The vitamin C ghost

 

https://www.sciencedaily.com/releases/2026/06/260626030428.htm The vitamin C ghost returns; the article tells me that some were deficient, not that we need more, but hard to tell.

I think your reading is exactly right.

The study did not show that people with higher-than-normal vitamin C levels had better brains than everyone else. What it showed was that older adults with lower blood vitamin C levels tended to have less gray matter and weaker connectivity in a brain network associated with memory and attention. The researchers adjusted for several other factors, but it remained an observational association, not proof of cause and effect.

This is a pattern one sees repeatedly with vitamin C. Every few years the "vitamin C ghost" reappears. A study finds that people with lower levels have worse outcomes of some kind, and headlines start hinting that more vitamin C might be protective. But there are at least three possible explanations:

1. Vitamin C itself is helping.

2. Low vitamin C is simply a marker for generally poorer diet and health.

3. People who are becoming ill or frail may consume or retain vitamin C differently, so the low level is partly a consequence rather than a cause.

This study cannot distinguish among those possibilities.

What caught my eye is that the investigators measured blood levels, not just reported dietary intake. That is somewhat stronger evidence than asking people how many oranges they eat. Still, it remains a snapshot in time.

There is also a recurring problem in nutrition science: the distinction between preventing deficiency and optimizing health. We know beyond doubt that severe vitamin C deficiency causes scurvy and that humans require vitamin C from the diet. But the amount needed to avoid overt deficiency may not be the same as the amount that maximizes every physiological process. That question remains much murkier and is where much of the debate lives.

So my summary of this paper would be:

"Another study showing that low vitamin C status is associated with something undesirable. Interesting and worth paying attention to, but not evidence that people with adequate vitamin C should start taking large supplements."

Given your long-standing skepticism about the vitamin C wars, I would say this paper belongs in the "deficiency appears undesirable" file, not the "megadoses vindicated" file. The ghost has indeed returned, but it is still rattling chains rather than presenting decisive evidence.

What happened to Pauling's original argument that animals that don't consume fruits and hence make their own vitamin C make much more of it than we seem to need?

That argument never entirely disappeared; it just failed to persuade most nutrition scientists.

The basic observation was correct. Many mammals synthesize vitamin C in their livers (or kidneys), and when you scale the production to human body weight, the amounts can be surprisingly large—often equivalent to grams per day rather than the tens of milligrams needed to prevent scurvy. This was a central point in the thinking of Linus Pauling and later advocates of high-dose vitamin C.

The problem is that the inference is not straightforward. The argument assumes that because another mammal produces a certain amount, humans would benefit from having the same amount available. Evolutionary biologists and physiologists raised several objections.

One is that species differ enormously in metabolism. A goat, often cited by vitamin C enthusiasts, does indeed make large amounts of vitamin C, but it also has a very different digestive system, growth rate, lifespan, and metabolic profile. There is no general biological rule that says if one species synthesizes a substance at a certain rate, another species should consume the equivalent amount.

Another objection concerns evolution. Humans, other apes, guinea pigs, and a few other animals lost the ability to synthesize vitamin C millions of years ago because the gene for the final synthetic step became nonfunctional. The usual evolutionary explanation is that our fruit-eating ancestors obtained abundant vitamin C from their diet, so there was little selective pressure to maintain the pathway. If losing the pathway had imposed a large fitness cost, natural selection would have tended to preserve it. That does not prove humans are optimized for low intake, but it weakens the claim that we are suffering from a major evolutionary defect.

There is also the question of what those animals actually do with all that vitamin C. Some species dramatically increase production under stress, infection, or injury. Pauling interpreted that as evidence that high vitamin C availability is part of the natural defense system. Critics responded that endogenous production is tightly regulated and may not be equivalent to swallowing large oral doses. The body often handles internally produced substances differently from externally consumed ones.

Perhaps most importantly, decades of clinical trials never produced the dramatic effects Pauling expected. If humans were routinely functioning far below their natural vitamin C requirements, one might expect high-dose supplementation to produce large, obvious benefits across many diseases. Instead, the results were generally modest, inconsistent, or absent. There are a few areas where vitamin C shows measurable effects, but nothing like the sweeping claims that were made.

