VDC test kit slider
VDC test kit slider
sperti logo 1
Text size A A A
High contrast on off

Information on the latest vitamin D news and research.

Find out more information on deficiency, supplementation, sun exposure, and how vitamin D relates to your health.

New study on vitamin D levels and hospital mortality

I have always recommended a 25(OH)D level between 40 and 80 ng/ml, or about in the middle of the reference range (30-100 ng/ml).  The traditional reason the reference range goes up to 100 ng/ml is because sunbathers and lifeguards occasionally have such levels. To my knowledge, the highest recorded 25(OH)D level from sun exposure alone is from a pregnant free-living hunter-gatherer in modern day Tanzania whose level was 105 ng/ml.

Luxwolda MF, Kuipers RS, Kema IP, van der Veer E, Dijck-Brouwer DA, Muskiet FA. Vitamin D status indicators in indigenous populations in East Africa. Eur J Nutr. 2013 Apr;52(3):1115-25.

Now evidence exists that levels above 60 ng/ml may be harmful. First author Dr Karin Amrein, of the Medical University of Graz in Austria, studied 90 day mortality in more than 24,000 patients admitted to one of two Boston teaching hospitals between the years 1993 and 2011.

Amrein K, Quraishi SA, Litonjua AA, Gibbons FK, Pieber TR, Camargo CA Jr, Giovannucci E, Christopher KB. Evidence for a U-shaped relationship between pre-hospital vitamin D status and mortality: a cohort study. J Clin Endocrinol Metab. 2014 Jan 13:jc20133481.

The authors separated the patients (average age 61) into seven categories of 25(OH)D levels. The lowest was < 10 ng/ml and the highest was > 70 ng/ml with every 10 ng/ml increment in between. Then they compared the 25(OH)D levels to 90 day mortality. The lowest odds ratio of dying was .94, which was seen in those with 25(OH)D levels between 50-60 ng/ml. However, those with levels below 30 ng/ml and above 60 ng/ml had a statistically significant increased risk of dying in 90 days.

Of the 24,000 patients, 390 (1.6% of total sample) had 25(OH)D levels between 60 and 70 ng/ml and 299 patients (1.2% of the total sample) had levels above 70 ng/ml. Patients with these two levels had odds ratios of dying in 90 days of 1.5 and 1.7, respectively. Of course, those with levels below 30 ng/ml also had an increased risk of dying.

Eight percent of the 390 patients with levels between 60 and 70 ng/ml died and 8% with levels above 70 ng/ml died in 90 days. The lowest death rate (5%) was in those with levels between 30 and 60 ng/ml. The highest death rate (10%) was in those with levels <10 ng/ml.

Unfortunately, the source of the vitamin D necessary to get levels above 60 ng/ml was not identified. Through most of those years it had to come from one of three sources: the sun, prescription ergocalciferol, or cod liver oil. During the majority of those years, vitamin D3 supplements were only available in 400 IU preparations, not enough to explain levels above 60 ng/ml. Food cannot get levels above 60 ng/ml unless one eats reindeer meat, seagull eggs, lichen and whale blubber.

There was a seasonal difference in those with high levels, but it does not look like enough sun was obtained to explain those high levels. The season of blood draw’s influence was not different for those with levels between 60 and 70 ng/ml and only slightly different for those with levels above 70 ng/ml. So sunshine cannot explain the high levels.

That means the patients with high levels during most of those years were probably either taking prescription ergocalciferol for osteoporosis or a cod liver oil supplement. Readers know about the potential problems with cod liver oil and its high vitamin A content.

Letter to FNB about subclinical vitamin A toxicity signed by 18 experts.

What about ergocalciferol? A recent subgroup analysis of the most recent Cochrane meta-analysis of randomized controlled trials and mortality suggested that vitamin D2 may increase mortality.

Bjelakovic G, Gluud LL, Nikolova D, Whitfield K, Wetterslev J, Simonetti RG, Bjelakovic M, Gluud C. Vitamin D supplementation for prevention of mortality in adults. Cochrane Database Syst Rev. 2014 Jan 10;1:CD007470.

So in all likelihood, these patients were either using cod liver oil or taking prescription ergocalciferol, both of which may be dangerous.

  About: John Cannell, MD

Dr. John Cannell is founder of the Vitamin D Council. He has written many peer-reviewed papers on vitamin D and speaks frequently across the United States on the subject. Dr. Cannell holds an M.D. and has served the medical field as a general practitioner, emergency physician, and psychiatrist.

10 Responses to New study on vitamin D levels and hospital mortality

  1. Rita and Misty says:

    Dr. Cannell,

    I proudly support the VDC because of your courage to call for vitamin D levels between 40 ng/ml–80 ng/ml. I wish other organizations would do the same. Then I could support these organizations as well. To utilize the levels of 40ng/ml–60 ng/ml (to me) is unethical, as the health benefits between 60 ng/ml–80 ng/ml are tremendous (again in my opinion).

