Has science proven that the minimal acceptable blood level of vitamin D, in the form of 25(OH)D, is above 50 ng/mL (125 nmol/L)? No. If you’ve been trying to maintain your levels this high because you thought this was the case, I’m sorry to break the news. There is, on the contrary, good evidence that 25(OH)D levels should be at least 30-35 ng/mL (75-88 nmol/L). Much higher levels may be better, or they could start causing harm, especially in the absence of adequate vitamins A and K2. Once we leave the land of 30-35 ng/mL, however, we enter the land of speculation.
The idea that science has proven we need to maintain 50 ng/mL as a minimum comes from Dr. John Cannell of the Vitamin D Council. In his article “Am I Vitamin D Deficient?” he writes the following:
Thanks to Bruce Hollis, Robert Heaney, Neil Binkley, and others, we now know the minimal acceptable level. It is 50 ng/ml (125 nmol/L). In a recent study, Heaney, et al expanded on Bruce Hollis’s seminal work by analyzing five studies in which both the parent compound (cholecalciferol) and 25(OH)D levels were measured. They found that the body does not reliably begin storing cholecalciferol in fat and muscle tissue until 25(OH)D levels get above 50 ng/ml (125 nmol/L). The average person starts to store cholecalciferol at 40 ng/ml (100 nmol/L), but at 50 ng/ml (125 nmol/L) virtually everyone begins to store it for future use. That is, at levels below 50 ng/ml (125 nmol/L), the body uses up vitamin D as fast as you can make it, or take it, indicating chronic substrate starvation—not a good thing. 25(OH)D levels should be between 50–80 ng/ml (125–200 nmol/L), year-round.
There are a few problems with this argument. To begin with, Drs. Hollis, Heaney, Binkley, and the other authors of this study rightly made very different conclusions from their own data. In the report they wrote for the American Journal of Clinical Nutrition, they wrote the following:
One could plausibly postulate that the point at which hepatic 25(OH)D production becomes zero-order [this is the point at which the enzymes converting vitamin D to 25(OH)D are saturated with vitamin D – CM] constitutes the definition of the low end of normal status. This value, as suggested from the equation in Figure 3 , is at a serum 25(OH)D concentration of 88 nmol/L (35.2 ng/mL) (the y-axis intercept of the linear portion of the equation in Figure 3 ). It is interesting that this estimate is very close to that produced by previous attempts to define the lower end of the normal range from the relations of serum 25(OH)D to calcium absorption (29) and to serum parathyroid hormone concentration (ie, 75–85 nmol/L, or 30–34 ng/mL) (30).
According to the authors of this study, then, the point at which the vitamin D enzymes are saturated and vitamin D “accumulates within the body, both in serum and probably in body fat” is not 40 or 50 ng/mL (100 or 125 nmol/L) but rather 35 ng/mL (88 nmol/L).
The authors used a statistical approach that pooled together data from several studies. They presented most of their data in Figure 4, and the data from one other study in Figure 5. They did not determine the point at which vitamin D starts getting stored in body fat in particular individuals. On the contrary, they used a statistical approach to infer the point at which this occurs in their entire study population. Now, if you compare Figures 4 and 5, looking for the point at which the slope of the line dramatically changes, you will see that it changes at a higher level of 25(OH)D in Figure 5. Dr. Cannell seems to have used the data from Figure 5 to say when vitamin D gets stored in body fat in “virtually everyone” as opposed to “the average person,” but in fact the authors stated that they did not use the data from Figure 5 to determine this point because a different and apparently inferior method of measuring vitamin D levels was used in that data set.
So, we are back to the authors’ original conclusions, that vitamin D saturates its activation enzymes and starts getting stored in body fat when 25(OH)D levels reach 35 ng/mL (88 nmol/L).
The second problem is that this study does not “prove” or “show” or “demonstrate” what the optimal or minimal blood level of vitamin D is. The authors state that one could plausibly postulate that the minimum acceptable blood level is the point at which the enzymes are saturated and vitamin D is stored in body fat, but they never state that “we now know the minimal acceptable level.”
The most definitive way to determine the ideal 25(OH)D level would be to conduct a randomized, controlled trial with different levels of vitamin D supplementation targeted at reaching specific blood levels of 25(OH)D and to test the effects of the different levels of supplementation on clinical outcomes, such as bone mineral density, fracture rate, insulin resistance, glucose tolerance, cancer, or heart disease.
