Little of this common plant pigment reaches the blood unchanged. The clearest human signal is arteries widening better; a narrower one is restored blood sugar handling where it is already impaired. Blood fat, weight, liver, memory and recovery claims fade once all trials are counted. Safety unremarkable; stomach upset and mislabeled products are the practical concerns. (Full Review)
| Marker | Target | Why |
|---|---|---|
| High-sensitivity C-reactive protein | Below 0.5 mg/L | Primary inflammation target and the strongest predictor of who responds |
| Fasting glucose | 75–86 mg/dL | Tracks the best-supported metabolic effect |
| Hemoglobin A1c | 4.8–5.3% | Three-month average blood sugar, the endpoint moved in animal work |
| Fasting insulin | 2–5 µIU/mL | Detects insulin-sensitivity change earlier than glucose |
| Apolipoprotein B | Below 80 mg/dL | Tests the contested lipid claim honestly, counting particles rather than cholesterol mass |
| Alanine aminotransferase | 10–26 U/L (men), 10–19 U/L (women) | Liver signal that moved in the fatty liver pilot |
| Ferritin | 30–100 ng/mL | Guards against the iron-binding interaction |
| Blood pressure | Below 120/80 mmHg | Cheapest read on vascular effect, though pooled trials show no change |
| Flow-mediated dilation | No established target; track change from the individual’s own baseline | Directly measures the best-supported benefit |
Cadence: Baseline panel in the week before the first dose, repeated at 12 weeks, then every 6 to 12 months on continued use. Daily capillary glucose for the first 4 weeks on glucose-lowering drugs; clotting-time rechecks at 1 and 2 weeks on warfarin.