銀杏 Ginkgo Biloba × 失智症

結論:證據分歧

失智症必須拆成「預防」與「治療」兩個問題:預防失智的證據已塵埃落定為陰性(GEM RCT n=3,069 + GuidAge + 預防型 MA OR 1.

D 🔴 D 反證據 證據分歧 ⚠️ medium — moderate promotional content
⚠️ 標記 ⚠️ stale 🇹🇼 台灣在地警示 💊 檢驗 / 藥物交互作用

失智症必須拆成「預防」與「治療」兩個問題:預防失智的證據已塵埃落定為陰性(GEM RCT n=3,069 + GuidAge + 預防型 MA OR 1.05,Examine 自身的 Alzheimer's Disease Risk 也僅 D),這部分屬 D 反證據。

治療既有失智的症狀性效益存在訊號(Tan 2015 MA 顯示 EGb 761 240mg/day 對認知/ADL 有改善),但 Cochrane Birks 2009 明評為「不一致、不可靠」,且正面 MA 多集中於 EGb 761 廠商相關之歐洲試驗、有 COI 疑慮——故治療面屬薄弱且具爭議的 C。

綜合判 C:訊號存在但僅限治療、品質受爭議;不給 B 是因為唯一夠強的證據是 prevention-null(反向),且五家權威學會/機構(NCCIH、Mayo、Cleveland、Harvard、alz.org+AAN)對失智一致不支持。

⚖️

評分透明度

所有分數由 7 層證據引擎計算,過程公開可查
原始分數 0.39
D
C
B
A
S
← 反證據 / 無效有效 / 強證據 →
最終評級
D · 證據分歧
信心度
80%
證據方向大致一致
證據層級
E1
Cochrane 高品質系統性回顧/統合分析

各層「支持此療效」的程度

分數越低=該層越不支持
L5 臨床機構權威立場
0.35
L2 PubMed原始文獻
0.45
L3 機轉生理合理性
0.45
L1 Examine國際基準
0.50
L11 AI 複核獨立判讀
0.50
不支持 中性 / 混合 支持
查看完整決策路徑(audit trail)
  1. compute_raw_score — 加權公式: L2×0.30 + L3×0.25 + L5×0.25 + L11×0.10 + L1×0.10 = 0.391
  2. tier_from_score — 依分數區間映射至 tier letter
  3. apply_hec_rules — 高品質 SR/MA 顯示 positive (1 篇 > 0 negative)
  4. tier_strict_requirement_check — Tier 條件達標,未降階
  5. detect_disputes — 偵測到 1 個 hard + 0 個 soft dispute
  6. decide_status — 依 tier + dispute 結果決定 status

Ginkgo biloba for prevention of dementia: a randomized controlled trial (GEM Study)
PMID: 19017911 2008 RCT (double-blind) n = 3,069
結論:Ginkgo did NOT prevent dementia: all-cause dementia HR 1.12 (95% CI 0.94-1.33, p=.21), AD HR 1.16 (95% CI 0.97-1.39, p=.11) over median 6.1 years in adults aged >=75.
🟢 高品質 政府資助 效應量:HR 1.12 all-cause dementia (null)
前往 PubMed
Ginkgo biloba for prevention of dementia: a systematic review and meta-analysis
PMID: 26058281 2015 統合分析 n = 5,889
結論:Pooling 2 RCTs (5,889 participants), no difference in dementia incidence between Ginkgo and placebo (347/2,951 vs 330/2,938; OR 1.05, 95% CI 0.89-1.23); no convincing evidence Ginkgo prevents dementia.
學術資助 效應量:OR 1.05 incident dementia (null)
前往 PubMed
Ginkgo biloba for cognitive impairment and dementia (Cochrane review, Birks & Grimley Evans)
PMID: 19160216 2009 系統性回顧 n = 2,016
結論:36 trials included (9 six-month trials, n=2,016 analysed in detail); more recent trials showed inconsistent results for cognition and ADL; evidence of clinically significant benefit judged inconsistent and unreliable.
🟢 高品質 學術資助
前往 PubMed
Efficacy and adverse effects of ginkgo biloba for cognitive impairment and dementia: a systematic review and meta-analysis (Tan et al.)
PMID: 25114079 2015 統合分析 n = 2,561
結論:Pooling 9 RCTs (n=2,561), EGb 761 240 mg/day favored over placebo for cognition (WMD -2.86, 95% CI -3.18 to -2.54), ADL (SMD -0.36, 95% CI -0.44 to -0.28) and global change (Peto OR 1.88, 95% CI 1.54-2.29), especially with neuropsychiatric symptoms.
效應量:Cognition WMD -2.86; ADL SMD -0.36 (favoring EGb 761)
前往 PubMed

