The Vitamin You Aren't Missing
A vitamin trial is not supposed to end with a warning that the vitamin is hurting people. This one did.
CARET gave beta-carotene, the antioxidant that makes carrots orange, plus a dose of vitamin A, to more than eighteen thousand people at high risk of lung cancer: heavy smokers and men who had breathed asbestos for a living. The theory was clean and hopeful. Antioxidants mop up the molecular damage that leads to cancer; people who eat more carotene-rich vegetables get less lung cancer; so give the pill and prevent the disease. In January 1996 the researchers stopped the study twenty-one months early, because the group on the supplement was developing more lung cancer, not less, and dying at a higher rate. The thing meant to protect them was doing the opposite.
Hold onto that, because it breaks the first thing everyone believes about vitamins.
We file supplements under “can’t hurt, might help.” A little insurance. Even if the pill does nothing, the reasoning goes, you’re only out the money. CARET is the clean refutation of the “can’t hurt” half. But the more useful problem is buried in the word itself. “Supplement” quietly staples together two acts that have almost nothing to do with each other.
The first act is correcting a real deficiency. Here the evidence is not just good, it’s some of the most solid in all of medicine. Folic acid before and during early pregnancy cuts the risk of neural-tube defects like spina bifida by roughly half. Vitamin B12 rescues vegans, and the large slice of people over sixty whose stomachs no longer absorb it, from nerve damage that can become permanent. Iron fixes the anaemia of a heavy-bleeding or poorly-fed body. Vitamin D genuinely helps people who are genuinely short of it. Nobody sensible argues with any of this. It is replacement, and replacement works.
The second act is topping up a tank that is already full. This is the daily multivitamin, the fish-oil capsule, the vitamin D and the “full-spectrum” mineral blend, swallowed by well-fed people with no measured shortage of anything. It is where most of the roughly fifty billion dollars a year Americans spend on supplements actually goes. And it is where the big trials keep coming back empty-handed. VITAL followed nearly twenty-six thousand people on vitamin D and fish oil and found no drop in cancer or heart attacks. The long multivitamin studies show the same flat line. In 2022 the US Preventive Services Task Force, after reading the lot, could not find enough benefit to recommend a multivitamin to the general public for preventing cancer or heart disease at all.
No benefit is one thing. The harder question is why topping up isn’t at least harmless.
The answer is that your body runs brakes on the nutrients it’s designed to get from sun and food, and a pill takes the brakes off. Your skin makes vitamin D and then, when it has had enough, destroys the excess in place; you cannot overdose on sunshine. Vitamin C saturates your tissues at about the level a few oranges provide, and the rest leaves in your urine within hours. Selenium arrives in food alongside the feeling of having eaten. Swallow the isolated compound at ten or fifty times the dietary dose and none of those governors apply. So the dose-response curve for a lot of these nutrients isn’t a rising line where more is better. It’s a U. The bottom of the U is “enough,” and both ends are worse.
Which is exactly the shape the trials keep drawing. Vitamin E, taken by healthy men in the SELECT trial, raised prostate cancer by about 17%. Beta-carotene killed smokers in CARET. High-dose fish oil has been linked, in pooled trials, to a higher rate of atrial fibrillation, the risk climbing with the dose, though I’ll be honest that this one is still contested and other data muddy it. The pattern is not that vitamins are poison. It’s that a nutrient your body wasn’t missing isn’t a gift. It’s just a dose, and doses have two directions.
Give the other side its due, because the core logic is sound. If a substance is essential and you’re short of it, replacing it should help, and it does. The trouble is the aim. Most well-fed adults are not short of the things the shelves sell them, while the people who are short, the housebound pensioner, the exhausted new vegan, the woman running on empty iron, often have no idea and are taking nothing. The supplement you actually need tends to be the one nobody markets to you, because it’s cheap and specific and requires knowing your own numbers.
And notice what CARET’s own premise got wrong. Beta-carotene from carrots really does track with less cancer. The molecule pulled out of the carrot and pressed into a tablet did the reverse. Nutrients seem to work in a crowd, wrapped in fibre and cofactors and a dozen other compounds, not as soloists blasting at concert volume. This is the dull, repeatable finding under a thousand flashy ones: eat the vegetable beats extract the compound, almost every time it’s tested.
So the practical rule is smaller and less fun than the supplement aisle wants it to be. Don’t dose blind. If you have a real reason to suspect a gap, a dark northern winter, a vegan diet, sixty-plus years, heavy periods, a pregnancy, the honest move is cheap: get the one relevant thing checked, then fill that specific gap. The proven wins are a short, boring list, and they are worth taking seriously. Everything past that list is mostly buying a feeling, and a slice of it is quietly buying risk.
Think back to the smokers in CARET. They did the most human thing there is: reached for a vitamin as a hedge against a danger they were living with. That instinct, that a natural, healthy-sounding molecule can only help, is precisely the one the evidence keeps punishing. The question worth carrying to the pharmacy shelf isn’t the hopeful one, “could this help me?” It’s the awkward one the marketing is built to keep you from asking: am I actually short of this, and how would I know?