We do not yet have this type of data. We do, however, have some strong support for raising 25(OH)D levels to at least 35 ng/mL (88 nmol/L). For example, as the authors of the study we have been looking at pointed out, similar attempts to use statistical approaches to define the 25(OH)D level that maximizes calcium absorption, maximally suppresses parathyroid hormone (which leaches calcium from bone), or maximizes bone mineral density have suggested similar results. A recent randomized, placebo-controlled trial showed that supplementing insulin-resistant women with 4,000 IU of vitamin D per day for six months reduced insulin resistance and had the most powerful effect in women whose 25(OH)D level was raised to over 32 ng/mL (80 nmol/L).
What about higher levels? The evidence is conflicting, and some of it indicates possible harm. For example, a study in the American Journal of Medicine published in 2004 found that in Americans aged over 50, the maximal bone mineral density (BMD) occurs around 32-40 ng/mL (80-100 nmol/L). Among Mexican Americans, BMD continues to rise a little after this point, but for whites it plateaus and begins dropping off around 45 ng/mL (110 nmol/L) and for blacks it begins dropping off even before 40 ng/mL (100 nmol/L).
If 50 ng/mL (125 nmol/L) is our minimal acceptable level, this study would seem to suggest that those of us who have “acceptable” levels of 25(OH)D would have lower bone mineral density than those of us who are moderately deficient. And that just doesn’t make sense.
Another study published in the European Journal of Epidemiology in 2001found that South Indians with 25(OH)D levels higher than 89 ng/mL (223 nmol/L) were three times more likely to have suffered from ischemic heart disease than those with lower levels – and of course with such a dramatic elevation of heart disease risk, the risk may have begun increasing at levels substantially lower than 89 ng/mL.
Neither of these studies were designed to show that high levels of 25(OH)D cause decreases in bone mineral density or increases in heart disease risk, but it is possible. As I especially emphasized in my Wise Traditions and Medical Hypotheses articles on vitamin K2, bone resorption and blood vessel calcification are prominent symptoms of vitamin D toxicity in animal experiments. I also emphasized the role of vitamins A and K2 in protecting against vitamin D toxicity. So, even if these levels are in fact harmful, they may only be harmful or may be primarily harmful in the absence of adequate vitamins A and K2. The presence of the other fat-soluble vitamins could even turn these levels from harmful to beneficial.
Nevertheless, what we need in order to show that levels higher than 50 ng/mL are helpful or harmful are vitamin D supplementation trials comparing the effect of different doses that result in different blood levels on clinical health outcomes, and similar studies examining the interactions between vitamin D and the other fat-soluble vitamins.
Lifeguards in the tropics can reach blood levels in the 50’s and 60’s naturally from sun exposure, suggesting these levels are “natural,” although lifeguards in Israel have 20 times the rate of kidney stones as the general population. Kidney stones may be the most sensitive indicator of vitamin D toxicity and are a symptom of vitamin A and K2 deficiency. Thus, I suspect these levels are healthful in the context of a diet rich in vitamins A and K2, and if my levels were to reach this high in the summer sun while I were eating such a diet I certainly would not worry.
But if you are trying desperately to maintain year-round 25(OH)D status between 50-80 ng/mL using vitamin D supplements, you have entered the land of speculation. Enter at your own risk.
Read more about the author, Chris Masterjohn, PhD, here.
🖨️ Print post
Thanks for a factual perspective on the vitamin D mania that seems to be sweeping the country right now. I get upset when I hear about health care practitioners putting people on 10,000 IU/d of D indefinitely, without testing. In the typical American fashion, people always seem to assume that more of a good thing is better. As you pointed out, this may not be the case with vitamin D.
Chris,
I live in downstate New York and am able to regularly get 15-30 minutes of full body sun exposure (just wearing shorts) around mid day (from mid-to-late Spring until early Fall), thanks to a break in my schedule at that time. Since I know where this exposure takes my blood levels, would this be a reasonable target to use as a baseline in terms of what to shoot for/where to cap my levels at if I ever ended up unable to regularly get sun exposure and had to resort to supplemental Vitamin D?
On a side note, prior to finding your blog, I had followed some of the work of Dr. Cannell, and had actually purchased one of the tanning lamps suggested by the Vitamin D council for use during the Winter time. Last Winter I used it a total of 2-3 times a week for around 5 minutes total per session. I supplemented with no oral vitamin D during this time period (save for 500 IU’s of cholecalciferol present in the very basic multi-vitamin I was taking and whatever amounts I picked up from periodically eating wild Alaskan salmon and egg yolks). Given your current thoughts on vitamin D levels, do you feel that this practice of moderate use of that lamp may not be such a good idea on my part? (as an aside, one thing I found beneficial was that I had a mild tan that enabled me to have a bit more of a natural buffer built up for when I did resume frequent sun exposure)
And on one final note, what’s your own personal strategy for ensuring adequate amounts of vitamin K2 in your diet (and vitamin A) for that matter? While I strive for a varied diet, I don’t find myself regulary consuming the handful of foods that are good sources of k2.
