Cosmos study
Executive summary
The COcoa Supplement and Multivitamin Outcomes Study (COSMOS) is a large, randomized, double-blind, placebo-controlled trial of daily cocoa flavanol extract and a standard multivitamin in older U.S. adults that enrolled about 21,442 participants to test effects on cardiovascular disease, cancer, cognition and other aging outcomes [1] [2]. Key published results show no clear reduction in total cardiovascular events from cocoa extract but a 27% reduction in cardiovascular death in primary analyses and signal-consistent cognitive benefits from multivitamin use in multiple COSMOS cognitive substudies, while investigators and funders stress the need for replication and careful interpretation [3] [4] [5].
1. Trial design and scale: why COSMOS matters
COSMOS randomized roughly 21,442 generally healthy men and women, aged 60+ for men and 65+ for women, in a 2×2 factorial design that tested a 500 mg/day cocoa flavanol extract and a commonly used multivitamin against placebo, with planned ancillary studies for cognition, eye disease and other endpoints—making it one of the largest long-term dietary supplement intervention trials to date [1] [6] [2].
2. Cardiovascular and mortality findings: mixed signals, guarded optimism
Primary published analyses found that cocoa extract supplementation did not significantly reduce total cardiovascular events, although the trial reported a 27% reduction in cardiovascular death and supportive signals in per-protocol analyses that the authors called warrant further study rather than definitive proof [3] [7]. COSMOS investigators and affiliated write-ups urge caution because confidence intervals, multiplicity of comparisons, and per-protocol vs intention-to-treat differences complicate interpretation and because authors explicitly called for additional research to clarify effects [3] [6].
3. Cognitive results: replicated benefits for multivitamins within COSMOS
Across COSMOS cognitive ancillary studies—including a web-based battery and an in-person clinic subcohort meta-analysis—daily multivitamin-mineral (MVM) supplementation showed statistically significant benefits for global cognition and episodic memory, with effect sizes equivalent to roughly 2–3 years less cognitive aging and preserved benefits over up to three years of follow-up in some analyses [4] [8] [5]. Authors describe these as the first evidence from a large randomized trial that typical daily multivitamins may improve memory and thinking in older adults, but they also call for replication in more diverse populations and biomarker-driven studies [4] [5].
4. Safety, generalizability and unresolved questions
COSMOS reports that both interventions were well tolerated and that long-term MVM use showed no notable adverse effects in the datasets referenced, yet investigators emphasize limits: the cohort skews older and healthier, ancillary cognitive samples are smaller, and findings may not generalize broadly without confirmatory trials that include more diverse populations and mechanistic biomarkers [5] [6] [9].
5. Funding, industry involvement and potential biases
COSMOS received investigator-initiated support from Mars Edge (including study pills and packaging) and in-kind support from Pfizer Consumer Healthcare for pills and packaging, alongside NIH grants; COSMOS and multiple reporting outlets acknowledge this support and recommend careful interpretation while maintaining that study oversight included institutional review and standard trial governance [7] [10] [11]. Industry funding and provision of study materials do not invalidate findings but are a relevant context for assessing potential implicit agendas and the need for independent replication [7] [10].
6. Bottom line and next steps for science and consumers
COSMOS contributes high-quality randomized evidence that a common multivitamin may modestly slow cognitive aging in older adults and that cocoa flavanol extract shows promising but not definitive cardiovascular signals, yet both conclusions require replication, wider population testing, and mechanistic work using biomarkers and omics to identify who benefits most; investigators and funders uniformly call for follow-up trials rather than immediate clinical mandates [4] [3] [5].