Effectiveness of Semmes–Weinstein monofilament examination for diabetic peripheral neuropathy screening

Journal of Diabetes and Its Complications 19 (2005) 47 – 53

Effectiveness of Semmes–Weinstein monofilament examination for diabetic peripheral neuropathy screening Nozomu Kamei, Kiminori Yamane *, Shuhei Nakanishi, Yasuyo Yamashita, Tomoko Tamura, Kayo Ohshita, Hiroshi Watanabe, Rumi Fujikawa, Masamichi Okubo, Nobuoki Kohno Department of Molecular and Internal Medicine, Division of Clinical Medical Science, Programs for Applied Biomedicine, Graduate School of Biomedical Sciences, Hiroshima University, Hiroshima, Japan Received 28 July 2003; received in revised form 12 December 2003; accepted 18 December 2003

Abstract Objective: To evaluate the effectiveness of the Semmes – Weinstein monofilament examination (SWME) for diabetic peripheral neuropathy (DPN) screening at an outpatient clinic. Methods: Eighty-two type 2 diabetic outpatients were questioned about 13 subjective symptoms and tested for deep reflexes. They were examined using two types of quantitative sensory testing: SWME (using 4.31/2 g and 5.07/10 g monofilaments) and vibration testing (using the C-64 quantitative tuning fork). The patients were diagnosed with DPN when two of the following three criteria from both legs were met: (1) the presence of either pain, numbness, or paraesthesia; (2) a diminished or absent ankle reflex; or (3) a scale below 4 for their vibration perception thresholds (VPT). Results: The most common subjective symptoms were muscle cramps and numbness in the extremities. SWME 4.31/2 g at the great toe or the plantar aspect of the fifth metatarsal was the most useful diagnostic test for DPN, providing 60.0% sensitivity and 73.8% specificity. SWME 4.31/2 g was correlated with lower limb VPT ( P = .029) and three subjective symptoms including fainting. SWME 5.07/10 g was correlated with lower limb VPT ( P = .011), the ankle reflex ( P = .013), the knee reflex ( P = .031), and two subjective symptoms. However, the sensitivity to diagnose DPN was low (30.0%). Conclusions: The abnormalities indicated by SWME correlated well with those identified by lower limb VPT. Moreover, SWME 4.31/2 g correlated with fainting, suggesting that it could detect abnormalities in the small fibers. Considering the minimal time for this examination, SWME 4.31/2 g is a useful examination in an outpatient setting along with lower limb VPT and ankle reflex testing. D 2005 Elsevier Inc. All rights reserved. Keywords: Semmes – Weinstein monofilament; Quantitative sensory testing; Vibration testing; Diabetic neuropathies; Outpatient clinic

1. Introduction Diabetic peripheral neuropathy (DPN) is not always clinically evident, and in many cases it is subclinical (American Diabetes Association, 1996). Therefore, if the physician only pays attention to subjective symptoms during outpatient visits, the detection of DPN will inevitably be delayed, thus affecting the treatment and prognosis. With its high rate of successful DPN detection, nerve conduction testing is certainly an effective approach for identifying DPN (Dyck, Davies, Litchy, & O’Brien, 1997). However, it is time-consuming and expensive, making this method

* Corresponding author. 1-2-3 Kasumi, Minami-ku, Hiroshima 7348551, Japan. Tel.: +81-82-257-5196; fax: +81-82-255-7360. E-mail address: [email protected] (K. Yamane). 1056-8727/05/$ – see front matter D 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.jdiacomp.2003.12.006

inappropriate for DPN screening in an outpatient setting (Perkins & Bril, 2003). Therefore, many alternative quantitative sensory testing (QST) protocols that can be performed in a limited time have been proposed, and their effectiveness has been reported (Dyck, Dyck, Larson, O’Brien, & Velosa, 2000; Kahn, 1992; Lunetta, Le Moli, Grasso, & Sangiorgio, 1998; Valk, Grootenhuis, van Eijk, Bouter, & Bertelsmann, 2000). The Semmes – Weinstein monofilament examination (SWME) has been used for detecting the loss of protective sensations in the diabetic foot (Abbott et al., 2002; Kumar et al., 1991; McGill, Molyneaux, Spencer, Heng, & Yue, 1999; Olmos et al., 1995; Sosenko et al., 1999). Recently, in an attempt to apply this method for DPN screening, it has been tested in combination with other QST protocols (Cheng et al., 1999; Dimitrakoudis & Bril, 2002; Jirkovska, Boucek, Woskova, Bartos, & Skibova, 2001; Olaleye,

