A plant compound from pomegranates, walnuts and berries that barely enters the blood and appears to act mostly through what gut bacteria convert it into. About half of adults lack those bacteria. Clearest human findings: blood fats, liver fat and liver enzymes; blood sugar is disputed. Safety is quiet at tested doses, and the evidence base is thin. (Full Review)
| Marker | Target | Why |
|---|---|---|
| Triglycerides | Below 80 mg/dL | Most consistent responder across both meta-analyses |
| HDL cholesterol | Above 60 mg/dL | The second lipid marker that moved in pooled human data |
| LDL cholesterol | 70 to 100 mg/dL | Detects the conflicting upward signal seen with polyphenol mixtures |
| Fasting glucose | 75 to 86 mg/dL | Safety check for additive lowering alongside glucose-lowering drugs |
| HbA1c | 4.8 to 5.3% | Confirms whether any glucose change is real or day-to-day noise |
| Fasting insulin | 2 to 5 µIU/mL | Tracks insulin resistance, the endpoint the two meta-analyses dispute |
| High-sensitivity C-reactive protein | Below 0.5 mg/L | The inflammatory marker that fell most consistently in trials |
| ALT | 10 to 26 U/L for men, 8 to 22 U/L for women | Primary liver readout in the fatty liver trials |
| GGT | Below 20 U/L for men, below 15 U/L for women | Second liver enzyme, and a sensitive marker of oxidative stress |
| Urinary urolithin A | No established target; track presence or absence instead | Classifies producer status and predicts whether response is possible |
Cadence: Baseline, then 8 to 12 weeks, then 6 months, then every 6 to 12 months if use continues