- Email: [email protected]
Effectiveness and feasibility of nailfold microcirculation test to screen for diabetic peripheral neuropathy
diabetes research and clinical practice
1 3 1 ( 2 0 1 7 ) 4 2 –4 8
Contents available at ScienceDirect
Diabetes Research and Clinical Practice journa l home page: www.e lse vier.com/locate/diabres
Effectiveness and feasibility of nailfold microcirculation test to screen for diabetic peripheral neuropathy Guotao Hu a, Fanglong Zhai a, Feifei Mo a, Li He b, Weiya Shen a, Hailan Wang a,* a b
Department of Endocrinology, Zunyi Medical College Affiliated Shenzhen Longgang Central Hospital, Shenzhen 518116, Guangdong, China Department of Nursing, Zunyi Medical College Affiliated Shenzhen Longgang Central Hospital, Shenzhen 518116, Guangdong, China
A R T I C L E I N F O
A B S T R A C T
Article history:
Aims: The nailfold microcirculation index (MI) is a non-invasive, objective, and highly sen-
Received 25 February 2017
sitive blood capillary detection method. This study evaluated the diagnostic efficiency of
Received in revised form
the nailfold MI relative to the more subjective vibration perception threshold (VPT) exam-
19 May 2017
ination for early diagnostic screening of diabetic peripheral neuropathy (DPN).
Accepted 9 June 2017
Methods: From February 2015 to February 2016, 227 diabetes mellitus patients and 39
Available online 13 June 2017
healthy individuals were enrolled. Each subject underwent the MI test and the VPT examination.
Keywords: Diabetes mellitus Diabetic peripheral neuropathy Nailfold microcirculation index Vibration perception threshold Neuropathy scoring system
Results: MI was more closely associated with DPN than age, diabetic duration, smoking, drinking, systolic pressure, serum creatinine, 24-h urinary protein, hypoxia-inducible factor-1a (HIF1A), vascular endothelial growth factor (VEGF), the VEGF receptors Flt-1 and Flt-4, ankle branchial index (ABI), DPN, or DPN stage (OR = 11.819). Both the MI and VPT closely correlated with age, diabetic duration, serum creatinine, 24-h urinary protein, HIF1A, VEGF, Flt-1, Flt-4, ABI, DPN, and DPN stage. By the receiver operating characteristic (ROC) curve, the MI diagnostic cutoff point was 2.56, where the corresponding Youden’s index was maximum and the area under ROC curve was 0.943. The diagnostic efficiency of MI and VPT were similar. MI and VPT indicated similar percentages of diabetic patients with DPN at the most severe stage, while MI achieved a higher diagnostic rate for the earliest stages. Conclusions: The nailfold MI is a feasible method for clinical early diagnostic screening of DPN in diabetic patients, and is more objective and reliable than VPT. Ó 2017 Elsevier B.V. All rights reserved.
1.
Introduction
Diabetic peripheral neuropathy (DPN) as a symmetrical sensorimotor multiple neuropathy is attributed to chronic hyperglycemia and the metabolic and microvascular changes due to cardiovascular risk factors [1]. DPN is the most common complication of diabetes mellitus, with a prevalence between * Corresponding author. E-mail address: [email protected] (H. Wang). http://dx.doi.org/10.1016/j.diabres.2017.06.017 0168-8227/Ó 2017 Elsevier B.V. All rights reserved.
17% and 66%, depending on diagnostic criteria and demography [2]. DPN involves the sensory, autonomic, and motor nerves, resulting in impaired sensation, autonomic symptoms, and movement disorders [3]. DPN is difficult to diagnose at onset. Many diabetic patients remain asymptomatic during the long incubation period, and symptoms when they occur are often not related
diabetes research and clinical practice
to damage severity. When first detected clinically, demyelination of the peripheral nerves is generally irreversible, as are other pathological changes [1]. Thus, early diagnosis of DPN is important. The nerve conduction velocity assay is considered the gold standard for diagnosing DPN [4], but can efficiently detect only lesions of large myelinated nerve fibers, and is not sensitive to the neuropathy of small nerve fibers. Thus, patients with early diabetes without obvious clinical symptoms may have small nerve fiber damage that is not detected by the nerve conduction velocity assay. In addition, the assay is expensive and time-consuming, and therefore is not suitable for wide diagnostic screening of DPN. Other diagnostic screening methods for DPN include the Toronto Clinical Scoring System (TCSS) [5], Michigan Neuropathy Screening Instrument (MNSI) [6], and Diabetic Neuropathy Symptom Score (DNS) [2]. However, the TCSS, MNSI, and DNS methods for diagnostic screening of DPN rely on subjective judgments, and are further limited by poor replicability [7]. The vibration perception threshold (VPT) examination evaluates the function of myelinated nerve fibers according to the patient’s awareness of vibration sensation. It is an effective method for screening DPN, as well as evaluating the risk of foot ulcer in diabetes patients [8]. By enabling early detection for timely intervention and prevention, the diabetic patient’s quality of life is improved [9]. The VPT examination has been widely used in clinical screening in Europe and the United States and has been recommended for early diagnosis of DPN by the American Diabetes Association [2]. Although the definitive pathophysiology of DPN remains elusive, accumulating evidence indicates the source of a variety of diabetic complications is microvascular lesion [2]. Our previous study and others have shown that microcirculation dysfunction is closely association with DPN [2,10]. A microcirculation examination may have be able to assess the risk of DPN and screen early [11–13]. Nailfold capillaroscopy is commonly used to investigate skin microcirculation in the clinic [2]. It is noninvasive, simple, fast, and economical, and effectively identifies peripheral microvascular changes. In the present study, we evaluated the effectiveness and feasibility of the nailfold microcirculation test for diagnostic screening of DPN, comparing it with the VPT, which is widely used in clinical screening for diabetic neuropathy.
1 3 1 ( 2 0 1 7 ) 4 2 –4 8
43
Patients with non-diabetic neuropathy due to any cause were excluded. Such neuropathies included any of the following: pernicious anemia; vitamin B6 intoxication; vitamin B12 deficiency; alcoholism; uremia; chemical toxins; nerve entrapment; squeezing of the nerves due to benign causes; hepatitis; idiopathic neuropathy; various congenital sensorimotor neuropathy; paraneoplastic syndromes; syphilis; HIV/ AIDS; drugs such as chemotherapy or isoniazid; cervical and lumbar spondylosis including nerve root compression, stenosis, or cervical and lumbar degenerative diseases; cerebral infarction; Guillain-Barre Syndrome; or severe arteriovenous vascular lesions (e.g., venous thrombosis, lymphangitis, or arteriovenous fistula). The control group in this study consisted of 39 healthy individuals, with no history of any of the following: diabetes mellitus; hypertension; mental disorder; diseases of the coronary, nervous, hematologic, or cardiovascular and cerebrovascular systems; or autoimmune disease. For each individual in the control group, routine blood and urine examination, stool test, liver and kidney function, and blood lipids were within normal levels, and there were no abnormalities in their chest radiography, electrocardiogram, abdominal ultrasonography, fundus examination, or physical examination.
2.2.
General data collection
The following data for each patient with diabetes mellitus were collected: age, gender, diabetes duration, smoking, drinking, height, weight, blood pressure, body mass index (BMI), and ankle branchial index (ABI). The ABI was assessed using an ES-100V3 Doppler flow detector (HADECO, Japan). In accordance with the criteria of the American Diabetes Association, the normal ABI range was from 0.91 to 1.30 [15].
2.3.
Biochemical tests
The Ethics Committee at Shenzhen Longgang Center Hospital independently approved this study. All the patients enrolled in this study provided written informed consent to participate.
The following biochemical levels were tested: fasting bloodglucose (FBG); hemoglobin A1c (HbAlc); total cholesterol; triglyceride; high-density lipoprotein (HDL); low-density lipoprotein (LDL); serum creatinine; blood uric acid; hypoxia-inducible factor-1a (HIF1A); vascular endothelial growth factor (VEGF); the VEGF receptors Flt-1 and Flt-4; and 24-h total urinary protein. Immunoturbidimetry was used for testing 24-h total urinary protein. HbAlc was tested using the VARIANT II glycated hemoglobin test system (Bio-Rad, America). HIF1A, VEGF, Flt-1, and Flt-4 were tested using enzyme-linked immunosorbent assay (ELISA) kits (TSZ, USA). The other chemical tests were performed with a biochemical autoanalyzer (Olympus AU600, Olympus, Japan).
2.1.
2.4.
2.
Materials and methods
Inclusion and exclusion criteria
From February 2015 to February 2016, 227 outpatients and inpatients with diabetes mellitus were recruited at the Department of Endocrinology, Shenzhen Longgang Center Hospital. Type 2 diabetes mellitus was diagnosed according to the 2017 Standards of Medical Care in Diabetes of the American Diabetes Association [14]. Only patients with neuropathy due to type 2 diabetes mellitus were included.