    I haven’t yet read the article upon which today’s blog is based. I tend to think along parallel lines with you–cod liver oil or prescription ergocalciferol might be the culprits. Or perhaps these individuals were already very ill prior to commencing vitamin D supplementation?

    As you and I both know: there is no cure for death. But, we can choose to live healthy, productive and happy lives (whatever happy means). To me, an optimal 25(OH)D level plays an integral role in my health.

    Be well, and know that I do appreciate your efforts,

  2. Ian says:

    Their conclusion:

    “Though previous reports have suggested an association between low vitamin D status and mortality, these data raise the issue of potential harm from high serum 25(OH)D levels, provide a rationale for an upper limit to supplementation, and emphasize the need for caution in the use of extremely high doses of vitamin D among patients.”

    Like you say the likely source of vitamin D is cod liver oil, especially in Austria. Has the letter to the FNB been acted upon? I see so many people, some medical and pharmaceutical advising people to take cod liver oil. It is a heavily advertised product. The NIH office of dietary supplements put cod liver oil as the number one food source but makes no mention of vitamin A or toxicity due to excess vitamin A.

    I know of many older people who take daily cod liver oil. I also wonder as people get older and sicker they may increase their dose in the belief it may improve their health especially when you see the marketing messages and where they come from. So when a cohort study like this is done it taps biased data in two ways:
    Older people taking an “old” remedy and sicker people taken more of an “old” remedy. Another confounding factor could be the diet of older Austrians, Do they eat a lot of liver and other organ meats, they certainly add them to their sausages.

  3. Rita and Misty says:

    Ian–60 ng/ml is not the upper limit of an optimal 25(OH)D level. It isn’t. Certainly, since I am not a researcher or physician many here may scoff at my comments as being anecdotal. They would be foolish, regardless of their supposed knowledge or educational status.

    Look to the success stories in Brazil (Dr. Coimbra I think) with multiple sclerosis. The 25(OH)D levels this physician looks to for good health are well above 60 ng/ml.

    We are overrun with autoimmune disease, are we not?

    I am being redundant, for sure, but my health only improved once my 25(OH)D level was raised to 74 ng/ml. I choose to keep it higher currently. I have experienced almost every issue associated with being born in mid-March, also known as the very height of vitamin D winter, with the lowest 25(OH)levels of the year. Now, I am healthy.

    My good health is not idiopathic—as mainstream medicine termed my ill heath.

    My good health is due to an optimal 25(OH)D level.

    Look to hunter/gatherer societies, and aim for between 46 ng/ml–105 ng/ml.

    John Cannell, keep up the good work.

  4. Ian says:

    Some people are very quick to use these studies (above and the Lancet meta-analysis) to substantiate their “position” on vitamin D supplementation and the importance of vitamin D level in the prevention of disease: Like Mark Bolland of the University of Auckland:


    It would be good if they did some decent research instead of these meta-analyses to show pre-ordained U-shaped relationships and “futility”.

  5. Magic says:

    I have been studying the benefits of D3 for about 8 years and I am convinced that Dr Cannell and others are absolutely correct in their findings. It is inexpensive and harmless in amounts up to 10,000 a day and probably much higher.

    Hospitals and nursing homes are a most dangerous place to reside. WHY aren’t 5000 units a day an automatic prescription??? It is not a profit center.

  6. Adelle Sperling says:

    A physician friend (largely influenced by me) now routinely has his patients who reside in nursing homes take 10,000 IU Vitamin D3 daily. They are not able to sun-bathe and of course are elderly.

  7. IAW says:

    To: Adelle Sperling
    “Bravo” to your physician friend for being open minded, smart and caring!

  8. There seems to be little or no effort to control for underlying disease states nor causes of death, making this entire treatise worse than merely misleading. This kind of “study” publication is very likely what Mark Twain had in mind when he added the third case to his statement, “There are three kinds of lies: lies, damned lies and statistics.”
    While indeed there can be too much of any “good thing”, this article has no merit in the debate.

  9. In mulling over my initial post, looking for flaws in my perception, I came across this thought: Were the individuals with the higher levels who died, in the process of chronic dying (as opposed to an acute version) and their bodies’ intrinsic systems fighting to avoid that demise by increasing their D levels in the face of that inflammatory/insulting process? If such a response is innate, does it fail to be as effective at greater ages due to the failure in other biochemical pathways of those oldest subjects, perhaps the body simply losing the ability, through the known aging processes, to “rise” to the occasion? Are the slightly less-aged cohorts surviving better because D compensates better in their systems?

  10. Magic says:


    Alzheimers disease…

    As usual vitamin d3 has been found to be “helpful” but the test was “too small.” It is a joke.