文献速递血液净化与透析

日本糖尿病血液透析患者糖化血红蛋白独特分布及其与死亡率的关系

Unique hemoglobin A1c level distribution and its relationship with mortality in diabetic hemodialysis patients.
PubMed 文献 · Kidney Int · 2017-04-08

本研究基于日本透析预后与实践模式研究(JDOPPS)2-5期2300例糖尿病血液透析患者,采用Cox回归分析(校正基线年龄、性别、透析龄、12种一般合并症、血红蛋白、白蛋白、肌酐水平及胰岛素使用,并按JDOPPS分期分层及考虑机构聚集性),发现HbA1c与死亡率呈U形关系;以死亡率最低组(HbA1c 6.0-7.0%)为参照,HbA1c<5.0%、5.0-6.0%、7.0-<8.0%和≥8.0%的校正风险比分别为1.56(95%CI 1.05-2.33)、1.26(0.92-1.71)、1.23(0.79-1.89)和2.10(1.32-3.33)。日本患者HbA1c分布与西方国家显著不同,提示需制定基于种族和国家的血糖控制指南。

要点

  • 糖尿病血液透析患者HbA1c与死亡率呈U形关系,最低死亡率组为6.0-7.0%。
  • HbA1c<5.0%和≥8.0%与死亡率升高独立相关。
  • 日本患者HbA1c分布与西方国家不同,需本土化指南。

核心结论

  • 糖尿病血液透析患者HbA1c与死亡率呈U形关系,HbA1c 6.0-7.0%组死亡率最低。

研究设计与样本

  • 基于JDOPPS 2-5期2300例日本糖尿病维持性血液透析患者,采用Cox回归分析,校正基线年龄、性别、透析龄、12种一般合并症、血红蛋白、白蛋白、肌酐水平及胰岛素使用,并按JDOPPS分期分层及考虑机构聚集性。

主要发现

  • 54%的患者HbA1c<6.0%,其中14%HbA1c<5.0%;HbA1c较高者胰岛素或口服降糖药使用较少。
  • 以HbA1c 6.0-7.0%为参照,HbA1c<5.0%、5.0-6.0%、7.0-<8.0%和≥8.0%的校正风险比分别为1.56(95%CI 1.05-2.33)、1.26(0.92-1.71)、1.23(0.79-1.89)和2.10(1.32-3.33)。
  • JDOPPS第5期数据中HbA1c水平与自我报告的低血糖事件无关。

关键数值与推荐

  • HbA1c<5.0%:校正HR 1.56(95%CI 1.05-2.33),以HbA1c 6.0-7.0%为参照。
  • HbA1c≥8.0%:校正HR 2.10(95%CI 1.32-3.33),以HbA1c 6.0-7.0%为参照。
  • HbA1c 5.0-6.0%:校正HR 1.26(95%CI 0.92-1.71),以HbA1c 6.0-7.0%为参照。
  • HbA1c 7.0-<8.0%:校正HR 1.23(95%CI 0.79-1.89),以HbA1c 6.0-7.0%为参照。

临床怎么用

  • 对糖尿病血液透析患者,HbA1c控制在6.0-7.0%可能死亡率最低,应避免<5.0%和≥8.0%。
  • 需注意日本患者HbA1c分布与西方国家不同,血糖控制目标应结合本土指南。

局限性

  • 研究仅纳入日本患者,结果外推至其他种族需谨慎。
  • HbA1c仅在入组时测定一次,未反映长期血糖控制情况。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

原文

Diabetic hemodialysis patients with hemoglobin A1c (HbA1c) levels below 6.5% and over 8.0% face a higher mortality risk. To determine the optimal glycemic control in Japanese patients, we examined the association between HbA1c and mortality in 2,300 Japanese diabetic patients on maintenance hemodialysis with HbA1c levels determined at enrollment in the Japanese Dialysis Outcomes and Practice Patterns Study (JDOPPS) phases 2-5, using Cox regression analysis with adjustment for baseline age, sex, dialysis vintage, 12 general comorbidities, hemoglobin, albumin and creatinine levels, and insulin use; stratification by JDOPPS phase; and facility clustering taken into account. Overall, 54% of patients had HbA1c levels under 6.0, including 14% with HbA1c levels under 5.0. Insulin or oral diabetes medications were used less frequently in patients with higher HbA1c levels. The dependence of mortality on HbA1c level was U shaped. When the group with the lowest mortality (HbA1c 6.0-7.0) was used as a reference, the hazard ratios for HbA1c categories under 5.0, 5.0-6.0, 7.0 to under 8.0, and 8.0 and greater were, respectively, 1.56 (95% confidence interval, 1.05-2.33), 1.26 (0.92-1.71), 1.23 (0.79-1.89), and 2.10 (1.32-3.33) in the adjusted model. The HbA1c level was not associated with self-reported hypoglycemic episodes in JDOPPS phase 5. The HbA1c levels in diabetic hemodialysis patients differ considerably between Japan and those reported from Western countries. Thus, our findings highlight the importance of domestic guidelines for glycemic control by race and country.

来源:Kidney Int,查看原文。版权归原作者所有。

引用

  1. Hoshino J, Larkina M, Karaboyas A, et al. Unique hemoglobin A1c level distribution and its relationship with mortality in diabetic hemodialysis patients. Kidney Int. 2017. PMID: 28396123. doi:10.1016/j.kint.2017.02.008

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。