Story
Multivitamin · #1 supplement worldwide
Last updated
In one pass If your diet is roughly in order, the daily multivitamin mostly has nothing to do. Not this — Take a multivitamin daily 'just in case' — In the PHS-II trial, a multivitamin did nothing for cardiovascular disease and cut total cancer by just 8%. The gap was never a pill.
Educational content, not medical advice — consult a clinician.
Story path
Chapter 1
Why more stops helping
If your diet is roughly in order, the daily multivitamin mostly has nothing to do. In decades of large randomized trials, it did not give healthy people less heart disease, and for cancer only one trial, in male doctors, saw a very small drop.
It was never designed for people who are already well. When vitamin C runs out, the step in your body that twists collagen into cables stalls: the cables cannot be wound tight, blood-vessel walls and gums lose their support, and so gums bleed and old wounds reopen. Put vitamin C back and the step starts up again, and the person recovers quickly. Repairing the gap from none to enough was the whole battlefield where vitamins first proved themselves.
The key point is that it is a switch, not a dial. From zero to enough, the body changes dramatically; from enough upward, the curve is already flat. So filling a gap and taking more to be healthier are two different things, and most of the large trial results that follow can be read from this.
If there are children in your home, lock up any multivitamin that contains iron. If you suspect a child has swallowed iron-containing tablets, do not wait at home for symptoms: contact poison control or go to the emergency department now.
It was never designed for people who are already well. When vitamin C runs out, the step in your body that twists collagen into cables stalls: the cables cannot be wound tight, blood-vessel walls and gums lose their support, and so gums bleed and old wounds reopen. Put vitamin C back and the step starts up again, and the person recovers quickly. Repairing the gap from none to enough was the whole battlefield where vitamins first proved themselves.
The key point is that it is a switch, not a dial. From zero to enough, the body changes dramatically; from enough upward, the curve is already flat. So filling a gap and taking more to be healthier are two different things, and most of the large trial results that follow can be read from this.
If there are children in your home, lock up any multivitamin that contains iron. If you suspect a child has swallowed iron-containing tablets, do not wait at home for symptoms: contact poison control or go to the emergency department now.
Mechanism · Why the curve flattens once you have enough
The way the body handles vitamins has a ceiling built in. Why does the effect curve go flat once you go past enough? The answer comes in two halves, and the two halves run in opposite directions.The water-soluble half (the B vitamins and vitamin C) runs into two gates. The first is in the small intestine: each vitamin is carried into the gut-lining cells by its own set of transporter proteins, and there are only so many transporters. Flood in too much at once, and whatever cannot get in line simply moves on down the gut. The second gate is in the kidney: after the kidney's filters strain these molecules out of the blood, the tubules pull back the part the body can still use, and that pull-back also has a limit, so the excess stays in the urine. So with a large dose of a water-soluble vitamin, a good share goes in through the mouth and out in the urine, spending only a few hours in the blood on the way. Riboflavin (vitamin B2) is bright yellow itself, and the yellow urine after a multivitamin is it being excreted. That shows neither good absorption nor total waste; it only shows that your intake has passed what the body needs right now.
The fat-soluble half (vitamins A, D, E, and K) has neither gate. They do not dissolve in water and travel the route fat takes. In the small intestine they first have to be wrapped by bile acids into tiny oil droplets (mixed micelles) before they can be unloaded at the gut wall. Once inside the gut-lining cells, they do not go straight into blood vessels either: they are packed into chylomicrons, large fat-carrying particles, take a detour through the lymph, and only then join the blood, and most of them end up stored in the liver and in body fat. What can be stored cannot be cleared fast. That one sentence explains two things that look unrelated: why fat-soluble vitamins can carry you through a stretch without sunshine or greens, and why, at high doses over a long time, they are the main class most likely to build up in the body to a harmful level.
Put the two halves together and you get the difference between a switch and a dial. In someone with a real gap, every bit supplemented is used at once and symptoms improve. In someone who already has enough, the water-soluble surplus is in the toilet within hours and the fat-soluble surplus stays in the liver. The first is repairing a broken line; the second is only adding pressure to a line that is already open, and the flow through a pipe is set not by the pressure but by its narrowest stretch.
Background · How vitamins were found
A single vitamin curing a specific deficiency disease was a genuine miracle in its day. The multivitamin business grew out along the line extending from that miracle.The story starts earlier. In 1747, James Lind ran one of the first controlled trials in history aboard the British navy ship HMS Salisbury: 12 sailors with scurvy were split into 6 groups, each given something different, and the group given oranges and lemons recovered fastest, able to return to duty within days. It was the first demonstration that something in food could cure a disease (though he did not isolate vitamin C; that had to wait for Szent-Györgyi in 1932).
In the first half of the twentieth century, a handful of physicians and chemists found the vitamins one after another. In 1912, the Polish chemist Casimir Funk gave this class of trace molecules essential for life the name vitamine, and proposed that scurvy, pellagra, beriberi, and rickets were each caused by missing one such factor. By the 1920–1940s, vitamins C, D, A, the B vitamins, K, and B12 had been isolated one by one, their structures worked out, and then made industrially.
The problem came with the next step, the generalization: if adding one vitamin cures one disease, would adding many at once prevent many chronic diseases? From the 1940s on, that hope carried the modern multivitamin industry, and later Pauling and others pushed it toward megadoses.
Funk's generation was not wrong: deficiency diseases are real, and filling the gap cures them. What was wrong was the extension, take a little more and you will get sick less. Over the next 50 years that sentence was tested again and again in (RCTs), and most of it did not hold. Looking back, its mistake was taking the steep slope at the far left of the curve for the slope of the whole curve.
Background · The name, and food fortification
Funk's insight in 1912, and a small naming mistake:He was studying the factor in rice bran that prevents beriberi (what later became vitamin B1, thiamine)He found that this class of substances contained nitrogen and belonged to the aminesSo he combined vital amine into one word, vitamine, as a name for this whole class of organic trace molecules essential for lifeIt later turned out that not every vitamin is an amine (vitamin C is not, and neither is vitamin D), so in 1920 Drummond proposed dropping the final e, which gave today's vitamin
Funk's other insight was to group the era's deficiency diseases into one class. Food fortification programs in many countries later followed that line of thought; in the United States, for example:
Iodized salt (1924, Michigan): to prevent goiterVitamin D–fortified milk (1933): to prevent ricketsEnriched flour (1941): iron, niacin, thiamine, and riboflavin added to prevent beriberi, pellagra, and iron-deficiency anemiaFolic acid–fortified grain (1998): to prevent neural-tube defects (NTD)
This is often counted among the most successful public-health measures of the 20th century: at very low cost, it nearly eliminated the classic deficiency diseases in wealthy countries.
