| Abstract | 目的 应用2种客观营养状态评分CONUT、PNI评估急性心肌梗死患者入院时的营养状态，分析营养状态与预后的相关性。
方法 本试验为回顾性单中心队列研究，研究对象为急性心肌梗死并完成了经皮冠状动脉介入（PCI）术的患者，以全因死亡为终点事件，随访时间中位数为36（33，36）个月。根据CONUT评分的营养状态分为3组：正常组（n=304例），轻度营养不良组（n=476例）与中重度营养不良组（n=58例）。分析三组患者死亡事件的差异，并应用生存分析Kaplan-Meier法，采用Pairwise log-rank检验，分析CONUT评分与死亡事件的关系，应用同样的方法分析PNI评分与死亡事件的关系；通过Pearson相关性分析探讨CONUT、PNI与临床指标之间的相关性；分析两种营养状态评分与死亡事件的相关性应用Cox回归分析。
结果 共入选838例，根据CONUT评分有大部分急性心肌梗死患者处于营养不良状态（63.7%），死亡51例（6.08%）。Kaplan-Meier法分析揭示随着CONUT评分[Pairwise log-rank 0-1 v.s. 2-4 P=0.021，0-1 v.s. 5-12 P<0.001]的营养状态越差，死亡风险显著增加；随着PNI评分的营养状态越差[Pairwise log-rank >38 v.s. 5-38 P<0.001，>38 v.s. <35 P<0.001]，死亡风险亦显著增加；Cox比例风险回归模型分析结果表明在未调整的模型中，得出[CONUT=2-4 v.s. CONUT=0-1：HR(95%CI), 2.423 (1.111-5.286), P=0.026; CONUT=5-12 v.s. CONUT=0-1：HR(95%CI), 9.167 (3.799-22.120) , P<0.001]及[PNI=5-38 v.s. PNI>38：HR(95%CI),3.365(1.674-6.763),P=0.001; PNI<35 v.s. PNI>38：HR(95%CI), 6.359(3.359-1.674) , P<0.001]。调整了年龄、吸烟、低密度脂蛋白胆固醇、左室射血分数等因素后，亦得出[CONUT=2-4 v.s. CONUT=0-1：HR(95%CI), 3.328 (1.399-7.919), P=0.007; CONUT=5-12 v.s. CONUT=0-1：HR(95%CI), 8.331 (2.376-29.211) , P=0.001]及[PNI=5-38 v.s. PNI>38：HR(95%CI),1.795(0.751-4.293), P=0.188；PNI<35 v.s. PNI>38：HR(95%CI), 3.169(1.545-6.501) , P=0.002]，提示营养状态越差，死亡风险显著增加，同时也提示CONUT评分及PNI评分的营养不良均可预测死亡事件，CONUT评分优于PNI评分。
结论 大部分急性心肌梗死患者处于营养不良状态，CONUT及PNI评分的营养不良均可以作为急性心肌梗死患者死亡的独立预测因子，且CONUT评分优于PNI评分。 |
| Other Abstract | Objective To evaluate the nutritional states of acute patients with myocardial infarction (AMI) in hospital with two kinds of objective nutritional indices includes controlling nutritional (CONUT) score and prognostic nutritional index (PNI) and to investigate and compare the correlation between objective nutritional indices and prognosis in these patients.
Method We performed a monocentric retrospective cohort study in patients with acute myocardial infarction after percutaneous coronary intervention(PCI) and the median follow-up was 36(33, 36 interquartile range) months in our center. The endpoit was all cause of death. Information was obtained from hospital discharge and telephone. All the cases were divided into three groups according to the CONUT score: normal(n=304), mild(n=476) and moderate-severe(n=58). The differences of mortality in every groups by CONUT and PNI score were compared with Kaplan-meier. In addition, the relationship of objective nutritional indices and clinical index were analyzed with the Pearson correlation coefficient. The correlation of nutritional states and mortality of AMI patients were analyzed with the Cox proportional hazard analyzes.
Results A total of 838 cases were selected and 51 patients were died for all cause of death. A major of patients (63.7%) were in malnutrition state with CONUT score. Kaplan-meier curves indicated that the patients with high CONUT score [Pairwise log-rank 0-1 v.s. 2-4 P=0.021，0-1 v.s. 5-12 P<0.001] , low PNI score[Pairwise log-rank >38 v.s. 5-38 P<0.001，>38 v.s. <35 P<0.001] was significantly correlated with mortality. The Cox poportional hazard analyzes revealed that both the CONUT [CONUT=2-4 v.s. CONUT=0-1：HR(95%CI), 2.423 (1.111-5.286), P=0.026; CONUT=5-12 v.s. CONUT=0-1：HR(95%CI), 9.167 (3.799-22.120) , P<0.001]and PNI [PNI=5-38 v.s. PNI>38：HR(95%CI),3.365(1.674-6.763),P=0.001; PNI<35 v.s. PNI>38：HR(95%CI), 6.359(3.359-1.674) , P<0.001] score were associated with increased risk of all cause of death in unajusted model. We still have the similar result in CONUT[CONUT=2-4 v.s. CONUT=0-1：HR(95%CI), 3.328 (1.399-7.919), P=0.007; CONUT=5-12 v.s. CONUT=0-1：HR(95%CI), 8.331 (2.376-29.211) , P=0.001] and PNI [PNI=5-38 v.s. PNI>38：HR(95%CI), 1.795(0.751-4.293), P=0.188；PNI<35 v.s. PNI>38：HR(95%CI), 3.169(1.545-6.501) , P=0.002] which revealed that the CONUT is better than PNI.
Conclusions Malnutrition is universal in AMI patients. Objective nutritional indices CONUT score could be a better predictor of mortality in AMI patients than PNI. |