DPN diagnosis
DPN was diagnosed based on the MNSI and neurological deficit score (NDS) for neuropathy [6]. The MNSI symptoms index was used to score the symptoms of patients, and the MNSI signs index and NDS index were used to score the signs of the patients. A diagnosis of DPN met the following criteria: MNSI symptoms index 4 and/or MNSI signs index 2; MNSI symptoms index 4 and NDS index 3; or NDS index 6.
44
diabetes research and clinical practice
The diabetic patients were divided into 4 groups based on Dyck’s 4 stages of DPN severity [16]: 0, no objective evidence of diabetic neuropathy; 1, no symptoms, but abnormal signs; 2, with symptoms, signs, and ankle dorsiflexion; and 3, disabling neuropathy, diabetic foot ulcer, and others.
2.5. VPT examination examination
and
nailfold
microcirculation
VPT examination was performed using a Bio-Thesiometer quantitative sensory tester (Beijing Dimeider Technology, China). The nailfold microcirculation examination was performed using a ZL102 nailfold capillaroscope (Xuzhou Medical Instrument, China). Briefly, the subject was seated in a quiet environment at 20–24 °C. Two droplets of cedar oil were dropped on the nailfold skin of the third finger on the left or right hand. Then the nailfold microcirculation was evaluated by nailfold capillaroscopy, as in a previous report [17]. Nailfold microcirculation was classified as one of 4 grades as follows: normal, index < 1; almost normal, 1 index 2; light abnormality, 2 < index < 4; median abnormality, 4 index < 8; and severe abnormality, 8 index.
2.6.
Receiver operating characteristic (ROC) curve
We used the ROC curve [18] to determine the diagnostic cutoff point and diagnostic efficiency of the microcirculation index (MI) and VPT. Based on the shift of the cutoff point in the ROC curve, multiple paired sensitivity and false positive rates (1 – specificity) could be obtained. Youden’s index was obtained by an equation (i.e., Youden’s index = sensitivity + specificity – 1), with the maximum Youden’s index considered the optimum diagnostic cutoff point. At any particular cutoff point, if the area under the ROC curve was 0.7, 0.7– 0.9, or >0.9, the diagnostic value was defined as low, medium, and high, respectively.
2.7.
Statistical analysis
Statistical analyses were performed using SPSS software (Version 21.0; SPSS, Chicago, IL, USA). Measurement data are shown as mean ± standard deviation, and count data are shown as a percentage. The DPN severities were divided into 4 stages according to Dyck’s score [13]. In each stage, measurement data met the standard normal distribution, but count data (gender, smoking, and drinking) did not meet the standard normal distribution. Qualitative data was analyzed using the chi-squared test. Comparisons between the 2 groups were analyzed using the independent sample Student’s t-test. Comparisons among multiple groups were performed using one-way analysis of variance and the least significant difference. Correlations of numerical and qualitative data were analyzed, respectively, using Pearson’s and Spearman’s correlation analyses. The determination of the risk factors of DPN was performed using univariate analysis by non-conditional logistic regression analysis. The ROC curve was used to determine the diagnostic point and efficiency of the MI and VPT. A P-value < 0.05 was considered statistically significant.
1 3 1 ( 2 0 1 7 ) 4 2 –4 8
3.
Results
3.1.
General features of patients
The study population comprised 227 diabetes mellitus patients (140 men, 87 women; aged 50.93 ± 12.16 y) and a control group of 39 healthy individuals (24 men; 15 women; aged 55.50 ± 7.21 y; Table 1). The patient and control groups were comparable in age and gender ratio (P = 0.756 and P = 0.363, respectively). The nailfold MI of the left and right hands of the 2 groups were statistically similar, as was the VPT of the left and right hands (P > 0.05; Supplementary Tables 1 and 2). We used the mean nailfold MI and VPT in this study. The mean nailfold MI and VPT of the diabetic patients (3.64 ± 1.91 and 15.27 ± 9.62, respectively) were each significantly higher than that of the control group (0.57 ± 0.17 and 10.40 ± 2.49; P < 0.001). Of the 227 patients, 133 (58.59%) received a diagnosis of DPN (including stages 1–3; Table 1). The patients were further grouped by stage of DPN (0, 1, 2, or 3). With each stage of higher number; the following were significantly higher: age; diabetic duration; systolic pressure; 24-h urinary protein; HIF1A; VEGF; Flt-1; Flt-4; serum creatinine; VPT; and nailfold MI (P < 0.05, each). Also with each stage, the ABI was significantly lower (P < 0.001).
3.2.
Nailfold MI as risk factor for DPN
For the univariate unconditional logistic regression analysis of the risk factors of DPN, DPN was considered the dependent variable (Table 2). The following characteristics were associated with DPN, indicating that they were potential risk factors for DPN (P < 0.05): age; diabetic duration; smoking; drinking; systolic pressure; 24-h urinary protein; HIF1A; VEGF; Flt-1; Flt-4; serum creatinine; VPT; and nailfold MI. In these diabetic patients, the nailfold MI was likely the main risk factor for DPN (odds ratio [OR] = 11.819).
3.3.
Correlations between nailfold MI and DPN risk factors
The VPT is widely used in clinical screening for diabetic neuropathy. We investigated correlations between the VPT and the DPN risk factors by correlation analysis (Table 3). VPT correlated linearly with each of the following: age; diabetic duration; serum creatinine; 24-h urinary protein; HIF1A; VEGF; Flt1; Flt-4; ABI; MI; DPN; and DPN stage (P < 0.05). VPT correlated negatively with ABI, and positively with other DPN risk factors. In particular, the VPT correlated closely with the MI (R = 0.757) and DPN stage (R = 0.829). Highly consistent with the VPT, the MI correlated positively with the following (Table 3): age, diabetic duration, serum creatinine, 24-h urinary protein, HIF1A, VEGF, Flt-1, Flt-4, ABI, MI, DPN, and DPN stage. The MI also correlated negatively with ABI and positively with other DPN risk factors. MI correlated closely with VPT (R = 0.757) and DPN stage (R = 0.840).
3.4.
Correlation between MI and DPN severity
Nailfold microcirculation was classified with reference to the grade index as follows: normal, <1; almost normal, 1 or 2; light
diabetes research and clinical practice
45
1 3 1 ( 2 0 1 7 ) 4 2 –4 8
Table 1 – General characteristics of diabetic patients by DPN stage of severity.
Subjects, n Male, % Age, y Duration, y BMI, kg/m2 Systolic pressure, mmHg Diastolic pressure, mmHg FBG, mmol/L HbA1c, % TC, mmol/L TG, mmol/L HDL-C, mmol/L LDL-C, mmol/L Serum creatinine, mmol/L Serum uric acid, mmol/L 24-h urinary protein, mg HIF1A, ng/L VEGF, pg/mL Flt-1, pg/mL F1t-4, pg/mL ABI VPT, V MI Smoking, % Drinking, % a
0
1
2
3
P value
94 63 (67.02%) 45.01 ± 10.53 1.74 ± 2.90 24.04 ± 3.61 127.97 ± 19.51 78.86 ± 11.76 12.09 ± 13.83 11.34 ± 2.68 5.05 ± 1.73 2.77 ± 6.24 1.11 ± 0.37 3.21 ± 1.13 72.76 ± 24.41 320.47 ± 107.51 17.90 ± 31.99 122.5 ± 5.24 771.67 ± 48.57 571.04 ± 34.69 1066.75 ± 95.99 1.18 ± 0.041 8.29 ± 3.68 2.21 ± 0.59 36 (38.30%) 24 (25.53%)
48 34 (70.83%) 50.48 ± 12.03 5.51 ± 3.83 23.57 ± 3.64 137.56 ± 22.82 81.02 ± 13.83 10.55 ± 3.45 11.16 ± 2.31 5.21 ± 1.64 2.39 ± 3.60 1.15 ± 0.33 3.28 ± 1.30 69.66 ± 26.15 285.17 ± 86.13 36.44 ± 65.47 144.19 ± 11.19 968.04 ± 58.12 685.02 ± 74.90 1505.20 ± 165.43 1.12 ± 0.05 13.16 ± 3.27 3.27 ± 0.52 16 (33.33%) 7 (14.58%)
65 33 (50.77%) 57.14 ± 10.68 10.54 ± 6.29 23.39 ± 2.99 132.78 ± 21.82 75.66 ± 13.00 10.71 ± 3.63 10.76 ± 2.49 4.96 ± 1.62 4.86 ± 22.65 1.11 ± 0.41 3.25 ± 1.07 91.29 ± 101.49 322.71 ± 112.10 123.80 ± 345.98 162.49 ± 19.46 1190.19 ± 72.03 812.36 ± 90.2 2980.00 ± 3723.89 1.08 ± 0.10 21.05 ± 7.35 4.97 ± 1.72 9 (13.85%) 2 (3.08%)
20 10 (50.00%) 59.65 ± 9.33 11.9 ± 7.17 22.94 ± 1.82 142.85 ± 22.04 75.10 ± 13.97 10.54 ± 2.69 10.29 ± 2.34 4.69 ± 1.22 1.43 ± 0.61 1.34 ± 0.26 3.01 ± 1.04 126.21 ± 168.67 346.10 ± 90.06 258.90 ± 433.33 192.30 ± 3.49 901.33 ± 165.29 658.55 ± 108.22 1139.44 ± 138.51 1.06 ± 0.57 34.37 ± 8.73 6.95 ± 1.81 3 (15.00%) 3 (15.00%)
– – <0.001a <0.001a 0.454 0.009a 0.1 0.714 0.258 0.678 0.628 0.071 0.846 0.017a 0.096 <0.001a <0.001a <0.001a <0.001a <0.001a <0.001a <0.001a <0.001a – –
Comparison among DPN groups.
