Long Covid Patient Microbiome Analysis

The Microbiome and COVID have strong relationships. The microbiome prior to COVID impacts the severity. The severity of the symptoms correlates with the microbiome changes. This leads naturally to Long COVID being a continuation of this theme.

 One study suggests that a core microbiota could predict COVID-19 severity in healthy subjects.27 Another study shows that the composition of the intestinal microbiota in the Chinese cohort is different between COVID-19 infected and un-infected controls, with symptom severity correlating with specific bacterial taxa.2

The gut microbiome of COVID-19 recovered patients returns to uninfected status in a minority-dominated United States cohort [2021]

A new study used fecal samples were collected at least 38 days following diagnosis. By common belief, the patients are fully recovered — except their microbiome are not!  What is the difference? It depends on how COVID presented.

  • positive detection of SARS-COV-2 RNA from the respiratory tract, defined as respiratory positive (RP) 
  • failure to detect in the respiratory tract, but had covid, is negative

They found changes in “13 phyla, 18 classes, 44 orders, 88 families, 234 genera, 1 phylum, 1 class, and 1 order were significant”. To put it simply, look at the changes below — they are NOT minor but major shifts!

Reversion of Gut Microbiota during the Recovery Phase in Patients with Asymptomatic or Mild COVID-19: Longitudinal Study [2021]

Our Long COVID Patient

Their summary:

  • End of March 2020: covid
  • Tachycardia until July 2020.
  • MRI showed pericarditis.Tachycardia stopped once I resumed H1 blockers
    • I had stopped out of paranoia when acutely ill, and B started antacid prescribed by cardiologist.
  • Severe constipation started at around April 2020. So we’re talking about over a year ago.
  • I had a break from constipation issue for about 6 months – november 2020 to may 2021.
  • Other ongoing symptoms are pain on the left side under the ribcage and internal vibrations, numbing of sensation “down there” (don’t feel much the need to go to the toilet,
  • Sex life is hampered too, nerve damage very likely according to myself and also gynecologist who thinks it’s postviral).
  • The entire first half of 2021 i adopted a diet based on green smoothies.
    • Other than 80% veg based smoothies and flaxseed i ate some veg stir fries and fresh salmon and some crisps (in small quantities!!! but every day a little pack).
    • Before covid i ate only crap, all sort of crap, so that this diet was for me a huge sacrifice.
    • But since covid junk food made me feel bad anyway, so slowly i accepted to change my diet. 
    • I’m also taking a very long list of supplements. I had high cholesterol before covid, had it at 18 already despite being slim, but after covid it remained high and my sugar level got very high (not yet in the red). [This may no longer be true — the measurements were from a year ago]

Pro Forma Analysis

First, we have two lists of bacteria available, the number of studies are few but slowly increasing.

Bacteria Out of Range

I see 34 Outliers using the Kaltoft-Moldrup ranges (which are usually bigger ranges than most testing labs use). This person mentions 50+ out of range from their lab. Well, that huge number is precisely what the study above reported. This is not a typical microbiome disruption.

End Product Out of Range

Three items were LOW out of range, Vitamin D, Phosphoamidase, and a-Galactosidase. For Vitamin D we have the following literature:

KEGG Bacteria Products Out of Range

Every single one of a list of 47 was low. Not enough being produced

KEGG Modules Out of Range

Nothing reported

KEGG Enzymes Out of Range

As above, a list of 48 items with every item being low

Kegg Suggestions

Where there are so many items with issues, I usually do not bother looking at them individually. Instead, I look at what can be computed to address them. Because every item is low, we do not need to look at trying to reduce anything — just add,

KEGG Suggested Probiotics

This is done by seeking out probiotic bacteria producing enzymes etc that are not being produced enough by existing bacteria. These can be viewed as a biological supplement producing items not available as regular supplements. The retail probiotics Sun Wave Pharma/Bio Sun Instant and Prescript Assist appear to be good choices (if available). The fall back by species are:

A common mistake is to slip into a homeopathic thinking, “oh, I am taking some — that is enough”. In general I recommend starting low and increasing to the maximum dosages used in clinical studies.