But the fact that fortified food plus an adequate diet made deficiency diseases almost vanish also means that for an ordinary adult in a wealthy country in the 21st century, the benefit a multivitamin can offer was limited from the start. That is one reason the later large trials found so little: when most people's blood nutrient levels are already in the adequate range, adding a bit more does not make them healthier.
Chapter 2
What's actually in the bottle
Tip a multivitamin into your palm and you see one tablet. To your small intestine, it is dozens of nutrients crowding into the same stretch of gut in the same minute.
The trouble is in that word crowding. Iron, zinc, copper, and manganese are all metal ions carrying two positive charges, and to get into the gut-lining cells some of them use the same or similar channels, so they compete. When they arrive together, whichever is at higher concentration takes the doorway first, and the rest slide past and leave in the stool. Calcium takes a different route, but when there is a lot of it, iron and zinc absorption still gets pushed down.
That is why the minerals in the bottle are often only a fraction of the recommended amounts. Half the reason is space (at full recommended doses the tablet would be too big to swallow); the other half is this competition, since the higher the doses are stacked, the worse the crush at the door. Taking them together and taking them apart have never been the same thing.
The trouble is in that word crowding. Iron, zinc, copper, and manganese are all metal ions carrying two positive charges, and to get into the gut-lining cells some of them use the same or similar channels, so they compete. When they arrive together, whichever is at higher concentration takes the doorway first, and the rest slide past and leave in the stool. Calcium takes a different route, but when there is a lot of it, iron and zinc absorption still gets pushed down.
That is why the minerals in the bottle are often only a fraction of the recommended amounts. Half the reason is space (at full recommended doses the tablet would be too big to swallow); the other half is this competition, since the higher the doses are stacked, the worse the crush at the door. Taking them together and taking them apart have never been the same thing.
Mechanism · Who crowds whom in the same pill
Several metals competing for the same stretch of entry is the most underrated design problem in a multivitamin.On the side of the small-intestine lining that faces the gut, there is a protein called the divalent metal transporter (DMT1), which carries metal ions with two positive charges into the cell. Iron (in its two-charge form), manganese, cobalt, and others can all be picked up by it, and the more kinds it accepts, the easier it is to jam. Swallow these together and they dissolve in the same stretch of gut and reach the same doorway at the same time; whichever is at higher concentration gets carried in more often, and the rest move on down and leave in the stool. Zinc and copper each have their own main channels, but the two compete with each other in a different way, described in the first point below.
The competition in this stretch explains several things on the label that look contradictory:
Too much zinc can cause copper deficiency. Once zinc gets into a gut-lining cell, it makes the cell produce a sticky molecule called metallothionein. Copper binds it even more tightly than zinc does, so copper gets stuck inside the cell. When that cell is naturally shed into the gut a few days later, the copper is thrown out with it. This is exactly how long-term high zinc leads to copper deficiency, and note the time lag: the extra zinc you take today costs you copper a few days later.Calcium and iron should not go in together. Calcium does not use this channel, but when there is a lot of it in the gut, iron absorption still drops. That is one reason multivitamins keep calcium low, and where the old advice to take calcium tablets and iron supplements at different times comes from.Vitamin C and iron, by contrast, are a pair. Iron in plant foods is mostly in the three-charge form, which this doorway cannot carry; vitamin C converts it to the two-charge form in the gut and holds on to it so other things cannot make it drop out, and only then can the iron get in. Vitamin C and iron in the same tablet is one of the few combinations that really help each other.Polyphenols in tea and coffee, and phytate in whole grains, bind metal ions directly in the gut into clumps that cannot be carried in, so none of them get through. That is a separate problem from which nutrient crowds which inside the tablet, but the losses add up: a strong cup of tea with an iron-containing multivitamin means both losses at once.
So the advice to take things separately is not superstition. If you really do need iron, taking it apart from the calcium-containing meal or tablet, with a little vitamin C, makes a clear difference to how much actually reaches your body. A has-everything multivitamin does exactly the opposite: it pins everything that interferes with everything else to the same minute. That is also why it cannot replace targeted supplements. It is not that the doses are too small to print on the label; it is that sharing one tablet capped each nutrient's dose from the start.
In practice · The label lines worth reading
Turn over a typical men's 50+ formula and the back of the label is roughly these groups piled together:Thirteen vitamins: A, C, D, E, and K, plus the full set of B vitamins (B1 through B12), folate, biotin, and pantothenic acidTen to twenty minerals: calcium, magnesium, iron, zinc, copper, manganese, selenium, chromium, molybdenum, iodine, potassium, and othersA few extras, such as lutein and lycopene, and blends labeled for eye or heart healthA pile of fillers and binders: magnesium stearate, microcrystalline cellulose, silicon dioxide, colorings, and a sugar coating, used to press all of the above into one swallowable tablet
There is an easy-to-miss imbalance in the doses. Most vitamins are set at 100% of the US Recommended Dietary Allowance (), and a few B vitamins and vitamin D go somewhat above it. Minerals, though, are almost always set low: they are heavy and bulky, a full RDA dose would make the tablet too big to swallow, and they also react with each other in the bottle. So a typical tablet often has only 200–300 mg of calcium (under a third of the RDA) and only 50–100 mg of magnesium (under a quarter). That is why a multivitamin cannot replace targeted calcium or magnesium supplements.
