Relationship between ultrasonographic nerve morphology and severity of diabetic sensorimotor polyneuropathy

Transcription

1 ORIGINAL ARTICLE Relationship between ultrasonographic nerve morphology and severity of diabetic sensorimotor polyneuropathy T. Arumugam a, S. N. O. Razali a, S. R. Vethakkan b, F. I. Rozalli c and N. Shahrizaila a a Neurology Unit, Department of Medicine, University of Malaya, Kuala Lumpur; b Endocrinology Unit, Department of Medicine, University of Malaya, Kuala Lumpur; and c Department of Radiology, Faculty of Medicine, University of Malaya, Kuala Lumpur, Malaysia Keywords: cross-sectional areas, diabetes mellitus, diabetic sensorimotor polyneuropathy, nerve conduction studies, Toronto Clinical Scoring System, ultrasound Received 14 June 2015 Accepted 4 August 2015 European Journal of Neurology 2015, 0: 1 7 doi: /ene Introduction Diabetic sensorimotor polyneuropathy (DSP) is a common complication of diabetes mellitus (DM) with a reported prevalence of about 50% [1]. The diagnosis of DSP is made based on the presence of characteristic symptoms and signs and is confirmed by abnormalities seen on nerve conduction studies (NCSs) [2]. However, NCSs have several limitations including poor patient tolerance, and in patients with more Correspondence: N. Shahrizaila, Neurology Laboratory, 6th Floor, South Block, University Malaya Medical Centre, Kuala Lumpur, Malaysia (tel.: ; fax: ; nortina7@gmail.com). Background and purpose: In the current study, the aim was to characterize the nerve ultrasound cross-sectional areas (CSAs) of type 2 diabetic patients with diabetic sensorimotor polyneuropathy (DSP) of different severities. Methods: A hundred symptomatic DSP patients and 40 age-matched healthy controls were prospectively recruited. DSP severity was ascertained through the Toronto Clinical Scoring System (TCCS). Nerve electrophysiology and ultrasound were performed on both lower limbs and the non-dominant upper limb. Results: The sural nerve was inexcitable in 19.1% of mild, 40.0% of moderate and 69.0% of severe DSP groups. In contrast, CSAs were measurable in all nerves of DSP patients and were significantly larger compared to controls. Patients with severe DSP had significantly larger ulnar, peroneal, tibial and sural nerves compared to mild DSP patients. By receiver operating characteristic curve analysis, the cut-off value for the sural nerve at 2 mm 2 was a good discriminator (area under the curve 0.88) between the presence and absence of DSP (sensitivity 0.90; specificity 0.74) but performed less well in discriminating between the severity of DSP (cut-off 2.75 mm 2 ; area under the curve 0.62; sensitivity 0.59; specificity 0.73). Significant correlations were demonstrated between TCSS scores, most neurophysiology parameters and CSAs of the ulnar, peroneal, tibial and sural nerves. Conclusion: Nerve ultrasound in DSP reveals enlarged CSAs and these changes worsen with increasing disease severity, thus serving as a useful diagnostic tool especially when neurophysiology is unrevealing. advanced DSP nerve action potentials are often inexcitable [3]. In recent years, peripheral nerve ultrasound has emerged as an additional tool in the assessment of peripheral nerve disorders demonstrating morphological changes in patients with different forms of neuropathy. Several ultrasound studies have demonstrated enlarged nerves, assessed by cross-sectional areas (CSA), in DSP patients in comparison to controls [4,5]. In one study, the CSA of the posterior tibial nerve with a threshold value of mm 2 was found to distinguish DM patients with and without DSP [6]. In the current study, the nerve CSA through ultrasound and NCSs in DM patients with different severities of DSP, as determined by the Toronto Clini- EUROPEAN JOURNAL OF NEUROLOGY 1