48

N. Kamei et al. / Journal of Diabetes and Its Complications 19 (2005) 47 – 53

Perkins, & Bril, 2001; Perkins, Olaleye, Zinman, & Bril, 2001; Rahman, Griffin, Rathmann, & Wareham, 2003; Vinik et al., 1995), or using filaments of thinner calibers (Nagai, Sugiyama, Abe, & Nomura, 2001). However, the applications of SWME for early DPN detection have been fully discussed. In this study, we studied whether SWME could be useful for outpatient DPN screening. With reference to the established diagnostic criteria of DPN defined by subjective symptoms, ankle reflexes, and vibration perception thresholds (VPT) of the lower limbs (Japanese Study Group on Diabetic Neuropathy, 2001), we examined the sensitivity and specificity of two different monofilaments (4.31/2 g and 5.07/10 g). We then evaluated the results of SWME in terms of their correlation with the subjective symptoms, vibratory sensation, or deep reflexes.

Fig. 1. SWME was conducted at three sites on each foot (Site 1, the great toe; Site 2, the plantar aspect of the first metatarsal; and Site 3, the plantar aspect of the fifth metatarsal) using either 4.31/2 g or 5.07/10 g monofilament, following the practical guidelines on the management and the prevention of the diabetic foot (The International Working Group on the Diabetic Foot, 1999).

2. Subjects and methods 2.1. Patients Among 95 type 2 diabetic patients who visited the Hiroshima University Hospital, we excluded those cases with arteriosclerotic obliterans, neuropathy under treatment, or neuropathy of other etiologies without diabetes. We then selected 82 patients (44 men and 38 women) who provided their informed consent to participate in this study (Table 1). The average age was 61.6F11.0 years, and the study group contained 35 patients over 65 years old. 2.2. Questionnaire on neuropathic symptoms and QST We surveyed the patients with a questionnaire concerning their subjective symptoms. The following 13 questions were prepared: (1) pain in hands; (2) pain in feet; (3) numbness in extremities; (4) paraesthesia of both feet; (5) cold sensation in the extremities; (6) burning sensation in the extremities; (7) constipation; (8) diarrhea; (9) muscle cramps; (10)

Table 1 Characteristics of study subjects n Sex (men/women) Age (years) Duration of diabetes (years) BMI (kg/m2) HbA1c (%) Smoking (yes/no) Alcohol drinking (yes/no) Retinopathy (NDR/NPDR/PPDR/PDR) Proteinuria ( /F/+) Therapy (diet/oral agents/insulin)

82 44/38 61.6F11.0 14.1F10.4 22.7F3.2 7.9F1.5 35/47 34/48 47/21/2/12 54/15/13 19/36/27

(27 – 87) (1 – 41) (17.1 – 31.6) (5.2 – 13.0)

Data are n or meansFS.D. (ranges). NDR, nondiabetic retinopathy; NPDR, nonproliferative diabetic retinopathy; PPDR, preproliferative diabetic retinopathy; PDR, proliferative diabetic retinopathy.

fainting; (11) dizziness; (12) urination difficulties; and (13) foot ulceration or nail discoloration. The symptoms were assessed by the following four categories: none, mild, moderate, or severe. For each symptom, we grouped the patients into two groups: none or positive. With respect to the clinical functional examinations, we conducted SWME on both lower limbs, VPT testing for both upper and lower limbs, and deep reflex testing (ankle reflex and knee reflex) for both lower limbs. All tests were preformed by one of two investigators (S. Nakanishi and Y. Yamashita). Following the practical guidelines on the management and prevention of the diabetic foot (The International Working Group on the Diabetic Foot, 1999), SWME was conducted at three noncallused sites on each foot (Site 1, the great toe; Site 2, the plantar aspect of the first metatarsal head; and Site 3, the plantar aspect of the fifth metatarsal head) using either 4.31/2 g or 5.07/10 g monofilament (Fig. 1). The SWME procedure is as follows (The International Working Group on the Diabetic Foot, 1999): (1) Place a filament perpendicular to the skin surface avoiding callosity and push the filament with sufficient force until it bends or twists. (2) Ask the patient if they feel anything touching the skin and if it is their left or right foot. (3) Repeat the examination three times for each site, and include at least one fake examination, in which the filament is not actually placed on the skin. (4) If the patient gives incorrect answers two or more times out of the three examinations per site, the site is considered as positive. On the other hand, if an incorrect answer occurs once or less, the site is considered as negative. (5) Conduct the examination at all six sites, with the order of examinations decided randomly each time. VPT was assessed on both sides of the upper and lower limb lateral condyles using a C-64 quantitative tuning fork (Takano Manufacturing, Nagoya, Japan) (Hotta et al., 1994). The tuning fork is based on an eight-point scale providing quantitative and objective data, and has been