Table 2 – Risk factor analysis for DPN by logistic analysis.
Duration of disease Age Smoking Drinking Systolic pressure HIF1A VEGF Flt-1 F1t-4 24-h urinary protein VPT MI
OR (95% CI)
P value
1.444 (1.306, 1.596) 1.085 (1.055, 1.116) 2.328 (1.292, 4.194) 0.236 (0.110, 0.507) 1.018 (1.005, 1.032) 1.748 (1.411, 2.165) 1.031 (1.022, 1.041) 1.042 (1.030, 1.057) 1.013 (1.010, 1.017) 1.014 (1.005, 1.023) 1.544 (1.372, 1.738) 11.819 (6.148, 22.72)
<0.001 <0.001 0.005 <0.001 0.006 <0.001 <0.001 <0.001 <0.001 0.002 <0.001 <0.001
CI, confidence interval; OR, odds ratio.
abnormality, 2 or 3; median abnormality, 4–8; and severe abnormality, 8 and higher. DPN patients were divided into 0, 1-, 2-, or 3-stage groups. A nonparametric correlation analysis was performed to analyze a correlation between the degree of microcirculation abnormality and DPN severity. A highly positive correlation was determined between microcirculation abnormality and DPN severity (R = 0.763), similar to the correlation between VPT and DPN stage.
3.5. Efficiency of nailfold microcirculation test for DPN diagnosis With the diagnostic reference of the DPN score, the ROC curve was plotted to analyze the cutoff points of MI and VPT for DPN diagnosis (Fig. 1; Table 4). When the diagnostic cutoff
Table 3 – Nailfold MI and VPT correlations with DPN risk factors. MI
Disease duration Age HIF1A VEGF Flt-1 F1t-4 24-h urinary protein ABI MI VPT DPN DPN stage Serum creatinine a b
VPT
R
P value
R
P value
0.656a 0.420a 0.697a 0.490a 0.464a 0.259 0.391 0.510a – 0.757a 0.681b 0.840b 0.341a
<0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 – <0.001 <0.001 <0.001 0.090
0.556a 0.422a 0.698a 0.451a 0.432a 0.166 0.230a 0.459a 0.757a – 0.729b 0.829b 0.251a
<0.001 <0.001 <0.001 <0.001 <0.001 0.012 0.004 <0.001 <0.001 – <0.001 <0.001 0.001
Pearson’s correlation analysis. Spearman’s correlation analysis.
point of MI was 2.56, its corresponding Youden index was at the maximum and the corresponding area under curve was 0.943. This indicated that the MI had good diagnostic efficiency for DPN. With a VPT diagnostic cutoff point of 12.73 V, the corresponding Youden index was at the maximum and the corresponding area under curve was 0.927, indicating good diagnostic efficiency for DPN. After we found the diagnostic cutoff point of MI, we further analyzed the efficiency of MI for diagnosing DPN, compared with the DPN score system and the VPT (Supplementary Table 3). According to the diagnosis criteria
46
diabetes research and clinical practice
1 3 1 ( 2 0 1 7 ) 4 2 –4 8
To determine the effectiveness and feasibility of MI for diagnostic screening of DPN, we compared the rates of diagnosis of MI and VPT according to the DPN stage of severity (Table 5). In the 39 healthy individuals, the rate of diagnosis of DPN according to the MI and VPT cutoff points was 0.00%. For DPN patients at stage 3, the rate of diagnosis of DPN according to the cutoff point for MI was similar to that of the VPT. However, the diagnostic rate when using MI at stages 0–2 was higher than when the VPT was applied.
4.
Fig. 1 – ROC curve to investigate the cutoff points of DPN diagnosis for MI and VPT.
Table 4 – Sensitivity, specificity, and Youden index of the diagnostic cutoff points for MI and VPT.* Cutoff point Sensitivity Specificity Youden index MI
2.43 2.48 2.56 2.70 2.78 VPT 12.53 12.63 12.73 12.88 13.05 *
0.932 0.925 0.925 0.917 0.910 0.812 0.805 0.805 0.789 0.774
0.830 0.851 0.862 0.862 0.872 0.904 0.904 0.915 0.926 0.926
0.762 0.776 0.787 0.779 0.782 0.716 0.709 0.720 0.715 0.700
Determined by ROC curve and area under the ROC curve.
of the DPN score system, 133 patients had DPN among the 227 diabetes mellitus patients (58.59%). Using the diagnostic cutoff points described above, that is, MI > 2.56 and VPT > 12.73 V, the DPN was diagnosed in 59.90% and 50.70% of the patients, respectively. Combining the MI and VPT, DPN was diagnosed in 66.5% of the patients. According to the chi-squared test, the rate of diagnosis of DPN using the MI was higher than when using either the neuropathy score or the VPT. The rate of diagnosis using MI and VPT (combined) was higher than that of the other diagnostic methods applied singly.
Discussion
To develop a rapid, effective, and economical screening method for the early diagnosis of DPN, we investigated the effectiveness and feasibility of the MI for diagnostic screening. We compared the diagnostic rate of the MI with that of the widely used VPT for a subject population of 227 patients with diabetes mellitus and 39 healthy individuals. The prevalence of DPN in our patients was 58.59%, similar to the rate published in a study in 2011 [2]. We found that a high MI was a factor that was closely associated with DPN (OR = 11.819), in addition to age, diabetic duration, smoking, drinking, systolic pressure, and other risk factors. Consistent with the results of the VPT, the MI closely correlated positively with age, diabetic duration, serum creatinine, 24-h urinary protein, HIF1A, VEGF, Flt-1, Flt-4, ABI, DPN, and DPN stage. In addition, the MI closely correlated positively with the DPN stage of severity. According to the ROC curve, we found that an MI diagnostic cutoff point of 2.56 resulted in good diagnostic efficiency for DPN. In this study, we also validated the effectiveness and feasibility of MI for DPN diagnosis in patients at each stage of DPN severity. Numerous risk factors are associated with DPN onset and progression. Ashok et al. [19] showed that age and the duration of diabetes mellitus were associated with DPN. Another study found that the prevalence of DPN was about 10% at the onset of diabetes mellitus, but 25 years later had increased up to 50% [20]. The ABI is the ratio of the systolic pressures of the ankle artery and brachial artery. The American Diabetes Association has recommended that the ABI should be routinely used for screening for peripheral arterial disease in patients with diabetes mellitus who are older than 50 years [21]. In addition, studies have shown that HIF1A, VEGF, Flt-1, and Flt-4 are closely associated with the progression of diabetic vascular complications [2]. In the present study, we found that the DPN
Table 5 – Diagnostic rates determined by MI and VPT of the control group and by DPN stage of severity in diabetic patients.a
Subjects, n MI >2 >2.56 VPT, V >15 >12.73 a
Control group
0
1
2
3
39 0.57 ± 0.17 0 0 10.40 ± 2.49 0 0
94 2.21 ± 0.59 54 (57.45%) 2 (12.77%) 8.29 ± 3.68 3 (3.19%) 5 (5.32%)
48 3.27 ± 0.52 47 (97.92%) 43 (89.58%) 13.16 ± 3.27 17 (35.42%) 30 (62.50%)
65 4.97 ± 1.72 64 (98.46%) 60 (92.31%) 21.05 ± 7.35 50 (76.92%) 47 (87.69%)
20 6.95 ± 1.81 20 (100%) 20 (100%) 34.37 ± 8.73 20 (100%) 20 (100%)
Reported as n (%) unless indicated otherwise.