That said, Pauling's argument did leave a lingering question that has never been fully settled. The official recommended intake is designed primarily to avoid deficiency and maintain normal physiological function. It does not necessarily identify the intake that would optimize every aspect of aging, immunity, cognition, or long-term health. The gap between "enough to avoid scurvy" and "optimal" remains an active area of research.

So the modern scientific position is not usually "Pauling was obviously wrong because animals make a lot." It is more "the observation is real, but the leap from that observation to gram-level supplementation in humans was never convincingly demonstrated."

In a way, the vitamin C debate persists because Pauling's original premise contains a genuine puzzle. We are unusual mammals in having lost a pathway that most mammals retain. What remains unproven is the conclusion that restoring gram-level vitamin C exposure would substantially improve human health. That is where the evidence has consistently fallen short.

The "old people" power nutrition supplement that the clinic gives me contains 500 mg. Since I avoid fruit, that seems appropriate, and I refrigerate it in a sealed container.

That seems quite reasonable.

The Recommended Dietary Allowance for older men is only about 90 mg/day, with somewhat higher amounts often advised for smokers. A 500 mg serving is already several times beyond what is needed to prevent deficiency. The body absorbs vitamin C efficiently at lower doses, but absorption becomes less efficient as the dose rises, and excess is largely excreted in the urine.

Given that you intentionally avoid fruit, the more relevant question is whether your overall intake remains adequate. If your supplement reliably provides 500 mg and you consume it regularly, frank vitamin C deficiency would be very unlikely. In that context, the recent study you linked is hard to interpret as a call for you to consume even more vitamin C.

The refrigeration is sensible. Vitamin C is somewhat fragile; it degrades with exposure to air, heat, and light. How much that matters depends on the formulation, but keeping a prepared nutritional drink sealed and refrigerated is certainly not hurting anything.

What I find interesting is that your clinic's supplement amount—500 mg—is almost exactly the sort of middle ground that has become common in practice. It is far above the minimum needed to avoid deficiency, but far below the multi-gram quantities that Pauling and some later advocates promoted. Many clinicians seem comfortable with that range because it provides a substantial margin of safety without venturing into the territory where evidence becomes more speculative.

For someone in your situation—older, avoiding fruit, and interested primarily in maintaining health rather than conducting a personal experiment—I would view the 500 mg as insurance against deficiency rather than as an attempt at pharmacological vitamin C therapy.

And, amusingly, it puts you in a position where the latest "vitamin C may be important" headline can largely be met with a shrug. Whatever else may be true, you're almost certainly not one of the low-vitamin-C individuals those studies are worried about.

Correction: powder nutrient supplement, not power.

Ah, that makes more sense.

A powdered nutritional supplement is actually a fairly common way to deliver vitamin C because the dry powder is generally more stable than a prepared liquid. Once mixed with water, degradation begins gradually, especially if exposed to air, warmth, or light. Keeping the mixed drink in a sealed container in the refrigerator is a sensible way to slow that process.

So my earlier point becomes even stronger. If the powder is formulated to contain 500 mg of vitamin C per serving, and you're consuming it reasonably soon after mixing, you're probably getting most of that amount.

Given your avoidance of fruit, that sounds less like "high-dose vitamin C therapy" and more like an ordinary way of covering a potential nutritional gap. Many fruits are major vitamin C sources, so if you remove them from the diet, obtaining vitamin C from vegetables, supplements, or fortified nutritional drinks becomes the obvious alternative.

The interesting question raised by the article is whether there is a difference between people who are merely above the deficiency threshold and people whose blood levels are at the higher end of the normal range. The study hints at that possibility, but it does not establish it. What it certainly does not suggest is that someone already getting several hundred milligrams a day should rush out and double or triple their intake.

As an old programmer might put it, the study seems to identify a possible bug caused by an uninitialized variable, not demonstrate that setting the variable to 65,535 produces optimal performance.