic foot risk classification system of the international working group on the diabetic foot, Diabetes Care 24 (2001) 1442–1447. [21] B.E. Sumpio, Foot ulcers, N. Engl. J. Med. 343 (2000) 787– 793.
Relationship of skin temperature to sympathetic dysfunction in diabetic at-risk feet Pi-Chang Sun a,b, Hong-Da Lin c, Shyh-Hua Eric Jao a,1, Yan-Chiou Ku d, Rai-Chi Chan a, Cheng-Kung Cheng a,* a
Orthopedic Biomechanics Laboratory, Institute of Rehabilitation Science and Technology, National Yang Ming University, No. 155, Sec. 2, Li-Nung St, Taipei 11221, Taiwan b Division of Physical Medicine and Rehabilitation, Taoyuan Veteran Hospital, Taoyuan, Taiwan c Division of Endocrinology and Metabolism, Taipei Veteran General Hospital, Taipei, Taiwan d Department of Nursing, Taoyuan Veteran Hospital, Taoyuan, Taiwan Received 25 April 2005; received in revised form 28 October 2005; accepted 16 December 2005 Available online 17 February 2006
Abstract The relationship of plantar skin temperature to diabetic neuropathy was studied using clinical, nerve conduction and autonomic evaluations. The sympathetic skin response (SSR) was found present in both feet of 25 control subjects and 29 (out of 69) diabetic patients (SSR+ group). For those diabetic patients absent with the SSR in both feet, 18 patients (at-risk group) had preulcerative skin lesions (dry and fissured skin) and 22 did not (SSR group). The at-risk group showed significantly higher mean foot temperature (30.2 1.3 8C) than the SSR (27.9 1.7 8C), the SSR+ (27.1 2.0 8C) and the control group (26.8 1.8 8C). The SSR group had smaller temperature differences (7.2 1.7 8C versus 8.6 1.6 8C, p < 0.05) and smaller normalized temperature (referencing to the forehead temperature) (0.19 versus 0.24, p < 0.01) than the SSR+ group. Although the three diabetic groups had no significant differences in clinical and cardiovascular abnormalities, the at-risk group showed more nerve conduction abnormalities than the SSR and SSR+ groups (55% versus 23% and 14%, p < 0.02). This study indicated that the thermoregulatory sweating abnormality signified early sympathetic damage in diabetic feet. Assessing skin conditions and sudomotor activities should help identify small fiber neuropathy in diabetic patients with at-risk feet conditions. # 2006 Elsevier Ireland Ltd. All rights reserved. Keywords: Skin temperature; Diabetic foot; Sympathetic skin response
1. Introduction Foot diseases have been known as a frequent complication of diabetes mellitus. Severe foot complications such as ulcers and amputations result in higher mortality and medical costs for people with diabetes. Risk
* Corresponding author. Tel.: +886 2 28267020; fax: +886 2 28202519. E-mail address: [email protected] (C.-K. Cheng). 1 Current address: School of Engineering, University of Mississippi, USA.
identification is fundamental in the preventive healthcare management of the diabetic patients having potential foot problems [1]. The autonomic neuropathy in the lower limbs of diabetic patients leads to vasomotor disturbance, reduced sweating and abnormal skin conditions including anhydrosis, fissures and blisters [2]. Peripheral autonomic dysfunction was found to be associated with the development of foot lesions in diabetic subjects [3]. Traditional autonomic tests with evaluation mainly on cardiovascular reflex yield an assessment more of the central autonomic function. Clinical examinations (e.g. monofilament and tuning fork) and nerve conduction
0168-8227/$ – see front matter # 2006 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.diabres.2005.12.012
42
P.-C. Sun et al. / Diabetes Research and Clinical Practice 73 (2006) 41–46
studies do not adequately investigate the degree of peripheral autonomic involvement [4]. The sympathetic skin response (SSR) involving multiple levels of the nervous system can be used to assess the sudomotor function of the extremity by reflecting the integrity of unmyelinated sympathetic fibers [5]. It was reported that most diabetic patients with generalized autonomic symptoms revealed absent sympathetic skin response [6]. However, the decreased sudomotor activity has not yet been investigated using the SSR assessment in diabetic patients at risk of foot complications. Skin temperature has been used to detect acute tissue damage in neuropathic joint and ulcer [7,8]. Thermal images can be an adjuvant in the diagnosis of these major foot lesions. Nevertheless, little is known of the range of abnormal thermoregulation in the diabetic patients with minor skin lesions. The skin temperature was not well studied in diabetic feet with varying degrees of polyneuropathy. It might be of interest to determine whether thermal asymmetry in neuropathic feet is related to their clinical or electrophysiological abnormalities. The purpose of this study is to characterize the distribution of the diabetic foot skin temperature and seek correlation of the mean skin temperature with autonomic dysfunction as detected by sympathetic skin response as well as cardiovascular reflex tests. The correlation of the mean skin temperature with peripheral somatic neuropathy by using nerve conduction studies and clinical examinations was also studied.