diabetes research and clinical practice
stage of severity was associated with age, diabetic duration, smoking, drinking, systolic pressure, 24-h urinary protein, HIF1A, VEGF, Flt-1, Flt-4, ABI, and serum creatinine. This suggests that these associated characteristics are risk factors for DPN. Notably, we found that the MI was the most important risk factor for DPN. The nailfold microcirculation test is a non-invasive and highly sensitive method for visualizing capillary circulation, and is widely used clinically in microvascular diseases such as systemic sclerosis [22]. As early as 1999, nailfold capillaroscopy was used for image analysis in diabetic patients, as a predictor of diabetic microvascular complications [23]. In the present study, we found that the nailfold microcirculation of diabetic patients was obviously abnormal relative to that of healthy individuals, in that the capillaries on the nailfold skin of the diabetic patients were more obscure, smaller, of variable length and disordered arrangement, malformed, and with slower blood flow. Diabetic patients, with or without DPN, had a higher MI compared with the healthy individuals. In addition, microcirculation damage correlated closely with the DPN stage of severity. These results indicate that the MI can potentially be used for advantageous early diagnostic screening of DPN. The VPT is well proved to indicate early neuropathy [24]. A prospective study of diabetes foot ulceration showed that VPT could diagnose with high sensitivity subclinical and mild DPN at the early stage [25]. Thus the VPT examination has been widely used in clinical screening for diabetic neuropathy [2]. In the present study, the VPT was used as the reference to investigate the effectiveness and feasibility of the MI for early screening of DPN. While highly consistent with the VPT, the MI also closely correlated with the risk factors for DPN, such as age, diabetic duration, serum creatinine, 24-h urinary protein, HIF1A, VEGF, Flt-1, Flt-4, ABI, and DPN severity. This indicates that the MI test, like the VPT, could potentially be used for the early diagnosis of DPN. We used the ROC curve to determine the diagnostic cutoff point and diagnostic efficiency of the MI and VPT. In our study, when the VPT was 12.73 V, the corresponding Youden’s index was at its maximum and the area under the ROC curve was 0.927. Therefore, this VPT had good diagnostic efficiency. In the present study, when the MI was 2.56, the corresponding Youden’s index was its maximum, and the area under the ROC curve was 0.943, indicating that at this point the diagnostic efficiency of MI was similar to that of the VPT. Furthermore, the MI indicated a diagnosis of DPN (all stages) at a rate that was comparable to that of the VPT for stage 4 (severe) DPN only, and the rate for stages 0 and 1 using the MI was even higher. Conventional diabetic neuropathy scoring systems include the nerve conduction velocity assay, TCSS, MNSI, and DNS [2]. These are only half-quantitative assessments of DPN severity, and are highly dependent on the patient’s subjective feelings. However, many patients cannot accurately describe their symptoms, and subjective factors can also compromise standard reportage of physical signs, making it difficult to evaluate the severity of DPN accurately. The VPT examination evaluates the function of myelinated nerve fibers, but not the function at location, and relies too much on subjective judgment. However, the MI test utilizes a capillaroscope to
1 3 1 ( 2 0 1 7 ) 4 2 –4 8
47
observe, directly and dynamically, the nailfold microcapillaries and blood flow, with more objectivity and reliability for earl diagnostic screening of DPN. Further advantages of the nailfold MI test for early screening of DPN include safety, non-invasiveness, convenience, fast performance, visually detailed information, and easy acceptance by patients. The early onset of microvascular lesions and peripheral neuropathy in diabetic patients is occult, without obvious clinical symptoms. The nailfold MI can non-invasively detect the capillary anomalies in patients, which is helpful for early diagnosis of microvascular lesions on other areas of body. DPN in the early stages mainly begins in the nerve terminal and small nerve fibers, resulting in pain and temperature hypoesthesia [26]. When sympathetic nerve dysfunction is caused by local hyperemia and large myelinated nerve fiber is damaged, the VPT can be used to detect neuropathy [27]. Thus, nailfold MI diagnostic screening is very helpful for early discovery and treatment of diabetic patients. There were some limitations in our study. This study was a cross-sectional study and the sample size was small. In addition, long-term follow-up observation is lacking. A prospective study with a large sample size is needed to investigate an association between the MI and DPN progression, and confirm the advantages of the MI for DPN screening.
Authors’ contributions Guotao Hu tested the nailfold MI and VPT, performed the statistical analysis, and drafted the manuscript. Fanglong Zhai participated in the design of the study. Feifei MO, Li He, and Weiya Shen collected the general clinical data. Hailan Wang conceived the study, and participated in its design and coordination and helped to draft the manuscript. All authors read and approved the final manuscript.
Conflict of interest The authors declare that there are no conflicts of interest associated with this article.
Acknowledgments This work was supported by the Science and Technology Development Fund of Longgang District, Shenzhen (No. 201406073001004).
Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.diabres. 2017.06.017.
R E F E R E N C E S
[1] Gibbons CH, Freeman R, Veves A. Diabetic neuropathy: a cross-sectional study of the relationships among tests of neurophysiology. Diabetes Care 2010;33:2629–34.
48
diabetes research and clinical practice
[2] Gordois A, Scuffham P, Shearer A, Oglesby A, Tobian JA. The health care costs of diabetic peripheral neuropathy in the US. Diabetes Care 2003;26:1790–5. [3] Juster-Switlyk K, Smith AG. Updates in diabetic peripheral neuropathy. F1000Res 2016; 5. [4] Taylor-Stokes G, Pike J, Sadosky A, Chandran A, Toelle T. Association of patient-rated severity with other outcomes in patients with painful diabetic peripheral neuropathy. Diabetes Metab Syndr Obes 2011;4:401–8. [5] Perkins BA, Olaleye D, Zinman B, Bril V. Simple screening tests for peripheral neuropathy in the diabetes clinic. Diabetes Care 2001;24:250–6. [6] Bril V, Perkins BA. Validation of the toronto clinical scoring system for diabetic polyneuropathy. Diabetes Care 2002;25:2048–52. [7] Feldman EL, Stevens MJ, Thomas PK, Brown MB, Canal N, Greene DA. A practical two-step quantitative clinical and electrophysiological assessment for the diagnosis and staging of diabetic neuropathy. Diabetes Care 1994;17:1281–9. [8] Mete T, Aydin Y, Saka M, Cinar Yavuz H, Bilen S, Yalcin Y, et al. Comparison of efficiencies of michigan neuropathy screening instrument, neurothesiometer, and electromyography for diagnosis of diabetic neuropathy. Int J Endocrinol 2013;2013:821745. [9] Rathur HM, Boulton AJ. The diabetic foot. Clin Dermatol 2007;25:109–20. [10] Peng L, Liu W, Zhai F, He L, Wang H. Microvessel permeability correlates with diabetic peripheral neuropathy in early stage of streptozotocin-induced diabetes rats. J Diabetes Complications 2015;29:865–71. [11] Boulton AJ. The diabetic foot: from art to science. The 18th Camillo Golgi lecture. Diabetologia 2004;47:1343–53. [12] Tomesova J, Gruberova J, Lacigova S, Cechurova D, Jankovec Z, Rusavy Z. Differences in skin microcirculation on the upper and lower extremities in patients with diabetes mellitus: relationship of diabetic neuropathy and skin microcirculation. Diabetes Technol Ther 2013;15:968–75. [13] Deng W, Dong X, Zhang Y, Jiang Y, Lu D, Wu Q, et al. Transcutaneous oxygen pressure (TcPO(2)): a novel diagnostic tool for peripheral neuropathy in type 2 diabetes patients. Diabetes Res Clin Pract 2014;105:336–43. [14] Stirban A. Microvascular dysfunction in the context of diabetic neuropathy. Curr Diab Rep 2014;14:541.
1 3 1 ( 2 0 1 7 ) 4 2 –4 8
[15] Wang YQ, Ge QG, Wang J, Niu JH, Huang C, Zhao YY. The WHO near miss criteria are appropriate for admission of critically ill pregnant women to intensive care units in China. Chin Med J (Engl) 2013;126:895–8. [16] American Diabetes A. Peripheral arterial disease in people with diabetes. Diabetes Care 2003;26:3333–41. [17] Dyck PJ. Detection, characterization, and staging of polyneuropathy: assessed in diabetics. Muscle Nerve 1988;11:21–32. [18] Kastenbauer T, Sauseng S, Sokol G, Auinger M, Irsigler K. A prospective study of predictors for foot ulceration in type 2 diabetes. J Am Podiatr Med Assoc 2001;91:343–50. [19] Curcio F, Ceriello A. Decreased cultured endothelial cell proliferation in high glucose medium is reversed by antioxidants: new insights on the pathophysiological mechanisms of diabetic vascular complications. In Vitro Cell Dev Biol 1992; 28a: 787–90. [20] Ashok S, Ramu M, Deepa R, Mohan V. Prevalence of neuropathy in type 2 diabetic patients attending a diabetes centre in South India. J Assoc Physicians India 2002;50:546–50. [21] Feldman EL, Russell JW, Sullivan KA, Golovoy D. New insights into the pathogenesis of diabetic neuropathy. Curr Opin Neurol 1999;12:553–63. [22] Standards of medical care in diabetes–2014. Diabetes Care 2014; 37 Suppl 1: S14–80. [23] Sebastiani M, Manfredi A, Cassone G, Giuggioli D, Ghizzoni C, Ferri C. Measuring microangiopathy abnormalities in systemic sclerosis patients: the role of capillaroscopy-based scoring models. Am J Med Sci 2014;348:331–6. [24] Hu Q, Mahler F. New system for image analysis in nailfold capillaroscopy. Microcirculation 1999;6:227–35. [25] Garrow AP, Boulton AJ. Vibration perception threshold–a valuable assessment of neural dysfunction in people with diabetes. Diabetes Metab Res Rev 2006;22:411–9. [26] Jensen TS, Bach FW, Kastrup J, Dejgaard A, Brennum J. Vibratory and thermal thresholds in diabetics with and without clinical neuropathy. Acta Neurol Scand 1991;84:326–33. [27] Shy ME, Frohman EM, So YT, Arezzo JC, Cornblath DR, Giuliani MJ, et al. Quantitative sensory testing: report of the therapeutics and technology assessment subcommittee of the American academy of neurology. Neurology 2003;60:898–904.