It seems like there are no simple answers when it comes to reductionist views of vitamins. Maybe because these views are too simplistic. Evidence you have presented in the past about the complex interactions between vitamins A, D, and K2 appear to be critical in understanding optimal health. Our best hypotheses for optimal health should be based a historic and prehistoric dietary perspective on food and sunshine and tested based on real food intake and sunshine exposure and not on refined supplement intake.
Bryan,
Thanks for your comment. I mostly agree that we need to study real foods; using supplements and other forms of reductionism, however, can be useful for understanding the mechanisms of how things work, which is of academic interest to some people and is of general utility in better understanding how to restore people to health who have lost it, which might require more than mere return to a traditional diet sometimes. What we really need, however, is research comparing refined supplements to equivalent nutrient intakes from whole foods, as this would help gain the support of the scientific and medical communities behind real food (something currently beginning to happen in the nutrition world, in part thanks to the influence of Michael Pollan’s writings).
Lou, I think you are on track that the “target” level should be what you get when getting a reasonable level of sun exposure. Most humans through history have not been inside all day, nor have they been lifeguards, so the ideal level is probably somewhere in between. I think the lamp might be a good idea if it doesn’t raise your levels way beyond those of moderate sunshine, especially since UV light produces many other goodies in the skin. I think eating liver regularly, grass-fed animal fats, green and orange vegetables, palm oil, and cod liver oil are the best ways to get A and K2.
Chris, good points, and thanks for your comments!
I am huge fan of TESTING Vitamin D levels and supplementing if necessary. I had been following a WAPF lifestyle and taking cod liver oil daily when I first tested my Vitamin D levels. My level was at 25 ng/mL which was surprising to me. I did push my levels above 50, but it appears that I have overcompensated. I recently read that you should take Vitamin D supplementation with your largest meal to increase absorption – do you know if that is true?
http://amoderatelife.com/2010/06/friday-thoughts-from-the-middle-of-the-road/
I have long believed that supplementation was not a good idea, and this only confirms that for me. Thanks for all you do! I shared this with my readers.
Corganic,
I am a fan of testing in principle, but the problem is that the optimal 25(OH)D level has not been characterized in the context of vitamins A, D, and K2, and that 25(OH)D levels represent not only input from the sun and diet but also activation to the fully active calcitriol.
Is the ideal 25(OH)D level different when one eats a diet rich in A and K2; if so, is it higher or lower?
If someones 25(OH)D is low because of a deficient input of vitamin D, then they should supplement or change their diet or lifestyle to increase vitamin D input. But what if their 25(OH)D is low because they have excessive activation to calcitriol? In that case vitamin D supplementation could be harmful.
Does vitamin A lower 25(OH)D by increasing the utilization of vitamin D? If so, that throws a wrench in the logic of testing — maybe in this case lower 25(OH)D is an index of the efficacy of vitamin A in helping D do its job.
And beyond that, the current recommendations about what is optimal are primarily derived from the agreement of different kinds of observational studies. Experimental studies are beginning and are starting to offer a little support but even in the absence of the above caveats the characterization of the optimal level of vitamin D is still in its infancy in a large way.
I have not looked at absorption with meals specifically, but it makes sense as generally fat-soluble vitamins are absorbed more efficiently with increasing levels of dietary fat.
Alex, thanks!!!
Chris
I think not enough people are aware that too much of some vitamins can be very dangerous.
About a year ago, after attending (and briefly speaking at the Institute of Medicine meeting on vitamin D supplementation), I co-wrote a post which pointed to additional liabilities with pro-supplementation arguments:
http://bacteriality.com/2009/08/10/iom/
If you want to read something more technical, the above post links to a peer-reviewed paper I co-wrote along these same lines.
Would someone help me out and tell me the equivalent in IU’s? Thanks!
Chris,
The 2004 study in the American Journal of medicine which showed that bmd plateaus between 32 and 40 ng/ml serum D3 was not available (no free access), but a 2009 study in the Journal of Clinical Endocrinology Metabolism (PMID 19158198) showed no plateau up to 52 ng/ml, the upper limit of the graph, and showed continuous improvement in physical performance up to 60 ng/ml.
