Rathmann W, Schloot N C, Kostev K, Reaney M, Zagar A J, Haupt A
German Diabetes Center, Institute for Biometry and Epidemiology, Leibniz Center for Diabetes Research, Heinrich-Heine University, Duesseldorf, Germany.
Lilly Deutschland GmbH, Bad Homburg, Germany.
Exp Clin Endocrinol Diabetes. 2014 Feb;122(2):92-9. doi: 10.1055/s-0033-1363684. Epub 2014 Feb 20.
To investigate the risk of macro- and microvascular complications in patients with type 2 diabetes receiving rapid-acting insulin analogues (IA) or human regular insulin (HI).General practice diabetes patients with continuous prescription of any IA or HI for ≥3 years were selected from the German Disease Analyzer database (IMS Health). Logistic and Cox regression models were applied to analyze the incidence and time to onset of vascular outcomes (IA vs. HI).2764 patients on IA (insulin lispro, glulisine, aspart) and 4193 patients on HI were included (age, mean [SD]: 61.0 [11.3] and 64.7 [10.5] years, follow-up [Q1,Q3]: 4.6 [3.7,6.1] and 4.7 [3.7,5.9] years). No significant differences were detected between IA and HI regarding the incidence of vascular complications (OR [95%CI]: macrovascular 0.92 [0.72-1.18], microvascular 0.95 [0.77-1.17]) or regarding time to their onset, after adjustment for sex, age, comorbidities and time on IA/HI, or by propensity-score-based matching. However, in an additional short-term analysis (median [Q1,Q3] follow-up (IA 2.9 [1.2,4.6], HI 2.4 [0.8,4.4] years) of a larger sample (no continuous insulin treatment required) with more comorbidities, time to onset of macrovascular complications was significantly longer for AI than HI (HR 0.88 [0.81-0.97], p=0.009; microvascular complications: no difference).After long-term continuous treatment with IA or HI under real-life conditions, there was no different risk of macro- or microvascular complications, contradicting previous short-term analyses. Further prospective studies are needed to clarify whether selection bias may have been introduced by using strict entry criteria.
研究接受速效胰岛素类似物（IA）或人常规胰岛素（HI）治疗的2型糖尿病患者发生大血管和微血管并发症的风险。从德国疾病分析数据库（艾美仕市场研究公司）中选取持续处方任何IA或HI≥3年的全科糖尿病患者。应用逻辑回归和Cox回归模型分析血管结局（IA与HI）的发生率和发病时间。纳入了2764例使用IA（赖脯胰岛素、谷赖胰岛素、门冬胰岛素）的患者和4193例使用HI的患者（年龄，均值[标准差]：61.0[11.3]岁和64.7[10.5]岁，随访[第一四分位数，第三四分位数]：4.6[3.7,6.1]年和4.7[3.7,5.9]年）。在调整性别、年龄、合并症以及使用IA/HI的时间后，或者通过倾向得分匹配后，未发现IA和HI在血管并发症发生率（比值比[95%置信区间]：大血管0.92[0.72 - 1.18]，微血管0.95[0.77 - 1.17]）或发病时间方面存在显著差异。然而，在一项针对更大样本（无需持续胰岛素治疗）且合并症更多的额外短期分析中（中位数[第一四分位数，第三四分位数]随访时间（IA 2.9[1.2,4.6]年，HI 2.4[0.8,4.4]年））发现，IA发生大血管并发症的发病时间显著长于HI（风险比0.88[0.81 - 0.97]；微血管并发症：无差异）。在现实生活条件下长期持续使用IA或HI治疗后，大血管或微血管并发症的风险没有差异，这与之前的短期分析结果相矛盾。需要进一步的前瞻性研究来阐明使用严格的纳入标准是否可能引入了选择偏倚。
