Intervention
Cocoa extract
Research Observatory
Evidence Update
Current evidence
Cocoa extract supplying 500 mg/day flavanols was compared with placebo in COSMOS, including cognitive outcomes in 2,262 COSMOS-Mind participants and cardiovascular outcomes in the 21,442-person parent trial. The multivitamin assignment is separate within the factorial design. COSMOS cocoa extract did not significantly improve the prespecified cognitive or total cardiovascular primary outcomes; favorable secondary cardiovascular mortality is not a positive trial-wide verdict.
What changed
Cocoa extract did not meet the selected cognitive or total cardiovascular primary endpoints; cardiovascular-mortality and per-protocol signals remain secondary context.
What did not change
Small score shifts or age-equivalent calibrations are not demonstrated rejuvenation.
Why it matters
Separates ingredient attribution, endpoint hierarchy and participant denominators.
Main limitation
Supplement is not ordinary chocolate; secondary and per-protocol analyses cannot silently replace intention-to-treat primary outcome.
Next proof
Confirm clinical signal independently with prespecified hierarchy.
Summary
Cocoa extract supplying 500 mg/day flavanols was compared with placebo in COSMOS, including cognitive outcomes in 2,262 COSMOS-Mind participants and cardiovascular outcomes in the 21,442-person parent trial. The multivitamin assignment is separate within the factorial design. COSMOS cocoa extract did not significantly improve the prespecified cognitive or total cardiovascular primary outcomes; favorable secondary cardiovascular mortality is not a positive trial-wide verdict.
Intelligence Take
Separates ingredient attribution, endpoint hierarchy and participant denominators.
Limitations
Supplement is not ordinary chocolate; secondary and per-protocol analyses cannot silently replace intention-to-treat primary outcome.