A Reasonable Approximation

Latest posts

Chris Masterjohn on Coronavirus, Part 1

Update (April 28th 2020): Part 2 published. Also a minor update to the ACE2 section.

Update (April 5th 2020): Chris released version two of his recommendations on April 2nd. There are some fairly major revisions. In particular, Chris now thinks Coronavirus infects the throat, not just the lungs. He was already making recommendations just in case this turned out to be true, so he hasn't specifically added anything, but it's still an important update. The other thing of note is that he no longer recommends using copper spray or taking echinacea until you get sick, except (for the copper) before and after potential exposure. I've updated this post accordingly. You can still read version 1.

Chris Masterjohn is a nutritionist who has some advice on supplements to take to help protect against Covid-19, and some to avoid. The raw advice is available for free, but the full report with explanation and references costs $10. I bought a copy. (The public advice has not been updated to reflect version 2.)

Should we trust him? On the one hand, "nutritionist" is not a profession I necessarily hold much respect for (it's not a protected term). Nor do I tend to think highly of people emphasizing that they have a PhD. Also, his website looks super sketchy to my eyes. Also also, something that costs money gets fewer eyeballs than something free, and so mistakes are less likely to be discovered.

(Only the last one of those should be considered a problem with the report; the rest are just priors.)

On the other hand, Chris previously turned me on to zinc so he has some credibility with me. Also, although I'm out of my depth in actually evaluating the science, I do think I have a decent bullshit detector, and the report is superficially very convincing to me.1 I get the sense that he actually really knows what he's doing, is weighing the evidence appropriately and making claims that are justified by it. He admits to uncertainty, and some of his recommendations are "this probably won't help, but just in case"; but he does speak precisely enough to be falsifiable. This doesn't mean he's right, of course. But I think he's worth some people paying some attention to.

My intention here is mostly to summarize his reasoning, in more detail than is on the linked page but less detail than is in the full report. You can make up your own minds. Much of it I haven't even checked how actionable it is right now (i.e. are these things available in stores or for delivery, and how much do they cost). I do have some questions and commentary of my own that I've scattered about, possibly not always well-flagged.

The report makes fifteen recommendations about fifteen supplements (some supplements are in multiple recommendations, and some recommendations are about multiple supplements):

"Potential exposure" is quite broad, including:

  • Going into public spaces to perform essential errands.
  • Accidentally encountering someone you don’t live with face-to-face within six feet, or making physical content with such a person.
  • Touching any surfaces that someone you don’t live with may have touched in the last nine days without any personal protective equipment.
  • Incorrect use of personal protective equipment (for example, taking gloves off your hands in any other way than explained here or reusing a mask without following these guidelines).
  • Putting your fingers into your mouth or nose, regardless of whether you washed them beforehand.

I think he's mostly expecting these won't happen very often. Personally, I'm doing most of them about once a week to go shopping, and the last one probably several times a day.

It includes any physical contact with someone you don't live with, coming face-to-face with such a person within six feet, and putting your fingers in your mouth or nose even if you washed them first.

In this post I'm going to focus on elderberry, zinc and copper; my current plan is to do the rest in future posts.

So far, I've followed this advice to the extent of:

If you're going to follow any of it yourself, you should at least read his public post, and ideally also the full report.

General info: ACE2

SARS-CoV-2 (the virus that causes the disease Covid-19) is not like the common cold or the flu, and things which work against those may not be helpful here. On the other hand, it's very like SARS-CoV (the virus that causes SARS). The genomes are 80% identical, and "the majority of its proteins are 85-100% homologous, with an average homology of 87%".

(Question: That seems like a really low average, given that range? Though I guess it includes the ones below 85%.)

The two main things the report focuses on are ACE2 and interferon.

ACE2 is an enzyme used in regulating blood pressure. SARS-CoV-2 enters our cells by docking to ACE22. It has this in common with just two other human coronaviruses, SARS-CoV and HCoV-NL63. So heightened levels of ACE2 could increase susceptibility to SARS-CoV-2.

(Question: "The binding of SARS-CoV to ACE2 is thought to downregulate ACE2 expression, leading to a loss of the anti-proliferative and anti-fibrotic actions of ACE2 in the lung, thereby contributing to the lung damage that occurs in SARS." That seems to point in favor of increasing levels of ACE2, so that it getting downregulated is not such a big deal. How to weigh these two concerns? I haven't read it closely, but this sounds like NL63 only affects ACE2 on infected cells, and doesn't so much "downregulate" it as "obliterate" it. So my super tentative take is: even if we start with high levels, that may not be sufficient to be protective if the infection gets bad; it would be more important to focus on stopping it from getting bad. But followup questions would be: what percent of cells express ACE2? What percent of those get infected, in a (light/moderate/bad) infection? When we do things that increase/decrease ACE2, does that increase the amount of ACE2 on individual cells, or the number of cells with any ACE2, or both? If a cell has lots of ACE2, does it do more of the sorts of things that ACE2 is good for?)