1 3 1 ( 2 0 1 7 ) 4 2 –4 8
Contents available at ScienceDirect
Diabetes Research and Clinical Practice journa l home page: www.e lse vier.com/locate/diabres
Effectiveness and feasibility of nailfold microcirculation test to screen for diabetic peripheral neuropathy Guotao Hu a, Fanglong Zhai a, Feifei Mo a, Li He b, Weiya Shen a, Hailan Wang a,* a b
Department of Endocrinology, Zunyi Medical College Affiliated Shenzhen Longgang Central Hospital, Shenzhen 518116, Guangdong, China Department of Nursing, Zunyi Medical College Affiliated Shenzhen Longgang Central Hospital, Shenzhen 518116, Guangdong, China
A R T I C L E I N F O
A B S T R A C T
Article history:
Aims: The nailfold microcirculation index (MI) is a non-invasive, objective, and highly sen-
Received 25 February 2017
sitive blood capillary detection method. This study evaluated the diagnostic efficiency of
Received in revised form
the nailfold MI relative to the more subjective vibration perception threshold (VPT) exam-
19 May 2017
ination for early diagnostic screening of diabetic peripheral neuropathy (DPN).
Accepted 9 June 2017
Methods: From February 2015 to February 2016, 227 diabetes mellitus patients and 39
Available online 13 June 2017
healthy individuals were enrolled. Each subject underwent the MI test and the VPT examination.
Keywords: Diabetes mellitus Diabetic peripheral neuropathy Nailfold microcirculation index Vibration perception threshold Neuropathy scoring system
Results: MI was more closely associated with DPN than age, diabetic duration, smoking, drinking, systolic pressure, serum creatinine, 24-h urinary protein, hypoxia-inducible factor-1a (HIF1A), vascular endothelial growth factor (VEGF), the VEGF receptors Flt-1 and Flt-4, ankle branchial index (ABI), DPN, or DPN stage (OR = 11.819). Both the MI and VPT closely correlated with age, diabetic duration, serum creatinine, 24-h urinary protein, HIF1A, VEGF, Flt-1, Flt-4, ABI, DPN, and DPN stage. By the receiver operating characteristic (ROC) curve, the MI diagnostic cutoff point was 2.56, where the corresponding Youden’s index was maximum and the area under ROC curve was 0.943. The diagnostic efficiency of MI and VPT were similar. MI and VPT indicated similar percentages of diabetic patients with DPN at the most severe stage, while MI achieved a higher diagnostic rate for the earliest stages. Conclusions: The nailfold MI is a feasible method for clinical early diagnostic screening of DPN in diabetic patients, and is more objective and reliable than VPT. Ó 2017 Elsevier B.V. All rights reserved.
1.
Introduction
Diabetic peripheral neuropathy (DPN) as a symmetrical sensorimotor multiple neuropathy is attributed to chronic hyperglycemia and the metabolic and microvascular changes due to cardiovascular risk factors [1]. DPN is the most common complication of diabetes mellitus, with a prevalence between * Corresponding author. E-mail address: [email protected] (H. Wang). http://dx.doi.org/10.1016/j.diabres.2017.06.017 0168-8227/Ó 2017 Elsevier B.V. All rights reserved.
17% and 66%, depending on diagnostic criteria and demography [2]. DPN involves the sensory, autonomic, and motor nerves, resulting in impaired sensation, autonomic symptoms, and movement disorders [3]. DPN is difficult to diagnose at onset. Many diabetic patients remain asymptomatic during the long incubation period, and symptoms when they occur are often not related
diabetes research and clinical practice
to damage severity. When first detected clinically, demyelination of the peripheral nerves is generally irreversible, as are other pathological changes [1]. Thus, early diagnosis of DPN is important. The nerve conduction velocity assay is considered the gold standard for diagnosing DPN [4], but can efficiently detect only lesions of large myelinated nerve fibers, and is not sensitive to the neuropathy of small nerve fibers. Thus, patients with early diabetes without obvious clinical symptoms may have small nerve fiber damage that is not detected by the nerve conduction velocity assay. In addition, the assay is expensive and time-consuming, and therefore is not suitable for wide diagnostic screening of DPN. Other diagnostic screening methods for DPN include the Toronto Clinical Scoring System (TCSS) [5], Michigan Neuropathy Screening Instrument (MNSI) [6], and Diabetic Neuropathy Symptom Score (DNS) [2]. However, the TCSS, MNSI, and DNS methods for diagnostic screening of DPN rely on subjective judgments, and are further limited by poor replicability [7]. The vibration perception threshold (VPT) examination evaluates the function of myelinated nerve fibers according to the patient’s awareness of vibration sensation. It is an effective method for screening DPN, as well as evaluating the risk of foot ulcer in diabetes patients [8]. By enabling early detection for timely intervention and prevention, the diabetic patient’s quality of life is improved [9]. The VPT examination has been widely used in clinical screening in Europe and the United States and has been recommended for early diagnosis of DPN by the American Diabetes Association [2]. Although the definitive pathophysiology of DPN remains elusive, accumulating evidence indicates the source of a variety of diabetic complications is microvascular lesion [2]. Our previous study and others have shown that microcirculation dysfunction is closely association with DPN [2,10]. A microcirculation examination may have be able to assess the risk of DPN and screen early [11–13]. Nailfold capillaroscopy is commonly used to investigate skin microcirculation in the clinic [2]. It is noninvasive, simple, fast, and economical, and effectively identifies peripheral microvascular changes. In the present study, we evaluated the effectiveness and feasibility of the nailfold microcirculation test for diagnostic screening of DPN, comparing it with the VPT, which is widely used in clinical screening for diabetic neuropathy.
1 3 1 ( 2 0 1 7 ) 4 2 –4 8
43
Patients with non-diabetic neuropathy due to any cause were excluded. Such neuropathies included any of the following: pernicious anemia; vitamin B6 intoxication; vitamin B12 deficiency; alcoholism; uremia; chemical toxins; nerve entrapment; squeezing of the nerves due to benign causes; hepatitis; idiopathic neuropathy; various congenital sensorimotor neuropathy; paraneoplastic syndromes; syphilis; HIV/ AIDS; drugs such as chemotherapy or isoniazid; cervical and lumbar spondylosis including nerve root compression, stenosis, or cervical and lumbar degenerative diseases; cerebral infarction; Guillain-Barre Syndrome; or severe arteriovenous vascular lesions (e.g., venous thrombosis, lymphangitis, or arteriovenous fistula). The control group in this study consisted of 39 healthy individuals, with no history of any of the following: diabetes mellitus; hypertension; mental disorder; diseases of the coronary, nervous, hematologic, or cardiovascular and cerebrovascular systems; or autoimmune disease. For each individual in the control group, routine blood and urine examination, stool test, liver and kidney function, and blood lipids were within normal levels, and there were no abnormalities in their chest radiography, electrocardiogram, abdominal ultrasonography, fundus examination, or physical examination.
2.2.
General data collection
The following data for each patient with diabetes mellitus were collected: age, gender, diabetes duration, smoking, drinking, height, weight, blood pressure, body mass index (BMI), and ankle branchial index (ABI). The ABI was assessed using an ES-100V3 Doppler flow detector (HADECO, Japan). In accordance with the criteria of the American Diabetes Association, the normal ABI range was from 0.91 to 1.30 [15].
2.3.
Biochemical tests
The Ethics Committee at Shenzhen Longgang Center Hospital independently approved this study. All the patients enrolled in this study provided written informed consent to participate.
The following biochemical levels were tested: fasting bloodglucose (FBG); hemoglobin A1c (HbAlc); total cholesterol; triglyceride; high-density lipoprotein (HDL); low-density lipoprotein (LDL); serum creatinine; blood uric acid; hypoxia-inducible factor-1a (HIF1A); vascular endothelial growth factor (VEGF); the VEGF receptors Flt-1 and Flt-4; and 24-h total urinary protein. Immunoturbidimetry was used for testing 24-h total urinary protein. HbAlc was tested using the VARIANT II glycated hemoglobin test system (Bio-Rad, America). HIF1A, VEGF, Flt-1, and Flt-4 were tested using enzyme-linked immunosorbent assay (ELISA) kits (TSZ, USA). The other chemical tests were performed with a biochemical autoanalyzer (Olympus AU600, Olympus, Japan).
2.1.
2.4.
2.
Materials and methods
Inclusion and exclusion criteria
From February 2015 to February 2016, 227 outpatients and inpatients with diabetes mellitus were recruited at the Department of Endocrinology, Shenzhen Longgang Center Hospital. Type 2 diabetes mellitus was diagnosed according to the 2017 Standards of Medical Care in Diabetes of the American Diabetes Association [14]. Only patients with neuropathy due to type 2 diabetes mellitus were included.
DPN diagnosis
DPN was diagnosed based on the MNSI and neurological deficit score (NDS) for neuropathy [6]. The MNSI symptoms index was used to score the symptoms of patients, and the MNSI signs index and NDS index were used to score the signs of the patients. A diagnosis of DPN met the following criteria: MNSI symptoms index 4 and/or MNSI signs index 2; MNSI symptoms index 4 and NDS index 3; or NDS index 6.