2. Materials and methods Subjects were recruited from the members of a diabetes association, representing a diabetic population with fairly good medical conditions. Patients with history of foot ulcer, overt neurological symptoms or peripheral vascular diseases (ankle brachial pressure index < 0.9) were excluded from this
study. Informed consents were obtained from all the participants. There were 69 type 2 diabetic patients and 25 nondiabetic control subjects chosen for this study. The diabetic patients were classified depending on the presence of minor skin lesions and the result of the sympathetic skin response test in both feet. There were 29 diabetic patients who showed presence of the SSR and were classified as the SSR+ group. Among the remaining 40 diabetic patients with absent SSR, 18 patients having preulcerative foot lesions such as dry and fissured skin were considered as the at-risk group, and other 22 patients did not have the minor skin lesions (SSR group). Note that the SSR was found present over both feet in all the control subjects as expected. The criterion used to screen for minor skin lesions were based on the Seattle Wound Classification System graded 1.2–1.3, i.e., cracked or fissured skin involving only epidermis or dermis layers [9]. The clinical characteristics of all subjects were shown in Table 1. Skin temperatures of both feet were assessed using medical thermal imaging radiometer system (Spectrum 9000 MB; Biovision Technologies, Inc., Taipei, Taiwan). The operator who measured and stored the thermal data was blinded as to either the grouping or the condition of the feet of all the subjects. All measurements were performed in the morning to eliminate the influence of diurnal temperature variation on the subjects. All subjects were asked to remain in a seated position with bare feet for 15–20 min to achieve equilibration with the constant ambient temperature (21 1 8C). A cool test environment was required for reliable data acquisition during the thermographic measurement as reported by prior study [10]. The plantar thermal image was divided into six regions of interest as shown in Fig. 1, and the mean foot temperature was calculated from these six regions. The group temperature average was obtained from the average value of both feet of the subject. The forehead temperature was also measured as a reference in order to conduct a data process of temperature normalization by the forehead temperature as given below. DT TN ¼ ; TR where TN is the normalized temperature (no unit), DT = TR TMF the temperature difference (8C), TR the cor-
Table 1 Clinical characteristic
Age Men/women BMI (kg/m2) DM duration Treatment (I:O) FBG (mg/dl) HbA1c (%)
At risk (N = 18)
SSR (N = 22)
SSR+ (N = 29)
Control (N = 25)
56.6 7.9 11/7 25.4 3.7 18.3 7.1* 3:15 134 35 6.8 1.4
54.7 5.5 13/9 27.1 4.4 12.5 5.3 4:18 125 22 7.2 1.3
58 6.3 17/12 26.3 4.2 9.8 2.9 4:25 142 35 7.7 1.5
55.8 5.6 16/9 25.8 3.9 – – – –
Data are expressed as mean S.D. N: numbers of subjects; SSR: absent sympathetic skin response; SSR+: present sympathetic skin response; treatment (I:O): insulin:oral hypoglycemic agent + diet; FBG: fasting blood glucose. * p < 0.01, at-risk group compared with SSR and SSR+ groups.
P.-C. Sun et al. / Diabetes Research and Clinical Practice 73 (2006) 41–46
43
Fig. 1. Experimental set-up and graphic representation of designated areas in both feet: A and G, hallux; B and H, lesser toes; C and I, forefoot; D and J, arch; E and K, lateral sole; F and L, heel.