A few ingredient forms on the label are worth a glance:
Folate is usually the cheap synthetic form, folic acid. People who carry two copies of the TT variant at the C677T site (a share that varies widely by ancestry; NIH ODS gives about 25% of Hispanic people, about 10% each of White and Asian people, and about 1% of African Americans) convert it to the active form less efficiently. The clinical significance is limited, though, so there is no need to pay more for a version labeled as active folate ().Vitamin D is usually (cholecalciferol); a few products still use the less potent D2 (ergocalciferol).Vitamin K is K1 (phylloquinone) by default; premium brands switch to (menaquinone), which is fine, but not worth paying extra for.Iron is usually left out of men's and 50+ formulas. Adult men and postmenopausal women usually are not short of it, and the body has no active way to get rid of iron, so any excess can only be stored in the liver, heart, and pancreas. Two warnings here matter more than taking too much. If there are children in the home, the bottle with iron must be locked away: children swallowing adults' iron tablets is one of the most common causes of fatal acute poisoning in young children, which is why US regulations require a warning on every solid iron-containing product. If you suspect a child has swallowed iron-containing tablets, do not wait at home for symptoms: contact poison control or go to the emergency department now. And people with hereditary hemochromatosis (most often caused by mutations in both copies of the HFE gene) already absorb too much iron, so taking iron blindly adds fuel to the fire; the first step is testing and transferrin saturation.Beta-carotene: some formulas use it to supply part of the vitamin A. The amount in an ordinary formula is far below what the ATBC and CARET trials used, but smokers should not take extra high-dose beta-carotene on top (the reason is in the chapter Why the big trials failed).
Finally, the prominent 100% Daily Value (DV) on the bottle is easy to misread. The DV is a single reference figure for everyone, and for a long time it was still based on the 1968 RDA. It is not the latest nutrition advice, and it is even less a measure of what you personally lack right now. It describes a statistical average person, and that person does not exist.
nih-ods-ironnih-ods-folate
In practice · Which meal to take it with
The same tablet, swallowed on an empty stomach or with a meal that has some fat, can put very different amounts into your blood, and the whole difference is in the four fat-soluble vitamins.Vitamins A, D, E, and K do not dissolve in water. In the gut they first have to be wrapped by bile acids into tiny oil droplets (mixed micelles) before they can be delivered to the surface of the gut-lining cells and unloaded. Bile is not released all the time: it is squeezed out only after you eat fat and the first part of the small intestine senses it. Swallow a multivitamin on an empty stomach, and little bile comes out, the micelles cannot form, and a good share of these four vitamins simply passes through you. You paid full price and got only a discounted part.
The water-soluble group is the other way round. The B vitamins and vitamin C dissolve in the gut fluid and are carried in by their own transporters, with no need for fat at all. Their limit is at the other end: once the transporters are saturated, the extra never reaches the blood, and whatever does reach the blood beyond what the kidney tubules can pull back ends up in the urine within hours.
Put those two facts together and you get a practical tip that is almost never printed on the bottle: take your multivitamin with the fattiest meal of the day. Same pill, same money, same dose; only the timing changes, and the four fat-soluble vitamins actually arrive.
While we are here, a common misunderstanding runs the other way. Some people take fat-soluble vitamins on an empty stomach on purpose because they fear buildup. That only stops them from being absorbed; it does not make them absorbed more safely. What decides the risk of buildup is how much you take each day, not when you swallow it. To control the risk, adjust the dose; do not use absorption as a brake.
Myth · Are natural or pricier vitamins better?
That natural vitamins are better than synthetic ones is a marketing line, and in most cases it does not hold:Vitamin C: ascorbic acid is ascorbic acid; whether it comes from oranges or a factory, the molecule is the same. Contains rose hips or contains acerola is just marketing; the vitamin C that actually comes from rose hips in one pill is tiny (on the order of mg).Vitamin E: here there really is a difference:d-alpha-tocopherol (natural) and dl-alpha-tocopherol (synthetic) really do differ in biological activity (roughly 1.4 to 1.0)More important, though, the vitamin E in food is a family of 8 forms (four tocopherols and four tocotrienols), while a supplement usually gives only 1 of them, alpha-tocopherol, which cannot stand in for the whole familyWhole foods (almonds, sunflower seeds, spinach) naturally contain alpha, beta, gamma, and delta tocopherols along with tocotrienols, a more complete setVitamin D: (cholecalciferol) is clinically equivalent whether it comes from sheep's wool grease, lanolin (natural), or chemical synthesisB vitamins: the difference between synthetic and natural is essentially negligible
Expensive equals good is often wrong too:
Independent testing has shown that cheap brands such as Costco's Kirkland and Walgreens do about as well as premium brands on whether the label content is accurateProducts labeled organic, natural, or whole-food often cost 5–10 times as much, with little nutritional difference
A few cases where paying more for quality is genuinely worth it:
Third-party certification (USP, NSF, ConsumerLab): confirms the label content is accurate and there is no heavy-metal contaminationA clear need for higher purity (pregnancy, or sensitivity to particular ingredients)Specific forms (such as rTG fish oil, or B12 in the methylcobalamin form), though the great majority of people do not need them
A pharmacy's own brand versus an international brand makes no meaningful difference for ordinary use; putting the money into a varied diet pays back more.
Chapter 3
Who actually benefits
Multivitamins do nothing and multivitamins save lives can both be true, because the two sentences are about different people.
The dividing line is not in the pill; it is whether one of your supply lines is actually broken. B12 in vegans is the clearest example. Only bacteria make this molecule; animals store it in their liver and muscle, and plants contain almost none. After a long cutoff, the myelin that wraps nerve fibers like insulation can no longer be repaired, the hands and feet start to tingle and go numb, walking feels unsteady, and this damage may not fully reverse even after B12 is restored. In that situation, the tablet you swallow is reconnecting a broken line.
The line can also break at the absorption end: too little stomach acid, part of the small intestine bypassed by surgery, or a long-term medicine that keeps dragging one nutrient down. Plenty goes in, but little reaches the blood.
In someone who eats a varied diet and digests normally, the line is open. Add more on top, and what you add has nowhere to go.
The dividing line is not in the pill; it is whether one of your supply lines is actually broken. B12 in vegans is the clearest example. Only bacteria make this molecule; animals store it in their liver and muscle, and plants contain almost none. After a long cutoff, the myelin that wraps nerve fibers like insulation can no longer be repaired, the hands and feet start to tingle and go numb, walking feels unsteady, and this damage may not fully reverse even after B12 is restored. In that situation, the tablet you swallow is reconnecting a broken line.
The line can also break at the absorption end: too little stomach acid, part of the small intestine bypassed by surgery, or a long-term medicine that keeps dragging one nutrient down. Plenty goes in, but little reaches the blood.
In someone who eats a varied diet and digests normally, the line is open. Add more on top, and what you add has nowhere to go.