2 2 T. ARUMUGAM ET AL. cal Scoring System (TCSS), were assessed. The aim was to comprehensively determine the relationship between nerve ultrasound and NCS parameters as well as to determine if ultrasound can reliably discriminate between the different grades of severity of DSP. Subjects Materials and methods A hundred type 2 DM patients attending the outpatient DM clinic at University Malaya Medical Centre, Kuala Lumpur, Malaysia, were prospectively recruited. Patients with other potential causes of peripheral neuropathy such as chronic alcohol consumption, toxic exposure, inflammatory diseases, longstanding chronic renal failure or nutritional disorders were excluded. Demographic data such as height and weight were collected at the time of the study. Disease severity was assessed with TCSS (Table S1): mild, score of 6 8; moderate, score of 9 11; severe (12 19). Other relevant information acquired includes HbA1c values and DM disease duration. The study received ethics approval from University Malaya Medical Centre Medical Research Ethical Committee and all patients signed an informed consent form. Nerve ultrasound Ultrasonography was performed using a 12-MHz linear array transducer (E Logic book Ò ; General Electric Healthcare, Waukesha, WI, USA) by a single assessor (SNOR) who was blinded to the severity of DSP. The CSA recordings of each nerve were measured at standardized anatomical sites. Ultrasound was performed in both lower limbs and one non-dominant upper limb. The CSAs of the median and ulnar nerves were assessed at the distal wrist crease, mid-forearm, elbow and mid-arm. In the lower limbs, the peroneal nerve was assessed at the fibular head and popliteal fossa, the tibial nerve at the medial malleolus, and the sural nerve at 10 cm above the lateral malleolus. CSA measurements were traced inside the hyperechoic rim of the nerve using an electronic tracer. The data were compared to 40 healthy controls. Nerve conduction studies Nerve conduction studies were performed by a single assessor (NS) who was blinded to the patient s DSP severity at the time of the study. Standard techniques of supramaximal percutaneous stimulation and surface electrode recording were applied and limb temperature was maintained at 32 C. Sensory studies were done in the non-dominant median, ulnar and radial nerves as well as both sural nerves. Motor studies were performed in the non-dominant median, ulnar and both peroneal and tibial nerves. Reference values were derived from previously established normal ranges at our laboratory. A diagnosis of DSP was made based on existing criteria [7]. A diagnosis of coexisting median nerve entrapment across the wrist and ulnar neuropathy at the elbow was made according to previously described criteria [8 10]. Statistical analysis Statistical analyses were performed using IBM SPSS Statistics for Windows, Version (Armonk, NY, USA). Normality of data was tested using the Kolmogorov Smirnov normality test. Demographic data and CSAs were compared between diabetic patients and control groups using the independent t test for parametric variables and the Mann Whitney test for non-parametric variables. Comparative studies between groups were done with ANOVA for parametric variables or Kruskal Wallis for non-parametric variables. Receiver operating characteristic (ROC) curve analysis was done to determine the nerve CSA cut-off values that best predict the presence of DSP as well as to differentiate between severe and non-severe DSP, based on TCSS values. Correlation studies were done with Spearman s rank correlation. Statistical significance was established at P < Results A total of 100 type 2 DM patients with DSP (mild, 34; moderate, 30; severe, 36) were recruited. The total number of nerves assessed was 199 lower limb nerves (one patient had a left below-knee amputation) and 100 upper limb nerves. Detailed demographic data for DM patients and control subjects are shown in Table 1. There were no significant differences in age, height, weight, gender or ethnicity ratios between controls and DM patients. In the DM cohort, the mean disease duration, HbA1c and TCSS score were 14.5 years, 7.87% and 10.5 respectively. There were no significant differences in the mean disease duration or HbA1c between the different DSP severity groups. The frequencies of coexisting upper limb entrapment neuropathies were 77.0% median nerve entrapment at the wrist and 12.0% ulnar neuropathy across the elbow. Looking specifically at the lower limb nerve conduction parameters, these were unrecordable in 43.2% sural, 13.1% peroneal and 8.0% tibial nerve potentials. There were significant differences (P < 0.001) between absence of sural