KEGG Suggested Supplements

We similarly identify supplements that are available retail (defined as being available on Amazon.com)

  • beta-alanine
  • D-Ribose
  • iron
  • L-Histidine
  • L-Lysine
  • L-Phenylalanine
  • L-Tryptophan
  • magnesium


One unique feature of Microbiome Prescription is that it not only identifies candidate issue areas, it also makes suggestions based solely on studies from the US National Library of Medicine. These suggests factor in side-effects on other bacteria. Every other site, has a blinkered thinking with their suggestions and do not consider side effects. Of course, there is one layer of side effects that only your MD can help — medical conditions you have. A suggestion may suggest peanut butter and you have an allergy to peanuts!

Checking against COVID and LONG COVID Profiles

The studies report on the US National Library of Medicine and are coded for averages being statistically high higher or lower than controls. This does not mean that the values are extremes. Statistically, this presents some challenges. I decided to explore how many matches happened with different definitions (Kaltoft-Moldrup ranges, top/bottom 3,6,9,12,15 %ile) for COVID and LONG COVID


For those who wish to do it themselves, go to advance suggestions and do settings like below.

We will be change Bacteria Slection and Explicit Bacteria

Then click the suggestions at the bottom. On the suggestion page, click Bacteria Details to see the bacteria that are picked

The results are below by bacteria. As we reduce how extreme values that are needed to be deemed “high” or “low”, we have more and more matches. C – active COVID; L – Long Haul COVID / Post COVID

Suggestions – 3 approaches

After viewing the table above, I decided to do 3 approaches:

  • 12%ile and COVID-19
  • 12%ile and Long COVID
  • Quick Kaltoft-Moldrup suggestions

I expect all to be similar but with some differences. I will cut off suggestions around .425 to prevent information overload (which happens easily with the microbiome)

12%ile and COVID-19

The lack of a fine graduation of Confidence implies that we do not have that many applicable studies for the bacteria identified as important. Also this is what they had, not currently have.

12%ile and Long COVID

This has the graduation of Confidence values that I like to see.

Quick Kaltoft-Moldrup suggestions

Reminder – The WHY for suggestions

On the suggestion line, you may see a ๐Ÿ“š. Clicking it will show the source of the recommendation and why. Remember the more positive impact (by number of studies reporting the same), the greater the confidence shown. It is the confidence that it will shift in the desired direction. It is not which works better. Microbiome Prescription strives to be open on the basis of it’s logic and allow easy verification by people who are interested.

Putting Suggestions together

Remember that the purpose of the site is to create prescriptions — suggestions to correct microbiome shifts. The suggestions attempt to be adjusted for side-effects on other bacteria. Labs suggestions are based on blinkered analysis, “You are too high in X, Z reduces X so we recommend it” – which often ignores the fact that X also increases Y which is also too high.

Suggestions are computed in two different ways with no overlap of source data (KEGG based on genes, and studies where substances were tested). Items that are on both sets of recommendations are definitely things to consider. There are items that may be only on KEGG suggestions because no one has done studies on them.

So our very top suggestions are:

Go to https://microbiomeprescription.com/Library/Dosages , then search for the probiotic of interest. then click on ๐Ÿ“ Studies and Trials beside it to see the dosages

A second criteria is to eliminate any items where there are contradictions (“playing it safe), the following are my top suggestions

At this point I should mention that a lot of items often used for microbiome issues appear contraindicated (i.e. AVOID). For example: neem,  quercetin,resveratrol,  triphala, resveratrol (grape seed/polyphenols/red wine),  glycyrrhizic acid (licorice),  melatonin supplement.

Remember, you can get opinions on over 3000 items in our database by going to the bottom of this list:

Iron was suggested by KEGG. I wanted to check it’s impact using the data from studies and was pleased with the result.