44
diabetes research and clinical practice
The diabetic patients were divided into 4 groups based on Dyck’s 4 stages of DPN severity [16]: 0, no objective evidence of diabetic neuropathy; 1, no symptoms, but abnormal signs; 2, with symptoms, signs, and ankle dorsiflexion; and 3, disabling neuropathy, diabetic foot ulcer, and others.
2.5. VPT examination examination
and
nailfold
microcirculation
VPT examination was performed using a Bio-Thesiometer quantitative sensory tester (Beijing Dimeider Technology, China). The nailfold microcirculation examination was performed using a ZL102 nailfold capillaroscope (Xuzhou Medical Instrument, China). Briefly, the subject was seated in a quiet environment at 20–24 °C. Two droplets of cedar oil were dropped on the nailfold skin of the third finger on the left or right hand. Then the nailfold microcirculation was evaluated by nailfold capillaroscopy, as in a previous report [17]. Nailfold microcirculation was classified as one of 4 grades as follows: normal, index < 1; almost normal, 1 index 2; light abnormality, 2 < index < 4; median abnormality, 4 index < 8; and severe abnormality, 8 index.
2.6.
Receiver operating characteristic (ROC) curve
We used the ROC curve [18] to determine the diagnostic cutoff point and diagnostic efficiency of the microcirculation index (MI) and VPT. Based on the shift of the cutoff point in the ROC curve, multiple paired sensitivity and false positive rates (1 – specificity) could be obtained. Youden’s index was obtained by an equation (i.e., Youden’s index = sensitivity + specificity – 1), with the maximum Youden’s index considered the optimum diagnostic cutoff point. At any particular cutoff point, if the area under the ROC curve was 0.7, 0.7– 0.9, or >0.9, the diagnostic value was defined as low, medium, and high, respectively.
2.7.
Statistical analysis
Statistical analyses were performed using SPSS software (Version 21.0; SPSS, Chicago, IL, USA). Measurement data are shown as mean ± standard deviation, and count data are shown as a percentage. The DPN severities were divided into 4 stages according to Dyck’s score [13]. In each stage, measurement data met the standard normal distribution, but count data (gender, smoking, and drinking) did not meet the standard normal distribution. Qualitative data was analyzed using the chi-squared test. Comparisons between the 2 groups were analyzed using the independent sample Student’s t-test. Comparisons among multiple groups were performed using one-way analysis of variance and the least significant difference. Correlations of numerical and qualitative data were analyzed, respectively, using Pearson’s and Spearman’s correlation analyses. The determination of the risk factors of DPN was performed using univariate analysis by non-conditional logistic regression analysis. The ROC curve was used to determine the diagnostic point and efficiency of the MI and VPT. A P-value < 0.05 was considered statistically significant.
1 3 1 ( 2 0 1 7 ) 4 2 –4 8
3.
Results
3.1.
General features of patients
The study population comprised 227 diabetes mellitus patients (140 men, 87 women; aged 50.93 ± 12.16 y) and a control group of 39 healthy individuals (24 men; 15 women; aged 55.50 ± 7.21 y; Table 1). The patient and control groups were comparable in age and gender ratio (P = 0.756 and P = 0.363, respectively). The nailfold MI of the left and right hands of the 2 groups were statistically similar, as was the VPT of the left and right hands (P > 0.05; Supplementary Tables 1 and 2). We used the mean nailfold MI and VPT in this study. The mean nailfold MI and VPT of the diabetic patients (3.64 ± 1.91 and 15.27 ± 9.62, respectively) were each significantly higher than that of the control group (0.57 ± 0.17 and 10.40 ± 2.49; P < 0.001). Of the 227 patients, 133 (58.59%) received a diagnosis of DPN (including stages 1–3; Table 1). The patients were further grouped by stage of DPN (0, 1, 2, or 3). With each stage of higher number; the following were significantly higher: age; diabetic duration; systolic pressure; 24-h urinary protein; HIF1A; VEGF; Flt-1; Flt-4; serum creatinine; VPT; and nailfold MI (P < 0.05, each). Also with each stage, the ABI was significantly lower (P < 0.001).
3.2.
Nailfold MI as risk factor for DPN
For the univariate unconditional logistic regression analysis of the risk factors of DPN, DPN was considered the dependent variable (Table 2). The following characteristics were associated with DPN, indicating that they were potential risk factors for DPN (P < 0.05): age; diabetic duration; smoking; drinking; systolic pressure; 24-h urinary protein; HIF1A; VEGF; Flt-1; Flt-4; serum creatinine; VPT; and nailfold MI. In these diabetic patients, the nailfold MI was likely the main risk factor for DPN (odds ratio [OR] = 11.819).
3.3.
Correlations between nailfold MI and DPN risk factors
The VPT is widely used in clinical screening for diabetic neuropathy. We investigated correlations between the VPT and the DPN risk factors by correlation analysis (Table 3). VPT correlated linearly with each of the following: age; diabetic duration; serum creatinine; 24-h urinary protein; HIF1A; VEGF; Flt1; Flt-4; ABI; MI; DPN; and DPN stage (P < 0.05). VPT correlated negatively with ABI, and positively with other DPN risk factors. In particular, the VPT correlated closely with the MI (R = 0.757) and DPN stage (R = 0.829). Highly consistent with the VPT, the MI correlated positively with the following (Table 3): age, diabetic duration, serum creatinine, 24-h urinary protein, HIF1A, VEGF, Flt-1, Flt-4, ABI, MI, DPN, and DPN stage. The MI also correlated negatively with ABI and positively with other DPN risk factors. MI correlated closely with VPT (R = 0.757) and DPN stage (R = 0.840).
3.4.
Correlation between MI and DPN severity
Nailfold microcirculation was classified with reference to the grade index as follows: normal, <1; almost normal, 1 or 2; light
diabetes research and clinical practice
45
1 3 1 ( 2 0 1 7 ) 4 2 –4 8
Table 1 – General characteristics of diabetic patients by DPN stage of severity.
Subjects, n Male, % Age, y Duration, y BMI, kg/m2 Systolic pressure, mmHg Diastolic pressure, mmHg FBG, mmol/L HbA1c, % TC, mmol/L TG, mmol/L HDL-C, mmol/L LDL-C, mmol/L Serum creatinine, mmol/L Serum uric acid, mmol/L 24-h urinary protein, mg HIF1A, ng/L VEGF, pg/mL Flt-1, pg/mL F1t-4, pg/mL ABI VPT, V MI Smoking, % Drinking, % a
0
1
2
3
P value
94 63 (67.02%) 45.01 ± 10.53 1.74 ± 2.90 24.04 ± 3.61 127.97 ± 19.51 78.86 ± 11.76 12.09 ± 13.83 11.34 ± 2.68 5.05 ± 1.73 2.77 ± 6.24 1.11 ± 0.37 3.21 ± 1.13 72.76 ± 24.41 320.47 ± 107.51 17.90 ± 31.99 122.5 ± 5.24 771.67 ± 48.57 571.04 ± 34.69 1066.75 ± 95.99 1.18 ± 0.041 8.29 ± 3.68 2.21 ± 0.59 36 (38.30%) 24 (25.53%)
48 34 (70.83%) 50.48 ± 12.03 5.51 ± 3.83 23.57 ± 3.64 137.56 ± 22.82 81.02 ± 13.83 10.55 ± 3.45 11.16 ± 2.31 5.21 ± 1.64 2.39 ± 3.60 1.15 ± 0.33 3.28 ± 1.30 69.66 ± 26.15 285.17 ± 86.13 36.44 ± 65.47 144.19 ± 11.19 968.04 ± 58.12 685.02 ± 74.90 1505.20 ± 165.43 1.12 ± 0.05 13.16 ± 3.27 3.27 ± 0.52 16 (33.33%) 7 (14.58%)
65 33 (50.77%) 57.14 ± 10.68 10.54 ± 6.29 23.39 ± 2.99 132.78 ± 21.82 75.66 ± 13.00 10.71 ± 3.63 10.76 ± 2.49 4.96 ± 1.62 4.86 ± 22.65 1.11 ± 0.41 3.25 ± 1.07 91.29 ± 101.49 322.71 ± 112.10 123.80 ± 345.98 162.49 ± 19.46 1190.19 ± 72.03 812.36 ± 90.2 2980.00 ± 3723.89 1.08 ± 0.10 21.05 ± 7.35 4.97 ± 1.72 9 (13.85%) 2 (3.08%)
20 10 (50.00%) 59.65 ± 9.33 11.9 ± 7.17 22.94 ± 1.82 142.85 ± 22.04 75.10 ± 13.97 10.54 ± 2.69 10.29 ± 2.34 4.69 ± 1.22 1.43 ± 0.61 1.34 ± 0.26 3.01 ± 1.04 126.21 ± 168.67 346.10 ± 90.06 258.90 ± 433.33 192.30 ± 3.49 901.33 ± 165.29 658.55 ± 108.22 1139.44 ± 138.51 1.06 ± 0.57 34.37 ± 8.73 6.95 ± 1.81 3 (15.00%) 3 (15.00%)
– – <0.001a <0.001a 0.454 0.009a 0.1 0.714 0.258 0.678 0.628 0.071 0.846 0.017a 0.096 <0.001a <0.001a <0.001a <0.001a <0.001a <0.001a <0.001a <0.001a – –
Comparison among DPN groups.