3 NERVE ULTRASOUND IN DIABETIC NEUROPATHY 3 Table 1 Patient characteristics Type 2 diabetic patients Healthy controls Table 2 Comparison of nerve cross-sectional areas (CSAs) between patients and healthy controls (excluding entrapment sites) Diabetic patients Healthy controls N Age (years) (8.76) (7.11) Gender (male:female) 41:59 24: Ethnicity (Malay: 33:33:34 14:14: Chinese:Indian) Height (cm) (7.47) (8.99) Weight (kg) (16.15) (12.39) Disease duration (years) (9.43) HbA1c (%) 7.87 (1.61) TCSS score (3.53) Neuropathy severity (according to TCSS), %(n) Mild 34 (34) Moderate 30 (30) Severe 36 (36) Median nerve 77 entrapment (%) Symptomatic median 60 nerve entrapment (%) Ulnar neuropathy across elbow (%) 12 Data are presented as mean (standard deviation). Age (8.76) (7.10) Height (cm) (7.47) (8.99) Weight (kg) (16.09) (12.39) Nerve CSA (mm 2 ) Median Mid-forearm 5.52 (1.45) 5.18 (0.90) Wrist forearm ratio 1.65 (0.57) 1.30 (0.34) Elbow 7.63 (2.15) 7.05 (1.72) Mid-arm 7.92 (1.91) 7.45 (1.28) Ulnar Wrist 4.54 (1.33) 4.12 (0.94) Mid-forearm 5.20 (1.41) 4.60 (1.06) Mid-arm 6.29 (1.60) 5.88 (1.56) Common peroneal Knee 8.53 (2.06) 7.78 (1.87) Tibial Ankle (3.01) (2.04) Sural Ankle 2.59 (0.96) 1.40 (0.59) <0.001 Radial Mid-forearm 1.35 (0.56) 1.15 (0.36) Values are reported as mean (standard deviation). Significant values are highlighted in bold. sensory potentials in the mild (19.1%), moderate (40.0%) and severe (69.0%) groups, in keeping with worsening DSP. Having excluded entrapment sites, nerve CSAs based on ultrasound were significantly larger in DM compared to healthy controls in the following nerves: ulnar nerve at mid-forearm (P = 0.016), radial nerve at mid-forearm (P = 0.039), peroneal nerve at the knee (P = 0.026), tibial nerve (P = 0.003) and sural nerve at the ankle (P < 0.001) (Table 2). Looking specifically at possible median nerve entrapment at the wrist, it was demonstrated that DM patients had a significantly higher wrist to forearm ratio compared to healthy controls. The wrist to forearm ratio (cut-off value 1.4) had a sensitivity of 73% and specificity of 41% in differentiating DM patients who were symptomatic and those who were not for carpal tunnel syndrome. Comparison of nerve CSAs between the three DM severity groups were statistically significant in peroneal nerve at the knee (P = 0.049) and tibial and sural nerve at the ankle (P = 0.006, P = 0.008) (Table 3). The area under the ROC curve (AUC) analysis was done to differentiate between patients with DSP and without DSP. The sural nerves were found to perform the best and a cut-off value of 2mm 2 had an AUC of 88%, sensitivity of 90% and specificity of 74%. The AUC that best discriminates nerve CSAs of DM patients with severe DSP (TCSS 12) from non-severe DSP (TCSS < 12) was also analysed. The majority of cut-off CSA values were extremely poor discriminators (AUC < 60%). The tibial nerve CSA at mm 2 and sural nerve CSA at 2.75 mm 2 had AUC values of 69% and 62% respectively. Based on these values, the tibial nerve could discriminate between DSP severity at 65% sensitivity and specificity and the sural nerve at 59% sensitivity and 73% specificity. Correlation studies between nerve CSAs and NCS parameters revealed a significant positive correlation between median distal motor latency and CSA at the wrist (P < 0.001, r = 0.453). The median compound action potential was negatively correlated with CSA at the wrist (P = 0.011, r = 0.254). Ulnar