Mechanism · How the absorption end breaks
Swallowed is not the same as absorbed; there is a whole assembly line in between that can break. B12's line is the longest in the body, which makes it the clearest example:① Unhooking in the stomach: stomach acid and the enzyme pepsin first pry B12 off the food protein it is bound to. When acid is low (atrophic gastritis, which is common in older adults, or long-term acid-suppressing drugs), this step fails. A counterintuitive conclusion follows: B12 in a supplement is already free and does not need this step, so a person with low stomach acid actually gets more from a tablet than from meat.② Changing escorts: cells in the stomach wall release a protein called intrinsic factor that grabs B12 and escorts it the rest of the way. When the immune system destroys those cells, the result is pernicious anemia; then no amount of oral B12 can connect, because there is no escort, so this situation calls for injections that bypass the whole gut.③ Delivery at the end of the ileum: the cells in the last stretch of the small intestine carry receptors that recognize intrinsic factor, and only there does B12 actually enter the body. If that stretch was cut out in surgery or damaged by inflammation (Crohn's disease, some weight-loss operations), getting the first two steps right still does no good.
Why weight-loss surgery means supplementing for life comes down to the same question of where the line broke. The stomach is made smaller, so both acid and intrinsic factor drop; the duodenum and the first part of the jejunum are bypassed, and that is exactly where iron and calcium absorption is most concentrated. This is not a problem of eating less; it is a problem of the road being rerouted. So what is needed after surgery is a targeted formula, not a general one off the shelf, and for life, because the detour will not grow back.
Medicines that drag nutrients down also each act at their own point. Acid-suppressing drugs block step one. Metformin, by current explanations, interferes with the ileal receptor's uptake of the intrinsic-factor complex, a step that depends on calcium ions. Diuretics wash electrolytes and water-soluble vitamins out in the urine together. Anti-seizure drugs speed up metabolic enzymes in the liver so that vitamin D is broken down faster. Four drugs, four completely different points, which is why the blanket line people on medication should take vitamins is not much use: which nutrient to supplement depends on which step the drug blocks.
So should I supplement? is the wrong question. The right one is where is my line broken? Broken at intake, changing your food is enough; broken at absorption, you need the right form and dose; broken at loss, you also have to ask whether you can leak less while you supplement.
Clinical · Who actually benefits
For these people, supplementing reconnects a broken line:Strict vegetarians and vegans: B12 is a must (it occurs naturally only in animal foods); long-term deficiency causes irreversible nerve damage. Iron, zinc, calcium, vitamin D, and often need attention too.Pregnancy and breastfeeding: folate (folic acid 400–800 µg) from before conception through early pregnancy. The evidence starts with the MRC randomized trial published in the Lancet in 1991: in women at high risk of a repeat because a previous pregnancy had a neural-tube defect (NTD), starting folic acid before pregnancy (at a higher treatment dose) cut the risk of it happening again by about 72%, while a mixture of seven other vitamins in the same trial showed no protective effect. Iodine (220 µg) supports the baby's brain development; iron needs rise throughout pregnancy; and vitamin D serves both the baby's bones and the mother's calcium balance. Most guidelines recommend a prenatal formula, not an ordinary multivitamin, and the difference is not only in dose. There is also a firm contraindication: preformed vitamin A (retinol, retinyl palmitate) in ordinary multivitamins and cod liver oil is an established cause of birth defects in early pregnancy. In a prospective cohort, pregnant women taking more than about 10,000 a day of preformed vitamin A from supplements had a clearly higher risk of birth defects arising from the cranial neural crest (Rothman 1995, NEJM; an observed association). Beta-carotene does not have this problem: the gut-lining cells convert it to vitamin A as needed, the body's need decides how much, and the rest is left unconverted. So prenatal formulas either replace vitamin A with beta-carotene or keep the preformed part very low, and that bottle of men's complete formula on the shelf, plus a spoonful of cod liver oil, is exactly where people most easily cross this line. Outside pregnancy there is a ceiling too: the tolerable upper intake of preformed vitamin A for adults is 3000 µg a day, and going above it for a long time may harm the liver and raise fracture risk.After weight-loss surgery (gastric bypass, sleeve gastrectomy): long-term malabsorption (B12, iron, calcium, vitamin D, and others) means a specialized post-surgery formula is needed for life.Older adults (75+): atrophic gastritis plus acid-suppressing drugs () lowers B12 absorption (this applies to about 15–20% of people over 60); the skin makes vitamin D less efficiently; and appetite itself shrinks. Cognition is not an extension of this B12 gap. The large trial of the last decade that tested a multivitamin on cognition is COSMOS-Mind (Baker 2023): average age 73, three years of telephone testing, a 0.07 difference in the global cognition z-score. The details of that figure are in the chapter Why the big trials failed: a small effect, a pre-specified secondary endpoint, and not dementia incidence. Do not read it as a reason for every older person to take a pill to stay sharp.Chronic illness and long-term medication: long-term PPIs pull down B12, calcium, and magnesium; long-term metformin pulls down B12 (in about 10–30% of people on it for more than 5 years); diuretics wash out potassium, magnesium, and vitamin B1; anti-seizure drugs pull down vitamin D, folate, and calcium; and chemotherapy, radiotherapy, and long-term steroids call for broad nutritional support.Severe dietary restriction, food allergies, eating disorders: very-low-calorie diets open several micronutrient gaps at once.
These people do not need it:
Healthy adults who eat a varied diet and digest normally (roughly the eating pattern of the US MyPlate guidelines).Can a daily multivitamin help healthy people live longer, get fewer chronic diseases, or have more energy? Large randomized trials mainly tested the heart and cancer questions, and the conclusion is no clinically meaningful benefit in sight. In PHS-II, a trial in male doctors, cardiovascular events did not differ; the closest thing to a positive cancer result was a very small 8% drop in total cancer in the same trial (both are in the chapter Why the big trials failed). Cognition is a separate line opened in the last decade: COSMOS-Mind found a small positive result on a secondary endpoint, which did not turn heart disease or cancer positive and did not show that dementia can be prevented. Claims like more energy were never endpoints in these trials at all: they have not been disproved, because they were never seriously tested in the first place.