N. Kamei et al. / Journal of Diabetes and Its Complications 19 (2005) 47 – 53

described in detail elsewhere (Hotta et al., 1994). A VPT score below 4 was judged to be abnormal (Martina, Koningsveld, Schmitz, Meche, & Doorn, 1998). The deep reflexes were assessed by striking the Achilles tendon and patellar tendon of both lower limbs using a percussion hammer, and were then classified into the following four categories: normal, diminished, absent, or exaggerated. The ankle reflex was tested in a position in which the patient knelt facing a wall. The knee reflex was conducted with the patients sitting on a chair with their feet hanging down. 2.3. Diagnostic criteria for DPN If two or more of the following three criteria were satisfied, the patient was diagnosed with DPN (The Japanese Study Group on Diabetic Neuropathy, 2001): (1) either pain, numbness, or paraesthesia in both feet; (2) ankle reflex in both legs is either diminished or absent; or (3) VPT score was below 4 for both lower limbs. The diagnostic criteria was modified from those proposed by the Japanese diabetic neuropathy study group in that a C-64 quantitative tuning fork was used instead of a 128-Hz tuning fork. 2.4. Statistical analysis The data were expressed as meansFstandard deviations or percentiles. Student’s t test was used for the continuous variables, and chi-square statistics were used for the categorical variables. Linear regression analysis was used between the age and the scale of VPT. A P value less than .05 was considered statistically significant. The data analysis was performed using SAS software Ver. 8.2 (SAS Institute, Cary, NC).

3. Results Based on the diagnostic criteria described in the methods, 48.8% of the patients were suffering from DPN. An analysis of each criterion revealed that (1) 51.2% of the patients

Table 2 Sensitivity and specificity of SWME 4.31/2 g or SWME 5.07/10 g at individual site in detecting DPN Site 1

Site 2

Site 3

Evaluator size

4.31

5.07

4.31

5.07

4.31

5.07

Right foot Sensitivity Specificity

47.5 81.0

22.5 95.2

32.5 85.7

5.0 92.9

40.0 76.2

15.0 97.6

Left foot Sensitivity Specificity

42.5 71.4

17.5 88.1

40.0 76.2

7.5 92.9

42.5 73.8

12.5 92.9

Data are %, n = 82.

49

Table 3 Sensitivity and specificity of SWME 4.31/2 g or SWME 5.07/10 g at a combination of sites in detecting DPN (right foot) Sites 1 and 2 Sites 1 and 3 Sites 2 and 3 Sites 1, 2, and 3 Evaluator size 4.31 5.07 4.31 5.07 4.31 5.07 4.31 5.07 Sensitivity 47.5 Specificity 76.2

22.5 92.9

60.0 73.8

30.0 92.9

50.0 73.8

15.0 92.9

60.0 71.4

30.0 92.9

Data are %, n = 82.