motor conduction velocity and CSA at the elbow were also negatively correlated (P = 0.043, r = 0.206) (Table 4). Comparative studies of the NCS parameters between the three DSP severity groups are presented in Table S2. Patients in the most severe category demonstrated significant changes in their NCS parameters. Significant positive correlations were demonstrated between TCSS and nerve CSA for the following nerves: ulnar nerve at mid-forearm (r = 0.310, P = 0.010) and elbow (r = 0.300, P = 0.006), peroneal nerve at fibula head (r = 0.301, P = 0.002), tibial nerve (r = 0.200, P = 0.005) and

4 4 T. ARUMUGAM ET AL. Table 3 Comparison of nerve cross-sectional areas (CSAs) between diabetic patients with different severities (excluding entrapment sites) Mild N = 34 Moderate N = 30 Severe N = 36 (between three groups) (mild versus severe) (moderate versus severe) (mild versus moderate) Age (7.89) (9.39) (9.24) Height (cm) (8.72) (6.62) (6.71) Weight (kg) (18.4) (12.53) (16.68) HbA1c (mmol/mol) 7.94 (1.58) 7.77 (1.64) 7.89 (1.66) Disease duration (years) (8.25) (7.36) (11.47) Nerve CSA (mm 2 ) Median Mid-forearm 5.38 (1.42) 5.50 (1.61) 5.67 (1.37) Elbow 7.35 (1.92) 7.20 (1.85) 8.25 (2.47) Mid-arm 7.65 (1.70) 7.73 (1.53) 8.33 (2.32) Ulnar Wrist 4.26 (0.99) 4.58 (1.73) 4.80 (1.20) Mid-forearm 4.88 (1.51) 5.03 (1.22) 5.64 (1.40) Mid-arm 6.12 (1.67) 6.17 (1.32) 6.56 (1.75) Common peroneal Knee 7.96 (2.46) 8.45 (1.28) 9.18 (2.06) Tibial Ankle (2.24) (2.22) (3.77) Sural Ankle 2.24 (0.62) 2.55 (0.72) 2.99 (1.26) Radial Mid-forearm 1.26 (0.45) 1.37 (0.49) 1.42 (0.69) Values are reported as mean (standard deviation). Significant values are highlighted in bold. sural nerve at the ankle (r = 0.335, P = 0.008) (Table S3). Significant positive correlations were also demonstrated between TCSS and most neurophysiology parameters (Table S4). A significant correlation was demonstrated between disease duration and HbA1c value (P = 0.013, r = 0.248) as well as disease duration and nerve CSA at median wrist, ulnar elbow and arm. However, significant correlation between nerve CSA and HbA1c was only demonstrated in the median nerve at the wrist (Table S3). Table 4 Correlation studies of nerve cross-sectional areas (CSAs) and electrophysiology in diabetic patients NCS parameter Sonographic measurements (CSA/site) r value DML Median wrist <0.001 dcmap DML Ulnar wrist dcmap CV Ulnar elbow DML Tibial ankle dcmap SNAP amplitude Sural ankle SNAP CV DML, distal motor latency (ms); dcmap, distal compound muscle action potential; CV, motor conduction velocity; SNAP, sensory nerve action potential. Significant values are highlighted in bold. Discussion In the current study, the nerve morphology through nerve ultrasound CSA of DM patients with different DSP severities as assessed by the clinical tool TCSS was investigated. It was found that the nerve CSAs of DM patients were larger as the severity of DSP progressed. To our knowledge there have been no previous studies that have described such a relationship utilizing nerve CSA values. In one study, the authors investigated glycaemic control based on glycated haemoglobin and nerve morphology based on sural nerve biopsy [11]. The latter is an invasive method and unlikely to be a feasible measure of nerve morphology in a clinical setting. Traditionally, NCSs have been the gold standard for the diagnosis of DSP. However, patients with advanced DSP are likely to have inexcitable nerves thus making an objective assessment of injured nerves challenging. NCSs are also limited to assessing large nerve fibres whereas small nerve fibres are typically the first to be affected in DSP [12]. Previous studies have found assessment of small nerve fibres such as the stimulated skin wrinkling test with EMLA more sensitive than NCSs as a diagnostic tool for DSP [13]. Nerve ultrasound has the advantage of assessing nerves of varying sizes and has been shown to be a useful diagnostic tool in peripheral neuropathies including DSP.