This list has one thing in common that is worth a moment: for every item in the top half, you can point to where the line broke, whether there is no source, absorption is bypassed, a drug drags it down, or intake is too low. The bottom half sees no benefit not because nutrition does not matter, but because the line is open, and there is no break to find.
rothman-vitamina-1995nih-ods-vitamin-amrc-1991-folate-ntd
Clinical · What the USPSTF said in 2022
The 2022 recommendation of the US Preventive Services Task Force (USPSTF) was based on an evidence report that pooled 84 studies (N > 700,000). Those 84 were mostly , but not all: to assess serious harms, the report also included observational cohort studies.It applies to ordinary adults living in the community (not pregnant or breastfeeding women, and not people with a diagnosed nutritional deficiency):
Multivitamins (including minerals) to prevent heart disease and cancer: an I statement, meaning the evidence is insufficient to weigh benefits against harms. There is no evidence that it prevents them, and no evidence that it harms.Beta-carotene to prevent heart disease or cancer: the USPSTF gives a D recommendation, advising against use; it has moderate certainty that the harms outweigh the benefits (lung cancer risk rose in smokers in ATBC and CARET).Vitamin E to prevent heart disease or cancer: the USPSTF likewise gives a D recommendation, advising against use, with moderate certainty that it has no net benefit (in SELECT's extended follow-up, prostate cancer even rose in men taking vitamin E alone).
The I and D here are the USPSTF's own grades: I says whether the evidence is enough to reach a conclusion; D says there is enough certainty to advise against use.
In other words, this US preventive-medicine body does not recommend that healthy people take multivitamins to prevent chronic disease, and it explicitly advises against beta-carotene and vitamin E for prevention.
The contrast with the marketing:
Worldwide annual multivitamin sales are over $50 BAbout 1/3 of US adults and about 50% of older adults take a multivitaminBehind this is the insurance mindset: it can't hurt anywayBut the randomized trials of beta-carotene and vitamin E did measure real harm, and high doses of iron and calcium carry their own risks
Practical advice:
Healthy adults: the money for a multivitamin is better spent on vegetables, fruit, whole grains, and good-quality protein; a varied diet beats any bottle of multivitamins.People with a clear gap (pregnancy, a vegan diet, after weight-loss surgery, long-term use of certain drugs): choose the matching formula or a single-nutrient supplement, rather than making do with a bottle of men's complete.Taking a multivitamin for life as insurance: unnecessary, and in some cases possibly slightly harmful.
Chapter 4
Why the big trials failed
Antioxidants sound as if they can only help: free radicals are bad, so clearing them out should make you healthier? Large trials kept giving the opposite answer, and the snag is in that premise: free radicals are not only enemies; they are also signals the body is using.
After one of your white blood cells swallows a bacterium, it burns it up with a burst of reactive oxygen species (, highly reactive oxygen-containing molecules that can damage other molecules). After a workout, muscle also needs that burst of reactive oxygen before the programs for repair and adaptation switch on. By this mechanism, keeping the signal turned down year after year could mean what should grow does not, and what should be killed is killed more slowly, though the evidence for this step in people is still limited.
Beta-carotene may go one step further and flip sides. The lungs are where oxygen levels in the body are highest, and smoke adds heavy oxidative stress on top. One explanation is that in this setting, high-dose beta-carotene, once oxidized, switches from catching free radicals to releasing them. That helps make sense of the two trials in smokers that follow, but it has not been measured directly in people.
After one of your white blood cells swallows a bacterium, it burns it up with a burst of reactive oxygen species (, highly reactive oxygen-containing molecules that can damage other molecules). After a workout, muscle also needs that burst of reactive oxygen before the programs for repair and adaptation switch on. By this mechanism, keeping the signal turned down year after year could mean what should grow does not, and what should be killed is killed more slowly, though the evidence for this step in people is still limited.
Beta-carotene may go one step further and flip sides. The lungs are where oxygen levels in the body are highest, and smoke adds heavy oxidative stress on top. One explanation is that in this setting, high-dose beta-carotene, once oxidized, switches from catching free radicals to releasing them. That helps make sense of the two trials in smokers that follow, but it has not been measured directly in people.
Evidence · Five large trials, one at a time
Put the large vitamin trials of the past few decades side by side, and the conclusion is strikingly consistent: in the general population, antioxidant supplements have shown no clear benefit, and some doses pushed risk up. Here are the numbers, one trial at a time.Start with the two heaviest, both in smokers. Finland's ATBC trial (1985–1993, NEJM 1994) randomly split 29,133 male smokers into four groups: beta-carotene 20 mg a day, vitamin E 50 mg a day, both, or placebo, for 5–8 years, in the hope of reducing lung and other cancers. Instead, men taking beta-carotene had 18% more lung cancer and 8% higher death from any cause. It was the first hard evidence that beta-carotene raises cancer in smokers, and it overturned the prevailing expectation. The US CARET trial (1985–1996, NEJM 1996) saw the same direction in 18,314 smokers, former smokers, and asbestos-exposed workers: beta-carotene 30 mg plus 25,000 of retinyl palmitate (a form of vitamin A). After an average of 4.0 years, lung cancer was up 28% ( 1.28, 95% 1.04–1.57) and death from any cause up 17% (RR 1.17, 1.03–1.33), and the trial was stopped 21 months earlier than planned. The time scale deserves a pause: the harm did not jump out in year two; it took an average of four years of exposure to build up the difference. Dose times time is exactly what this chapter is trying to teach. One mechanistic explanation is that high-dose beta-carotene in a smoker's lungs, where reactive oxygen species () are already high, flips from antioxidant to pro-oxidant; this has not been measured directly in people. Since then, high-dose beta-carotene is no longer recommended for lowering cancer risk, and in 2022 the USPSTF explicitly advised against using it to prevent heart disease or cancer.
Vitamin E on its own did not do much better. The HOPE trial (NEJM 2000) gave 9,541 patients at high cardiovascular risk 400 IU a day for 4.5 years, with no benefit at all on the main cardiovascular endpoint. The SELECT trial split 35,533 men into four groups: selenium 200 µg, vitamin E 400 IU, both, or placebo, hoping to prevent prostate cancer. It is worth reading twice. At its first report in 2009 (JAMA), after a median follow-up of 5.46 years, prostate cancer was 13% higher in the vitamin E group but not statistically significant ( 1.13, 99% CI 0.95–1.35, P=0.06), and the conclusion then was only that it did not prevent it. The bad news landed after extended follow-up: the 2011 update (JAMA) reported that prostate cancer in the vitamin E group was up 17%, and now significant (HR 1.17, 99% CI 1.004–1.36, P=0.008). Same people, same intervention, a few more years of watching, and a non-significant hint grew into a confirmed harm. The 17% you see quoted elsewhere comes from the 2011 report, not the 2009 paper.
