比較限時進食與其他飲食策略對於代謝健康之影響：系統性文獻回顧與網絡統合分析

Abstract

研究背景：「代謝性疾病」是全球所面臨的健康議題，而「健康的飲食模式」被認為是促進及改善代謝健康地的最佳策略。限時進食 (Time-restricted eating, TRE) 是近年備受矚目的飲食模式，其限制一天的進食時間需在10 小時以內，並依據進食/禁食的時間分為：早時段 (early TRE)、中時段 (mid TRE)和晚時段 (late TRE) 限時進食。雖然已有許多文獻針對限時進食進行探討，但其代謝效果是否優於其他飲食策略，例如連續熱量限制 (Continuous energy restriction, CER)仍存在不一致性，並且也缺乏直接比較不同型式限時進食的研究。因此本研究想利用網絡統合分 (Network meta-analysis)，評估不同種限時進食與其他飲食策略相比對代謝健康的效果差異。

材料與方法：本研究從Embase、Cochrane Library和PubMed數據庫之創始時期至2022年12月搜尋限時進食與代謝相關之研究，並納入體位測量（體重、身體質量指數、脂肪質量、無脂肪質量/瘦肉組織、腰圍）、血糖代謝（空腹血糖、空腹胰島素、糖化血色素、胰島素阻抗指數）、血壓（收縮壓和舒張壓）、血脂分析（總膽固醇、低密度膽固醇、高密度膽固醇、三酸甘油酯）的結果。以隨機效應的網絡統合分析評估不同飲食策略對代謝結果的相對有效性，並藉由P-score對不同飲食策略進行排名。

結果：共有六十六篇符合標準的試驗納入本研究中，這些試驗的參與者來自不同人群，總計有3016位受試者。在體位測量結果方面，自由選擇禁食時段之限時進食、中時段限時進食和20:4限時進食被認為是最佳飲食策略的比例較高。在血糖代謝測量方面，早時段、中時段和20:4限時進食被認為是最佳飲食策略的比例較高。在血壓測量方面，早時段限時進食、自由選擇禁食時段之限時進食、18:6和16:8限時進食被認為是最佳策略的比例較高。在血脂分析方面，自由選擇禁食時段之限時進食和12:12限時進食被認為是最佳飲食策略的比例較高。與常規飲食和連續性斷食比較中，大多數分析結果顯示限時進食能有利的改善代謝健康。儘管限時進食與無脂肪質量/瘦肉組織、糖化血色素和所有血脂分析指標皆未達到統計上顯著差異，但與常規飲食或連續型斷食比較下並未有更差的效益。透過使用網絡統合分析可信度測量 (Confidence in network meta-analysis, CINeMA)，針對每個網絡統合分析之證據可信度進行評估，結果顯示76％的直接性證據被歸類為低或非常低的可信度質量。在所有納入探討服從性的試驗中，紀錄了無論是早時段、中時段、晚時段、還是自由選擇禁食時段之限時進食，其服從性平均皆超過70％。此外，在66篇試驗中，有20篇研究報告了有關不良事件發生率，其中有九篇明確指出試驗期間無發生不良事件；而另11篇研究，報告限時進食受試者和對照組中皆有發生輕微副作用。

結論：根據網絡統合分析結果，在進行限時進食時，無論自由選擇禁食時段之限時進食、早時段限時進食和午時段限時進食，似乎與常規飲食相比都為代謝健康指標，如體重、身體質量指數、脂肪質量、腰圍、空腹血糖、空腹胰島素、胰島素阻抗指數等帶來更好的效果。然而，本研究暫無法明確得知任何一種時間長短進行之限時進食是在對於改善代謝結果方面擁有最佳或較差的效益。本研究顯示限時進食是一種具有潛力的飲食策略，至少在改善代謝健康方面表現出與常規飲食和連續熱量限制飲食有更好的效果，且服從性偏高。然而，本研究也顯示大部分直接證據的可信度偏低，因此未來需要更進一步進行高質量的試驗來確認限時進食對代謝健康的療效。

關鍵字：限時進食、代謝健康、網絡統合分析

Background: Metabolic diseases increase the risk of serious health conditions globally. Dietary changes are usually the first-line therapy employed to improve metabolic health. Time-restricted eating (TRE) involves restriction of daily eating window to 10 hours or less. TRE can be divided into three forms: early TRE, mid TRE, and late TRE. It is still controversial whether the effect of TRE on metabolic health can be greater than other dietary strategies such as continuous energy restriction (CER). In addition, there is a lack of studies comparing the different forms of TRE. Therefore, we aim to conduct a network meta-analysis (NMA) to evaluate the pooled relative effect of TRE forms compared to other dietary interventions.

Methods and Materials: We search for potential studies involving TRE from Embase, Cochrane Library, and PubMed from databases inception to December 2022. Outcomes included anthropometric measurements (body weight (BW), body mass index (BMI), fat mass (FM), fat-free mass/lean mass (FFM/LM), waist circumference (WC)), glucose metabolism (fasting blood glucose (FBG), fasting insulin (FI), hemoglobin A1c (HbA1c), homeostasis model assessment-insulin resistance index (HOMA-IR)), blood pressure (systolic blood pressure (SBP) and diastolic blood pressure (DBP)), lipid profile (total cholesterol (TG), low density lipoprotein cholesterol (LDL), high density lipoprotein cholesterol (HDL), triglycerides (TG)). We used NMA with random effects to estimate the relative effectiveness of the different dietary strategies for each outcome and rank them based on P-score.

Results: There were 66 eligible trials included, and the participants enrolled in these studies were 3016 individuals from various populations. Overall, self-selected TRE and TRE 20:4 are considered the best TRE approach for anthropometric measurements according to the P-score ranking. For glucose metabolism outcomes, early TRE, mid TRE, and TRE 20:4 are considered the best overall. For blood pressure outcomes, early TRE self-selected TRE, TRE 18:6, and TRE 16:8 are considered the best TRE forms. For lipid profile measurements, self-selected TRE and TRE 12:12 are considered the best TRE forms. TRE demonstrated a more favorable effect in most of the outcomes when compared to a usual diet and CER. Although there was no significant improvement in FFM/LM, HbA1c, and all the outcomes of lipid profile, they were not worse than the usual diet or CER. Through Confidence in Network Meta-analysis (CINeMA) to rate the confidence of evidence for each NMA, 76% of the direct evidence available was classified as low or very low quality. The included studies (>70%) noted excellent adherence rate regardless of the types of TRE (early TRE, mid TRE, late TRE, or self-selected TRE). In addition, out of the 66 studies analyzed, 20 studies reported information about adverse events. Among these 20 studies, nine specifically stated that no adverse events occurred during the trials. Eleven studies reported only minor side effects occurring in both participants following a TRE diet or other strategies.

Conclusion: According to the NMA, late TRE seemed to yield inferior effects compared to early TRE, mid TRE and self-selected TRE. However, in the different eating windows of TRE, none of the TRE durations consistently demonstrated a superior or inferior effects on improving metabolic outcomes. TRE appears as a promising dietary strategy that, at the very least, shows comparable effectiveness to usual diet and CER in improving metabolic health such as body weight, body mass index, fat mass, waist circumference, fasting blood glucose, fasting insulin, and HOMA-IR. However, since most evidence was of low or very low certainty, further high-quality trials are warranted to confirm the relative efficacy of TRE on metabolic health.

Key words: time-restricted eating, metabolic health, network meta-analysis