What Is the Optimal Vitamin D Level?

If everyone took 2,000 units of vitamin D a day, it could shift the curve from average blood levels in the mid-50s to about a 110 nmol/L, which some estimate could add years to our life expectancy. Data derived from randomized clinical trials have convinced some influential experts, such as Harvard’s Chair of Nutrition, that we should shoot for this kind of range, levels that about nine out of ten people fail to reach because it may necessitate taking 1,800 to 4,000 units of vitamin D a day.

The Institute of Medicine (IOM), however, considered blood levels of 50 nmol/L to be sufficient and therefore recommended only 600 to 800 units a day for those with little or no sun exposure. Why so low? Because the IOM was only considering bone health. Even if we cared just about our bones and not our lifespan, we’d still probably want to shoot for a 75 nmol/L threshold, because there’s evidence from hundreds of autopsies of people who died in car accidents, for instance, showing osteomalacia, or softening of the bones, in 18 to 39 percent of people who reach the IOM target of 50 nmol/L, but failed to make it to 75 nmol/L.

There has even been a charge that the Institute of Medicine simply made a mistake in its calculations, and, based on its own criteria, should be recommending thousands of units a day, as well. However, the mere absence of soft bones “can hardly be considered an adequate definition either of health or of vitamin D sufficiency.” It’s like saying you only need 10 mg of vitamin C to avoid scurvy—yes, but we need way more than that for optimal health.

The Institute of Medicine took the position that the burden of proof “fell on anyone who claimed benefits for intakes higher than the panel’s [minimal] recommendations,” which is a good approach for drugs. For unnatural substances, less is more until proven otherwise. For nutrients, however, shouldn’t the starting point at least be the natural levels to which our bodies have become finely tuned for millions of years? I explore this question in my video The Optimal Dose of Vitamin D Based on Natural Levels.

The target level of 75 nmol/L only sounds high compared to average levels today, but in modern times, we often practice unnatural activities like working at a desk job, or even wearing clothes! We evolved running around naked in equatorial Africa getting sun all day long. If we measure vitamin D levels in those living traditional lives in the cradle of humanity, a normal vitamin D level would be over 100 nmol/L. So, maybe that should be the starting point until proven otherwise—a concept, regrettably, many guideline committees seem to have ignored.

The natural level, however, isn’t necessarily the optimal level. Maybe the body would have thrived with less, so we still have to look at what levels correspond to the lowest disease rates. And, when we do, the higher levels do indeed seem to correlate with less disease.

When I was doing pediatrics, it always struck me that breastfed babies required vitamin D drops. Shouldn’t human breast milk be a perfect food? Of course, for the medical profession, the solution is simple: Provide the baby supplements, the vitamin D drops. But it seems like we shouldn’t have to. If we measure human breast milk these days, however, it has virtually no vitamin D and would cause rickets unless the mom has vitamin D levels up around the level natural for our species, which of course makes total sense. The way we live in our modern world is like an environmental mismatch. It helps to think of vitamin D as what it truly is: a hormone, not a vitamin. If you think of it as a hormone, then it would be reasonable to have normal levels. We physicians try to maintain blood pressure and all sorts of parameters within normal limits, “but why so little attention is paid to the status of the hormone ‘vitamin D’?”


If one is going to make an evolutionary argument for what a “natural” vitamin D level may be, how about getting vitamin D in the way nature intended—that is, from the sun instead of supplements? That’s the subject of my video The Best Way to Get Vitamin D: Sun, Supplements, or Salons?.

For the other videos in this series, check out:

I also explore vitamin D as it relates to specific diseases:

In health,
Michael Greger, M.D.

PS: If you haven’t yet, you can subscribe to my free videos here and watch my live, year-in-review presentations:

The Amount of Vitamin D Supplementation I Recommend

Randomized, controlled clinical trials have found that vitamin D supplements extend one’s lifespan. What is the optimal dose? What blood level is associated with living longest? In my nine-part video series on vitamin D from 2011, I noted that the relationship between vitamin D levels and mortality appeared to be a U-shaped curve, meaning low levels of vitamin D were associated with increased mortality, but so were levels that were too high, with the apparent sweet spot around 75 or 80 nmol/L based on individual studies. (See Vitamin D and Mortality May Be a U-Shaped Curve for more on this.)