suffered either pain, numbness, or paraesthesia in both feet; (2) 56.1% experienced their ankle reflex being diminished or absent in both legs; and (3) 36.6% had scales below 4 for VPT on both lower limbs. Among the clinical symptoms, muscle cramps (40%), numbness in the extremities (39%), urination difficulties (33%), cold sensation in the extremities, constipation (32% each), and fainting (31%) were found at high frequency. Numbness in the extremities, muscle cramps (12% each), and pain in the feet (11%) were scored by a high proportion of patients as moderate or severe. In all cases, the sites that were positive using a 5.07/10 g monofilament were also found to be positive using a 4.31/ 2 g monofilament. Using the diagnostic criteria for DPN described above as a ‘‘gold standard,’’ the sensitivity of SWME 4.31/2 g was 32.5 –47.5% at Sites 1 – 3, whereas the sensitivity of SWME 5.07/10 g was 5.0 – 22.5% at the same sites (Table 2). In terms of combining more than one site, the combination of Sites 1 and 3 was the most sensitive at 60.0% for a 4.31/2 g monofilament and 30.0% for a 5.07/10 g monofilament. The specificity was also favorable for this combination at 73.8% and 92.9%, respectively, for 4.31/2 g and 5.07/10 g monofilaments (Table 3). We did not find significant differences between the left and right feet for Sites 1 and 3, but there was a greater difference for Site 2 (data not shown). Therefore, in the following examinations, the patients who had positive results for Site 1 or 3 of the right foot were considered to be abnormal. The rate of abnormality was 43% by SWME 4.31/2 g and 18% by SWME 5.07/10 g (Table 4). In both groups, we did not find any relationship with age, the duration of diabetes, or the treatment method. However, we found a correlation between alcohol intake and SWME 5.07/10 g (odds ratio [OR] = 3.58 [95% confidence interval 1.10– 11.71], P = .029) (Table 5). In terms of the relationship with subjective symptoms, SWME 4.3/2 g was correlated with paraesthesia of the feet (OR = 8.67 [2.23 – 33.62], P < .001), numbness in the extremities (OR = 2.50 [1.004 – 6.21], P = .048), and fainting (OR = 2.78 [1.06 – 7.30], P = .037). SWME 5.07/10 g was correlated with paraesthesia of the feet (OR = 5.64 [1.64 – 19.36], P = .004) and urination difficulties (OR = 4.08 [1.27 – 13.10], P = .014) (data not shown). Vibration testing resulted in an average value of 5.52F 1.40 for the right upper limb and 3.67F1.58 for the right

50

N. Kamei et al. / Journal of Diabetes and Its Complications 19 (2005) 47 – 53

Table 4 Characteristics of study subjects according to classifications of SWME and deep reflexes

n (%) Age (years) Duration of diabetes (years) BMI (kg/m2) HBA1c (%) VPT of upper limbs VPT of lower limbs

SWME 4.31/2 g

SWME 5.07/10 g

Ankle reflex

Normal

Abnormal

Normal

Abnormal

Normal

Abnormal

Normal

Knee reflex Abnormal

47 (57%) 60.2F11.4 13.3F10.9

35 (43%) 63.5F10.5 15.2F10.1

67 (82%) 61.5F11.8 14.0F11.1

15 (18%) 62.1F7.4 14.5F7.3

34 (42%) 61.9F11.5 10.9F10.0

47 (58%) 61.5F10.9 16.2F10.5*

46 (58%) 60.2F10.9 11.8F10.4

33 (42%) 63.3F11.4 17.4F10.1*

22.7F3.2 7.77F1.34 5.71F1.36 4.00F1.60

22.8F3.3 8.04F1.63 5.27F1.45 3.23F1.48*

22.9F3.4 7.78F1.27 5.52F1.38 3.88F1.62

21.9F2.5 8.36F2.14 5.53F1.60 2.73F1.03*

22.7F3.2 7.68F1.39 5.96F1.42 4.12F1.81

22.8F3.2 8.06F1.52 5.22F1.35* 3.34F1.36*

22.7F3.4 7.82F1.45 5.65F1.42 3.89F1.69

22.8F3.0 8.01F1.56 5.42F1.39 3.45F1.44

Data are n (%) or meansFS.D. Diagnosis of normal or abnormal is shown in the text. * P<.05 (t test between normal and abnormal).

lower limb, indicating a significant reduction of the threshold in the lower limb ( P < .001). Since we did not find any statistically significant differences between the right and left limbs (data not shown), we used the results from the right side for the following examinations. With simple linear regression analysis between the patient’s age and VPT, there were negative correlations for both the upper (b = .028/year, r = .223, P = .044) and lower limbs (b = .052/year, r = .360, P < .001). Deep reflex testing did not reveal any significant differences between the right and left legs (data not shown). We therefore divided the results from the right deep reflex testing into two groups: normal and abnormal (i.e., diminished or absent). Patients with exaggerated

Table 5 Odds ratio between SWME and categorical variables by chi-square statistics SWME 4.31/2 g

Sex (men/women) Smoking (+/ ) Alcohol intake (+/ ) Hypertension (+/ ) Hyperlipidemia (+/ ) Retinopathy (NPDR or PPDR or PDR/NDR) Proteinuria (F or +/ ) Diabetic therapy (oral agents or insulin/dietary) Ankle reflexes (abnormal/normal) Knee reflexes (abnormal/normal) OR, odds ratio.

4. Discussion

SWME 5.07/10 g

OR [95% CI]

P

OR [95% CI]

P

1.05 [0.43 – 2.51] 0.83 [0.34 – 2.01] 0.90 [0.37 – 2.19] 1.36 [0.56 – 3.30] 0.96 [0.40 – 2.30] 1.52 [0.63 – 3.69]

N.S.

