What specific multivitamin formulations and doses were used in the COSMOS cognitive substudies?
Executive summary
The COSMOS cognitive substudies tested a daily commercial multivitamin‑mineral (MVM) capsule versus placebo alongside a cocoa flavanol arm; the trial documents identify the MVM as a commercially available formulation (Centrum Silver was used as the reference product for COSMOS) with modest formulation changes from earlier trials, including added lutein and lycopene for the COSMOS formulation [1] [2] [3]. The public papers and trial website describe the product as a once‑daily multivitamin but the detailed per‑nutrient doses are reported only in the AJCN supplemental materials and the trial protocol rather than in the headline text of the published results [3] [4].
1. What was given: “a commercial multivitamin‑mineral” — the label and dosing schedule
COSMOS randomized participants to take a daily commercial multivitamin‑mineral supplement or matching placebo; published trial summaries and the COSMOS website repeatedly describe the intervention as a once‑daily MVM given for the duration of follow‑up (median ~3.6 years in the full trial, with 2–3 year cognitive follow‑ups in the substudies) [2] [5] [1]. The COSMOS‑Mind and COSMOS‑Clinic reports explicitly refer to a “commercial multivitamin‑mineral (MVM) versus placebo” given daily, confirming the operational dose regimen was one tablet per day [1] [6].
2. Which product formulation was used and how it differed from past trials
Investigators state that the COSMOS MVM corresponds to a Centrum Silver formulation and that the COSMOS formulation included subtle differences from prior trials such as the Physicians’ Health Study II (PHS II), notably the addition of lutein and lycopene for COSMOS [3]. The AJCN article cites Supplemental Table 2 to document those formulation distinctions, indicating the trial team relied on a standard high‑quality, multi‑micronutrient commercial product rather than designing a bespoke combination for the study [3].
3. What the public sources do not provide—nutrient‑by‑nutrient doses
While multiple COSMOS publications, the trial website, and press summaries confirm a once‑daily commercial MVM (and note the Centrum Silver link and the lutein/lycopene additions), the specific per‑nutrient doses (milligrams or IU for each vitamin and mineral) are not presented in the main text excerpts available in these sources; the AJCN paper refers readers to a supplemental table for precise formulation details, and the full ClinicalTrials.gov protocol archives that ingredient lists should be consulted for exact dosing [3] [4]. Therefore, the publicly cited summaries alone are insufficient to enumerate exact microgram/milligram amounts for each vitamin and mineral without consulting the AJCN supplemental materials or the full protocol documents.
4. Where to find the exact formulation and why it matters
The original AJCN publication notes that precise differences versus PHS II are detailed in Supplemental Table 2, which is the most direct source for exact doses used in COSMOS [3]; similarly, the COSMOS study protocol and ClinicalTrials.gov entry host large‑document supplements where intervention details are often archived [4] [2]. Knowing exact doses matters because prior null or positive results for single vitamins or different MVMs have been attributed to differences in formulation (for example, presence/absence of lutein and lycopene or variations in B‑vitamin levels), and COSMOS authors themselves flag formulation distinctions as a plausible explanation for differing cognitive results across trials [3].
5. Context, caveats and competing readings
Reports and investigator statements frame the COSMOS MVM as “typical” and accessible, and meta‑analyses across COSMOS substudies concluded modest benefits on global cognition and episodic memory from the daily multivitamin [7] [6], yet at least one commentator cautioned about methodological interpretation even as investigators emphasized the promise of an affordable intervention [8]. Importantly, the research community has not isolated which specific micronutrients — if any — drive cognitive benefit; COSMOS papers acknowledge that the multi‑nutrient package was tested as a whole and that individual nutrient effects cannot be determined from these analyses alone [9] [3].