The recent article “Association of Plasma Concentration of Vitamin B12 With All-Cause Mortality in the General Population in the Netherlands,” reported in the January 15, 2020 issue of JAMA Network Open, although well intended, is very misleading for many reasons. Unfortunately, this study, title, and abstract may deceive health care providers, reporters, and the public to fear vitamin B12. Below, I unravel this study and conclude that their findings are misguided because of the following flaws and unknown variables:

1) Failure to realize there is a difference between cyanocobalamin verses hydroxocobalamin, methylcobalamin, and adenosylcoblamin. Cyanocobalamin has been found to be harmful in patients with renal impairment. This is caused by thiocyanate accumulation from cyanide in cyanocobalamin. Therefore, hydroxocobalamin and methylcobalamin should be used in renal impairment, not cyanocobalamin.

2) Failure to state or ascertain what form of B12 the participants were taking (cyanocobalamin vs. methylcobalamin or hydroxocobalamin).

3) Failure to ascertain how much B12 in mcg they were taking? Were they taking a supplement or getting it from food or both? Were they consuming fortified cereals or energy drinks with cyanocobalamin?

4) What percent of the 5,571 participants had renal impairment? The authors stated they took 7,768 people with renal impairment and 3,395 without renal impairment. This adds up to 11,163 participants. They then used data from 6,894 participants and excluded 1,265 people with missing B12 results and 58 people on B12 injections, leaving 5,571 participants to study. They failed to state what percent of the 5,571 people had renal impairment — (this is important because we know that taking cyanocobalamin with impaired renal function causes thiocyanate accumulation as pointed out in #1). This study may have had 50% or greater with renal impairment, but authors failed to disclose this, which is problematic and will affect the results.

5) The range where the authors defined serum B12 as highest vs. lowest categories is problematic. The range is very similar. The lowest group, B12 less than 338.85 pg/mL and the highest group, B12 greater than 455.41 pg/mL. This is only 116.56 pg/mL difference. These researchers are proposing that if one’s serum B12 is more than 455.41 pg/mL they are greater risk of all-cause mortality without knowing what percent had renal insufficiency, who was on cyanocobalamin (which I suspect may be nearly 100%), and what percent actually had metabolic B12 deficiency (defined as a serum B12 < 542 pg/mL and one abnormal marker [methylmalonic acid, total homocysteine, holotranscobalamin]).

This study concluded that there were 33.8 deaths/10,000 with B12 less than 338.85 pg/mL and 65.7 deaths/10,000 with B12 greater than 455.41 pg/mL. Therefore, they are claiming a difference of 116.56 pg/mL of B12 in one’s blood increased risk of death, which is preposterous. The cause of death in the 7.7 to 8.9 year follow up of these 5,571 participants is also not disclosed. How many died from causes completely unrelated to their “higher” B12 level (i.e. trauma, motor vehicle accidents, falls, suicide, homicide, stroke, myocardial infarction, cancer, etc.)?

6) People can have higher serum B12 concentrations because of genetic mutations. None of these participants were screened for this which could alter serum B12 levels in those not even taking B12. These same patients could actually be B12 deficient if other markers had been checked (methylmalonic acid, total homocysteine, holotranscobalamin). They did screen for homocysteine, but did not disclose what percent had elevated levels.

In addition to the above, do the researchers of this study know that only 6% to 20% of B12 is active? Reading the title of this article or skimming it over quickly and not understanding other facts can cause panic. Health care providers and the public may conclude, wrongly, that they should not have a serum B12 level over 442 pg/mL. This is very troublesome. However, as I have pointed out in my books years ago, we should be consuming methylcobalamin, adenosylcoblamin, or hydroxocobalamin and NOT cyanocobalamin. And we should be administering hydroxocobalamin or methylcobalamin injections and NOT cyanocobalamin.  This further supports my position, from recent research revealing that cyanocobalamin should not be used in people with renal insufficiency or chronic kidney disease (CKD) and renal failure.  Remember, the prevalence of CKD in the adult U.S. population is 11% and the prevalence in persons over age 60, is much higher at around 39.4%.

The following is further recent evidence for the need of higher serum B12 levels and for consuming or administering the proper form of B12. It is taken from numerous articles by J. David Spence, M.D., Professor of Neurology & Clinical Pharmacology at Western University in London, Ontario. He is the Director of the Stroke Prevention & Atherosclerosis Research Centre at the Robarts Research Institute. Dr. Spence is a Gold Fellow in the Stroke Council of the American Heart Association, and has served on leadership committees of both that Council & the Peripheral Vascular Disease Council of the AHA. He was a site investigator for the CLOSURE, RESPECT and REDUCE trials of percutaneous closure of patent foramen ovale and has also participated in studies of aspirin, clopidogrel and ticagrelor.