2.83 [0.82 – 9.80] 2.37 [0.75 – 7.42] 3.58 [1.10 – 11.71] 1.30 [0.42 – 4.00] 1.41 [0.46 – 4.33] 1.22 [0.40 – 3.75]

N.S.

1.57 [0.63 – 3.95] 1.03 [0.37 – 2.91]

N.S.

2.69 [0.86 – 8.41] 1.25 [0.31 – 5.01]

N.S.

1.76 [0.71 – 4.35] 2.25 [0.90 – 5.62]

N.S.

6.12 [1.28 – 29.30] 3.57 [1.09 – 11.70]

.013

N.S. N.S. N.S. N.S. N.S.

N.S.

N.S.

reflexes (1 for ankle reflex and 2 for knee reflex) were excluded from this analysis. The abnormal group comprised 58% and 42% for the ankle and knee reflexes, respectively. In both cases, the duration of the diabetes was significantly longer than in the abnormal groups ( P = .024, P = .021, respectively) (Table 4). We then analyzed the relationship between SWME and various examinations. The abnormal groups for SWME 4.31/2 g and SWME 5.07/10 g had significantly lower values on their lower limb VPT than the normal groups ( P= .029, P = .011, respectively) (Table 4). We also found that SWME 5.07/10 g was significantly correlated with the ankle (OR=6.12 [1.28 – 29.30], P = .013) or knee reflex (OR = 3.57 [1.09 – 11.70], P = .031) (Table 5). In addition, the abnormal group for the ankle reflex had a lower value for the upper and lower limbs VPT than the normal group ( P = .021, P = .30, respectively) (Table 4).

N.S. .029 N.S. N.S. N.S.

N.S.

.031

The prevalence of neuropathic complications in diabetic patients varies depending on the reports (Young, Boulton, Macleod, Williams, & Sonksen, 1993; Ziegler, Gries, Spuller, Lessman, & the DiaCAN Multicenter Study Group, 1992), but in recent years it has been concluded to be about 30% (Tesfaye et al., 1996). Diabetic neuropathy affects the prognosis and quality of life as illustrated by its association with sudden death, orthostatic hypotension, and lower limb amputation. On the other hand, recovery from this complication can be achieved as long as it is detected early, making it essential to detect and treat this complication during outpatient visits. According to the recommendations of the American Diabetes Association (1996), the diagnosis of diabetic neuropathy requires the following: (1) clinical symptoms; (2) clinical examination; (3) electrodiagnostic studies (EDX); (4) QST; and (5) autonomic function testing. However, it is difficult to conduct EDX such as nerve conduction studies or electromyography as a screening examination in an outpatient setting. Therefore, the diagnosis has to depend on

N. Kamei et al. / Journal of Diabetes and Its Complications 19 (2005) 47 – 53

a combination of the remaining four evaluations. Among the QST, SWME is gaining increased attention after several effective trials for DPN screening by combining it with other QST (Cheng et al., 1999; Dimitrakoudis & Bril, 2002; Jirkovska et al., 2001; Olaleye et al., 2001; Perkins et al., 2001; Rahman et al., 2003; Vinik et al., 1995) or by modifying the target force of the filaments (Nagai et al., 2001). In this study, we conducted SWME using two kinds of monofilaments: 4.31/2 g and 5.07/10 g. SWME using a 5.07/10 g monofilament was correlated with lower limb VPT, the ankle reflex, the knee reflex, and two subjective symptoms. It is therefore an important examination, but the sensitivity for DPN diagnosis is low (30.0%), whereas the specificity is high (92.9%). This low sensitivity suggests that the abnormality will be recognized only in advanced DPN. SWME 5.07/10 g has been suggested to be effective for the detection of the loss of protective sensations in the diabetic foot. In DPN screening, however, the possibility of this method detecting a lesion that cannot be identified by deep reflex testing or VPT of the lower limbs is low, making it less attractive as a screening examination in addition to the preexisting methods. On the other hand, SWME using a 4.31/2 g monofilament showed a specificity of 73.8% and a sensitivity of 60.0% for DPN diagnosis, which was higher in sensitivity than SWME using a 5.07/10 g monofilament, but less sensitive than what we had expected. SWME 4.31/2 g was correlated with lower limb VPT, likely indicating an early abnormality in the large myelinated fibers. It also correlated with fainting that was not detected by vibration testing, raising the possibility that it could also detect an abnormality in the small nerve fibers. Furthermore, it might be possible to consider SWME 4.31/2 g as a useful test for asymptomatic peripheral neuropathy detection (Kumar et al., 1991; McGill et al., 1999). SWME 4.31/ 2 g may potentially detect cases with DPN that could not be diagnosed by the combination of subjective symptoms, ankle reflex, and lower limb VPT. It is also important to note that the combination of multiple QST can be effective in detecting those patients that might otherwise have been overlooked. Therefore, we think it is useful to conduct SWME 4.31/2 g in addition to the other examination protocols. In terms of the site of SWME testing, the combination of Sites 1 and 3 was the most sensitive without a significant difference between the right and left feet. Therefore, we suggest that this combination is the most useful. McGill et al. (1999) examined five sites on the right foot, including the three sites that we examined in this study. They concluded that the combination of Sites 2 and 3 was the most effective. However, the purpose of our study was to use SWME as an indicator of DPN, whereas they used it to detect insensate feet for prophylaxis against lower limb amputation. Furthermore, their study was different from ours in that they only used a 5.07/10 g monofilament and

