Astragalus for Kidney Disease
Astragalus taken alongside standard care shows consistent reductions in proteinuria and a possible slowing of kidney-function decline in chronic and diabetic kidney disease, but the underlying trials are low quality, so the benefit remains unproven and astragalus is not a substitute for guideline treatment.
Why this grade7-layer evidence engine
This earns a weak "Disputed" grade because the human research, though unusually large, is methodologically fragile. Four pooled analyses point the same direction: a 2014 Cochrane review (PMID 25335553, n=1323) found 24-hour proteinuria fell by about 0.53 g and creatinine improved mainly in more advanced disease; a 2019 meta-analysis (PMID 31034954, 66 RCTs, n=4785) and a 2024 meta-analysis combining astragalus with RAAS blockers (PMID 38836372, 32 RCTs, n=2462) both reported less proteinuria; and a 2024 oral-astragalus RCT (PMID 38810556, n=118) showed eGFR declining about 4.6 mL/min/year slower, though albuminuria did not improve.
Quality is the problem. Every source rated the evidence LOW, with extreme heterogeneity (I-squared = 95%), clear publication bias, and almost all trials run in China with poor randomization and blinding. Most used intravenous astragalus injections unavailable outside mainland China, and no study reported hard outcomes such as death or progression to dialysis.
Regulators and clinics offer no support: the US FDA recognizes only tragacanth gum (a different Astragalus species) as a food additive, not astragalus supplements for kidney disease, and major patient-education sources (Mayo Clinic, Cleveland Clinic, Harvard Health, NIH Office of Dietary Supplements) are silent on this use. A drug-interaction flag also applies, which matters for kidney and diabetes patients who are often on multiple medications.
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 = 0.58
- 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 — | B→C 因 scope.conflation_risk=true 且 L11 獨評較低 (B7-2 tier cap)
- detect_disputes — 偵測到 1 個 hard + 0 個 soft dispute
- decide_status — 依 tier + dispute 結果決定 status