What COSMOS Found About Multivitamins and Cognition

Jordan Harbinger: I had heard that multivitamins were “expensive urine.” You said they may help with age-related cognition. What did the research find?

Dr. Rhonda Patrick: COSMOS was one large randomized factorial trial with several cognitive substudies. Older adults received a standard Centrum Silver multivitamin or placebo. The cognitive substudies used different testing methods and included overlapping parts of the same trial program, so they are not three independent replications.

Across the pooled cognitive analyses, the multivitamin group had a small average advantage in global cognition and episodic memory over two to three years. The pooled effect was about 0.07 standard deviation for global cognition and 0.06 standard deviation for episodic memory.

Investigators translated the global-cognition difference into about two years of age-related cognitive change. One episodic-memory comparison was translated into about 4.8 years. These are model-based age-equivalent comparisons of test scores. They do not mean that a brain became literally two or five years younger. The studies did not measure structural brain age, prevent dementia, or show longer life.

The result is encouraging because a standard multivitamin can help fill selected dietary gaps. It does not prove that every older adult has the same gap or that everyone over 65 should take one.

Most standard multivitamins provide vitamins and minerals. They do not usually provide meaningful essential fatty acids, protein, essential amino acids, or fiber. Centrum Silver also contains small amounts of selected carotenoids. A multivitamin cannot replace a varied diet.

Population surveys show that many U.S. adults consume less than estimated requirements for nutrients such as magnesium, calcium, and vitamins A and C. Intake below an estimated requirement is not the same as a diagnosed clinical deficiency, and prevalence differs by age, sex, diet, survey method, and whether supplements are counted.

Jordan Harbinger: The trial included thousands of people and used a placebo. Does that make a multivitamin an obvious choice for parents or grandparents?

Dr. Rhonda Patrick: The randomized design strengthens the small cognitive finding. It does not make a multivitamin harmless or necessary for everyone. The main COSMOS trial did not show significant reductions in invasive cancer, cardiovascular events, or mortality, and earlier multivitamin cognition research was mixed.

Safety depends on the formula and the person. Vitamin K can interact with warfarin. Iron, preformed vitamin A, vitamin B6, iodine, selenium, zinc, and other nutrients can become excessive when products are stacked. Pregnancy, smoking history, kidney or liver disease, diet, medicines, and laboratory findings can change what is appropriate.

Claims that depleted soil prevents people from obtaining adequate nutrients are too broad. Nutrient content varies by crop, cultivar, geography, agricultural practice, storage, and diet. This does not establish that everyone needs a supplement.

Jordan Harbinger: Do you recommend a specific product?

Dr. Rhonda Patrick: Product quality matters. Independent finished-product verification can reduce uncertainty about identity, potency, contaminants, and banned substances, but certification does not prove that one brand improves cognition or is clinically superior.

Iron should be taken for a demonstrated need or clinician-directed indication. Many adult men and postmenopausal women do not need routine supplemental iron, but either group can become deficient from blood loss, donation, low intake, malabsorption, or disease. Menstruating or athletic women do not automatically need iron. A complete blood count, ferritin, transferrin saturation, symptoms, and clinical context guide the decision.