L4a US FDA
謹慎
Most consider the main benefit of Ginkgo biloba to be its ability to prevent the worsening of memory problems due to … neurodegenerative disease. 來源↗
L4b EU EFSA
未表態
Some 'general function' claims (1548 botanical-related claims) have been put on hold pending for the Commission and Member States final consideration. 來源↗
L4c UK NHS
謹慎
Some people with dementia and their carers use complementary remedies, such as gingko biloba, curcumin or coconut oil. However, there's not enough evidence to say whether such remedies are effective. ... It's best to be wary of any products that claim to benefit people with dementia. If you're thinking of taking such a product or supplement, it's important to consult a doctor first. Some remedi… 來源↗
L4d TW TFDA / 衛福部
謹慎
銀杏果(俗稱白果)屬於一般食品,並無療效。而銀杏葉萃取物,在我國是以藥品管理。其效果僅限於改善末梢血液循環,對於老年失智、中風及動脈血管疾病,並未證實有治療效果。 來源↗
L4e WHO
中性
WHO monographs on selected medicinal plants - Volume 1 來源↗

L5a NIH Office of Dietary Supplements
謹慎
There's no conclusive evidence that ginkgo is helpful for any health condition. 來源↗
L5b Mayo Clinic
中性
— 本適應症無對應資料
L5c Cleveland Clinic
謹慎
This supplement is not intended to diagnose, treat, cure, or prevent any disease. 來源↗
L5d Harvard Health
謹慎
— 本適應症無對應資料
L5e Specialty Society (condition-mapped)
反對
A study funded by the National Institute on Aging (NIA) of more than 3,000 older volunteers found that ginkgo biloba was no more effective than a placebo in preventing or delaying Alzheimer's or other dementia. 來源↗

PTT · Dcard · Mobile01 彙整自公開論壇討論,非統計抽樣,僅反映社群風向。
廣告 / 業配密度 中度
📍立場總覽

社群討論量中等,整體偏懷疑。多數鄉民認為銀杏對改善腦力/失智幫助有限、近乎吃心安,常被酸是廣告誇大;少數人提到搭配 PS、DHA 後主觀有感但停用即退。普遍提醒銀杏體質差異大、不是每個人都適合,並有出血/腸胃不適與藥物交互作用警告。

💬社群實感

分歧偏負面(多數認為無感/吃心安,少數搭配 PS、DHA 主觀有感)

破解迷思 社群最常見的 3 個誤解
事實誤以為銀杏果/銀杏粉就有護腦效果(社群與藥師指出有效的是標準化銀杏葉萃取物 EGb761,銀杏果只是一般食材)
事實誤以為銀杏能『預防或治療』失智(多人提醒僅可能延緩部分症狀,無法逆轉,且證據不足、健保已取消給付改列指示藥)
事實誤以為保健品是安全『多吃無妨』,忽略銀杏與抗凝血藥物併用的出血風險
🩹 社群通報的副作用
  • 頭暈
  • 噁心/腸胃不適
  • 出血風險(與抗凝血藥、魚油、阿斯匹靈併用)
  • 曾有人提到長輩服用後昏倒(個案)

查看代表討論串 ↗

L10a · 廠商行銷話術 行銷語言
💬 通路如何宣傳

永信HAC 銀杏果錠 180錠 $1,300

代表來源 ↗
L10b · TFDA 法定身份 官方認定

銀杏葉萃取物,在我國是以藥品管理

來源 ↗

  • 膽鹼酯酶抑制劑(如 donepezil、rivastigmine、galantamine)
  • Memantine(NMDA 受體拮抗劑)
  • 規律身體活動(風險因子控管)
PMID 100% 反查全部經 NCBI Entrez 驗證
🔬 4 篇 L2 文獻 經多層 sub-agent 獨立評估
🇹🇼 含台灣社群分析L10c PTT / Dcard / Mobile01
aggregated_at: 2026-06-01 claim_version: v25 engine_version: v1.0 claim_id: CLM-COND-dementia-INT-ginkgo-biloba-001
查看 ClaimReview 結構化資料 (JSON-LD)
{
  "@context": "https://schema.org",
  "@type": "ClaimReview",
  "url": "https://gpt-dict.com/claim/CLM-COND-dementia-INT-ginkgo-biloba-001/",
  "datePublished": "2026-06-01",
  "author": {
    "@type": "Organization",
    "name": "gpt-dict.com",
    "url": "https://gpt-dict.com"
  },
  "claimReviewed": "銀杏能改善失智症",
  "inLanguage": "zh-TW",
  "itemReviewed": {
    "@type": "Claim",
    "appearance": []
  },
  "reviewRating": {
    "@type": "Rating",
    "ratingValue": 1,
    "bestRating": 5,
    "worstRating": 1,
    "alternateName": "🔴 D 反證據"
  }
}