Why might higher vitamin D levels be associated with higher risk? Well, the study I profile in my video How Much Vitamin D Should You Take? was a population study, so we can’t be sure which came first. Maybe the higher vitamin D higher risk, or perhaps higher risk led to higher vitamin D levels, meaning maybe those who weren’t doing as well were prescribed vitamin D. Maybe it’s because it was a Scandinavian study, where individuals tend to take a lot of cod liver oil as a vitamin D supplement, one spoonful of which could exceed the tolerable upper daily limit of intake for vitamin A, which could have negative consequences.

Anyway, the U-shaped curve is old data. An updated meta-analysis has shown that as population vitamin D levels go up, mortality appears to go down and stay down, which is good because then we don’t have to test to see if we’re hitting just the right level. Routine testing of vitamin D levels is not recommended. Why? Well, it costs money, and, in most people, levels come right up to where you want them with sufficient sun or supplementation, so what’s the point? As well, the test is not very good: Results can be all over the place. What happens when you send a single sample to a thousand different laboratories around the world? You’d perhaps expect a little variation, but results from the same sample ranged anywhere from less than 20 to over 100 nmol/L. Depending on what laboratory your doctor sent your blood sample to, the results could vary dramatically, so one could argue the test isn’t necessarily very helpful.

So, what’s a safe dose that will likely get us to the purported optimal level? A thousand units a day should get most people up to the target 75 nmol/L (which is 30 ng/mL), but by most people, researchers mean 50 percent of people. To get around 85 percent of the U.S. population up to 75 nmol/L would require 2,000 IU a day. Two thousand IU a day would shift the curve so that the average person would fall into the desired range without fear of toxicity. We can take too much vitamin D, however, but problems don’t tend to be seen until blood levels get up around 250 nmol/L, which would take consistent daily doses in excess of 10,000 IU.

Note that if you’re overweight, you may want to take 3,000 IU and even more than that if you’re obese. If you’re over age 70 and not getting enough sun, it may take 3,500 IU to get that same 85 percent chance of bumping up your levels above the target. Again, there’s no need for the average person to test and retest, since a few thousand IU per day should bring up almost everyone without risking toxicity.

Given this, why then did the Institute of Medicine set the Recommended Daily Allowance at 600 to 800 IU? In fact, official recommendations are all over the map, ranging from just 200 IU a day all the way up to 10,000 IU a day. I’ll try to cut through the confusion in my next post.


After all that work plowing through the new science, the same 2,000 IU per day recommendation I made in 2011 remains (for those not getting enough sun): http://nutritionfacts.org/2011/09/12/dr-gregers-2011-optimum-nutrition-recommendations/.

The other videos in this series include:

I also explore Vitamin D as it relates to specific diseases:

In health,
Michael Greger, M.D.

PS: If you haven’t yet, you can subscribe to my free videos here and watch my live, year-in-review presentations:

How Doctors Responded to Being Named a Leading Killer

In my video Why Prevention Is Worth a Ton of Cure, I profiled a paper that added up all the deaths caused by medical care in this country, including the hundred thousand deaths from medication side effects, all the deaths caused by errors, and so on. The author of the paper concluded that the third leading cause of death in America is the American medical system.

What was the medical community’s reaction to this revelation? After all, the paper was published in one of the most prestigious medical journals, the Journal of the American Medical Association, and was authored by one of our most prestigious physicians, Barbara Starfield, who literally wrote the book on primary care. When she was asked in an interview what the response was, Starfield replied that her primary care work had been widely embraced, but her findings on how harmful and ineffective healthcare could be received almost no attention.