Update 28th April 2020: Chris has written more about this, see section "ACE2: Friend or Foe?" He points out that differing ACE2 levels are likely to have much more effect on viral replication (which is an exponential process) than on normal bodily functioning (which isn't).

I'll talk about interferon in a future post, because it's not relevant to anything in this post.

Daily essential: Elderberry

"In rhesus monkey kidney cell culture, elderberry has virucidal, anti-plaque, anti-replication and anti-attachment activity toward HCoV-NL63". Most of the effect seems to come from caffeic acid. That binds directly to ACE2, which suggests elderberry would be similarly effective against SARS-CoV and SARS-CoV-2.

(Question: is that in vitro? Doesn't seem very reliable in general, but maybe the specific mechanism makes it moreso.)

As a bonus, elderberry is also effective against avian infectious bronchitis virus through compromising the lipid envelope. Since all coronaviruses have one of those, that effect might generalize to SARS-CoV-2.

Other foods include caffeic acid, but only black chokeberries have anything like a comparable amount. And elderberry extract is the thing that's been actually studied, so that's what Chris recommends.

There are studies using elderberry in humans to treats colds, the flu, and cardiovascular disease, but Chris doesn't mention their results. He just uses them to determine a safe dose.

Daily essential: Nutritional zinc

This is zinc from food or most supplements (including tablets or capsules).

Zinc "inhibits at least two proteins required for SARS-CoV replication, papain-like protease-2 and helicase." So it probably inhibits the homologous proteins in SARS-CoV-2. (In version 1 Chris thought it also inhibited the protein helicase, but he no longer believes that.)

(Question: how similar do proteins need to be for this to be a decent guess? This suggests proteins called "homologous" might be only 40% similar in sequence. If I guess that "homologous" means this is likely to be a decent guess; and that these ones are >85% similar (based on base rates of similarity between the viruses)… that suggests it's probably a pretty good guess? But I'm not at all cofident in this thinking.)

So we should try to deliver zinc to the relevant tissues. What are those?

The infection would begin somewhere between the nose or throat (where the virus mostly enters our body) and lungs (where it primarily infects), wherever the virus first encounters ACE2.

(Question: this seems to assume that the virus doesn't first infect something outside of this path, and then move on to the lungs. Is that a safe assumption? I would guess so.)

In version one, Chris thought the virus wouldn't find any ACE2 until it reached the lungs. To quote the previous version of this post:

There are two papers trying to answer this question, and they give different opinions. …

Hamming et al (2004)3 suggests that the virus wouldn't find any ACE2 until it reached the lungs. They did find ACE2 in the oral and nasal mucous membranes, but on the wrong side to be any use to the virus.

Xu et al (2020)4 argues that ACE2 is highly expressed through the surface of the mouth, especially the tongue.

The two used different methods. Xu had better resolution, down to single-cell. But Hamming could tell where on the cells the ACE2 was expressed. The second thing matters, and the first doesn't. Xu's results are what Hamming would have predicted, they just aren't relevant. (The symptoms are also what Hamming would predict: a cough suggests lung infection, and is present. A runny nose or other cold symptoms would suggest throat infection, but they aren't present.)

But now there's direct evidence that the virus infects the throat too. (Note that the "not-yet peer reviewed" paper is now peer reviewed.)

So it seems that although the main infection site is the lungs, the infection starts in the mouth, nose or throat. We can specifically target those with ionic zinc, for which see below.

We don't have a specific mechanism to target the lungs with zinc. We just have to take it orally in the normal way (that is, in food or supplements) and let it be distributed there.

Chris recommends 7-15mg of zinc four times a day, away from sources of phytate ("whole grains, nuts, seeds, and legumes") which can inhibit zinc intake.

(At one point he says 10-15mg, and at one point 7-10, but I think this is just bad proofreading. Mostly he says 7-15.)

Conventional wisdom says we can't absorb nearly that much, but Chris thinks we just need more dakka: "the relevant studies have been limited to less than 20 mg/d. Supplementation with 100 mg/d zinc sulfate has been shown to more than double total zinc absorbed from 4.5 to 10.5 mg/d, while decreasing the fraction absorbed from 43% to 9%."