51

assumed vibration testing using a biothesiometer as the gold standard. Site 2 is also expected to be more tylotic when compared with other sites. Even if areas with callosity were avoided for this measurement, such an effect cannot be completely excluded. Moreover, the plantar subcutaneous tissue of the first metatarsal head is the thickest of the five toes (Abouaesha, Van Schie, Griffths, Young, & Boulton, 2001). We believe this callosity and subcutaneous tissue thickness contributed to the reason why the measurements varied greatly between the rightand left-hand side, and also why the sensitivity was reduced at this site. Vibration testing can detect an abnormality before other clinical examinations or QST (American Diabetes Association, 1996). Therefore, it is an essential examination to diagnose DPN and, consequently, has been conducted in most of the reports concerning DPN (Cheng et al., 1999; Dimitrakoudis & Bril, 2002; Jirkovska et al., 2001; Nagai et al., 2001; Olaleye et al., 2001; Perkins et al., 2001; Rahman et al., 2003; Vinik et al., 1995). Perkins et al. (2001) used a 128-Hz tuning fork and reported that the on– off method was more useful than the timed method. Although it is the mainstream of vibration testing, quantification was difficult with the timed method using a 128-Hz tuning fork. Vibration testing using a C-64 quantitative tuning fork, which was performed in the current study, has been suggested to correlate with lower limb motor nerve conduction velocity, and represents an easy and objective method (Hotta et al., 1994; Martina et al., 1998). Deep reflexes were associated with the duration of diabetes, retinopathy, and diabetic therapy (data not shown). It is apparent that the deep reflexes worsened gradually as the duration of diabetes increased, and that it is an indicator of relatively advanced DPN. Since DPN results in disorders of neural fibers with various diameters such as large myelinated fibers, small myelinated fibers, and unmyelinated fibers, patients experience various subjective symptoms. In the current study, the prevalence of having paraesthesia of the feet was 16%, which is a relatively low percentage among the 13 questions. Nevertheless, it showed strong associations with SWME 4.31/2 g, SWME 5.07/10 g, lower limb VPT, and ankle reflex, suggesting a specific symptom to diagnose large fiber neuropathy. Hand pain, foot pain, and a burning sensation in the extremities were not correlated with the QST conducted in this study. In general, allodynia and defective warm thermal sensation are specific symptoms of small fiber neuropathy (Valk et al., 2000; Vinik, Park, Stansberry, & Pittenger, 2000). It is thought that the cold thermal threshold reflects Ay fibers, whereas the warm thermal threshold reflects C fibers (Hanson, Schumacker, Debugne, & Clerin, 1992). Small fiber neuropathy is known to result in a reduction of light touch sensation, and 4.31/2 g SWME testing could have detected this reduction. However, a correlation with the relevant symptoms in our results was not detected, suggesting that it is difficult to screen for

52

N. Kamei et al. / Journal of Diabetes and Its Complications 19 (2005) 47 – 53

allodynia or defective warm thermal sensation even if a filament of small caliber is used. Nevertheless, SWME 4.31/2 g was correlated with fainting, which may be one of the symptoms associated with the autonomic nervous system, suggesting the possibility that it could detect an abnormality in the small fibers. In this study, we did not include examinations of the autonomic nerves (B fibers) such as cardiac autonomic tests (R-R interval variation, etc.) (Ewing, Borsey, Bellavere, & Clarke, 1981), but we would include such tests in an actual outpatient examination. In view of the short time required and low cost, SWME using a 4.31/2 g monofilament is a useful examination for DPN screening in an outpatient clinic, along with lower limb VPT and ankle reflex testing. Moreover, it is recommended that the thermal threshold test and an autonomic test should be included in these examinations if possible.

