摘要目的:探索口腔颌面手术老年患者的血清脂质代谢物与术后神经认知恢复延迟（dNCR）的相关性。方法:择期行全麻下口腔颌面手术的老年患者102例，年龄≥65岁，美国麻醉医师协会分级Ⅰ~Ⅲ级，根据是否发生dNCR分为dNCR组（38例）、非dNCR组（64例）。记录患者一般资料（身高、术前血小板计数等）及dNCR发生率。术前及术后（术后24 h）各采集5 ml动脉血液样本，采用代谢组学的方法对血清脂质代谢物进行检测与筛选，并对差异脂质代谢物进行受试者操作特征（ROC）曲线分析及logistic回归分析，以探索口腔颌面手术老年患者血清脂质代谢物与dNCR的相关性。结果:102例患者中，38例（37.3%）发生dNCR。dNCR组身高高于非dNCR组、术前血小板计数低于非dNCR组（均 P<0.05）。术前两组差异脂质代谢物共24种，与非dNCR组比较，dNCR组血清甘油三酯（TG）、磷脂酰甘油（PG）、磷脂酰胆碱（PC）、磷脂酰丝氨酸（PS）水平较低，而溶血磷脂酰乙醇胺（LPE）、溶血磷脂酰胆碱（LPC）水平较高（均 P<0.05）。术后两组差异脂质代谢物共12种，与非dNCR组比较，dNCR组血清TG、PG水平较低，而心磷脂（CL）、磷脂酰乙醇胺（PE）水平较高。术前TG（54:2/18:2）［曲线下面积（AUC）=0.768， P<0.001］、TG（54:2/18:0）（AUC=0.754， P<0.001）对dNCR具有良好预测价值，且术后TG（54:2/18:0）仍保持显著预测效能（AUC=0.722， P<0.001）。校正年龄、性别和教育水平后，多因素logistic回归分析发现，术前及术后TG（54:2/18:2）和TG（54:2/18:0）低表达是dNCR发生的独立危险因素（均 P<0.05）。 结论:以TG为代表的脂质代谢物可作为潜在的dNCR早期标志物。

abstractsObjective:To explore the relationship between serum lipid metabolites and delayed neurocognitive recovery (dNCR) in elderly patients undergoing oral and maxillofacial surgery.Methods:One hundred and two elderly patients aged≥65 years, who were classified as American Society of Anesthesiologists grades Ⅰ?Ⅲ and underwent elective oral and maxillofacial surgery under general anesthesia were selected. According to the occurrence of dNCR, the patients were divided into two groups: a dNCR group ( n=38) and a non-dNCR group ( n=64). Their general information (including height and preoperative platelet count) and the incidence of dNCR were recorded. Arterial blood samples (5 ml) were collected before surgery and at postoperative (postoperative 24 h) . Serum lipid metabolites were analyzed and screened by metabolomics. Differential lipid metabolites were subjected to receiver operating characteristic (ROC) analysis and logistic regression to explore the relationship between serum lipid metabolites and dNCR in elderly patients undergoing oral and maxillofacial surgery. Results:Among 102 patients, 38 (37.3%) developed dNCR. The dNCR group showed a higher height and lower preoperative platelet count than the non-dNCR group (all P<0.05). Before surgery, there were 24 differential lipid metabolites between the two groups. Comparison with non-dNCR group, serum triglycerides (TG), phosphatidylglycerols (PG), phosphatidylcholine (PC), and phosphatidylserine (PS) levels were lower, while lysophosphatidylethanolamine (LPE) and lysophosphatidylcholine (LPC) levels were higher in the dNCR group (all P<0.05). After surgery, there were 12 differential lipid metabolites between the two groups. Comparison with non-dNCR group, serum TG and PG levels were lower, while cardiolipin (CL) and phosphatidylethanolamine (PE) levels were higher in the dNCR group. Preoperative TG (54:2/18:2) [area under curve (AUC)=0.768, P<0.001] and TG (54:2/18:0) (AUC=0.754, P<0.001) showed good predictive value for dNCR, and postoperative TG (54:2/18:0) also maintained significant predictive efficacy (AUC=0.722, P<0.001). After adjustment for age, sex, and education level by multivariate logistic regression, low expression of TG (54:2/18:2) and TG (54:2/18:0) before and after surgery were found as independent risk factors for the occurrence of dNCR（all P<0.05）. Conclusions:Lipid metabolites, represented by TG, can serve as potential early biomarkers for dNCR.