Iron was suggested by KEGG and appears to have a beneficial impact using studies data

I also confirmed magnesium was also a positive (and magnesium deficient, a negative)

This is a MODEL not a PROTOCOL

This is directed to people reading this post and saying “I will do what is described”. What is the difference? A Protocol comes from clinical experience and is a defined set of actions that are repeated for each patient. A model is a theoretical way to generate candidate actions that may help. This is not a model for Long COVID patients, it is a model for one person’s microbiome. Every Long COVID patient will have a different microbiome and thus different candidate actions. You can see this by looking at the next post on Long COVID microbiome.

An analogy, Long COVID can be compared to a headache. There are at least 17 types of headaches. You may need to see a dentist (tooth issues), or take a antihistamine (allergy) or take oxygen or …. Details drives the treatment.

For information on suitable 16s Microbiome Providers.

Bottom Line

Another post COVID person just contacted me, with their samples, so a second COVID post is also available Both this person’s sample, and the recent study confirmed my suspicion that Long COVID is a Post-Infection Syndrome. Post-Infection Syndromes are, IMHO, infection altered microbiomes that failed to return to normal.

Dialog Notes with User

Q: “I’ve got a huge issue/reservation with a part of the concept: the norm in the distribution of the data base might be far from normal. And even further, what is normal might not be optimal at all. What is normal reflects an average diet, but maybe an optimal diet would lead to an outlier sample, how do you address that issue….”

A: I do not use a bell curve, See this post for where I have evolved to. It’s based on Percentile and shape of distributions

Q:  As a laywoman looking at the data my immediate focus went to methane-sibo. This matches my current issues and I’m surprised you didn’t mention it. 

A: SIBO does not have a microbiome signature that is reliable. See this 2017 post reviewing the literature on SIBO

REMINDER: These are suggestions generated by an artificial intelligence program. Before implementing, they should be reviewed by your medical professional.

5 thoughts on “Long Covid Patient Microbiome Analysis

  1. I am the patient above. I spend a fair amount of time researching my condition and in recent days worked on the wonderful info unearthed by microbiome science, the software and Ken’s expertise. Two days ago I added all the recommended stuff (from the list I was already taking magnesium, vitamin D and L-Tryptophan, but now upped my vitamin D significantly) to my supplement routine. I also ate a lot of cocoa that wasn’t on Ken’s list but was suggested to me by the software with high confidence. Within hours i started feeling significantly worse, and this morning I was back to how i was a year ago, with shortness of breath, headaches and a increased heart rate. But I think i have already found the culprits! And I believe this experiment can be useful to many other long covid sufferers. In the summary i sent Ken about my symptoms and long-covid history, I mentioned my tachycardia resolved at the time i started taking H1 and H2-blockers. Over time i also slowly moved to a low histamine diet, and I have kept taking H1 blockers (ceterizine) and other natural antihistamine substances like quercetin. Antihistamines are becoming a known adjuvant to long-covid recovery, and are now even officially recommended on the UCLH website for example (university college london hospital). Back to my gut: it shows a lot of imbalances, and in order to rectify some of them the sofware recommends histidine and phenylalanine among other things. But why? Because my gut level of histamine is abnormally low: these are histamine precursors. The irony is that I painfully achieved this abnormally low level with hard work: it is actually the best i can currently achieve in terms of wellbeing. I am taking the time to write this up because i suspect that as a long covid patient I am rather typical in that respect ( and for example one avenue of research in long covid is to understand its link with MCAS). The challenge for me will be to continue improving, given my own limitations… I will now proceed with the list of recommended supplements, but leave out the alanine, histidine and phenylalanine. And out too is the cocoa binging… I can’t wait to update here with my progress. In the meanwhile three cheers to microbiome research in general and Ken in particular.

    1. Remember these are theoretical suggestions generated by an AI model. The goal is have better things (higher probability of success) to try than random internet friend suggestions.

      Suggestion should always be done one at a time, so if an adverse response occurs, you can identify that item, remove it and move on to the next suggestion.

  2. Interesting! Thank you for sharing.

    Dosages for the probiotics arenโ€™t available as they look to be a dead webpage link. Would it be possible to add as text or refresh the webpage link, please?

Leave a Reply

Your email address will not be published. Required fields are marked *