Closest to the everyday question is the PHS-II trial, the largest randomized trial of long-term multivitamin use in healthy people: 14,641 US male doctors aged 50 or older took one pill a day, with a median follow-up of 11.2 years. Its results were published as two papers, each covering one endpoint, and they are easy to blur into one:
The cancer paper (JAMA 2012, Gaziano et al.): the primary endpoint was total cancer, and the result was a significant but small drop of 8% (HR 0.92, 95% CI 0.86–0.998, P=0.04); in absolute terms, 17.0 versus 18.3 cases per 1000 person-years. Prostate cancer was a secondary endpoint and did not change (HR 0.98, P=0.76).The cardiovascular paper (JAMA the same year, Sesso et al., same participants): major cardiovascular events showed no benefit (HR 1.01).
So the accurate sentence is: in these men, the multivitamin did not touch heart disease, and on its primary endpoint, cancer, it achieved a small drop that just crossed the line of significance. Calling total cancer a secondary endpoint, or calling the result negative, both draw the conclusion for the reader. The real limits lie elsewhere: an 8% relative drop is small in absolute terms, and the participants were all male doctors, so how far it applies to women and younger people is limited.
What pulls these together is Bjelakovic 2013, a two-page evidence summary in JAMA of the same authors' Cochrane review (78 randomized trials, 296,707 people). It found that antioxidant supplements did not lower death from any cause. In the trials at low risk of bias, beta-carotene and vitamin E raised total deaths slightly; vitamin A's estimate also leaned toward harm without reaching significance, and higher doses looked worse. Together these are the commonly quoted 3–7% rise in total deaths. Vitamin C and selenium had no effect on death rates. Note that it analyzed by type of antioxidant (given alone or combined with other nutrients), so it does not answer the question of an ordinary multivitamin. The conclusion for multivitamins in healthy people has to come from PHS-II and the USPSTF; it cannot be carried over from this line.
After this string of numbers, one direction stands out: every harm came from a single nutrient at a dose far beyond what food can supply, while a multivitamin with doses close to the recommended amounts showed no clear harm, and a benefit so small it took tens of thousands of people followed for a decade to barely measure. The two statements do not conflict. They describe two stretches of the same dose curve: one flat, one already turning down.
yusuf-2000-hope
Mechanism · Why the trials kept coming up empty
The trials kept coming up empty not because they were badly designed, but because five possible mechanisms push in the same direction. Most come from mechanistic reasoning and observation; not every one has been measured directly.① The antioxidant paradox: mistaking a signal for noise. In a test tube, vitamin E and carotenes really do neutralize free radicals, so in theory they should slow aging and reduce cancer. In the body, free radicals are also working signals. The burst of reactive oxygen species () that muscle produces during exercise is one of the switches that starts the adaptation program for being stronger next time. And when cells sense oxidative stress, they release a switch called NRF2, which has the cell make its own full antioxidant system, glutathione and more. On this explanation, pouring in large doses of antioxidants from outside smooths the signal away in advance, and the body's own system gets the message no need to start work. The net effect may be that the outside supply goes up, the in-house supply shuts down, the total barely changes, and the two jobs that rely on the signal, training adaptation and killing bacteria, are weakened instead.
② The food matrix cannot be taken apart. One carrot holds beta-carotene, alpha-carotene, lycopene, lutein, vitamin C, fiber, potassium, and hundreds of other active compounds at once. They work together in the gut: some protect each other from oxidation, some are released slowly thanks to the fiber, and some can only be absorbed when another is present. Pull one out and put it in a tablet, and what you get is not concentrated carrot; it is a combination that has never existed in nature.
③ A higher dose is not better; it is a different part of the curve. The effect of almost every micronutrient is U-shaped: too little harms, enough is best, and too much harms again. Food naturally keeps the dose in the middle; it is hard to reach ATBC-level blood concentrations by eating carrots. Trials, to make an effect visible, usually used pharmacological doses, stepping straight onto the right-hand side of the curve, so what they measured was toxicity rather than nutrition.
④ People who already have enough cannot be topped up. Large trials mostly enrolled ordinary people in wealthy countries whose blood nutrient levels were already in the adequate range. The stretch from short to enough pays off a lot; the stretch from enough to more is flat. This is the switch-versus-dial idea from the chapter Why more stops helping, only this time checked by hundreds of thousands of people at once.
⑤ The healthy-user effect. In observational studies, people who take supplements often live longer, but they also tend to exercise more, eat more vegetables, and get checkups earlier. Once randomization removes that , the leftover effect disappears. The old finding that supplement users are healthier recorded habits, not pills.
Put the five together and you have a tool you can take away: when a supplement claims to benefit you, first work out whether it assumes you are short, or more is better. The first can be checked (a blood test will tell you); the second has almost never paid off in fifty years of trials.
Evidence · How big the cognition result is
In the last decade, the statistically significant positive result a multivitamin earned in a large trial measured neither heart attacks nor cancer, but cognition.COSMOS-Mind (Baker 2023, in *Alzheimer's & Dementia*) was a 2×2 factorial, 3-year pragmatic randomized trial: cocoa extract (500 mg of flavanols a day) versus placebo, and at the same time a commercial multivitamin-mineral versus placebo. It enrolled 2262 older adults, average age 73, 60% women, and 92% completed the baseline plus at least one yearly assessment. Cognition was tested by telephone, combining a telephone cognitive-status interview (TICS), word-list and story recall, oral trail-making, verbal fluency, number span, and other tests into one global cognition z-score (a score in units of standard deviation).
The endpoints have to be kept apart first. The primary endpoint was cocoa extract: over three years, the mean difference in global cognition was z = 0.03 (95% −0.02 to 0.08; P = .28), no effect. The multivitamin was the pre-specified secondary endpoint: compared with placebo, global cognition z = 0.07 (95% CI 0.02 to 0.12; P = .007). The authors also report benefits in the same direction on the memory and executive-function composites. The interaction between cocoa and the multivitamin was not significant.
A difference of 0.07 standard deviations crossed the significance line, but it is a small step on a composite score from telephone tests, not dementia prevented. This paper does not report the incidence of mild cognitive impairment or dementia; that belongs to a later analysis. People with a history of cardiovascular disease had a larger point estimate (0.14), but that subgroup's interval crosses zero (−0.02 to 0.31), and the interaction test is only nominally P = .01. So it cannot be read as if you have heart disease you should take it all the more.
