Cacao Flavanols: The Unexpected Cardiovascular Trial

← Back to articles

Cacao flavanols — the bitter polyphenols in dark chocolate and cocoa — have been studied for decades as putative cardiovascular protectants. In 2022, the COSMOS trial (n=21,442) provided the first large randomised, placebo-controlled evidence that a cocoa flavanol supplement reduces cardiovascular death. The result was modest and didn't hit the primary endpoint, but it shifted the evidence landscape.

The COSMOS results

Over a median of 3.6 years, participants randomised to 500 mg/day cocoa flavanols (including ~80 mg epicatechin) had a 10% reduction in the primary composite cardiovascular events outcome that did not reach statistical significance (HR 0.90, 95% CI 0.78–1.02). Cardiovascular death — a secondary endpoint — was reduced by 27% (HR 0.73, 95% CI 0.54–0.98). Sensitivity analyses adjusted for non-adherence pushed several outcomes toward stronger effects (Sesso 2022, COSMOS, American Journal of Clinical Nutrition; the trial registry record is NCT02422745).

Mechanism

Cacao flavanols, especially the catechin/epicatechin fraction, raise endothelial nitric-oxide signalling, improve flow-mediated dilation, modestly lower blood pressure, and have mild anti-platelet effects. A Cochrane review of 35 short-term trials concluded that cocoa products lower systolic BP by about 1.8 mmHg and diastolic by 1.8 mmHg in adults, with greater effects in hypertensives (Ried 2017; PMID 28439881; DOI 10.1002/14651858.CD008893.pub3). An earlier meta-analysis of cardiovascular surrogate endpoints reached similar conclusions on lipid and endothelial outcomes (Hooper 2012, American Journal of Clinical Nutrition).

Dark chocolate is not the same

COSMOS used a standardised cocoa flavanol supplement (500 mg flavanols, ~80 mg epicatechin per daily dose; the CocoaVia formulation). Commercial dark chocolate delivers a highly variable 50–500 mg flavanols per 100 g, and "Dutched" (alkalised) cocoa loses most of its flavanol content. Eating enough dark chocolate daily to match COSMOS would deliver several hundred kcal of sugar and fat — not a serious cardiovascular strategy.

Practical options

If you want what was tested, the standardised cocoa flavanol supplement is the only direct match. If you prefer food, choose minimally processed dark chocolate (>70% cocoa, no "processed with alkali"). Real-world dietary intake almost certainly produces a smaller effect than the trial because flavanol content varies and absorption depends on processing.

Who benefits most

Middle-aged and older adults with cardiovascular risk factors appear to be the population most likely to gain something. No major harms in COSMOS apart from mild GI upset and the calorie cost of any chocolate-based approach. Cocoa is not a statin substitute, but a reasonable adjunct for people not eligible for or not taking lipid-lowering therapy.