At such high doses, side effects have been observed. "Zinc at 50 mg/d has been shown to lower superoxide dismutase, and at 85 mg/d increased self-reported anemia. Both of these could probably have been averted by proper balancing with copper, which is addressed in the section below. However, the increased need for copper at high zinc intakes reflects increased expression of metallothionein, which can bind to many positively charged metals besides copper." I confess I'm not sure what this "however" is meant to mean. It kind of sounds like "so we still probably shouldn't go that high", but then we go that high anyway. I'm a bit concerned about this.

(I also confess that I have no idea what superoxide dismutase is.)

If you take zinc, you should balance it with copper.

Multi essential: Ionic zinc

This specifically means zinc from the kind of lozenges that work for a cold, or failing that from an ionic zinc throat spray.

(The mechanism that makes this work against a cold will not help with SARS-CoV-2.)

It delivers ionic zinc to the mouth, nose and throat tissues, like we couldn't do with the lungs. Chris recommends using one lozenge a day preventatively; plus several a day on the first sign of symptoms; plus an extra one before and after any potential exposure.

(Question: to be effective against the cold, this delivers ionic zinc to the surfaces of these tissues. Here we want it on the inside. Will that work?)

Daily essential: Copper

Copper surfaces work great against coronaviruses. This knowledge is not super helpful, since we are not copper surfaces.

It does suggest that copper ions in our cells might be toxic to the virus. But this has never been well studied.

Like zinc, copper inhibits papain-like protease 2 of SARS CoV. But it's much less effective at it.

The main reason to take copper is to keep the zinc-to-copper ratio acceptable. It should be between 10:1 and 15:1. (At one point he says 2:1 - 15:1, but again, I think that's bad proofreading.)

Essential if you get sick, before and after exposure: Copper spray

He doesn't actually have loads of detail on this in the report. That was less of a problem in version 1, when he considered it a hedge. Now that he calls it essential, it feels lacking. Fortunately, he's actually written about it publicly in more detail. (Though I think still less detail than the report usually contains.)

He recommends also using an ionic copper spray for the throat infections. This is similar to his recommendation for ionic zinc as well as nutritional zinc, but he doesn't say anything about copper lozenges. I assume those don't exist and/or wouldn't work for some reason. The goal in this case is to get the copper to the surface of the relevant tissues, in hopes that this destroys the virus in much the same way copper surfaces do. It's not clear whether he has specific evidence for this effect (even against other viruses), or just expects it to work based on theory (and if so how robust he considers that theory). Based on what he's written I don't think I'd put this in "essentials" (even acknowledging that "essential" is already an exaggeration).

He doesn't recommend using it daily. Partly because he thinks the effects wouldn't last long enough for that to be worthwhile. (Question: why is zinc different here?) But also partly because he's not convinced it's safe long term. Stop if you notice irritation.

He recommends a specific brand (named in the above link) that seems to have some evidence for short-term safety.

  1. There are some things that kind of smell like bullshit to me. Most notably, I feel like at times, the report goes into a lot of detail on things that aren't super relevant, like the renin-angiotensin system that ACE2 plays a part in. As far as I've seen so far, the precise mechanics of that don't really matter. Meanwhile, a lot of the important claims are speculative - necessarily so, because things are moving too fast to have good evidence here, but speculative all the same. In combination, this can kind of feel like… "throw a lot of impeccably researched, uncontroversial and unimportant science at the reader; then try to sneak in the difficult bits under the radar"? I'm sure there's a term for this that I'm forgetting.

    I don't actually think this is what's happening. My sense is that it's more likely to be bad editing, and I'm not even confident it's that. But it seemed important to note.

    I've also noticed some inconsistencies that I chalk up to bad proofreading. 

  2. Some papers talk about "ACE2 receptors". I currently think they're being imprecise, and there's no such thing.

    By comparison, the common cold is mostly caused by rhinoviruses, and most of those dock to ICAM-1. Some colds are caused by coronaviruses, but those dock to aminopeptidase N or sialic acid. The flu docks to sialic acid. So if something protects against those by preventing them from docking, it's likely to have no effect on Covid-19. 

  3. Hamming, I. et al. Tissue distribution of ACE2 protein, the functional receptor for SARS coronavirus. A first step in understanding SARS pathogenesis. J. Pathol. 203, 631–637 (2004). 

  4. Xu, H. et al. High expression of ACE2 receptor of 2019-nCoV on the epithelial cells of oral mucosa. Int. J. Oral Sci. 12, 8 (2020). 

Posted on 29 March 2020

Tagged: practical