Cinnamon for Type 2 Diabetes
Cinnamon is not a reliable treatment for type 2 diabetes. It may lower fasting blood sugar modestly, but its effect on long-term control (HbA1c) is small and inconsistent, and major health authorities do not recommend it for managing blood glucose.
Why this grade7-layer evidence engine
The grade is Weak (C) because the trial evidence pulls in different directions. The 2012 Cochrane review of 10 RCTs (PMID 22972104, n=577) found no significant effect on HbA1c, insulin, or post-meal glucose and called the fasting-glucose data inconclusive amid high risk of bias. A 2013 review (PMID 24019277, n=543) saw fasting glucose fall about 25 mg/dL but no HbA1c change, while later meta-analyses (PMID 37818728; PMID 38917435, n=3054) and a 2025 umbrella review (PMC12620228) report small significant drops in both. These conflicting results stem from different study pools and high heterogeneity.
Where there is a signal, it is limited. Fasting-glucose improvements are fairly consistent (the umbrella review rated them 'highly suggestive' at I-squared=0%) but clinically modest at roughly 15 mg/dL, and that same review flagged HbA1c and insulin-resistance findings as needing caution. A triple-blind RCT (PMID 33605836, n=160) found only a 0.2% HbA1c drop over three months. Optimal dose, formulation, and duration remain unclear.
Health authorities reinforce the cautious read. The US FDA treats cinnamon only as a flavoring agent with no approved health claim, and the WHO endorses it solely for poor appetite and indigestion, not blood sugar. The Mayo Clinic says the evidence is not strong enough to recommend it, with Cleveland Clinic, Harvard, and diabetes specialty guidance similarly unconvinced. Cassia cinnamon also contains coumarin, a liver toxin at high intake, and can interact with glucose-lowering drugs, so it should not replace proven diabetes care.
Scoring transparency
All scores computed by a 7-layer evidence engine — fully auditable▸View the full decision path (audit trail)
- compute_raw_score — 加權公式: L2×0.30 + L3×0.25 + L5×0.25 + L11×0.10 + L1×0.10 = 0.435
- tier_from_score — 依分數區間映射至 tier letter
- apply_hec_rules — 高品質 SR/MA 顯示 positive (4 篇 > 1 negative)
- tier_strict_requirement_check — Tier 條件達標，未降階
- detect_disputes — 偵測到 0 個 hard + 0 個 soft dispute
- decide_status — 依 tier + dispute 結果決定 status