Acknowledgments The authors thank Yasuyo Mimori, MD, PhD, for his invaluable comments. We also thank Minako Hirabayashi for her secretarial assistance.

References Abbott, C. A., Carrington, A. L., Ashe, H., Bath, S., Every, L. C., Griffiths, J., Hann, A. W., Hussein, A., Jackson, N., Johnson, K. E., Ryder, C. H., Torkington, R., Van Ross, E. R., Whalley, A. M., Widdows, P., Williamson, S., & Boulton, A. J. (2002). The NorthWest Diabetes Foot Care Study: incidence of, and risk factors for, new diabetic foot ulceration in a community-based patient cohort. Diabetic Medicine, 19, 377 – 384. Abouaesha, F., Van Schie, C. H., Griffiths, G. D., Young, R. J., & Boulton, A. J. (2001). Plantar tissue thickness is related to peak plantar pressure in the high-risk diabetic foot. Diabetes Care, 24, 1270 – 1274. American Diabetes Association (1996). Clinical practice recommendations 1996. Diabetic neuropathy. Diabetes Care, 19 (Suppl. 1), S67 – S71. Cheng, W. Y., Jiang, Y. D., Chuang, L. M., Huang, C. N., Heng, L. T., Wu, H. P., Tai, T. Y., & Lin, B. J. (1999). Quantitative sensory testing and risk factors of diabetic sensory neuropathy. Journal of Neurology, 246, 394 – 398. Dimitrakoudis, D., & Bril, V. (2002). Comparison of sensory testing on different toe surfaces: implications for neuropathy screening. Neurology, 59, 611 – 613. Dyck, P. J., Davies, J. L., Litchy, W. J., & O’Brien, P. C. (1997). Longitudinal assessment of diabetic polyneuropathy using a composite score in the Rochester Diabetic Neuropathy Study cohort. Neurology, 49, 229 – 239. Dyck, P. J., Dyck, P. J., Larson, T. S., O’Brien, P. C., & Velosa, J. A. (2000). Patterns of quantitative sensation testing of hypoesthesia and hyperalgesia are predictive of diabetic polyneuropathy: a study of three cohorts. Nerve growth factor group. Diabetes Care, 23, 510 – 517. Ewing, D. J., Borsey, D. Q., Bellavere, F., & Clarke, B. F. (1981). Cardiac autonomic neuropathy in diabetes: comparison of measures of R-R interval variation. Diabetologia, 21, 18 – 24.