responding forehead temperature (8C) and TMF is the mean foot temperature (8C). This temperature normalization was adopted in this study for its reported good reliability [10]. SSR was investigated using a widely available electromyographic machine (Dantec, Keypoint, Skovlunde, Denmark). The method detected voltage changes from the skin surface as a result of the sudomotor fiber activities triggered by median nerve electrical stimulations at wrist with duration of 0.1 ms and intensity between 10 and 50 mA. The responses were filtered at a band pass of 0.5–2 kHz and recorded with active electrodes placed on both soles and reference electrodes on dorsal surfaces [11,12]. Amplitude and latency of each response was not assessed in this study. Absent response was defined when no voltage was found after 10 consecutive stimulations. Two standardized cardiovascular reflex tests were used to assess their autonomic function. The postural blood pressure change was measured from sitting to standing position. The heart rate variation during deep breathing was assessed by measuring the difference between maximum and minimum heart rate. All tests were performed and interpreted according to the described methods and threshold of normality [13]. All diabetic patients underwent a quantitative neurological examination, followed by a set of nerve conduction studies to evaluate peripheral somatic nerve function. The 128 Hz tuning fork and Semmes–Weinstein monofilament were applied to the dorsum of the great toe, and the patients were asked to respond if they felt the vibration or light touch. The sensation, muscle strength and tendon reflex was scored separately according to the described criteria [14]. The sural, peroneal motor, median sensory and motor, and ulnar sensory nerves
were evaluated [4]. The severity of somatic neuropathy was graded according to the Michigan Diabetic Neuropathy Score [14]. Each patient was given a composite score and graded based on the sum of scores on the clinical examination and the number of abnormal nerve conductions. Diabetic patients with neuropathy score > 6 points were considered clinically abnormal, and patients having two or more abnormal conduction studies were considered electrophysiologically abnormal. 2.1. Statistical analysis All data were expressed as means S.D. and analyzed by the SPSS software package (SPSS, Inc., Chicago, IL). ANOVA was used to determine overall temperature differences among the four groups. When significant differences were seen in the multiple groups’ comparison, the Scheffe post hoc test was used to identify which group differences accounted for the significant p-value. Differences in the frequency of CVR, clinical and nerve conduction abnormalities between groups were compared using Fisher’s exact test. The difference was considered statistically significant when the p-value was less than 0.05.
3. Results There were no significant differences in gender, age and BMI among the four groups (Table 1). No statistical difference was found between the three diabetic groups in the mean HbA1c. The patients with minor foot lesions, i.e., the at-risk group, had longer duration of
Table 2 Skin temperature measurements among four groups At risk (N = 18) Mean Mean Mean Mean
TMF TR DT TN
*
30.2 1.3 35.1 1.2 4.9 1.3* 0.14*
SSR (N = 22)
SSR+ (N = 29)
Control (N = 25)
p
27.9 1.7 35.3 1.1 7.2 1.7y 0.19yy
27.1 2.0 35.6 1.3 8.6 1.6 0.24
26.8 1.8 34.9 1.2 8.4 2.0 0.23
<0.001 >0.05 <0.0001 <0.0001
TMF: mean foot temperature; TR: reference (forehead) temperature; DT: TR – TMF; TN = DT/TR normalized temperature. * p < 0.0001, at risk compared with SSR and SSR+. y p < 0.05, SSR compared with SSR+. yy p < 0.01, SSR compared with SSR+.
44
P.-C. Sun et al. / Diabetes Research and Clinical Practice 73 (2006) 41–46
Table 3 showed the numbers of patients with cardiovascular, clinical and nerve conduction abnormalities of the three diabetic groups. There were no significant differences in the percentage of cardiovascular and clinical abnormalities among the three groups. The at-risk group had a significantly higher rate (10 out of 18, p < 0.02) of abnormal nerve conductions than the other diabetic groups. The SSR group seemed have slightly higher percentage of abnormal nerve conductions (5 of 22; 23%) than the SSR+ group (4 of 29; 14%), but their difference was not statistically significant ( p > 0.05). Fig. 2. Temperature distribution in six plantar areas among four groups.
4. Discussions
diabetes than the other two diabetic groups ( p < 0.01). The mean foot temperature was obtained in each group as shown in Table 2. The at-risk group with preulcerative lesion had significantly higher (30.2 1.7 8C) temperature than all other groups ( p < 0.0001). Our results showed that overall temperature was increased unanimously over the six plantar areas in the group with feet at risk (Fig. 2). In terms of the mean foot temperature, the SSR group was higher (27.9 2.0 8C) than the SSR+ group (27.1 2.4 8C), which in turns was higher than the control group (26.8 2.3 8C). However, the differences were not statistically significant among these 3 groups ( p > 0.05). The mean DT as defined by subtracting the foot temperature from the forehead temperature was shown in Table 2. The four groups had pretty close mean forehead temperatures (TR). The mean foot temperature was lower (4.9 1.5 8C) than the mean forehead temperature in the at-risk group. The mean DT in the SSR group was significantly smaller than that in the SSR+ group (7.3 1.8 8C versus 8.6 2.1 8C, p < 0.05) and noted that there was no significant difference when looking at the mean foot temperature alone as mentioned above. The normalized temperature TN by the forehead temperature also showed significant difference between the SSR group and the SSR+ group (0.19 versus 0.24, p < 0.001).