5 NERVE ULTRASOUND IN DIABETIC NEUROPATHY 5 Nerve enlargement is seen in DSP possibly reflecting increased water content due to conversion of glucose into sorbitol in the nerve [14]. In the current study, our DM DSP group was initially compared with normal controls and it was found that the nerve CSAs were significantly larger at one or more sites imaged at the median, ulnar, radial, peroneal, tibial and sural nerves. Although there was significant overlap in the ranges of CSA values between the two groups, this overlap was less evident in the lower limb nerves reflecting a length-dependent pattern that is seen in DSP. In the clinical setting, it would be important to establish normative values for individual nerves to allow for a more valid comparison of their diseased state. It was also found that nerves were larger at common entrapment sites, i.e. wrist for the median nerve, elbow for the ulnar nerve and fibular head for the peroneal nerve. This is likely to reflect the increased susceptibility of patients with DSP to develop entrapment neuropathies [15]. This is supported by the significantly greater number of DM patients with abnormal median wrist to forearm nerve CSA ratio. In one study, a ratio of 1.4 was found to be sensitive at detecting carpal tunnel syndrome [16]. In the current study, this ratio was found to be sensitive but not specific for clinically symptomatic median nerve entrapment in DM patients with DSP. Having excluded entrapment sites, significantly larger nerves at the peroneal knee as well as the tibial and sural nerve at the ankle were found. Our findings are in keeping with previous studies that have found both tibial and sural nerves to be enlarged [2,6,17]. However, the mean CSA of the sural nerve in our study of 2.5 mm 2 is lower than previously reported (4.9 mm 2 ) [18]. This may be due to differences in the patients, who were older and heavier than the current cohort. Other possibilities may include factors inherent to the individual such as ethnicity. One study comparing median nerve CSAs between normal Indian and Dutch cohorts found significant differences with the Dutch cohort demonstrating larger CSAs [19]. In a different study, cooling of the limbs to a mean of approximately 17 C was also associated with significantly larger nerve CSAs [20]. The latter study raises the possibility that normal subjects who are exposed to cooler conditions may have larger nerves. The surface temperatures during nerve ultrasound were not recorded but the baseline surface temperature in NCSs was around 28 C in the majority of patients prior to being warmed to 32 C for the NCS. Our findings further highlight the importance of establishing normative data in individual cohorts before evaluating cut-off values to be used as diagnostic markers. In the current cohort, ROC curve analysis found the sural nerve CSA to be a good discriminator between the presence and absence of DSP using a cut-off value of 2mm 2. At this cut-off value, the sural nerve CSA had a sensitivity of 90% and specificity of 74%. Increasing the cut-off value to 3 mm 2 would improve the specificity of the test to 90% but compromise on its sensitivity at 47%. In view of the high prevalence of DSP in DM, a more sensitive test would be of higher diagnostic value and a sural nerve CSA of 2 mm 2 can be used as a screening test to detect DSP prior to subjecting patients to NCSs, which are less well tolerated [21]. When comparing nerve CSAs in the DSP groups based on severity, significant differences in the tibial and sural nerve CSAs were found. Both tibial and sural nerves showed progressively larger nerve CSAs with worsening severity. This suggests that both tibial and sural nerve CSAs can potentially be used as markers of disease severity in DSP. This is of particular importance as the sural nerve potentials were unrecordable in 69.0% of patients in the severe DSP group. AUC analyses revealed fair discriminatory cut-off CSA values for the tibial and sural nerves at and 2.75 mm 2 respectively to differentiate between severe and less severe DSP as determined by TCSS. Previous studies have suggested an optimum threshold value that identifies DSP without specifically focusing on severity. In one study, a cut-off value for the tibial nerve of 19.1 mm 2 was suggested to identify DSP [6]. In the same study, NCSs were utilized to detect the presence of neuropathy. Thus, whilst NCSs are useful as an objective marker of DSP, they may not be as useful in discriminating