Hanson, P. H., Schumacker, P., Debugne, T. H., & Clerin, M. (1992). Evaluation of somatic and autonomic small fibers neuropathy in diabetes. American Journal of Physical Medicine and Rehabilitation, 71, 44 – 47. Hotta, N., Sugiura, K., Tsuchida, I., Sano, T., Koh, N., Matsumae, H., & Sakamoto, N. (1994). Use of the C64 quantitative tuning fork and the effect of niceritol in diabetic neuropathy. Clinical Therapeutics, 16, 1007 – 1015. Japanese Study Group on Diabetic Neuropathy. (2001). Simplified diagnostic criteria for diabetic polyneuropathy. Peripheral Nerve, 12, 225 – 227. Jirkovska, A., Boucek, P., Woskova, V., Bartos, V., & Skibova, J. (2001). Identification of patients at risk for diabetic foot: a comparison of standardized noninvasive testing with routine practice at community diabetes clinics. Journal of Diabetes and Its Complications, 15, 63 – 68. Kahn, R. (1992). Proceedings of a consensus development conference on standardized measures in diabetic neuropathy. Quantitative sensory testing. Diabetes Care, 15, 1092 – 1094. Kumar, S., Fernando, D. J., Veves, A., Knowles, E. A., Young, M. J., & Boulton, A. J. (1991). Semmes – Weinstein monofilaments: a simple, effective and inexpensive screening device for identifying diabetic patients at risk of foot ulceration. Diabetes Research and Clinical Practice, 13, 63 – 67. Lunetta, M., Le Moli, R., Grasso, G., & Sangiorgio, L. (1998). A simplified diagnostic test for ambulatory screening of peripheral diabetic neuropathy. Diabetes Research and Clinical Practice, 39, 165 – 172. Martina, I. S., van Koningsveld, R., Schmitz, P. I., van der Meche, F. G., & van Doorn, P. A. (1998). Measuring vibration threshold with a graduated tuning fork in normal aging and in patients with polyneuropathy. European Inflammatory Neuropathy Cause and Treatment (INCAT) group. Journal of Neurology, Neurosurgery, Psychiatry, 65, 743 – 747. McGill, M., Molyneaux, L., Spencer, R., Heng, L. F., & Yue, D. K. (1999). Possible sources of discrepancies in the use of the Semmes – Weinstein monofilament. Impact on prevalence of insensate foot and workload requirements. Diabetes Care, 22, 598 – 602. Nagai, Y., Sugiyama, Y., Abe, T., & Nomura, G. (2001). 4-g monofilament is clinically useful for detecting diabetic peripheral neuropathy. Diabetes Care, 24, 183 – 184. Olaleye, D., Perkins, B. A., & Bril, V. (2001). Evaluation of three screening tests and a risk assessment model for diagnosing peripheral neuropathy in the diabetes clinic. Diabetes Research and Clinical Practice, 54, 115 – 128. Olmos, P. R., Cataland, S., O’Dorisio, T. M., Casey, C. A., Smead, W. L., & Simon, S. R. (1995). The Semmes – Weinstein monofilament as a potential predictor of foot ulceration in patients with noninsulindependent diabetes. The American Journal of the Medical Sciences, 309, 76 – 82. Perkins, B. A., & Bril, V. (2003). Diabetic neuropathy: a review emphasizing diagnostic methods. Clinical Neurophysiology, 114, 1167 – 1175. Perkins, B. A., Olaleye, D., Zinman, B., & Bril, V. (2001). Simple screening tests for peripheral neuropathy in the diabetes clinic. Diabetes Care, 24, 250 – 256. Rahman, M., Griffin, S. J., Rathmann, W., & Wareham, N. J. (2003). How should peripheral neuropathy be assessed in people with diabetes in primary care? A population-based comparison of four measures. Diabetic Medicine, 20, 368 – 374. Sosenko, J. M., Sparling, Y. H., Hu, D., Welty, T., Howard, B. V., Lee, E., & Robbins, D. C. (1999). Use of the Semmes – Weinstein monofilament in the strong heart study. Risk factors for clinical neuropathy. Diabetes Care, 22, 1715 – 1721. Tesfaye, S., Stevens, L. K., Stephenson, J. M., Fuller, J. H., Plater, M., Ionescu-Tirgoviste, C., Nuber, A., Pozza, G., & Ward, J. D. (1996). Prevalence of diabetic peripheral neuropathy and its relation to glycaemic control and potential risk factors: the EURODIAB IDDM complications study. Diabetologia, 39, 1377 – 1384. The International Working Group on the Diabetic Foot. (1999). Practical guidelines on the management and the prevention of the diabetic

N. Kamei et al. / Journal of Diabetes and Its Complications 19 (2005) 47 – 53 foot. Based upon the International Consensus on the Diabetic Foot. Amsterdam, the Netherlands. Valk, G. D., Grootenhuis, P. A., van Eijk, J. T. M., Bouter, L. M., & Bertelsmann, F. W. (2000). Methods for assessing diabetic polyneuropathy: validity and reproducibility of the measurement of sensory symptom severity and nerve function tests. Diabetes Research and Clinical Practice, 47, 87 – 95. Vinik, A. I., Park, T. S., Stansberry, K. B., & Pittenger, G. L. (2000). Diabetic neuropathies. Diabetologia, 43, 957 – 973. Vinik, A. I., Suwanwalaikorn, S., Stansberry, K. B., Holland, M. T.,

53

McNitt, P. M., & Colen, L. E. (1995). Quantitative measurement of cutaneous perception in diabetic neuropathy. Muscle and Nerve, 18, 574 – 584. Young, M. J., Boulton, A. J. M., Macleod, A. F., Williams, D. D. R., & Sonksen, P. H. (1993). A multicentre study of the prevalence of diabetic peripheral neuropathy in the United Kingdom hospital clinic population. Diabetologia, 36, 150 – 154. Ziegler, D., Gries, F. A., Spuller, M., Lessmann, F., the DiaCAN Multicenter Study Group. (1992). The epidemiology of diabetic neuropathy. Journal of Diabetes and Its Complications, 6, 49 – 57.