Zinc Supplementation for Eating Disorders

The topic of zinc supplementation came up in a thread on the forums and I want to address it more specifically to frame the clinical trial data for those undergoing recovery who are not in an inpatient or residential care setting.

Oregon State University: Flickr.com

Oregon State University: Flickr.com

The Science

Dr. Soodabeh Safai-Kutti hypothesized that zinc deficiency might be a player in the pathogenesis (the origins of onset) of anorexia nervosa. In an open trial of 20 patients Safai-Kutti placed the subjects on 45-90 mg of zinc sulphate. An open trial means that these patients were not being compared to a placebo group or a control group.

During the follow-up period of 8 to 56 months, 17 of the patients had achieved a 15% (or more) increase in body weight. 1 Additionally, patients did not lose weight and none developed bulimic symptoms.

Taking up the need for a more controlled trial, Dr. C. Laird Birmingham and his colleagues at the University of British Columbia, here in Vancouver, are responsible for a follow-up on Safai-Kutti’s initial work, using elemental zinc supplementation to support weight restoration for anorexic patients.

Birmingham performed a randomized, placebo-controlled, double-blind trial in 1994 on 35 hospitalized anorexic (AN) patients. Double-blinding in clinical trials means that neither those giving the substance being studied, nor those receiving the substance being studied are aware of whether the subject is receiving the real thing or a placebo (sugar-pill made to look exactly like the real thing). Double-blinding greatly lowers the possibility that the trial results are due to either experimenter or placebo effects.

Experimenter effect is when an experimenter knows the preferred outcome he or she would like; takes great pains to avoid having that become apparent to the subjects in the trial; and yet nonetheless the subjects are swayed by the need for a certain outcome. Lo and behold the hoped-for results are achieved. If you are interested in experimenter effects (they are everywhere), look out Robert Rosenthal’s amazing work on the topic. Placebo effect is something of which most of you are likely already aware, and it involves the subjects themselves hoping for an outcome that they can achieve merely through the belief that the outcome will occur.

The subjects in Birmingham’s trial were given a 100 mg of zinc gluconate each day (or the placebo) until they achieved a 10% increase in body mass index. The 16 patients who turned out to actually be receiving the zinc gluconate gained weight at twice the rate of those who were taking the placebo. There is a 97% certainty that these results are accurate and not the result of accident or chance. In data-speak the outcome is identified as the statistically significant result of p=0.03. 2

Back in the 80s other scientists had already determined that those with anorexia were prone to zinc deficiencies. 3

However the data as to whether zinc deficiencies in these patients (both AN and BN) actually sustain  eating disorder behaviors or not, are inconclusive. 

One trial of both bulimic and anorexic patients showed that zinc status improved in those who were receiving zinc supplementation compared to those receiving a placebo and the authors suggest that this zinc deficiency sustains restrictive eating behaviors. 4 However another multi-site trial indicated that there was absolutely no correlation (link) between zinc levels in patients and symptomatology of the restrictive eating behaviors. 5

Alright, what do these trials tell you if you are someone with an eating disorder; about to undergo re-feeding and weight restoration; and trying to decide whether you should take zinc supplements or not?

The Framing

First and foremost, if you do not have a zinc deficiency and begin taking it as a supplement, then you will likely generate a copper deficiency that can lead to anemia. So clearly taking these supplements should not be done in the absence of there being a clear deficiency in the first place. 

I don’t think that this point can be overstated: don’t supplement your diet in the absence of test-identified deficiencies.

Secondly, while Birmingham’s trial is ironclad from a scientific point of view, it necessarily looks at a very small piece of the overall recovery-to-remission process for those with eating disorders. In an inpatient setting, where in fact Dr. Birmingham worked at the time, the focus is always one of medical stabilization and not remission from the chronic condition in question.

Doubling the rate of weight restoration will necessarily halve the time that the patient must remain in a hospital setting. That is a tremendous outcome from a medical standpoint, but it tells us nothing about whether those 16 patients achieved remission or maintained remission.

Thirdly, 16 patients with double the weight restoration on zinc, without further repeat trials of the caliber of Birmingham’s original trial, reveal a compelling need for further research and not a definitive conclusion that zinc supplementation is something to apply automatically during a recovery effort.

Fourthly, as I mentioned in my tweet on this topic, supplementation presupposes not only an identifiable deficiency, but also an otherwise robust intake of food in the first place. You cannot supplement a restrictive intake of food because supplementation is defined as “a quantity added”.

All of the patients were under medical supervision and were eating recovery amounts of food. If you are not at recovery intake guidelines, then you are not adding to your diet, you are replacing it. These micronutrients are not designed as replacements for food.

When a deficiency of a micronutrient such as zinc is present, then if it is due to over all food intake restriction, then the hierarchy of rectifying the deficiency is to first redress the overall food intake. In fact, a zinc deficiency has to persist despite adequate food intake for it to be considered a clinical deficiency.

Working with your health care provider, you can screen for zinc deficiency.

If a deficiency is actually present, then you need to be honest with your health care provider as to whether you are still actively restricting either through taking in too little food and/or applying exercise sessions that create net energy deficits.

The eating disorder patients for whom zinc supplementation was beneficial were not restricting food intake. So there is no point in even beginning to apply 50-100 mg of zinc gluconate or zinc sulphate if you are not going to embrace an unrestricted intake of food.

And finally, members active on the forums are in vastly different community environments and in hugely variable states of health and illness. Your individual circumstances matter greatly when it comes to recovery, restoration and remission. 

The upshot?

Food: yes. 

Supplementation: not unless your medical team says so (and even then you may want to discuss with them that such interventions are based on very narrow clinical trial data as per above).

On other news the blog post that was part of the vote on the forums will be up in the next two weeks…leaving you in suspense on the winning topic.

1. Safai-Kutti, S., E. Selin, S. Larsson, R. Jagenburg, I. Denfors, G. Sten, and I. Kjellmer. "Zinc therapy in children with cystic fibrosis." Beitrage zur Infusionstherapie= Contributions to infusion therapy 27 (1990): 104-114.

2. Birmingham, Carl L., Elliot M. Goldner, and Rita Bakan. "Controlled trial of zinc supplementation in anorexia nervosa." International Journal of Eating Disorders 15, no. 3 (1994): 251-255.

3  Safai-Kutti, S., and J. Kutti. "Zinc therapy in anorexia nervosa." The American journal of psychiatry 143, no. 8 (1986): 1059.

4. McClain, Craig J., Mary A. Stuart, B. Vivian, M. McClain, R. Talwalker, L. Snelling, and L. Humphries. "Zinc status before and after zinc supplementation of eating disorder patients." Journal of the American College of Nutrition 11, no. 6 (1992): 694-700.

5. Røijen, S. B., U. Worsaae, and G. Zlotnik. "[Zinc in patients with anorexia nervosa]." Ugeskrift for laeger 153, no. 10 (1991): 721-723.