between DSP severities. For this, TCSS has been validated as an instrument that can indicate the presence and severity of DSP [22]. TCSS is also a universally accessible tool that is easily implemented in DM patients. In the current study, it was also possible to demonstrate significant associations between NCS parameters and nerve size in DSP patients. Distal motor latencies of the median nerve were positively correlated with nerve CSAs whereas compound motor action potential amplitudes were negatively correlated with nerve CSAs in the median and ulnar nerve. Ulnar motor nerve conduction velocity was also negatively correlated with nerve CSAs across the elbow. No significant correlations were detected in the lower limb NCS parameters and nerve CSAs. However, it is important to note that the lower limb nerve potentials were inexcitable in 43.2%, 13.1% and 8.0% of the total number of sural, peroneal and tibial nerves respectively. The unrecordable potentials were probably due to significant axonal degeneration resulting in

6 6 T. ARUMUGAM ET AL. too few remaining axons that are able to conduct an electrically quantifiable signal. Previous studies have suggested that unrecordable potentials reflect more severe neuropathy with an increased risk of developing foot ulcers [23 26]. The relationship between disease duration and glycaemic control (HbA1c) with DSP severity, NCS parameters and nerve CSAs was also investigated. It was not possible to demonstrate a significant relationship between glycaemic control and DSP severity or NCS and CSA parameters. This is similar to a previous study where no significant association was found between HbA1c and DSP severity [27]. However, a significant positive correlation between duration of DM and DSP severity was demonstrated. This is similar to other studies that have also found that disease duration was a better predictor of DSP severity [28,29]. A significant positive correlation was also seen between disease duration and nerve CSA of the median nerve at the wrist and ulnar nerve at the elbow. A previous study described similar findings to the morphological severity of sural nerve fibre density [11]. Some of the limitations to our study include the small number of patients in the individual DSP severity groups. However, it was possible to demonstrate that nerve ultrasound has a role in discriminating between DSP severities. Different cut-off values to previous studies were also found, further highlighting the limitations of studies on homogeneous populations and the importance of establishing normative values in individual cohorts. Future studies involving a larger and more diverse cohort may yield further useful information such as CSA cut-off values with better performance when discriminating between DSP severities. The use of more advanced ultrasound techniques incorporating colour Doppler imaging might also provide further insight. In conclusion, the current study demonstrates significantly larger nerves in patients with DSP compared to controls. Our findings also suggest that nerve CSA values have a role in determining the severity of DSP, particularly when other objective parameters such as nerve action potentials are inexcitable. Acknowledgements Dr Shahrizaila receives research funding from the University of Malaya (RG491/13HTM). Disclosure of conflicts of interest The authors declare no financial or other conflicts of interest. Supporting Information Additional Supporting Information may be found in the online version of this article: Table S1. Toronto Clinical Scoring System. Table S2. Comparative studies of neurophysiology parameters between the different DSP severity groups. Table S3. Correlation studies of nerve CSAs at different sites versus different disease markers. Table S4. Correlation studies of nerve conduction parameters versus different disease markers. References 1. Dyck PJ, Overland CJ, Low PA, et al. Signs and symptoms versus nerve conduction studies to diagnose diabetic sensorimotor polyneuropathy: Cl vs. NPhys trial. Muscle Nerve 2010; 42: Watanabe T, Ito H, Sekine A, et al. Sonographic evaluation of the peripheral nerve in diabetic patients: the relationship between nerve conduction studies, echo intensity, and cross-sectional area. J Ultrasound Med 2010; 29: Mallik A, Weir AI. Nerve conduction studies: essentials and pitfalls in practice. J Neurol Neurosurg Psychiatry 2005; 76(Suppl. 