The measurement of plantar skin temperature using infrared can discern small thermal difference beyond the palpable limit, particularly for the patients without gross inflammatory symptoms as in our experiment. Sympathetic dysfunction in the lower limbs leads to reduced sweating and dry skin, which is prone to crack and fissure. These dermatological conditions are highly related to thermoregulatory sweating abnormalities [1,15]. The diabetic patients in this study were categorized based on the degree of peripheral sudomotor dysfunctions. The concordance of results of the patients’ grouping and the skin temperature difference implies that small fiber neuropathy plays a preliminary role in the development of microcirculatory disturbance. The at-risk patients all showed absence of sympathetic skin response in both feet. The decreased sudomotor activity and concomitant thermoregulatory disturbance should be an early sign of sympathetic damage in diabetic feet [16]. Note that the peripheral small fiber neuropathy could not be accurately evaluated by the clinical and autonomic tests mentioned previously. The temperature and electrophysiological assessment conducted in this study proved to be useful for healthcare providers to identify the patients with sympathetic dysfunction in early stage to warrant their feet care. The sympathetic skin response is mediated by the postganglionic, unmyelinated small fibers via the
Table 3 The number and percentage of patients with abnormal studies
Cardiovascular reflex tests Postural blood pressure change Clinical score Nerve conduction *
At risk (N = 18)
SSR (N = 22)
SSR+ (N = 29)
3 1 4 10
5 0 4 5
4 1 3 4
p < 0.02, at risk compared with SSR and SSR+.
(16%) (5%) (22%) (55%)*
(23%) (0%) (18%) (23%)
(14%) (3%) (10%) (14%)
P.-C. Sun et al. / Diabetes Research and Clinical Practice 73 (2006) 41–46
sudomotor pathway, which is anatomically distinct from the vasomotor pathway [5]. Skin temperature is regulated by the sympathetic vasomotor and sudomotor functions. The neuropathic foot has elevated skin temperature primarily through increased arteriovenous shunt flow [17,18]. The absence of SSR indicated only sudomotor dysfunction to some degrees. However, The minimal temperature difference between the SSR and SSR+ group can be discerned by the normalized temperature, but not by the absolute plantar temperature. These findings show that the sympathetic skin response plays a certain role in the prediction of diabetic patients with thermoregulatory dysfunction. The SSR patients require further investigation to confirm if they have definite vasomotor problems. The peripheral neuropathy affects both somatic and sympathetic components in diabetic patients. Clinical and nerve conduction studies investigate only the somatic function of large myelinated fibers [4]. The atrisk group with longer diabetic duration is supposed to have more somatic nerves involved than the other diabetic groups. However, the clinical and nerve conduction examinations are not useful in discerning the two groups with and without SSR. The presence of SSR depends on intact innervations through the small unmyelinated fibers, which are frequently affected in diabetics [6]. Damage to the ummyelinated small fibers should be characteristic of neuropathic feet during its early stage [19]. The presence of peripheral small fiber neuropathy is more associated with temperature abnormality in the diabetic feet with subclinical symptoms than large fiber neuropathy. The SSR test presents itself as an adjunct method distinct from the traditional nerve conduction and autonomic studies in evaluating the peripheral nervous system. This study shows that the application of clinical examinations or nerve conduction study alone is not adequate in screening diabetic at-risk feet at early stage. Most previous studies emphasized the importance of identifying foot ulcerations in order to predict the risk of amputation [20,21], but they did not investigate on the roles of the preulcerative skin lesions in a cascade of worsening sequelae gradually leading to the ultimate disability of the lower extremity. The break of cutaneous barrier makes the skin more susceptible to undue mechanical injury [15]. The feet with elevated skin temperature should be indicative of early diabetic neuropathy as well reported in the literatures [1,18]. The potential arteriovenous shunting and compromised nutritive capillary flow could contribute to foot ulceration and amputation [17]. The minor skin lesions may be a pivotal precursor to subsequent foot problems,
