應用核磁共振氫譜技術探討台灣老年人代謝體特徵與認知功能障礙之關聯研究

Abstract

背景：過去的研究已發現與許多代謝物與認知功能相關，但單一的代謝物並不能解釋認知障礙複雜的病理機轉。代謝體學提供人體內動態且有多變項參與代謝反應的概觀。目前認知功能障礙的機制還不清，本研究旨在探討代謝體特徵和認知功能障礙的關聯。 方法：本研究為橫斷式研究。共招募了256位65歲及以上在台大醫院參加老人健康檢查的老人(2011-2013)。認知功能的評估則是使用中文版的蒙特利爾認知評估量表，總分為30分，小於24分為認知功能障礙，大於及等於24分為認知功能正常。運用核磁共振氫譜(1H-NMR)光譜儀測定受試者血漿中的代謝體特徵。隨後使用主成分分析(PCA)和最小乘方分析法(PLS-DA)將代謝物分組，再放入邏輯斯迴歸模型評估代謝物和認知功能障礙的關聯。 結果：研究結果顯示，主成分分析(PCA)和最小乘方分析法(PLS-DA)的散佈圖不易區分認知功能受損和正常老人。13個代謝物被辨認出來，其中VLDL&LDL, lipid (CH2)n (P = 0.02) 在認知功能障礙及正常老人之間有顯著的差異。依據ApoE ɛ4遺傳半型有無的狀態做分層後，未帶有ApoE ɛ4的老人，VLDL&LDL, lipid (CH2)n (AOR = 2.40, 95% CI = 1.14-5.04) 和N-acetyl glycoprotein (AOR = 5.01, CI = 1.10-22.72)濃度之升高和認知功能障礙風險的增加有關。 結論：本研究研究運用核磁共振氫譜技術的代謝體方法，成功地在未帶有ApoE ɛ4對偶基因的老人辨識出2個與認知功能障礙有關的代謝物(VLDL&LDL, lipid (CH2)n以及 N-acetyl glycoprotein)。未來大型的世代追蹤研究有助於確認本研究之發現。

Background. Previous studies relating the alteration of specific circulating metabolites to cognitive impairment was unable to capture of the complex pathogenesis of cognitive impairment. Metabolomics provided an overview of the dynamic and multiparametric metabolic response in the human body. The mechanism of cognitive impairment is complex. Therefore, this study was aimed to explore the association between metabolomic profiles and the risk of cognitive impairment. Methods. This was a cross-sectional study. A total of 256 participants, aged 65 years or older, were recruited from the annual Elderly Health Checkup (EHC) program at National Taiwan University Hospital between 2011 and 2013. Montreal Cognitive Assessment (MoCA), Chinese version, was used to assess global cognitive function. 1H nuclear magnetic resonance spectroscopy (NMR) was applied to investigate the perturbation of metabolome in the plasma sample. Principal components analysis (PCA) and partial least squares-discriminant analysis (PLS-DA) were used to differentiate metabolic profiling. Multivariable regression models were used to assess the association between identified metabolites and cognitive impairment. Results. The unsupervised PCA and supervised PLS-DA showed little visual separation between cognitive impaired and normal elders. Among 13 identified metabolites VLDL&LDL, lipid (CH2)n was significantly different between cognitive impaired and normal elders (P=0.02). After stratification by ApoE ɛ4 status, elevated VLDL&LDL, lipid (CH2)n [adjusted odds ratio (AOR) = 2.40, 95% confidence interval (CI) = 1.14-5.04] and N-acetyl glycoprotein [AOR = 5.01, CI = 1.10-22.72] were associated with the risk of cognitive impaired among ApoE ɛ4 non-carriers. Conclusion. This study uses 1H NMR-based metabolomic approach and successfully identified two metabolites (VLDL&LDL, lipid (CH2)n and N-acetyl glycoprotein) for cognitive impairment among ApoE ɛ4 non-carriers. Larger study with longitudinal design and prospective studies are needed to confirm our findings.