“High levels of tHcy markedly increase the risk of stroke…This is because a serum total B12 in the reference range does not define the adequacy of functional B12.  Only ∼6% – 20% of serum total B12 is active, so within the ‘normal range’ of serum B12 (∼217- 813 pg/mL), a large proportion of patients have inadequate metabolically active B12.  The serum B12 below which plasma levels of tHcy of methylmalonic acid become elevated is ∼ 542 pg/mL.  The combination of a serum B12 in the lower half of the reference range with elevation of plasma MMA (or in folate-replete patients, elevated tHcy) is called ‘metabolic B12 deficiency.’ ” (1)

To be confident one’s serum B12 is adequate it needs to be greater than 542 pg/mL. Dr. Spence advises it is necessary to perform confirmatory tests (methylmalonic acid, total homocysteine, or holotranscobalamin) when a patient’s serum B12 is less than 542 pg/mL. Dr. Spence also points out cyanocobalamin reduces risk of stroke in patients with normal renal function, but is harmful in patients with renal impairment caused by thiocyanate accumulation from cyanide in cyanocobalamin. He points out that methylcobalamin or hydroxocobalamin should be used in renal impairment.

It is critical to rule out and treat metabolic B12 deficiency according to Dr. Spence because metabolic B12 deficiency & elevated total homocysteine (tHcy) are so common among patients with stroke. All patients with stroke should have their serum B12 and tHcy measured, and should be appropriately treated (2).  Dr. Spence further notes that the treatment with B vitamins to lower homocysteine definitely reduces the risk of stroke, but results of early trials were clouded by harm from cyanocobalamin among patients with impaired renal function. Remember, Dr. Spence advises that patients with serum B12 levels less than 542 pg/mL need other adjunctive markers to rule out deficiency. In this study, however, they conclude that if your B12 is greater than 455.51 pg/mL you are more at risk of death because of too much B12, when in reality they could be deficient.

Patients who received high dose vitamins & could absorb B12 well (B12 > 436 pg/mL), “there was a 34% reduction of combined stroke, myocardial infarction and vascular death compared with patients who received low-dose vitamins & had baseline B12 below the median.”(1)  It is also apparent in early trials that harm from cyanocobalamin among those with renal impairment hid the benefit of B vitamins among those with normal renal function.

In JAMA (2010), the “Diabetic Intervention with Vitamin in Diabetic Nephropathy” (DIVINe) trial found cyanocobalamin to be harmful in patients with renal failure. In patients with diabetic nephropathy, a high dose of B vitamins (including 1,000mcg of cyanocobalamin) accelerated the decline of renal function and doubled cardiovascular events. All the events occurred in patients with GFR < 50 mL/min/1.73 m². (3)

Katsushi Koyama, MD, et al found “patients with renal failure have high levels of thiocyanate, which consumes hydrogen sulfide in the excretion of cyanide, and methylcobalamin lowered levels of both tHcy & asymmetric dimethylarginine (ADMA, a nitric oxide antagonist), whereas cyanocobalamin did not lower ADMA in the WENBIT trial in Norway. (3)

A 2017, meta-analysis reported in Lancet Neurology arranged by dose of cyanocobalamin and by renal function “CONFIRMED that B vitamins are beneficial in participants with good renal function, and HARM from cyanocobalamin in participants with poor renal function obscured this benefit.” (3)

In addition to the main flaws of this retrospective study, the authors themselves also disclosed that the following may have comprised their theory that higher plasma concentrations of B12 are associated with all-cause mortality, independent of traditional risk factors:

1) They did not have measurements of B12 plasma concentrations beyond baseline assessment, which limited them to evaluate the regression dilution of      B12.

2) They only had access to pharmacy records on injectable B12, but not for OTC tablets, which limits the implication of B12 supplementation.

3) That residual confounding variable are an important limitation in all observational studies.

4) Dietary patterns can also influence plasma concentration of B12 and risk mortality. In the PREVEND cohort, detailed dietary information was not                 available, which they considered a limitation.

5) That a noncausal association should not be discarded and deserves further investigation.

In conclusion, this is why one must read the entire article, and study it (not just the title and abstract).  Remember, as in anything, a headline can falsely sway one’s opinion.  


  1. Spence, JD. Cardioembolic stroke: everything has changed. Stroke Vasc Neurol, 2018 Jun; 3(2): 76-83.
  2. Ahmed, S, Bogiatzi C, Hackam DG et al. Vitamin B12 deficiency and hyperhomocysteinaemia in outpatients with stroke or TIA: a cohort study at an academic medical centre. BMJ Open, 2019; 9(1): e026564.
  3. Spence, JD.  The rest of the story on vitamin B for stroke prevention. Medpage Today, October 17, 2018.
  4. Spence, JD.  Metabolic vitamin B12 deficiency:  a missed opportunity to prevent dementia and stroke. Nutr Res. 2016 Feb;36(2): 109-16. doi: 10.1016/j.nutres.2015.10.003.
  5.  Spence, JD, Hachinski, V.  B vitamins for stroke prevention: interaction of low platelet count and high plasma total homocysteine. JACC, 2018 May; 71(19): 2147-8.