Table 2 – Risk factor analysis for DPN by logistic analysis.
Duration of disease Age Smoking Drinking Systolic pressure HIF1A VEGF Flt-1 F1t-4 24-h urinary protein VPT MI
OR (95% CI)
P value
1.444 (1.306, 1.596) 1.085 (1.055, 1.116) 2.328 (1.292, 4.194) 0.236 (0.110, 0.507) 1.018 (1.005, 1.032) 1.748 (1.411, 2.165) 1.031 (1.022, 1.041) 1.042 (1.030, 1.057) 1.013 (1.010, 1.017) 1.014 (1.005, 1.023) 1.544 (1.372, 1.738) 11.819 (6.148, 22.72)
<0.001 <0.001 0.005 <0.001 0.006 <0.001 <0.001 <0.001 <0.001 0.002 <0.001 <0.001
CI, confidence interval; OR, odds ratio.
abnormality, 2 or 3; median abnormality, 4–8; and severe abnormality, 8 and higher. DPN patients were divided into 0, 1-, 2-, or 3-stage groups. A nonparametric correlation analysis was performed to analyze a correlation between the degree of microcirculation abnormality and DPN severity. A highly positive correlation was determined between microcirculation abnormality and DPN severity (R = 0.763), similar to the correlation between VPT and DPN stage.
3.5. Efficiency of nailfold microcirculation test for DPN diagnosis With the diagnostic reference of the DPN score, the ROC curve was plotted to analyze the cutoff points of MI and VPT for DPN diagnosis (Fig. 1; Table 4). When the diagnostic cutoff
Table 3 – Nailfold MI and VPT correlations with DPN risk factors. MI
Disease duration Age HIF1A VEGF Flt-1 F1t-4 24-h urinary protein ABI MI VPT DPN DPN stage Serum creatinine a b
VPT
R
P value
R
P value
0.656a 0.420a 0.697a 0.490a 0.464a 0.259 0.391 0.510a – 0.757a 0.681b 0.840b 0.341a
<0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 – <0.001 <0.001 <0.001 0.090
0.556a 0.422a 0.698a 0.451a 0.432a 0.166 0.230a 0.459a 0.757a – 0.729b 0.829b 0.251a
<0.001 <0.001 <0.001 <0.001 <0.001 0.012 0.004 <0.001 <0.001 – <0.001 <0.001 0.001
Pearson’s correlation analysis. Spearman’s correlation analysis.
point of MI was 2.56, its corresponding Youden index was at the maximum and the corresponding area under curve was 0.943. This indicated that the MI had good diagnostic efficiency for DPN. With a VPT diagnostic cutoff point of 12.73 V, the corresponding Youden index was at the maximum and the corresponding area under curve was 0.927, indicating good diagnostic efficiency for DPN. After we found the diagnostic cutoff point of MI, we further analyzed the efficiency of MI for diagnosing DPN, compared with the DPN score system and the VPT (Supplementary Table 3). According to the diagnosis criteria
46
diabetes research and clinical practice
1 3 1 ( 2 0 1 7 ) 4 2 –4 8
To determine the effectiveness and feasibility of MI for diagnostic screening of DPN, we compared the rates of diagnosis of MI and VPT according to the DPN stage of severity (Table 5). In the 39 healthy individuals, the rate of diagnosis of DPN according to the MI and VPT cutoff points was 0.00%. For DPN patients at stage 3, the rate of diagnosis of DPN according to the cutoff point for MI was similar to that of the VPT. However, the diagnostic rate when using MI at stages 0–2 was higher than when the VPT was applied.
4.
Fig. 1 – ROC curve to investigate the cutoff points of DPN diagnosis for MI and VPT.
Table 4 – Sensitivity, specificity, and Youden index of the diagnostic cutoff points for MI and VPT.* Cutoff point Sensitivity Specificity Youden index MI
2.43 2.48 2.56 2.70 2.78 VPT 12.53 12.63 12.73 12.88 13.05 *
0.932 0.925 0.925 0.917 0.910 0.812 0.805 0.805 0.789 0.774
0.830 0.851 0.862 0.862 0.872 0.904 0.904 0.915 0.926 0.926
0.762 0.776 0.787 0.779 0.782 0.716 0.709 0.720 0.715 0.700
Determined by ROC curve and area under the ROC curve.
of the DPN score system, 133 patients had DPN among the 227 diabetes mellitus patients (58.59%). Using the diagnostic cutoff points described above, that is, MI > 2.56 and VPT > 12.73 V, the DPN was diagnosed in 59.90% and 50.70% of the patients, respectively. Combining the MI and VPT, DPN was diagnosed in 66.5% of the patients. According to the chi-squared test, the rate of diagnosis of DPN using the MI was higher than when using either the neuropathy score or the VPT. The rate of diagnosis using MI and VPT (combined) was higher than that of the other diagnostic methods applied singly.
Discussion
To develop a rapid, effective, and economical screening method for the early diagnosis of DPN, we investigated the effectiveness and feasibility of the MI for diagnostic screening. We compared the diagnostic rate of the MI with that of the widely used VPT for a subject population of 227 patients with diabetes mellitus and 39 healthy individuals. The prevalence of DPN in our patients was 58.59%, similar to the rate published in a study in 2011 [2]. We found that a high MI was a factor that was closely associated with DPN (OR = 11.819), in addition to age, diabetic duration, smoking, drinking, systolic pressure, and other risk factors. Consistent with the results of the VPT, the MI closely correlated positively with age, diabetic duration, serum creatinine, 24-h urinary protein, HIF1A, VEGF, Flt-1, Flt-4, ABI, DPN, and DPN stage. In addition, the MI closely correlated positively with the DPN stage of severity. According to the ROC curve, we found that an MI diagnostic cutoff point of 2.56 resulted in good diagnostic efficiency for DPN. In this study, we also validated the effectiveness and feasibility of MI for DPN diagnosis in patients at each stage of DPN severity. Numerous risk factors are associated with DPN onset and progression. Ashok et al. [19] showed that age and the duration of diabetes mellitus were associated with DPN. Another study found that the prevalence of DPN was about 10% at the onset of diabetes mellitus, but 25 years later had increased up to 50% [20]. The ABI is the ratio of the systolic pressures of the ankle artery and brachial artery. The American Diabetes Association has recommended that the ABI should be routinely used for screening for peripheral arterial disease in patients with diabetes mellitus who are older than 50 years [21]. In addition, studies have shown that HIF1A, VEGF, Flt-1, and Flt-4 are closely associated with the progression of diabetic vascular complications [2]. In the present study, we found that the DPN
Table 5 – Diagnostic rates determined by MI and VPT of the control group and by DPN stage of severity in diabetic patients.a
Subjects, n MI >2 >2.56 VPT, V >15 >12.73 a
Control group
0
1
2
3
39 0.57 ± 0.17 0 0 10.40 ± 2.49 0 0
94 2.21 ± 0.59 54 (57.45%) 2 (12.77%) 8.29 ± 3.68 3 (3.19%) 5 (5.32%)
48 3.27 ± 0.52 47 (97.92%) 43 (89.58%) 13.16 ± 3.27 17 (35.42%) 30 (62.50%)
65 4.97 ± 1.72 64 (98.46%) 60 (92.31%) 21.05 ± 7.35 50 (76.92%) 47 (87.69%)
20 6.95 ± 1.81 20 (100%) 20 (100%) 34.37 ± 8.73 20 (100%) 20 (100%)
Reported as n (%) unless indicated otherwise.