This inspires the recollection of “the dark dystopia of George Orwell’s 1984, where awkward facts are swallowed up by the ‘memory hole’ as if they had never existed at all.” Report after report has come out, and the response has been a deafening silence both in deed and in word, failing to even openly discuss the problem, leading to thousands of additional deaths. We can’t just keep putting out reports, we have to actually do something.

As I discuss in my video How Doctors Responded to Being Named a Leading Killer, the first report was published in 1978, suggesting about 120,000 preventable hospital deaths a year. The response? Silence for another 16 years until another scathing reminder was published. If we multiply 120,000 by those 16 years, we get 1.9 million preventable deaths, about which there was near total doctor silence. There was no substantial effort to reduce the number of those deaths. The Institute of Medicine (IOM) then released its landmark study in 1999, asserting that yet another 600,000 patients died during that time when providers could have acted.

Some things have finally changed. Work hour limits were instituted for medical trainees. Interns and residents could no longer be worked more than 80 hours a week, at least on paper, and the shifts couldn’t be more than 30 hours long. That may not sound like a big step, but when I started out my internship, I worked 36 hour shifts every three days, 117-hour work weeks.

When interns and residents are forced to pull all-nighters, they make 36% more serious medical errors, five times more diagnostic errors, and have twice as many “attentional failures.” That doesn’t sound so bad, until you realize that means things like nodding off during surgery.

The patient is supposed to be asleep during surgery, not the surgeon.

Performance is impaired as much as a blood alcohol level that would make it illegal to drive a car—but these overworked interns and residents can still do surgery. No surprise there were 300% more patient deaths. Residents consider themselves lucky if they get through training without killing anyone. Not that the family would ever find out. With rare exceptions, doctors are unaccountable for their actions.

The IOM report did break the silence and prompted widespread promises of change, but what they did not do is act as if they really believed their own findings. If we truly believed that a minimum of 120 people every day were dying preventable deaths in hospitals, we would draw a line in the sand. If an airliner was crashing every day, we’d expect that the FAA would step in and do something. The Institute of Medicine could insistently demand that doctors and hospitals immediately adopt at least a minimum set of preventive practices—for example, bar-coding drugs so there aren’t any mix-ups, like they do for even a pack of Tic Tacs at the grocery store. Rather than just going on to write yet another report, they could bluntly warn colleagues they would publicly censure those who resisted implementing these minimum practices, calling for some kind of stringent sanctions.

Instead, we get silence. But not for Barbara Starfield, who is unfortunately no longer with us. Ironically, she may have died from one of the adverse drug reactions she so vociferously warned us about. She was placed on aspirin and the blood-thinner Plavix to keep a stent she had to have placed in her coronary artery from clogging up. She told her cardiologist she was bruising more, bleeding longer, but those side effects are the risks you hope don’t outweigh the benefits. Starfield apparently hit her head while swimming and bled into her brain.

The question for me is not whether she should have been on two blood-thinners for that long or even whether she should have had the stent inserted. Instead, I question whether or not she could have outright avoided the heart disease, which is 96% avoidable in women.

The number-one killer of women need almost never happen.


For those curious about my time in medical training, you can read my memoir of sorts, Heart Failure: Diary of a Third Year Medical Student.

It isn’t just medical treatment that can be harmful. Even medical diagnosis can be dangerous, as I discuss in my video Cancer Risk From CT Scan Radiation.

And, just as we’re (finally) seeing some changes in training protocols, the times, they are a-changin’ with the emergence of the field of lifestyle medicine, as I present in several videos, including:

 I recently made some videos to give people a closer look at why I believe it’s so important for us to take responsibility for our own health. You can see all of them on our new Introductory Videos page.

I’m excited to be part of this revolution in medicine. Please consider joining me by supporting the 501c3 nonprofit organization that keeps NutritionFacts.org alive by making a tax-deductible donation. Thank you so much for helping me help so many others.

In health,

Michael Greger, M.D.

PS: If you haven’t yet, you can subscribe to my free videos here and watch my live, year-in-review presentations: