Tulsi (Holy Basil) for Type 2 Diabetes
A few small, mostly dated trials suggest tulsi (holy basil) might modestly lower fasting blood glucose, but the evidence is too thin and inconsistent to recommend it for type 2 diabetes. It should never replace proven diet, exercise, or medication.
Why this grade7-layer evidence engine
The positive signal rests on a single classic trial and one pooled analysis. A 1996 single-blind crossover RCT in 40 people with type 2 diabetes (PMID 8880292) found fasting glucose dropped about 21 mg/dl (roughly 18%) on holy basil leaves versus placebo. A 2018 meta-analysis (Jamshidi, Journal of Functional Foods; not PubMed-indexed) pooled trials to a fasting-glucose reduction of about 16 mg/dl, but with extreme heterogeneity (I-squared 91%), meaning the trials disagreed sharply.
The grade stays low because the underlying studies are small, mostly Indian, methodologically dated, and prone to publication bias — and critically, none used HbA1c, the standard long-term glucose marker, as a primary endpoint. A separate 2017 trial (PMID 28811698) improved insulin resistance but enrolled young, non-diabetic overweight adults and never measured glucose. A large 2023 network meta-analysis of antidiabetic herbs (PMID 37451111) did not even include tulsi, underscoring how much thinner its evidence is than fenugreek, cinnamon, or apple cider vinegar.
Authorities do not endorse it. The American Diabetes Association states there is insufficient evidence to support routinely using herbal supplements to improve glycemia; the WHO monograph notes traditional use for diabetes but calls the clinical data inadequate; and the FDA merely lists the plant without any health endorsement. Anyone taking glucose-lowering medication should be cautious and consult a clinician, given the theoretical added risk of low blood sugar.
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 = 0.547
- tier_from_score — 依分數區間映射至 tier letter
- apply_hec_rules — 高品質 SR/MA 顯示 positive (2 篇 > 0 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