The authors' own discussion is direct: this is the first evidence from a large, long-term, pragmatic trial that a multivitamin might improve cognition in older adults, and it still needs confirming in more diverse groups, with a mechanism found. 89% of these participants were non-Hispanic White.
So what this positive result shows is that a multivitamin is not zero on every endpoint. It does not show that one pill a day prevents dementia in older adults. For heart disease and cancer, the answer still comes from PHS-II and the USPSTF; this paper did not overturn them.
Chapter 5
What to do instead
If not a multivitamin, then what? The answer has two layers, and the order cannot be reversed.
The first layer is reconnecting a broken line. First find out whether you actually have a gap: a blood test, then supplement the one you are short of, and stop once you have enough. One specific gap matched with one targeted supplement is the main way of using supplements that has been confirmed again and again.
The second layer is widening the whole line, and a pill cannot do that. In a carrot, a fish, or a handful of nuts, dozens or hundreds of components arrive in your gut bound together: some protect each other from oxidation, some are released slowly thanks to fiber, and the doses are naturally held within the range the body has known for hundreds of thousands of years. All a pill can do is pull one of them out and magnify it to a level food never reaches, which is exactly where things went wrong in the large trials.
So spending the multivitamin money on groceries is not being stingy; it puts the money back into the layer that actually works.
The first layer is reconnecting a broken line. First find out whether you actually have a gap: a blood test, then supplement the one you are short of, and stop once you have enough. One specific gap matched with one targeted supplement is the main way of using supplements that has been confirmed again and again.
The second layer is widening the whole line, and a pill cannot do that. In a carrot, a fish, or a handful of nuts, dozens or hundreds of components arrive in your gut bound together: some protect each other from oxidation, some are released slowly thanks to fiber, and the doses are naturally held within the range the body has known for hundreds of thousands of years. All a pill can do is pull one of them out and magnify it to a level food never reaches, which is exactly where things went wrong in the large trials.
So spending the multivitamin money on groceries is not being stingy; it puts the money back into the layer that actually works.
In practice · What to take when there is a real gap
A supplement only has a job when you can say which nutrient I am short of, and how I know. Common indications and what to do about each:Vitamin , 1000–2000 a day: people in northern regions, indoor workers, older adults, people with darker skin, people with obesity; check blood to see whether it is really low (< 30 nmol/L is deficiency, 30–50 nmol/L may be inadequate, ≥ 50 nmol/L is enough for most people)Vitamin B12 (vegetarians, atrophic gastritis, long-term or metformin use): a 1000 µg tablet dissolved under the tongueIron (women, heavy periods, vegetarians): check blood (< 30 µg/L means iron deficiency)Calcium (postmenopausal women, people who eat no dairy): food first; a 500–600 mg tablet, taken with vitamin D. Calcium tablets and the heart have been argued over: a pooled analysis linked calcium tablets taken without vitamin D to slightly more heart attacks (Bolland 2010), while a 2016 guideline from the US National Osteoporosis Foundation and the American Society for Preventive Cardiology judged calcium neutral for the heart as long as intake stays under the upper limit. Read together, the safe course is food first and not too much at onceIodine (pregnancy, breastfeeding, no seafood): 150–220 µgThe omega-3s and (no fish, pregnancy): fish oil or algal oil, 250–500 mgFolate (planning a pregnancy, early pregnancy): 400 µg of folic acidRecovery from injury, serious illness: a short-term combination of protein, vitamin C, and zinc
The difference between this list and a has-everything tablet is not how many items it covers. It lies in two places: every line starts with an indication and only then gives a dose, and almost every dose is far above the matching cell in a multivitamin, because a multivitamin has to cram dozens of things into one tablet and can give each only a little (the chapter What's actually in the bottle explains how minerals packed into one tablet also hold each other back). So trying to fill a real gap with a multivitamin usually does not move it.
One sequence is often skipped: test first, then supplement, then test again. The blood test turns I am probably short into this is how short I actually am, and it also tells you when you can stop, and stopping once you have enough is something a year-round product like a multivitamin cannot do by design.
Evidence · Habits with better evidence than pills
The evidence for the habits below is more direct or more consistent than for a multivitamin. The type of evidence is given in each line:The Mediterranean diet: in the PREDIMED randomized trial, adults at high cardiovascular risk who ate a Mediterranean diet with added olive oil or nuts had fewer major cardiovascular events than those on a low-fat control diet. Results for diabetes, cognition, and depression come from its secondary analyses and other smaller studies, and carry less weight.The DASH diet: several randomized trials consistently show that it lowers blood pressure.Eating 30+ kinds of plants a week: in the American Gut Project, a large observational project, people who ate this way had more diverse gut bacteria (an observed association; it does not show that plant variety is the cause).Eating less ultra-processed food (category 4 in the NOVA classification): in several cohort studies, people who ate more of it had more deaths from any cause and more multiple chronic illnesses (an observed association).Getting enough protein, 1.0–1.6 g/kg: helps maintain muscle, and makes it easier to feel full.Oily fish 2–3 times a week: people who eat fish often have less heart disease (an observed association), while randomized trials of Fish Oil capsules have given mixed results. These are different kinds of evidence, so their strength cannot be compared directly.
Why can a whole diet beat a single ingredient? Half the answer is the food matrix explained in the chapter Why the big trials failed: the components are absorbed bound together, and pulled apart they are no longer the same thing. The other half is almost never mentioned: the substitution effect. Eat one more serving of vegetables and you usually eat one less serving of something else. A pill only adds, while a diet adds and subtracts at the same time, and for chronic disease what you subtract is often worth more than what you add.
That also explains something often misread: why eating more vegetables looks so good in cohort studies, while taking the vitamins found in vegetables shows nothing in randomized trials. The difference is not only the matrix but also what that serving of vegetables pushed off the plate, and a pill pushes nothing off; it just adds one more thing.
estruch-2018-predimedbolland-2010-bmj-calcium-mikopecky-2016-calcium-cvd-guideline
Myth · Products you do not need to buy
A few kinds of product that keep turning up in ads without the evidence to match:Hair, nail, and skin formulas: the main selling point is high-dose biotin. Its most certain effect is actually interfering with a set of common lab tests (including thyroid function and markers of heart-muscle damage), making results falsely high or low; for hair and nails themselves, people who are not short of biotin show no benefit.Anti-aging and : impressive animal data, thin human data, high prices.Whole-food multivitamins: a marketing term. They often cost 5–10 times as much as a synthetic formula, with measured vitamin content about the same or even lower.Blends that promise to protect capillaries, detox, or boost immunity: appealing names, and not a single behind them.