diabetes research and clinical practice
stage of severity was associated with age, diabetic duration, smoking, drinking, systolic pressure, 24-h urinary protein, HIF1A, VEGF, Flt-1, Flt-4, ABI, and serum creatinine. This suggests that these associated characteristics are risk factors for DPN. Notably, we found that the MI was the most important risk factor for DPN. The nailfold microcirculation test is a non-invasive and highly sensitive method for visualizing capillary circulation, and is widely used clinically in microvascular diseases such as systemic sclerosis [22]. As early as 1999, nailfold capillaroscopy was used for image analysis in diabetic patients, as a predictor of diabetic microvascular complications [23]. In the present study, we found that the nailfold microcirculation of diabetic patients was obviously abnormal relative to that of healthy individuals, in that the capillaries on the nailfold skin of the diabetic patients were more obscure, smaller, of variable length and disordered arrangement, malformed, and with slower blood flow. Diabetic patients, with or without DPN, had a higher MI compared with the healthy individuals. In addition, microcirculation damage correlated closely with the DPN stage of severity. These results indicate that the MI can potentially be used for advantageous early diagnostic screening of DPN. The VPT is well proved to indicate early neuropathy [24]. A prospective study of diabetes foot ulceration showed that VPT could diagnose with high sensitivity subclinical and mild DPN at the early stage [25]. Thus the VPT examination has been widely used in clinical screening for diabetic neuropathy [2]. In the present study, the VPT was used as the reference to investigate the effectiveness and feasibility of the MI for early screening of DPN. While highly consistent with the VPT, the MI also closely correlated with the risk factors for DPN, such as age, diabetic duration, serum creatinine, 24-h urinary protein, HIF1A, VEGF, Flt-1, Flt-4, ABI, and DPN severity. This indicates that the MI test, like the VPT, could potentially be used for the early diagnosis of DPN. We used the ROC curve to determine the diagnostic cutoff point and diagnostic efficiency of the MI and VPT. In our study, when the VPT was 12.73 V, the corresponding Youden’s index was at its maximum and the area under the ROC curve was 0.927. Therefore, this VPT had good diagnostic efficiency. In the present study, when the MI was 2.56, the corresponding Youden’s index was its maximum, and the area under the ROC curve was 0.943, indicating that at this point the diagnostic efficiency of MI was similar to that of the VPT. Furthermore, the MI indicated a diagnosis of DPN (all stages) at a rate that was comparable to that of the VPT for stage 4 (severe) DPN only, and the rate for stages 0 and 1 using the MI was even higher. Conventional diabetic neuropathy scoring systems include the nerve conduction velocity assay, TCSS, MNSI, and DNS [2]. These are only half-quantitative assessments of DPN severity, and are highly dependent on the patient’s subjective feelings. However, many patients cannot accurately describe their symptoms, and subjective factors can also compromise standard reportage of physical signs, making it difficult to evaluate the severity of DPN accurately. The VPT examination evaluates the function of myelinated nerve fibers, but not the function at location, and relies too much on subjective judgment. However, the MI test utilizes a capillaroscope to
1 3 1 ( 2 0 1 7 ) 4 2 –4 8
47
observe, directly and dynamically, the nailfold microcapillaries and blood flow, with more objectivity and reliability for earl diagnostic screening of DPN. Further advantages of the nailfold MI test for early screening of DPN include safety, non-invasiveness, convenience, fast performance, visually detailed information, and easy acceptance by patients. The early onset of microvascular lesions and peripheral neuropathy in diabetic patients is occult, without obvious clinical symptoms. The nailfold MI can non-invasively detect the capillary anomalies in patients, which is helpful for early diagnosis of microvascular lesions on other areas of body. DPN in the early stages mainly begins in the nerve terminal and small nerve fibers, resulting in pain and temperature hypoesthesia [26]. When sympathetic nerve dysfunction is caused by local hyperemia and large myelinated nerve fiber is damaged, the VPT can be used to detect neuropathy [27]. Thus, nailfold MI diagnostic screening is very helpful for early discovery and treatment of diabetic patients. There were some limitations in our study. This study was a cross-sectional study and the sample size was small. In addition, long-term follow-up observation is lacking. A prospective study with a large sample size is needed to investigate an association between the MI and DPN progression, and confirm the advantages of the MI for DPN screening.
Authors’ contributions Guotao Hu tested the nailfold MI and VPT, performed the statistical analysis, and drafted the manuscript. Fanglong Zhai participated in the design of the study. Feifei MO, Li He, and Weiya Shen collected the general clinical data. Hailan Wang conceived the study, and participated in its design and coordination and helped to draft the manuscript. All authors read and approved the final manuscript.
Conflict of interest The authors declare that there are no conflicts of interest associated with this article.
Acknowledgments This work was supported by the Science and Technology Development Fund of Longgang District, Shenzhen (No. 201406073001004).
Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.diabres. 2017.06.017.
R E F E R E N C E S
[1] Gibbons CH, Freeman R, Veves A. Diabetic neuropathy: a cross-sectional study of the relationships among tests of neurophysiology. Diabetes Care 2010;33:2629–34.
48
diabetes research and clinical practice
[2] Gordois A, Scuffham P, Shearer A, Oglesby A, Tobian JA. The health care costs of diabetic peripheral neuropathy in the US. Diabetes Care 2003;26:1790–5. [3] Juster-Switlyk K, Smith AG. Updates in diabetic peripheral neuropathy. F1000Res 2016; 5. [4] Taylor-Stokes G, Pike J, Sadosky A, Chandran A, Toelle T. Association of patient-rated severity with other outcomes in patients with painful diabetic peripheral neuropathy. Diabetes Metab Syndr Obes 2011;4:401–8. [5] Perkins BA, Olaleye D, Zinman B, Bril V. Simple screening tests for peripheral neuropathy in the diabetes clinic. Diabetes Care 2001;24:250–6. [6] Bril V, Perkins BA. Validation of the toronto clinical scoring system for diabetic polyneuropathy. Diabetes Care 2002;25:2048–52. [7] Feldman EL, Stevens MJ, Thomas PK, Brown MB, Canal N, Greene DA. A practical two-step quantitative clinical and electrophysiological assessment for the diagnosis and staging of diabetic neuropathy. Diabetes Care 1994;17:1281–9. [8] Mete T, Aydin Y, Saka M, Cinar Yavuz H, Bilen S, Yalcin Y, et al. Comparison of efficiencies of michigan neuropathy screening instrument, neurothesiometer, and electromyography for diagnosis of diabetic neuropathy. Int J Endocrinol 2013;2013:821745. [9] Rathur HM, Boulton AJ. The diabetic foot. Clin Dermatol 2007;25:109–20. [10] Peng L, Liu W, Zhai F, He L, Wang H. Microvessel permeability correlates with diabetic peripheral neuropathy in early stage of streptozotocin-induced diabetes rats. J Diabetes Complications 2015;29:865–71. [11] Boulton AJ. The diabetic foot: from art to science. The 18th Camillo Golgi lecture. Diabetologia 2004;47:1343–53. [12] Tomesova J, Gruberova J, Lacigova S, Cechurova D, Jankovec Z, Rusavy Z. Differences in skin microcirculation on the upper and lower extremities in patients with diabetes mellitus: relationship of diabetic neuropathy and skin microcirculation. Diabetes Technol Ther 2013;15:968–75. [13] Deng W, Dong X, Zhang Y, Jiang Y, Lu D, Wu Q, et al. Transcutaneous oxygen pressure (TcPO(2)): a novel diagnostic tool for peripheral neuropathy in type 2 diabetes patients. Diabetes Res Clin Pract 2014;105:336–43. [14] Stirban A. Microvascular dysfunction in the context of diabetic neuropathy. Curr Diab Rep 2014;14:541.
1 3 1 ( 2 0 1 7 ) 4 2 –4 8
[15] Wang YQ, Ge QG, Wang J, Niu JH, Huang C, Zhao YY. The WHO near miss criteria are appropriate for admission of critically ill pregnant women to intensive care units in China. Chin Med J (Engl) 2013;126:895–8. [16] American Diabetes A. Peripheral arterial disease in people with diabetes. Diabetes Care 2003;26:3333–41. [17] Dyck PJ. Detection, characterization, and staging of polyneuropathy: assessed in diabetics. Muscle Nerve 1988;11:21–32. [18] Kastenbauer T, Sauseng S, Sokol G, Auinger M, Irsigler K. A prospective study of predictors for foot ulceration in type 2 diabetes. J Am Podiatr Med Assoc 2001;91:343–50. [19] Curcio F, Ceriello A. Decreased cultured endothelial cell proliferation in high glucose medium is reversed by antioxidants: new insights on the pathophysiological mechanisms of diabetic vascular complications. In Vitro Cell Dev Biol 1992; 28a: 787–90. [20] Ashok S, Ramu M, Deepa R, Mohan V. Prevalence of neuropathy in type 2 diabetic patients attending a diabetes centre in South India. J Assoc Physicians India 2002;50:546–50. [21] Feldman EL, Russell JW, Sullivan KA, Golovoy D. New insights into the pathogenesis of diabetic neuropathy. Curr Opin Neurol 1999;12:553–63. [22] Standards of medical care in diabetes–2014. Diabetes Care 2014; 37 Suppl 1: S14–80. [23] Sebastiani M, Manfredi A, Cassone G, Giuggioli D, Ghizzoni C, Ferri C. Measuring microangiopathy abnormalities in systemic sclerosis patients: the role of capillaroscopy-based scoring models. Am J Med Sci 2014;348:331–6. [24] Hu Q, Mahler F. New system for image analysis in nailfold capillaroscopy. Microcirculation 1999;6:227–35. [25] Garrow AP, Boulton AJ. Vibration perception threshold–a valuable assessment of neural dysfunction in people with diabetes. Diabetes Metab Res Rev 2006;22:411–9. [26] Jensen TS, Bach FW, Kastrup J, Dejgaard A, Brennum J. Vibratory and thermal thresholds in diabetics with and without clinical neuropathy. Acta Neurol Scand 1991;84:326–33. [27] Shy ME, Frohman EM, So YT, Arezzo JC, Cornblath DR, Giuliani MJ, et al. Quantitative sensory testing: report of the therapeutics and technology assessment subcommittee of the American academy of neurology. Neurology 2003;60:898–904.