It certainly seems reasonable that increasing vitamin D3 intake without increasing intake of vitamin A and K2 would result in a significantly lower “optimum” vitamin D3 intake. I think it very unlikely that maintaining a vitamin D3 level of 50 ng/ml would result in reduced BMD as long as vitamin a and K2 intake is adequate.
Hi Jack,
Thanks for your comments. The PMID you gave must be wrong because it yielded a Russian article on a different topic. Nevertheless, I accept your point that there will be variation between populations and studies, which is one of the main points to draw from the NHANES data. I agree that 50 ng/mL may be safe and effective for everyone or near everyone if vitamin A and K2 status is sufficient, but I have no data to directly support that and, alas, you and I are both just guessing. Moreover, both of these studies are observational so we are just making an educated guess that high 25(OH)D will cause decreases in BMD to begin with.
The reality is that almost anything applying to the 50-100 ng/mL range is based on guessing.
Chris
I appreciate the quality of this website info. I personally have expressed for decades now, that, like in one Star Trek episode, we need a universal body analyzer for each person’s imbalances, so we can then simply give them what they need, exactly.
I am sure it will come one day, but we want it now, at least those of us who are somewhat awakened and would like to cruise on this planet a while longer and enjoy it. (grin)
Thanks for your specificity.
Eroca
The spline curve in Figure 1 of Melamed et al’s NHANES III study seems to suggest that about 35ng/ml may be optimal for reducing all cause mortality. Link : http://www.ncbi.nlm.nih.gov/pubmed/18695076?dopt=abstract
Of course, this is only an observation study, but it seems to me like a good place to start until randomized trials are conducted.
Hi Rob,
Thanks. I like that level. But there’s a lot more to do in characterizing the requirement, as you note!
Chris
I had my 25(OH)D levels tested by a vitamin D research group last week and it was 891ng/ml (yes, that’s right, 891 nanograms per milliliter). I nearly hit the floor in amazement at my recklessness. I had been an alcoholic for ten years and figured I would put myself on an aggressive therapeutic dosing program that started last month. It included retinyl palmitate, menatetrenone, and all eight kinds of vitamin E (and B vitamins and Zinc to support the retinol). I was recently diagnosed with mild to moderate autism so I guess I’m really not altogether in my head so I decided to quit this nonsense and just eat good food.
I have an appt with a physician for a physical in a few days and I’m hoping that maybe I took my last dose of D too close to the initial test. As for real symptoms, I do remember feeling nauseous and hyper-thirsty for a week a while back but I am feeling fine now. I am terrified that I’m going to find out that my kidneys are failing or something. I did take a bit more vitamin K2 and A in the hopes of balancing it out, but at the moment I’m just going to wait for the doctors tests and prognosis and not rely on my foolish self lest I end up killing myself inadvertently.
Is there anything I might want to do or watch out for while I wait for my appt.? I am trying to avoid oxalates as much as possible to prevent kidney stone formation.
Looks like I’m 11 years late to comment, but I’ll give it a shot anyway. Specifically with regard to the study of kidney stones in Israeli lifeguards, it looks like the authors of that study overlooked dehydration as the cause of both the elevated levels of kidney stones and hyperuricemia. I know of no connection between Vitamin D and uric acid concentration, so dehydration would be a singular cause of both observations. The dehydration is mentioned thusly: “Daily urine volumes of the 11 LG who worked in the warm and arid Gulf of Eilat had mean daily urinary volumes of 856±270 ml/24 hr, which was lower than in controls and lower than in LG and in controls from Northern Israel (p<0.01).”
According to https://kidneystones.uchicago.edu/the-low-flows/ “ Epidemiological links between becoming a stone former and urine volume have been discussed on this site already. Above above 2.5 liters daily is no longer associated with stone onset.” The mean volume of the lifeguards was 0.856 liters daily, well below the cited value. Note that web site also confirms dehydration as associated with elevated kidney stone risk. Other sources indicate dehydration as known and common risk factor for kidney stones.
And lifeguards, even ones who operate in indoor pools, suffer from dehydration per “Prevalence of Inadequate Hydration Levels in Aquatic Safety Personnel: A Pilot Study” http://faculty.usi.edu/media/1344/08_anderson-ijare_2015-0019_329-341.pdf
None of this proves high dehydration was causal of course, but I think any reasonable assessment would indicate the authors of that study found support for a pre-existing theory rather than an objective look at all the evidence they uncovered. Dehydration appears more explanatory of all the evidence than Vitamin D toxicity.
Dang. I’ve been taking supplements. Definitely need to do some testing. But, so happy the sun is about to come out – time to spend some time outdoors!