45
though their causal relationship has yet to be clearly established. Follow-up of this patient population is needed to determine whether the preulcerative lesion is a strong indicator of the development of severe diabetic foot complications later. In conclusion, the thermoregulatory disturbance and sweating dysfunction should be a sign of early sympathetic damage in diabetic feet. Assessing skin conditions and sudomotor activities shall help healthcare providers to identify the patients with small fiber neuropathy in early stage more accurately, in order to alert them of their feet conditions. References [1] P.J. Watkins, The diabetic foot, BMJ 326 (2003) 977–979. [2] P.J. Watkins, P.K. Thomas, Diabetes Mellitus and the nervous system, J. Neurol. Neurosurg. Psychiatry 65 (1998) 620–663. [3] J.E. Gilmore, J.A. Ellen, J.R. Hayes, Autonomic function in diabetic neuropathic patients with foot ulceration, Diabetes Care 16 (1993) 61–67. [4] J. Kimura, Principles and variations of nerve conduction studies, in: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice, Oxford University Press, New York, 2001, pp. 91– 129. [5] B.T. Shahani, J.J. Halperin, P. Boulu, J. Cohen, Sympathetic skin response—a method of assessing unmyelinated axon dysfunction in peripheral neuropathies, J. Neurol. Neurosurg. Psychiatry 47 (1984) 536–542. [6] B. Soliven, R. Maselli, J. Jaspan, Sympatheic skin response in diabetic neuropathy, Muscle Nerve 10 (1987) 711–716. [7] M. McGill, L. Molyneaux, T. Bolton, K. Ioannou, R. Uren, D.K. Yue, Response of Charcot’s arthropathy to contact casting: assessment by quantitative techniques, Diabetologia 43 (2000) 481–484. [8] E.J. Boyko, J.H. Ahroni, V.L. Stensel, Skin temperature in the neuropathic diabetic foot, J. Diabetes Complications 15 (2001) 260–264. [9] D.K. Litzelman, D.J. Marriott, F. Vinicor, Independent physiological predictors of foot lesions in patients with NIDDM, Diabetes Care 20 (1997) 1273–1278. [10] P.C. Sun, S.H. Jao, C.K. Cheng, Assessing foot temperature using infrared thermography, Foot Ankle Int. 26 (2005) 847–853. [11] T. Zgur, D.B. Vodusek, M. Krzan, M. Vrtovec, M. Denislic, B. Sibanc, Autonomic system dysfunction in moderate diabetic polyneuropathy assessed by sympathetic skin response and Valsalva index, Electromyogr. Clin. Neurophysiol. 33 (1993) 433–439. [12] C. Ionescu-Tirgoviste, S. Pruna, I. Mincu, Peripheral sympathetic neuropathy evaluated by recording the evoked electrodermal response using an impedance reactometer, Diabetes Res. Clin. Pract. 9 (1990) 201–209. [13] H.S. Chen, C.M. Hwu, B.I. Kuo, Abnormal cardiovascular reflex tests are predictors of mortality in type 2 diabetes mellitus, Diabet. Med. 18 (2001) 268–273. [14] E.L. Feldman, M.J. Stevens, P.K. Thomas, M.B. Brown, N. Canal, D.A. Greene, A practical two steps quantitative clinical electrophysiological assessment for the diagnosis and staging diabetic neuropathy, Diabetes Care 11 (1994) 1281–1289.
46
P.-C. Sun et al. / Diabetes Research and Clinical Practice 73 (2006) 41–46
[15] K. Springett, R.J. White, Skin changes in the at risk foot and their treatment, Br. J. Community Nurs. 12 (2002) 25–32. [16] R.D. Hoeldtke, K.D. Bryner, G.G. Horvath, R.W. Phares, L.F. Broy, G.R. Hobbs, Redistribution of sudomotor response is an early sign of sympathetic dysfunction in type 1 diabetes, Diabetes 50 (2001) 436–443. [17] L. Uccioli, L. Mancini, A. Giordano, A. Solini, P. Magnani, A. Manto, et al., Lower limb arterio-venous shunts, autonomic neuropathy and diabetic foot, Diabetes Res. Clin. Pract. 16 (1992) 123–130.
[18] M.D. Flynn, J.E. Tooke, Diabetic neuropathy and the microcirculation, Diabet. Med. 12 (1995) 298–301. [19] C.J. Tack, P.J. van GurP, C. Holmes, D.S. Goldstein, Local sympathetic denervation in painful diabetic neuropathy, Diabetes 51 (2002) 3545–3553. [20] E.J. Peters, L.A. Lavery, Effectiveness of the diabetic foot risk classification system of the international working group on the diabetic foot, Diabetes Care 24 (2001) 1442–1447. [21] B.E. Sumpio, Foot ulcers, N. Engl. J. Med. 343 (2000) 787– 793.