There is a general way to spot them: ask whether the promise is a measurable gap being filled, or a better state with no clear way to check it. The first you can confirm or rule out with a blood test. The second can never disappoint you, because it never set a condition under which it could fail. A promise that cannot even say what not working looks like cannot say what working means either.
Background · Why people keep taking it anyway
If the trials keep coming up empty, why do so many people still swallow that pill every day? This layer is not a nutrition question, but it decides whether all the evidence above does any good.Insurance thinking. The voice in your head says it can't hurt, and it might help. Behavioral economists call this kind of tendency ambiguity aversion: faced with a bad outcome whose odds are unclear (a heart attack, cancer), people will pay for anything that looks as if it lowers the risk, without asking too hard whether it works. It is cheap and takes three seconds a day, so it is easy to keep doing.
The comfort of having done something. The moment you swallow the pill, you really do get the feeling that I am managing my health, and compared with actually changing your diet or actually going out for a walk, it costs almost nothing. That is also where the side effect lies. A psychology experiment (Chiou 2011) found that people who believed they had just taken a supplement were then more likely to choose less healthy behavior: I already took my vitamins. A cheap stand-in action may crowd out the costly real one.
Sixty-plus years of advertising, plus a legal loophole. The US DSHEA law of 1994 placed dietary supplements outside drug regulation, with no need to prove they work before going on sale. So structure-function claims such as supports bone health can be written freely, even though in law that is a completely different thing from prevents osteoporosis; consumers just hear them as the same. The 60+ years of full of energy, shiny hair, stronger immunity were built up layer by layer this way.
Sunk cost. A common worry: I have been taking it for 5 years, so what if my health falls apart when I stop? But the body will not suddenly break down when you stop a multivitamin. That protection was never shown to exist, so there is nothing to lose.
The real costs come in four kinds: money (people in the US spend ~$370 each a year on supplements); behavioral substitution (settling for having done something instead of making a real change); rare but real harms (beta-carotene, vitamin E, and high doses of iron and calcium); and the most expensive one, reducing nutrition to taking pills instead of eating food and living your life.
chiou-2011-supplement-licensing
References · 9
- US Preventive Services Task Force. (2022). Vitamin, mineral, and multivitamin supplementation to prevent cardiovascular disease and cancer: recommendation statement. JAMA, 327(23), 2326–2333. 10.1001/jama.2022.8970
- Baker, L. D., Manson, J. E., Rapp, S. R., Sesso, H. D., Gaussoin, S. A., Shumaker, S. A., & Espeland, M. A. (2023). Effects of cocoa extract and a multivitamin on cognitive function: A randomized clinical trial. Alzheimer's & Dementia, 19(4), 1308–1319. 10.1002/alz.12767
- The ATBC Cancer Prevention Study Group. (1994). The effect of vitamin E and beta carotene on the incidence of lung cancer and other cancers in male smokers. NEJM, 330(15), 1029–1035. 29,133 male smokers aged 50-69 in south-western Finland; alpha-tocopherol 50 mg/day, beta-carotene 20 mg/day, both, or placebo; follow-up 5-8 years. Lung cancer: alpha-tocopherol -2% (-14 to 12%), beta-carotene +18% (3 to 36%). Fewer prostate cancers with alpha-tocopherol, more deaths from haemorrhagic stroke; total mortality 8% higher with beta-carotene (abstract, PMID 8127329). 10.1056/NEJM199404143301501
- Omenn, G. S., et al. (1996). Effects of a combination of beta carotene and vitamin A on lung cancer and cardiovascular disease (CARET). NEJM, 334(18), 1150–1155. 10.1056/NEJM199605023341802
- Gaziano, J. M., et al. (2012). Multivitamins in the prevention of cancer in men: the Physicians' Health Study II randomized controlled trial. JAMA, 308(18), 1871–1880. 10.1001/jama.2012.14641
- Sesso, H. D., Christen, W. G., Bubes, V., Smith, J. P., MacFadyen, J., Schvartz, M., Manson, J. E., Glynn, R. J., Buring, J. E., & Gaziano, J. M. (2012). Multivitamins in the prevention of cardiovascular disease in men: the Physicians' Health Study II randomized controlled trial. JAMA, 308(17), 1751-1760. 14,641 male US physicians aged 50+, median 11.2 years. A daily multivitamin had no effect on major cardiovascular events (hazard ratio 1.01, 95% CI 0.91-1.10), nor on myocardial infarction, stroke or cardiovascular mortality. 10.1001/jama.2012.14805
- Bjelakovic, G., Nikolova, D., & Gluud, C. (2013). Antioxidant supplements to prevent mortality. JAMA, 310(11), 1178–1179. 10.1001/jama.2013.277028
- Lippman, S. M., et al. (2009). Effect of selenium and vitamin E on risk of prostate cancer (SELECT): a randomized controlled trial. JAMA, 301(1), 39–51. 35,533 men (African American 50 or older, others 55 or older; PSA 4 ng/mL or less) randomised to selenium 200 micrograms/day (L-selenomethionine), vitamin E 400 IU/day, both, or placebo. Median follow-up 5.46 years at the 23 October 2008 data cut. Prostate cancer HR (99% CI): vitamin E 1.13 (0.95-1.35), selenium 1.04, combination 1.05 - none significant; no other prespecified cancer end point differed. Type 2 diabetes with selenium RR 1.07 (0.94-1.22), P = .16, not significant (abstract, PMID 19066370). 10.1001/jama.2008.864
- Klein, E. A., Thompson, I. M., Tangen, C. M., Crowley, J. J., Lucia, M. S., Goodman, P. J., et al. (2011). Vitamin E and the risk of prostate cancer: the Selenium and Vitamin E Cancer Prevention Trial (SELECT). JAMA, 306(14), 1549-1556. Among 35,533 men, vitamin E 400 IU/day significantly increased prostate cancer risk (about 17% relative increase; 76 vs 65 cases per 1,000). 10.1001/jama.2011.1437