臺灣老年人飲食型態、代謝質體特徵與認知功能下降之關聯性研究

Abstract

研究背景：隨著全球老化速度不斷上升，失智症議題已成為全球重要的公共衛生議題。在失智症中最主要的類型為阿茲海默氏症且尚無治癒的方法。因此，目前的研究希望能在失智症的臨床前期(輕度認知功能缺損)進行疾病的預測以便及早開始預防。過去研究顯示營養與認知功能之間有著密切的關係。相較於單一營養素，飲食型態提供了人體所需不同營養素與食物的整體概念。而代謝質體特徵則可以反映出疾病的內源性因素及外界環境的刺激。因此，本研究旨在探討臺灣老年人飲食型態、代謝質體特徵與認知功能下降之間的關聯性。 材料與方法：此前瞻性研究於2011至2013年間招募605位參加臺大醫院老人健檢計畫且年齡大於65歲的長者。整體認知功能利用臺灣版蒙特利爾認知評估量表(MoCA-T)進行評估；特定認知功能範疇的評估則分別採用以下的測驗：魏氏記憶量表第三版中的邏輯記憶測驗、空間記憶廣度測驗一、二和專注力測驗，及語意流暢度測驗-水果、蔬菜、魚(verbal fluency test)和路徑描繪測驗(trail making tests)。過去一年的飲食資料以44項半定量的飲食頻率問卷來收集。血漿的代謝質體圖譜透過500兆赫(mHz)的核磁共振儀進行分析。透過探索性的因素分析找出此老年族群的飲食型態與認知功能型態，接著用多變項線性迴歸模型與中介變項模式探討飲食型態、代謝質體特徵與認知功能下降之間的關聯性。 研究結果：本研究之前即找出3種飲食型態，分別為蔬菜飲食型態 (深綠色蔬菜、含油蔬菜、淺色蔬菜和根莖類作物)，肉類飲食型態 (豬肉、牛肉、家禽和燒烤肉類)，傳統飲食型態 (發酵食物和醃漬食物) (Chen et al., 2017)。攝取較多的蔬菜的老年人(T3)記憶功能比起吃得較少者(T1)的下降得慢[記憶型態：調整後的迴歸係數(β)=0.28，邏輯記憶-回憶一：調整後的迴歸係數(β)=0.18，邏輯記憶-回憶二：調整後的迴歸係數(β)=0.20]。中介變項模式顯示，高度攝取蔬菜保護記憶功能的部分作用可藉由特定的代謝物解釋(記憶型態：CH2CO、creatinine和pyruvate，邏輯記憶-回憶一：CH2CH2CO、creatinine和valine，邏輯記憶-回憶二：lactate)。相似地，當傳統飲食型態最高的組別(T3)，比起最低的組別(T1)，能顯著的延緩記憶的認知功能下降[邏輯記憶-主題二：調整後的迴歸係數(β)=0.19，邏輯記憶-回憶一：調整後的迴歸係數(β)=0.17，邏輯記憶-回憶二：調整後的迴歸係數(β)=0.18]。然而，高度攝取蔬菜會增加長者對執行功能下降的風險[執行功能型態：調整後的迴歸係數(β)=-0.27；路徑描繪測驗A：調整後的迴歸係數(β)=-0.23]。 結論：高度攝取蔬菜能延緩記憶功能之下降，特別是邏輯記憶-回憶一、邏輯記憶-回憶二。此保護作用部分可以藉由一些代謝物的機轉來解釋。本研究之前找出的飲食型態有助於預防臺灣長者認知功能下降。未來需要追蹤更長時間的前瞻性研究來驗證我們的發現。

Background: As population aging rapidly, dementia has become an important public health issue worldwide. Alzheimer’s disease (AD) is the leading type of dementia, which cannot be cured so far. Therefore, current studies are aiming at the preclinical stage of dementia, i.e., mild cognitive impairment (MCI) and hoping to prevent this disease at an early stage. Nutrition has been linked with cognitive impairment previously. In contrast to nutrients, dietary patterns (DP) provide an overview of nutrients and foods. The metabolic profile reflects the metabolites via influenced endogenous factors, underlying pathologies, and environmental stimuli. Therefore, this study aimed to explore the association between dietary patterns (DPs), metabolic profile with cognitive decline in Taiwanese elderly. Methods: This prospective cohort study recruited 605 Taiwanese elderly aged 65 or older from the annual Elderly Health Checkup (EHC) program at National Taiwan University Hospital (NTUH) during 2011-2013. Global cognition was assessed by Montreal Cognitive Assessment-Taiwan version (MoCA-T). Domain-specific cognition was assessed by the following tests: Wechsler Memory Scale-third edition was used to assess logical memory-recall I & II and thematic I & II, and digit span-forward & backward (attention domain). We also assessed verbal fluency test (fruits, vegetables, and fishes), and trail making test A and B. Dietary intake in the past year was assessed by a 44-item semi-quantitative food frequency questionnaire (FFQ). All plasma Nuclear Magnetic Resonance Spectroscopy (NMR) spectrum were acquired using an Avance-500 spectrometer. Exploratory factor analysis (EFA) was used to identify DPs and cognitive patterns. Multivariable linear regression model and mediation model were then used to explore the association of DPs, metabolic profile with cognitive decline. Results: Previously, our team identified three DPs as below (Chen et al., 2017). “Vegetable” DP was consisted of dark-green vegetables, vegetable with oil, light-colored vegetable, and tuber. “Meat” DP was characterized by meat, poultry, and roasted meats. “Traditional” DP included pickled vegetables and fermented foods. High-score “vegetable” DP (the highest tertile, T3) protected against the decline of memory pattern (β = 0.28), logical memory-recall I (β = 0.18) and logical memory-recall II (β = 0.20) compared with the lowest tertile of “vegetable” DP (T1). This protective effect may be partially explained by the mediation effect of metabolites (e.g., CH2CO, creatinine and pyruvate for memory pattern; CH2CH2CO, creatinine, and valine for logical memory-recall I; and lactate for logical memory-recall II). Similarly, high-score “traditional” DP (T3) protected against the decline of logical memory-thematic II (β = 0.19), logical memory-recall I (β = 0.17) and logical memory-recall II (β = 0.18) compared with the lowest tertile of “traditional” DP (T1). However, high-score “vegetable” DP (T3) increased the decline of executive function (β = -0.27), especially in trail making test A (β = -0.23). Conclusion: High intake of “vegetable” DP protected against the decline of memory pattern over 2 years, especially in logical memory-recall I and recall II. This protective effect may be mediated by some metabolites. DP identified previous in our team may be useful for preventing cognitive decline in Taiwanese elderly. Future prospective cohort study with longer follow-up is warranted to confirm our findings.