目的：探讨神经肌电图（electromyography，EMG）联合体感诱发电位（somatosensory evoked potential，SEP）在糖尿病神经系统并发症评估中的应用价值。

方法：将135 例符合病例选择标准的糖尿病患者分别按糖尿病病程（＜ 1 年、1 ～ 10 年、＞10 年）、有无糖尿病周围神经病（diabetic perilpheral neuropathy，DPN）、是否伴发缺血性卒中以及脑卒中部位进行分组。回顾性分析和比较各组患者的年龄、糖化血红蛋白水平、神经传导速度（nerve conduction velocity，NCV）、SEP、DPN 和缺血性卒中情况。

结果：135 例糖尿病患者中，50 例（37.0%）明确诊断为DPN，伴发缺血性卒中32 例（23.7%），EMG 发现NCV 异常56 例（41.5%）、F 波异常16 例（11.9%），SEP 异常27例（20.0%）。糖尿病病程1 ～ 10 年和＞ 10 年组患者的NCV 异常率、DPN 比例及缺血性卒中比例均显著高于糖尿病病程＜ 1 年组（P＜ 0.05）。有DPN 症状及体征患者的NCV 异常率（P ＜ 0.01）、F 波异常率（P ＜ 0.01）和SEP 异常率（P ＜ 0.01）均显著高于无DPN 症状及体征患者。伴发缺血性卒中的糖尿病患者与未伴发缺血性卒中的糖尿病患者相比，前者的年龄较大（P ＜ 0.01）、病程较长（P ＝ 0.01）、SEP 异常率较高（P ＝ 0.02）。Spearman 相关分析结果显示，NCV 异常（r ＝ 0.26，P ＜ 0.01）和缺血性卒中（r ＝ 0.22，P ＜ 0.01）与病程呈正相关；DPN 与病程（r ＝ 0.21，P ＜0.05）、年龄（r ＝ 0.19，P ＜ 0.05）、糖化血红蛋白水平（r ＝ 0.24，P ＜ 0.01）、NCV 异常（r ＝ 0.75，P ＜ 0.01）和F 波异常（r ＝ 0.22，P ＜ 0.01）均呈正相关；NCV 异常（r ＝ 0.28，P ＜ 0.01）和F 波异常（r ＝ 0.30，P ＜ 0.01）均与糖化血红蛋白水平呈正相关；SEP 异常与F 波异常（r ＝ 0.62，P ＜0.01）、缺血性卒中（r ＝ 0.20，P ＜ 0.05）和住院（7 ～ 14 d）前后美国国立卫生研究院卒中量表评分差值呈正相关（r ＝ 0.43，P ＜ 0.05）。

结论：EMG 是筛查和诊断DPN 的重要方法，EMG 联合SEP 可以筛查早期DPN，有助于提高亚临床DPN 的检出率，并且对糖尿病伴发中枢系统损害及预后的评估具有一定的应用价值。

Objective: To investigate the value of electromyography (EMG) combined with somatosensory evoked potential (SEP) in evaluation of diabetic neurologic complications. 

Methods: One hundred and thirty-five diabetic patients were grouped according to the course of diabetes (< 1 year, 1-10 years, > 10 years), diabetic perilpheral neuropathy (DPN) (yes or no), ischemic stroke (yes or no), and the locus of cerebral stroke (anterior circulation infarction or posterior circulation infarction). The age, glycosylated hemoglobin level, nerve conduction velocity (NCV), SEP, DPN and ischemic stroke were retrospectively analyzed and compared.

Results: Of the 135 diabetic patients, 50 (37.0%) were diagnosed of DPN, 32 (23.7%) with ischemic stroke, 56 (41.5%) with NCV abnormality, 16 (11.9%) with F wave abnormality, and 27 (20.0%) with SEP abnormality (20%). The rate of NCV abnormality, the proportion of DPN and the proportion of ischemic stroke in diabetes course 1-10-year and >10-year groups were significantly higher than those in the diabetes course < 1 year group (P < 0.05). The rate of NCV abnormality (P < 0.01), the rate of F wave abnormality (P < 0.01) and the rate of SEP abnormality (P < 0.01) in patients with DPN symptoms and signs were significantly higher than those in patients without DPN symptoms and signs. The age of diabetic patients with ischemic stroke was older than those without ischemic stroke (P < 0.01), and the course of diabetes was longer (P = 0.01) with a higher rate of SEP abnormality (P = 0.02). Spearman correlation analysis showed that NCV abnormality (r = 0.26, P < 0.01) and ischemic stroke (r = 0.22, P < 0.01) were positively correlated with the course of diabetes; DPN was positively correlated with the course of diabetes (r = 0.21, P < 0.05), age (r = 0.19, P < 0.05), glycosylated hemoglobin level (r = 0.24, P < 0.01), NCV abnormality (r = 0.75, P < 0.01), and F wave abnormality (r = 0.22, P < 0.01); NCV abnormality (r = 0.28, P < 0.01) and F wave abnormality (r = 0.30, P < 0.01) were both positively correlated with glycosylated hemoglobin level; SEP abnormality was positively correlated with F wave abnormality (r = 0.62, P < 0.01), ischemic stroke (r = 0.20, P < 0.05) and the difference in National Institute of Health stroke scale (NIHSS) score before and after hospitalization (duration: 7-14 d) (r = 0.43, P < 0.05). 

Conclusion: EMG is an important method for screening and diagnosing DPN. EMG combined with SEP can screen early DPN, which helps to improve the detection rate of subclinical DPN, and has a certain value in the assessment of central system damage and prognosis of diabetes.