2): ii23 ii Kele H. Ultrasonography of the peripheral nervous system. Perspect Med 2012; 1: Ishibashi F, Taniguchi M, Kojima R, Kawasaki A, Kosaka A, Uetake H. Morphological changes of the peripheral nerves evaluated by high-resolution ultrasonography are associated with the severity of diabetic neuropathy, but not corneal nerve fiber pathology in patients with type 2 diabetes. J Diabetes Investig 2015; 6: Riazi S, Bril V, Perkins BA, et al. Can ultrasound of the tibial nerve detect diabetic peripheral neuropathy? A cross-sectional study Diabetes Care 2012; 35: England JD, Gronseth GS, Franklin G, et al. Practice parameter: the evaluation of distal symmetric polyneuropathy: the role of autonomic testing, nerve biopsy, and skin biopsy (an evidence-based review) Report of the American Academy of Neurology, the American Association of Neuromuscular and Electrodiagnostic Medicine, and the American Academy of Physical Medicine and Rehabilitation.. PM R 2009; 1: Moon HI, Kwon HK, Kim L, Lee HJ, Lee HJ. Ultrasonography of palm to elbow segment of median nerve in different degrees of diabetic polyneuropathy. Clin Neurophysiol 2014; 125: Bahou YG, Elhadidy S. Ulnar neuropathy at Jordan University Hospital. Neurosciences (Riyadh) 2005; 10: England JD, Gronseth GS, Franklin G, et al. Distal symmetric polyneuropathy: a definition for clinical research Report of the American Academy of Neurology, the American Association of Electrodiagnostic Medicine, and the American Academy of Physical Medicine and Rehabilitation. Neurology 2005; 64: Perkins BA, Greene DA, Bril V. Glycemic control is related to the morphological severity of diabetic sensorimotor polyneuropathy. Diabetes Care 2001; 24:

7 NERVE ULTRASOUND IN DIABETIC NEUROPATHY Said G, Slama G, Selva J. Progressive centripetal degeneration of axons in small fibre diabetic polyneuropathy. Brain 1983; 106(Pt4): Ping Ng KW, Ong JJ, Nyein Nyein TD, et al. EMLA induced skin wrinkling for the detection of diabetic neuropathy. Front Neurol 2013; 4: Lee D, Dauphinee DM. Morphological and functional changes in the diabetic peripheral nerve: using diagnostic ultrasound and neurosensory testing to select candidates for nerve decompression. J Am Podiatr Med Assoc 2005; 95: Edwards JL, Vincent A, Cheng T, Feldman EL. Diabetic neuropathy: mechanisms to management. Pharmacol Ther 2008; 120: Hobson-Webb LD, Massey JM, Juel VC, Sanders DB. The ultrasonographic wrist-to-forearm median nerve area ratio in carpal tunnel syndrome. Clin Neurophysiol 2008; 119: Liu F, Zhu J, Wei M, Bao Y, Hu B. Preliminary evaluation of the sural nerve using 22-MHz ultrasound: a new approach for evaluation of diabetic cutaneous neuropathy. PLoS ONE 2012; 7: e Hobson-Webb LD, Massey JM, Juel VC. Nerve ultrasound in diabetic polyneuropathy: correlation with clinical characteristics and electrodiagnostic testing. Muscle Nerve 2013; 47: Burg EWV, Bathala L, Visser LH. Difference in normal values of median nerve cross-sectional area between Dutch and Indian subjects. Muscle Nerve 2014; 50: Ulasli AM, Tok F, Karaman A, et al. Nerve enlargement after cold exposure: a pilot study with ultrasound imaging. Muscle Nerve 2014; 49: Lalkhen GG, McCluskey A. Clinical tests: sensitivity and specificity. Contin Educ Anaesth Crit Care Pain 2008; 8: Bril V, Perkins BA. Validation of the Toronto Clinical Scoring System for diabetic polyneuropathy. Diabetes Care 2002; 25: Veves A, Malik RA, Lye RH, et al. The relationship between sural nerve morphometric findings and measures of peripheral nerve function in mild diabetic neuropathy. Diabet Med 1991; 8: Vinik AI, Bril V, Litchy WJ, Price KL, Bastyr EJ III. Sural sensory action potential identifies diabetic peripheral neuropathy responders to therapy. Muscle Nerve 2005; 32: Behse F, Buchthal F, Carlsen F. Nerve biopsy and conduction studies in diabetic neuropathy. J Neurol Neurosurg Psychiatry 1977; 40: Severinsen K, Andersen H. Evaluation of atrophy of foot muscles in diabetic neuropathy a comparative study of nerve conduction studies and ultrasonography. Clin Neurophysiol 2007; 118: Sachedina S, Toth C. Association of comorbidities with increasing severity of peripheral neuropathy in diabetes mellitus. World J Diabetes 2013; 4: Dyck PJ, Davies JL, Wilson DM, Service FJ, Melton LJ III, O Brien PC. Risk factors for severity of diabetic polyneuropathy: intensive longitudinal assessment of the Rochester Diabetic Neuropathy Study cohort. Diabetes Care 1999; 22: Mimi O, Teng CL, Chia YC. The prevalence of diabetic peripheral neuropathy in an outpatient setting. Med J Malaysia 2003; 58: