This email arrived and contributed in two changes:

• Warning on using Special Studies — suggestions can be erratic and should not be used if they conflict with the usual consensus building algorithms.
• Adding Modelled Food Suggestions — this should also partially address some challenges doing suggestions in Japan

Back Story

Love your work, I have read many of your posts on CFSremission, and based on that and my own research I think your microbiome-based view of ME/CFS is generally correct.

I have been suffering from cfs since 2009, I did a study abroad in South Korea and had a weird fever there, after that I got tired easily, often felt light-headed, head felt hot, but I was able to mostly live a normal life and actually spent a lot of time in the gym. (I think this supports the microbiome theory, my diet changed radically when I went to Korea)

I dealt with the fatigue with regular consumption of coffee/tea throughout the day, often going out drinking at night. I also used to lift weights in the gym almost every day. Full body lifts like squats, deadlifts, etc. During this time I drank about a quart of milk a day as part of my bodybuilding routine. My sleep always seemed unrefreshing.

Then in late 2019 I got sick with a EBV/Mono-type illness, swollen lymph nodes and tonsils, crushing fatigue, sore throat, that lasted a month. Sore throat resolved, but the tonsils were still a bit swollen and the lymph nodes got smaller but seemed to be permanently hard. I thought I might have thyroid issues or cancer, but multiple screenings ruled that out. 

I tried multiple times to go back to the gym, but my workouts were poor and I got hit with what I now understand to be PEM the next day. Eventually I had to stop trying to exercise. I’ve tried various supplements such as methyl-b12, doses of tumeric or curcumin, too many to count, then I discovered your website. I think it lines up with my experience and is a good model to explain the so-called “anomalous” way that some treatments work for some CFS sufferers and not for others.

So I am writing this email to you now hoping I can get some insight. I have read your blog posts about other CFS sufferers analyzing their samples, so I hope you could take a look at mine as well. Feel free to use this or parts of it as a blog post, but don’t use my name or email address obviously. 

Some background, I had to get a sample manually added from a lab here in japan, based on those suggestions I took two rounds of miyarisan, (I live in Japan, so it was the most easily obtained of the probiotics suggested) as well as added lots of inulin, oats, whole grains etc. to my diet. Similar to some of your other posts, including the suggestions to avoid Vitamin B supplements (the greedy bacteria taking the B-vitamins!) But, I feel like it made me worse. These days my legs are very heavy and tired. 

I had another sample taken between the two rounds of miyarisan which i sent to Biomesight (a much better choice.) It took over a month for the results to get back. When they came back, it suggested a totally different course, putting miyarisan and inulin into the strong-avoid category! With the inulin+oats+miyarisan diet, I am more tired and my libido dropped a lot.

So I made a new analysis based on the national and special studies for unrefreshing sleep, ME/CFS without IBS, cold intolerance, general fatigue, etc.

1. As attached, it suggests “alcoholic beverages” pretty high. How should I interpret this? Beer? Wine? Vodka shots every evening? Is there more context for this?

2. Not many foods with any strong suggestions, what can I eat realistically (here in Japan) off this list of suggestions? It suggests a milk diet, but whole milk is not suggested? Seems to have a lot of contradictory suggestions.

3. I decided to go with national + special studies, but the “general consensus” is totally different. I assume the studies are better for my condition?

Feel free to look at the data of both of my samples, or offer a different way of getting a consensus, I really need some guidance here!

Regards,

From a reader (with permission to post)

Initial Impressions

• “EBV/Mono-type illness” — EBV, HHV5 and other virus in CFS gives context, but this post, Viral Reactivation and the Microbiome gives context and cites “For EBV, viral load was significantly higher when 25(OH)D levels were low, demonstrating an inverse correlation between 25(OH)D levels and EBV load. ” [2018]. My first step would be to get 25D levels measures (likely low) and 1,25D levels (likely very high). It’s an easy issue to address with Vitamin D3 supplements (likely 20,000 IU/Day may be needed. See the following for more information
  • Vitamin 25D and 1,25D measure – one reduces symptoms and one indicates CFS state
  • In the past I cited Cholecalciferol loading dose guideline for vitamin D-deficient adults [2010], Vitamin D oral intermittent treatment (DO IT) study, a randomized clinical trial with individual loading regimen [2021] .. but there is an online calculator: NEW Loading Dose Vitamin D*Calculator!
    • My suggestion [2016] for target level is:  90-100 ng/mL (200-250 nmol/L). (of course, to be discussed with your medical professional)

The Key Problem with Suggestions is the picking of bacteria

Suggestions are based on several main factors:

• The bacteria you decide to alter (i.e. increase or decrease)
• The importance of each (sometimes called weight) when there is a trade off
• What substances has had any research. This is a nightmare – between contradictory results, small sample sizes, study done in the context of a specific diseases, etc. This is why I use fuzzy logic.

With the above stated, I walked thru this sample trying to first improve the bacteria selected (using my experience and statistical understanding), and then looking at the suggestions they generate.

This issue can be compounded with the depth of bacteria reported. “The disease is in the small details”. This is why more detailed and comprehensive (i.e. number of bacteria types reported) tests are a better starting point.

Contradictory Suggestions root issue

Facts in the database are based on what is specified in the study. A simple example: one study may use turmeric and a different study used curcumin. Curcumin(diferuloylmethane) is a main component of turmeric, but it also contains two other compounds demethoxycurcumin, and bisdemethoxycurcumin. In addition, volatile oils (tumerone, atlantone, and zingiberene) [Antiinflammatory Herbal Supplements, 2019]. The studies may result in a bacteria increasing in one and decreasing in the other. Both are right! It is the additional components that are significant. The worst case of “fuzziness” is with anything that has the word “diet”. Many people offering advice will deem them to be the same to simplify the facts that they need to remember; Dr. A.I. does not need to simplify — but that comes at a cost of confusion when things seem similar at a high level to the user. Another example: lactate, lactose, versus milk.

Analysis

Going to the My Profile / Health Analysis page, we see the two items that where he is at highest percentile (98%ile and 99%ile) are related and would agree with unrefreshing sleep.

• Sleep Apnea
• Insomnia

ADHD is high, but that seems common with ME/CFS. Dr. Jason Hawrelak Recommendations come in at the 89%ile. Going over to special studies, we see a lot of matches. The matches are not predictive — there are other factors (like DNA/SNP) before symptoms appear. They indicate simply increased risk.

Strategy to build Consensus

The two PubMed profiles above, Standard Lab Ranges (+/- 2 Std Dev),  Box Plot Whisker, and Kaltoft-Moltrup Normal Ranges gives 5 sets of suggestions. This will be my base set. I will also look at some special studies results (comparing those suggestions to

Simplified Suggestions

The majority of probiotics are to be avoided (not unusual for ME/CFS). The top suggestion was lactobacillus casei which is an easy one to get in Japan. The well studied one is sold as Yakult. The next ones are: lactobacillus gasseri, bifidobacterium breve. With most of the probiotics being negatives, you do not want to get them in probiotic mixtures.

In terms of vitamins, we see most of the B-vitamins are suggested (this is seen in one subset of ME/CFS patients, a different subset has it as an avoid). Vitamin D is a very mild avoid — but given the EBV issue above, I would ignore it and make that judgement call based on blood tests.

This also may apply to B-vitamins — none of the B vitamins are strong avoid, so a B-Complex is fine.

I have filtered the rest of the list to only the to take, coffee is sitting high in the list which appears to agree with “I dealt with the fatigue with regular consumption of coffee/tea throughout the day“. It helps in additional ways on shifting the microbiome.

Garlic		Take	4 gm/day	423.5
Hesperidin	Take	1.5 gm/day	418.1
NAC		Take	2400 mg/day	410.1
Melatonin	Take	10 mg/day	386.1
Luteolin		Take	400 mg/day	385.8
Coffee or non-herbal tea	Take	4+ Cups/day	313.4
Curcumin		Take	3 gm/day	239.2
Baicalin		Take		123
Glycine		Take	15 gm/day	119
Echinacea	Take	4 gm/day	109.5
Bilberry fruit extract	Take	40   g/day Vaccinium myrtillus powder	109.5
Piperine		Take		109.5
Tea Tree oil	Take		109.5
Fisetin		Take	20 mg/day	99
Polyphenols	Take	3 gm/day	95.3
Raspberry fruit (black raspberry)	Take	50 gm/day	61.5
Walnut		Take	75 gm/day	30.5
Olive leaf extract	Take	700 mg/day	24

Looking at the details of the Consensus Report

There are a few items not in the simplified list that are worth calling out:

• bacillus subtilis natto (probiotics) – is in the Japanese desert food called Natto. (it is a bit of an acquired taste [I have acquired that taste] and usually found only in some Asian markets in the US). Nattokinease is an extract from it.
• melatonin supplement – which goes with the sleep issues cited above.
• ascophyllum nodosum (sea weed) — various sea weeds are used in traditional Japanese cuisine.

Comparing Suggestions from PubMed Studies

• Insomnia — only 3 bacteria matched
• Sleep Apnea – only 3 bacteria matched
• ME/CFS without IBS – 7 bacteria matched – We get good agreement with the above

Comparing Suggestions from Special Studies

As a result of this email (and several others received this week), I looked at Special Studies suggestions for some specific people. My expectation when I did special studies was that the suggestions would converge tighter — so some people that is true. For other people, like this person is it false. For more information read the blog post: Caution: Special Studies Suggestions

I noticed that most of the results had inulin, etc – which are to avoid above and appears to make the person worse.

• Chronic Fatigue Syndrome (CFS/ME) – 52 bacteria some agreements and some disagreements
• Unrefreshed Sleep — 62 bacteria matched
• ME/CFS without IBS – 94 bacteria matched

I will be using this person sample to experiment with revised algorithms for Special Studies.

Probiotics

The top probiotic from KEGG data are E.Coli probiotics. Both Mutaflor and Symbioflor-2 (commercial E.Coli probiotics) are on the to take consensus list.

The next usually available item on the KEGG list is Bacillus subtilis, followed by Bacillus subtilis subsp. natto. As cited above, Natto is easily available in Japan.

Looking at Modelled Food Suggestions

Looking at the Flavonoids, we see just two things: Bananas and Barley. Rice is on the slight avoid (which may be a challenge for Japan).

We do not fair much better with Food Contents 🙁 A Guide to the Different Types of Tofu in Japanese Cuisine

This was an unexpected frustration. I wait back to the algorithm and made it less conservative. By this I mean, not eliminating items where there are contradictory results from studies. This resulted in in more suggestions, some Japan specific [there are 50 items acquired in Japan on the list]

I was going to leave the algorithm with the more relaxed condition. I then did cross-validation with the consensus and found that most of the items appearing as take using the relaxed filter were on the avoid list of the consensus. Tofu, which showed up with the strict criteria, is on the consensus to take list (as in Soy).

Why do we have any disagreements? The root problem is not sufficient studies – often with contradictory results — one may be on people with diabetes and another people with asthma. Existing conditions (and in some cases severity) can result in difference response. The microbiome is not a machine but a complex society that interacts with a lot of things

The Hail Mary Exploration

I redid the consensus and included prescription items. For many people, the pharmaceuticals are way down the to take list. Herbs and probiotic having a higher positive weight often. For this person, this was not the case. What was also unusual was none of the antibiotics that are recommended in other ME/CFS analysis (and which are used for ME/CFS – i.e. cross validation) are anywhere near the top of the list. What I found listed near the top of the pharmaceuticals were:

• atorvastatin (prescription) – treats  treat high cholesterol and triglyceride levels.
• clemizole hydrochloride,(prescription) – used for treating Hepatitis
• sertaconazole nitrate,(prescription) – an antifungal

My immediate, due diligence, suggestion is to get testing for cholesterol, hepatitis and fungal infections. Was the “EBV/Mono-type illness” perhaps hepatitis? Two of these can result in chronic fatigue. In terms of special studies, ME/CFS (three variations) were poor matches (one was the bottom of the list). Using US National Library of Medicine studies, he is at the 29%ile.

What I find interesting is that soy is a recommendation, and we find these studies..

• Effect of soy protein supplementation in patients with chronic hepatitis C: a randomized clinical trial [2012] – positive effect
• Inhibitory effect of Japanese rice-koji miso extracts on hepatitis A virus replication … 2018

In terms of having the typical microbiome for someone that has a ME/CFS diagnosis — he does not match. He is also at the 0%ile for hypercholesterolemia (High Cholesterol), 8%ile for GERDs (usually common for a subset of ME/CFS).

My gut feeling is that he was not sufficiently tested before the ME/CFS label was slapped on him. His microbiome is not a match. Nothing match his symptoms to the microbiome associated conditions that I have data on.

FEEDBACK FROM READER AFTER READING

Those are very interesting points. Since my initial symptoms in Korea, I was diagnosed with Gilbert’s syndrome, an excess of Bilburin. I did get my liver checked (due to slight yellowing of the eyes) out but they don’t seem to have found anything (Over a decade ago so my memory is pretty spotty.) I think over the years in college I got MRI’s and various bloodtests with obviously no real diagnosis. Even so, my life was very active over that decade. I lived in China for a few years, frequently attended the gym, also did a fair share of partying, have always had a big appetite. Minimal impairment throughout my life overall.

Since getting sick a few years ago in Japan, I got a lot of different tests, including thyroid and diabeties. If I rifle through all the different result sheets (I have quite a few), a few things of note:

Dec. 2020 I was positive for EBVVCA-IgG which my doctor told me meant that I had had EBV in fairly recent past. No treatment was suggested.

Feb.. 2020 I had slightly elevated liver enzymes (ALT, GTP) (Not surprising given my frequent social drinking in my 20s) which persisted until April 2021 and seemed to have resolved. This was explained by lowered alcohol consumption. Never received a suggestion for follow-up or differential diagnosis.  

Persistently higher tryclycerides (I have gained some weight since my illness, matches with the atorvastatin suggestion!) and CRP indicating inflammation. Suggested to lose weight and exercise (The latter being somewhat problematic for me.) Back in my weight-lifting days, I drank lots of milk and overall had a high-fat/protein diet. Since I recently stopped drinking so much whole milk, I wonder if that will resolve it or not.

Before I moved recently I had been going to a cfs clinic which conducted sleep studies (of which I have had several over the years, indicating that I do not have sleep apnea, although that was my thought for years.) and did some other tests indicating I had high stress/inflammation. He gave me vitamins, CoQ10 (which I ordered myself) and anti-depressants for sleep. (Interesting I score high for Depression (47%) but up until the last few years I had a very active lifestyle, I don’t think depression fits my physical symptoms at all. I’ve always been very social and generally on the adventurous side. Still am, just don’t have the physical stamina I used to. I requested a prescription for Piracetam (all of the -tams were recently regulated in Japan) and he refused.

I haven’t gone back there since as I felt like the approach he had was very symptoms based and surface level. Reading your recovery stories, I pretty much ruled out getting any kind of prescription anti-biotics or being able to utilize those out here. I don’t know how the legal system works here in Japan, but my own and other anecdotal experience tells me that doctors here don’t enjoy doing things off-label, it’s an extremely by-the-book conservative approach to medicine. Fortunately, if you have a specific, known medical problem, you can get extremely professional and competent care here.

It is very interesting that I do not fit in the ME/CFS microbiome profile. An infection with Hepatitis or a fungus at some point could fit, especially given that I’ve spent time in a variety of different environments. I think I could easily get tested for Hepatitis here, it will just have to wait until next month (Currently very busy as a **** .) As far as fungus goes, usually I hear about “Candida,” any suggestions of what kind of fungus I should perhaps be looking out for would be helpful.

I will have to think about how I will go about getting tested for fungal infections here, it is somewhat difficult to explain my issues to doctors and get them to take it seriously. Definitely a good idea to also pursue a potential alternative diagnosis though, the data certainly doesn’t rule it out.

You have given me a lot to think about, thanks again for all of your consideration!

Postscript – and Reminder

I am not a licensed medical professional and there are strict laws where I live about “appearing to practice medicine”.  I am safe when it is “academic models” and I keep to the language of science, especially statistics. I am not safe when the explanations have possible overtones of advising a patient instead of presenting data to be evaluated by a medical professional before implementing.

I cannot tell people what they should take or not take. I can inform people items that appears to have better odds of improving their microbiome as a results on numeric calculations. I am a trained experienced statistician with appropriate degrees and professional memberships. All suggestions should be reviewed by your medical professional before starting.

I use modelling and various mathematical technique to estimate forecasts when there is no hard data available.

I have been in dialog with a microbiome testing firm on incorporating my suggestions engine into their site. One of their requests was to provide more information on foods. Of course, if there are no studies than how can you make suggestions?

It is possible to model likely impact of foods by looking at what is in them. That information is available from:

• United States Department of Agriculture – Agricultural Research Service
• EU’s Phenol-Explorer

So the process using existing pubmed studies to identify the constituents of the food that impacts certain bacteria and then aggregate these constituents to get a modelled benefit from the food.

The results are two tables (depending on which decomposition approach is used).

This is on the Changes Tab

Appears to Cross Validate

I did some of the other suggestions methods, built a consensus report and then looked for Almonds on the same sample and found agreement. Same with Barley, Basil, and Buckwheat. Some suggestions from the second list have disagreements (not unexpected).

I would suggest that that you see how it works for your samples between these two approaches.

For your amusement, the second list is LONG (typically 800+ items) and for those that drink wine, may be amusing to scan..

Other Labs

This has been added to the define suggestions page. Both buttons open the suggestions in new windows so it is easy to switch back and forth.

If you have already uploaded a sample, a link has been added there.

Special Studies are a conceptual thought experiment. The logic is simple, identify the bacteria that are have major statistical significance from the reference. Then use these bacteria with the weight that each has being the z-score to generate suggestions.

The first reviews that I did using them had good results and agreement with my preferred trio to build consensus, namely:

• Standard Lab Ranges (+/- 2 Std Dev)
• Box Plot Whisker
• Kaltoft-Moltrup Normal Ranges

This last week I have gotten several emails from people who got counter-indicated suggestions. I have verified that for their samples, it produces contrary suggestions.

Digging into the mathematics and fuzzy logic being used, I see several possible failure points that I want to slowly investigate. The top failure points are:

• Using the z-score for the weight to give for each desired shifts. A different formula may resolve it. Two candidate formula are:
  • z-score * incidence of bacteria being seen
  • z-score * function(bacteria count) — with many possible functions
  • The z-score cut off is too low, I am using 5.0 at present, it may need to be raised to a higher value.
• The criteria for picking a bacteria to include may not be specific enough, so a lot of bacteria that are fine are included. This can result in excessive noise in the suggestions
• The data available for suggestions at the species levels that we are working are insufficient (and in some cases, may not exist). A lot of the species flagged are rarely seen in studies showing changes.

Bottom Line

Use the suggestions generated with great caution. If they compliment the suggestions from the three preferred consensus methods listed above — good. If they contradict, keep to the original consensus method — I have been getting consistent report that they work. The special studies suggestions are getting inconsistent results.

A special study with a z-score below 6.6 is very suspect and should be ignored.

A reader contacted me over a new post on Biomesight – How to reduce brain fog. He was concerned over the content (knowing that I have often researched and posted on brain fog), so I am doing this post to provide some clarity on brain fog. (Bad pun: Remove some fog from brain fog)

What is Brain Fog?

• The term brain fog is a vague term that has been defined in the literature as a combination of the following more accurate (and measurable by tests) conditions. A better term is executive dysfunction [2015] or Cognitive Fatigue [2014]. The literature goes back to at least 1989. I know from personal experience, I have taken them from professional psychologist, and other in the family has too.
• Some people will perform badly on all tests, other will perform poorly on certain tests only.

Related forms include

• Proverb comprehension impairments [2009]

IMHO, if you do not have the majority of the above, the term may be misapplied.

For different diseases, what constitues brain fog can vary, for example:

The clinical picture typically affects visuo-spatial immediate memory (g = − 0.55, p = 0.007), reading speed (g = − 0.82, p = 0.0001) and graphics gesture (g = − 0.59, p = 0.0001). Analysis also revealed difficulties in several processes inherent in episodic verbal memory (storage, retrieval, recognition) and visual memory (recovery) and a low efficiency in attentional abilities.

Systematic review and meta-analysis of cognitive impairment in myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS) [2022]

Common Conditions Having it

The list of conditions having issues somewhere in the executive function space is large, just a few are listed below.

• Disentangling the cognitive, physical, and mental health sequelae of COVID-19 [2022]
• Systematic review and meta-analysis of cognitive impairment in myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS). [2022]
• Executive function in fibromyalgia: Comparing subjective and objective measures [2016]
  • Fibrofog in Daily Life: An Examination of Ambulatory Subjective and Objective Cognitive Function in Fibromyalgia.[2020]
• Evidence for executive dysfunction in autism. [1994]
• An investigation of frontal executive dysfunction in attention deficit disorder subgroups. [1993]
• Executive function abilities in autism and Tourette syndrome: an information processing approach. [1994]
• Cognitive executive dysfunction in patients with obstructive sleep apnea syndrome (OSAS) after CPAP treatment. [1998]
• Executive Dysfunction in Patients With Alcohol Use Disorder: A Systematic Review. [2022]
• Pain and depression are associated with both physical and mental fatigue independently of comorbidities and medications in primary Sjögren‘s syndrome. [2019]
• Physical and mental fatigue in people with non-communicable chronic diseases. [2022]

Many of the above have distinct microbiome signatures and thus the hope of getting a universal microbiome signature for brain fog is an ideological belief. This appears to be confirmed in the analysis from Special Studies … Brain fog strongest z-score is just 5.2. This is lowest significance level of 26 items evaluated, the next lowest is General Fatigue. IMHO, there may be no true significance, the z-scores numbers were not adjusted for False discovery rate and incidence of reporting.

Most people will agree that there is no magic cause or microbiome signature for general fatigue — it could be an issue with iron levels, excessive lactic acid (impairment in clearing it), blood circulation issues, respiratory issues etc.

Brain fog could be described as mental fatigue and thus the same wide variety of issues can be involved. For ME/CFS, the dominant causes for brain fog, according to the literature are mentioned in some of my prior posts:

• Hypoperfusion of the brain in ME/CFS and Long Covid
• Approaches to D-Lactic Acidosis

For Long COVID, we should also consider damage to the lungs impacting oxygen levels.

Concerns about Biomesight – How to reduce brain fog

My first concern is simple, the belief of there being a common microbiome pattern is very questionable. There are likely patterns, for example, a microbiome pattern that results in higher d-lactic acid production; a pattern that results in lower d-lactic acid production; a pattern that inhibits one of the many steps in the coagulation cascade; a pattern that overloads one of the many steps in the coagulation cascade; a pattern that causes vascular constriction; a pattern that cause inflammation; a pattern that inhibits the absorption of iron…and on and on. There is not a single pattern that applies to all.

My second concern is a failure to cross validate/document. D-lactic acid is well associated with brain fog in the literature, ” D–lactic acidosis is characterized by brain fogginess (BF) and elevated D-lactate“[2018], Recent research indicates that d–lactic acid may inhibit the supply of energy from astrocytes to neurons involved with memory formation.  [2010]. It is not mentioned once in the post (as of this time).

When a statement like this is made “Unsurprisingly, high ethanol producers in the gut based on research findings (separate from what we are seeing from the Biomesight dataset) is associated with brain fog.” I would expect links to these research papers to be included for the reader to follow up. Ethanol is drinking alcohol, booze – which has me very curious about the links and especially if they are seen with many of the conditions cited above.

I am very curious because I use KEGG data to see if any compound production/consumption was statistically significant in Special Study: Neurocognitive: Brain Fog, and ethanol was not found to be statistically significant. Searching for ethanol on Pub Med, we find “Lactic acidosis and acute ethanol intoxication [1994]” and “SEVERE LACTIC ACIDOSIS SECONDARY TO ACUTE ALCOHOL INTOXICATION” [2021] but that was from explicit ethanol (alcohol) consumption.

This does lead me to a model for alcohol intolerance developing with ME/CFS, which I posted here [Alcohol Intolerance in ME/CFS – A Model].

How does special studies compared to Biomesight post

The table below shows no agreement between my special studies and their findings. We used different statistical process, but finding not a single agreement should be a red flag on relying on the data. While I have a smaller sample (approximately 1/3), the data processing to get the microbiome data was identical.

Bacteria	BiomeSight Post	Special Studies Z-Score
Escherichia coli (species)		5.3
Lactiplantibacillus pentosus (species)		5.1
Shuttleworthia (genus)		5.1
Escherichia (genus)		4.5
Veillonella (genus)		4.4
Veillonella dispar (species)		4.4
Staphylococcus pseudolugdunensis (species)		4.2
Clostridium cellulovorans (species)		4.1
Class Deltaproteobacteria	X
Species Bacteroides uniformis	X
Species Bacteroides cellulosilyticus	X
Species Phascolarctobacterium faecium	X
Genus Bacteroides	X
Species Anaerotruncus colihominis	X
Species Faecalibacterium prausnitzii	X
Genus Prevotella	X

Bottom Line

I do not believe that we can aggregate all microbiome samples reporting brain fog into a single set and find a universal pattern to address a priori. The numbers from Special Study: Neurocognitive: Brain Fog were the weakest of all special studies and, based on some other recent work in progressed, results may be adversely affected by sampling bias, sample quality, and false detection rate.

Back Story For Mother

“Lots of high % potential health stuff. In real time, mother’s very hormone sensitive . Gets like a Premenstrual dysphoric disorder (PMDD) state , feels depressed. Struggles with anxiety and depression in general . Notice she’s been getting allergies. During the birth of 2nd child two years ago; mother had to get vancomycin because she was strep B positive. When the family get sick, the mother gets the sickest of the 4 of us unfortunately. Mother caught COVID in Sep 2022. Mother’s cycles in 2021 started to be heavy where before that never happened. As a kid, mother was on Accutane a lot. When mothers results through biomesight I also got a message that said unusually small file size or something along those lines.”

Foreword – and Reminder

I am not a licensed medical professional and there are strict laws where I live about “appearing to practice medicine”.  I am safe when it is “academic models” and I keep to the language of science, especially statistics. I am not safe when the explanations have possible overtones of advising a patient instead of presenting data to be evaluated by a medical professional before implementing.

I cannot tell people what they should take or not take. I can inform people items that appears to have better odds of improving their microbiome as a results on numeric calculations. I am a trained experienced statistician with appropriate degrees and professional memberships. All suggestions should be reviewed by your medical professional before starting.

Mother Overview

• Jason Hawrelak Recommendations – 99.7%ile (better than most), High Roseburia, Blautia. The others are low

Checking on the two studies that I could find Premenstrual dysphoric disorder

• Characteristics of the Gut Microbiota in Japanese Patients with Premenstrual Syndrome [2022]
• Characteristics of the gut microbiota in women with premenstrual symptoms: A cross-sectional study [2022]

If you can locate any other studies, please send to me.

Tax Name	Shift	Percentile
Bacteroidia (NCBI:200643 )	Low	4%ile
Anaerotaenia (NCBI:1843206 )	High	not reported by lab
Bifidobacterium (NCBI:1678 )	High	80%ile
Blautia (NCBI:572511 )	High	98%ile
Butyricicoccus (NCBI:580596 )	Low	not reported by lab
Collinsella (NCBI:102106 )	High	0%ile
Extibacter (NCBI:1918452 )	Low	not reported by lab
Megasphaera (NCBI:906 )	Low	92%ile
Parabacteroides (NCBI:375288 )	Low	67%ile
Bacteroidetes (NCBI:976 )	Low	4%ile

This suggests that we have a general match with the literature and should include in our consensus, a hand picked set of suggestions.

We have an impressive list of bacteria deemed unhealthy. Note that counts are often small (agreeing with the abundance of bacteria at low levels seen above).

Name	Rank	Percentile	Count	Comment	More Info
Actinomyces	genus	87	230	Pathogen	Citation
Anaerotruncus colihominis	species	78	2380	Not Healthy Predictor	Citation
Bacteroides fragilis	species	74	5659	Pathogen	Citation
Blautia producta	species	75	1440	Not Healthy Predictor	Citation
Clostridium	genus	72	21469	Pathogen	Citation
Collinsella	genus	0	10	High COVID Risk	Citation
Corynebacterium	genus	89	1520	Pathogen	Citation
Dorea	genus	97	23040	Increased COVID risk	Citation
Eggerthella lenta	species	99	3469	Not Healthy Predictor	Citation
Finegoldia magna	species	91	830	Infectious bacteria	Citation
Granulicatella adiacens	species	84	80	Not Healthy Predictor	Citation
Leptospira	genus	92	160	Pathogen	Citation
Peptoniphilus harei	species	67	50	Infectious bacteria	Citation
Peptoniphilus lacrimalis	species	78	170	Infectious bacteria	Citation
Peptostreptococcus	genus	83	170	Pathogen	Citation
Peptostreptococcus stomatis	species	88	150	Not Healthy Predictor	Citation
Prevotella timonensis	species	95	1890	Infectious bacteria	Citation
Ruminococcus gnavus	species	99	62200	Not Healthy Predictor	Citation
Staphylococcus haemolyticus	species	72	40	Pathogen	Citation
Streptococcus vestibularis	species	65	370	Not Healthy Predictor	Citation
Veillonella atypica	species	94	859	Not Healthy Predictor	Citation

Potential Medical Conditions Detected had at 99%ile, Schizophrenia, which suggests some overlap with PMDD.

Going Forward

I added a new choice to the site that gets canned suggestions for Bacteria Deemed Unhealthy above because of the large number of bacteria flagged in this sample.

The suggestions for this person with this new option.

Building the Consensus

I am building from:

• Hand Picked (see above focused on PMDD) (4 picked)
• Unhealthy Suppression – just added (33 picked)
• Outside Range from Nirvana/CosmosId (5 picked)
• Outside Lab Range (+/- 1.96SD) (6 picked)
• Outside Box-Plot-Whiskers (60 picked)
• Outside Kaltoft-Moldrup (95 picked)

The top items to suggested to take are shown below (meat and B-vitamins are concentric):

• carboxymethyl cellulose (prebiotic)
• low carbohydrate diet
• melatonin supplement
• glycyrrhizic acid (licorice)
• l-glutamine
• quercetin,resveratrol
• thiamine hydrochloride (vitamin B1)
• gluten-free diet
• vegetable/fruit juice-based diets
• lard
• vitamin b3 (niacin)
• high animal protein diet
• high red meat
• Arbutin (polyphenol)
• diosmin,(polyphenol)
• pyridoxine hydrochloride (vitamin B6)

The top items to suggested to avoid are shown below

• walnuts
• lactobacillus rhamnosus gg (probiotics)
• inulin (prebiotic)
• Burdock Root
• arabinoxylan oligosaccharides (prebiotic)
• wheat
• saccharomyces boulardii (probiotics)
• brown rice
• lactobacillus reuteri (probiotics)
• fruit/legume fibre
• raw potato starch
• lactobacillus salivarius (probiotics)
• dietary fiber
• sodium butyrate

Probiotics

We have a challenge here, take lactobacillus rhamnosus gg (probiotics) by itself is a -409, but in combination with propionibacterium freudenreichii it is 190-263. This suggests that bacteria as a supplement ( Nutricology/Securil ) or lots of real Emmental cheese (which uses it).

Similarly lactobacillus salivarius (probiotics) is a negative -275 but Lactobacillus salivarius UCC118 is a positive 285. I would avoid it, instead of playing probiotic roulette.

The next one on the list as a reasonable candidate is lactobacillus fermentum (probiotics), then mutaflor escherichia coli nissle 1917 (probiotics).

Going over to KEGG suggestions, the top choices (not in Prescript Assist which is a negative) are (in order)

• Escherichia coli – top of the list
• Bacillus licheniformis
• Bacillus subtilis
• Weizmannia coagulans

Which points to microbiome labs/ megasporebiotic OR organic 3 / primal soil (see probiotic page for other choices)

My suggestion for rotation would be:

• Symbioflor-2 (since Mutaflor is not available for this person in the USA)
• Nutricology/Securil
• A probiotic containing Bacillus subtilis etc

“for curiosity’s sake, is there any way i can find the small number of actionable bacteria affected by all these foods?”

taxon	Tax Name	Tax Rank	Modifier that Impacts
186802	Eubacteriales	order	1421
816	Bacteroides	genus	1398
815	Bacteroidaceae	family	1396
543	Enterobacteriaceae	family	1384
28116	Bacteroides ovatus	species	1382
29523	Bacteroides sp.	species	1381
626931	Bacteroides oleiciplenus	species	1381
357276	Phocaeicola dorei	species	1381
817	Bacteroides fragilis	species	1381
820	Bacteroides uniformis	species	1381
387090	Phocaeicola coprophilus	species	1380
674529	Bacteroides faecis	species	1380
1365140	Bacteroides sp. J1511	species	1380
246791	Bacteroides sp. 35BE35	species	1380
371599	Bacteroides sp. XB12B	species	1380
1244208	Bacteroides sp. 2011_Ileo_VSA_D11	species	1380
151276	Bacteroides coprosuis	species	1380
93975	Bacteroides sp. AR29	species	1380
338188	Bacteroides finegoldii	species	1380

Modifier	Modifier type	Bacteria Impacted
berberine	Herb or Spice	2101
glycyrrhizic acid (licorice)	Herb or Spice	2067
Human milk oligosaccharides (prebiotic, Holigos, Stachyose)	Prebiotics and similar	2054
triphala	Herb or Spice	2008
gentamicin (antibiotic)s	Antibiotics, Antivirals etc	1961
lactobacillus plantarum (probiotics)	Probiotics	1856
azithromycin,(antibiotic)s	Antibiotics, Antivirals etc	1829
trimethoprim (antibiotic)s	Antibiotics, Antivirals etc	1810
lactobacillus reuteri (probiotics)	Probiotics	1795
iron	Vitamins, Minerals and similar	1738
Slippery Elm	Herb or Spice	1718
neomycin (antibiotic)s	Antibiotics, Antivirals etc	1717
zinc	Vitamins, Minerals and similar	1686
foeniculum vulgare (Fennel)	Herb or Spice	1686
folic acid,(supplement Vitamin B9)	Vitamins, Minerals and similar	1666
amoxicillin (antibiotic)s	Antibiotics, Antivirals etc	1661
resistant starch	Prebiotics and similar	1650
N-Acetyl Cysteine (NAC),	flavonoids, polyphenols etc	1624
streptomycin (antibiotic)s	Antibiotics, Antivirals etc	1620
cinnamon (oil. spice)	Herb or Spice	1586

I attended this series of talks and for the readers of this blog I link to the videos in terms of my preferences below. The first ones are very technical, Day Three is much easier watching.
Unfortunately, they did not post the videos, talk by talks..

• Dr. Katherine Amato on Day 3 (hour 3) is likely the best starting one.
• Dr. Poyet Day 3 (hour 1) on Industrialization Microbiome
• Dr. Iraola Day 3 (hour 0) – on the Microbiomes of South America etc
• Day 3 hour 2 — very interesting on the impact on the microbiome of anal intercourse…. there is significant impact.

This is a summary of items that may be the cause. This assumes that a lack of oxygen getting to tissue (or the brain) is the cause. Note that if the brain is tired(oxygen starved), that there may be false signals saying the body is exhausted because the brain does not have the energy to manage the body. The sources are typically from Chronic Fatigue Syndrome. Some of these may be related to the microbiome (lack of studies). It is suspected that many people are borderline (have no apparent issues) and then some event pushes them over the edge into fatigue.

This is intended as a checklist to review with your medical professional. Items should be objectively excluded by actual tests and not subjectively excluded (often a response if the professional does not know how to test). See Hypoxia[2022] and Hypoxia (medical)

Coagulation Issues

The typical scenario seems to be an inherited or acquired (epigenetic — sometimes from a virus infection) coagulation defect. Some step of the coagulation cascade gets “constipated” with the result being what is often called “sticky blood”. It only takes one step.

• Health Rising Reposts: Microbiome and Coagulation

If the blood gets thick, it gets slow and hence less oxygen gets to the body and the brain.

Most “blood thinners” only impacts one of these steps (with heparin being the main exception). Taking excessive supplements that impacts the same step may result in long bleeding time or easy bruising.

Note: We are talking about sub-clinical (i.e. not having a stroke or blood clot) levels. The lab levels may not be abnormal, just low (or high).

Low Iron Or Impaired Heme

Hemoglobin is the iron-containing oxygen-transportmetalloprotein in the red blood cells of all vertebrates  Hemoglobin in the blood carries oxygen from the respiratory organs (lungs or gills) to the rest of the body (i.e. the tissues). There it releases the oxygen to permit aerobic respiration to provide energy to power the functions of the organism in the process called metabolism. [wikipedia]. Hemogloblin can be inhibited by some chemicals (Carbon Monoxide [2022] and the pH of the blood (Bohr effect [2021]) are the most common. pH is influenced by the microbiome.

Iron insufficiency and hypovitaminosis D in adolescents with chronic fatigue and orthostatic intolerance. Restless leg syndrome (RLS) is  negatively correlated with ferritin / iron [2012] and RLS is also co-morbid with CFS (i.e. the less iron, the more likely you will have RLS).

For example, Staphylococcus aureus is known to adversely impact iron [2022]

Malformed Red Blood Cells

Back in 1997, Dr. Les Simpson in New Zealand wrote “Myalgic Encephalomyelitis (ME): A Haemorheological Disorder Manifested as Impaired Capillary Blood Flow” observed this in CFS patients. “So it has been proposed – first at the Cambridge Symposium on ME in 1990 and in another article in 1998, that ME is a dysfunctional state resulting from reduced rates of capillary blood flow due to the presence of shape-changed, poorly-deformable red cells” [source]

It should be noted that SPECT scans show the expected effects of shape-
changed, poorly deformable red cells in reducing cerebral blood flow in
regions which by chance have smaller than usual capillaries.

TheBMJ, Chronic fatigue syndrome or myalgic encephalomyelitis [2007]

Inflammatory hypoxia

Hypoxia and inflammation are frequently co-incidental microenvironmental features of sites of concentrated physiological or pathological immune activity.

Hypoxia activates hypoxia-inducible factor, which is a major regulator of multiple aspects of immune cell function. Consequently, hypoxia plays a key role in the regulation of immunity and inflammation.

The impact of hypoxia on immunity and inflammation is site-specific and cell type-specific.

Pharmacological hydroxylase inhibition, which activates hypoxia-sensitive pathways, is profoundly protective in multiple models of inflammation.

Regulation of immunity and inflammation by hypoxia in immunological niches [2017]

Low Blood Volume

David Bell, M.D. did a study in 1995 (twenty years ago) finding about low blood volume. Cort Johnson has an 2015 update by Dr. Bell that is worth reading. “In contrast, ME patients have a volume that can be as low as 50% of normal.”

Small Heart Size

“A considerable number of CFS patients have a small heart. Small heart syndrome may contribute to the development of CFS as a constitutional factor predisposing to fatigue, and may be included in the genesis of CFS.” [2008]

” Echocardiographic examination revealed that CFS patients with “small heart” had an actually small LV chamber and poor cardiac performance. Cardiac functional changes evaluated by repeated examinations appeared to be directly associated with the severity of their symptoms. Small heart syndrome with impaired cardiac function may contribute to the development of CFS through low cardiac output as a constitutional factor.” [2009]

Enzymes and thus Microbiome

“We’ve discovered that the muscles regulate oxygen consumption in a very precise way using the oxygen-sensitive enzyme FIH,”

How muscles regulate their oxygen consumption [2018]

” cytochrome P450 (CYP), monoamine oxygenase (MAO), and cyclooxygenase (COX). CYP enzymes are central players.. are important for therapeutic intervention and treatment of neurological and inflammatory diseases”

Analysis of activity and inhibition of oxygen-dependent enzymes by optical respirometry on the LightCycler system [2010]

Obstructive Sleep Apnea

Note that blood platelets activation is part of the coagulation cascade above.

Snoring, collapse of upper airways and intermittent hypoxia are main causes of smoldering systemic inflammation in patients suffering from obstructive sleep apnea. The systematic inflammation is considered one of the key mechanisms leading to significant cardiovascular complications. Blood platelets, formerly not even recognized as cells, are currently gaining attention as crucial players in the immune continuum. Platelet surface is endowed with receptors characteristic for cells classically belonging to the immune system, which enables them to recognize pathogens, immune complexes, and interact in a homo- and heterotypic aggregates.

Obstructive Sleep Apnea: From Intermittent Hypoxia to Cardiovascular Complications via Blood Platelets [2018]

For the last few months I have been working with someone that runs a Long COVID support group. This has resulted in more modifiers being added. One of the outcomes has just been added to the site. She requested that a simplified set of suggestions be added to the site to make her life easier. This consists of items she picked from her experience dealing with the group.

These simplified suggestions has been added everywhere (tell me if I missed a page). If you use GI Map or similar reports, it is available after you have transferred and return to adjust suggestions. You should see these options on most suggestion pages now. To get this new condensed report, just click the checkbox.

GI Map and similar

The report is a CDV file (loads into Excel or other spreadsheet programs)

After loading, into Excel, you will need to adjust column widths

If an item is a Take, then a clinical dosage is given (if known). Clinical Dosages are those reported from studies listed on the U.S. National Library of Medicine Clinical Trials site.

Other sections are shown below.

CAUTIONS

At the bottom of the page are some essential reminders.

Some of the clinical dosages above may be problematic with some medical conditions. Dosages should be reviewed by a medical professional before starting. The suggestions are based on a mathematical model. dosages are from clinical studies.

One of the purposes of these blog posts is to learn. My thinking and thoughts are there with the ability of people to correct, to comment and to learn. This post looks at GI-MAP and Biomesight report for the same person, taken at the same time. The key objectives are:

• Comparing appropriate results between the two (same bacteria only)
• Comparing the suggestions produced from each
  • GI-MAP using US National of Medicine Library studies
  • BiomeSight using our Special Studies

Foreword – and Reminder

I am not a licensed medical professional and there are strict laws where I live about “appearing to practice medicine”.  I am safe when it is “academic models” and I keep to the language of science, especially statistics. I am not safe when the explanations have possible overtones of advising a patient instead of presenting data to be evaluated by a medical professional before implementing.

I cannot tell people what they should take or not take. I can inform people items that appears to have better odds of improving their microbiome as a results on numeric calculations. I am a trained experienced statistician with appropriate degrees and professional memberships. All suggestions should be reviewed by your medical professional before starting.

Simple Back Story

” I got COVID at the end of Sept 2020 and have been much worse ever since.”

We have 2 samples from Biomesight April,2022 and Sept 2022 and will see if the microbiome shifts reflect getting worse

GI-MAP vs BiomeSight

We are comparing different lab results, so reading: The taxonomy nightmare before Christmas… is strongly recommended. While results may disagree, they may be both technically accurate given the testing methodologies.

Bacteria	GI MAP	BiomeSight Percentile
Enterobacter spp. ( Enterobacter)	High	78%ile
Bacteroidetes ( Bacteroidetes)	High	81%ile
Firmicutes ( Firmicutes)	High	8%ile
Bacillus spp. (Bacillus)	High	47%ile (seen in 66%)
Enterococcus faecalis	High	Not reported (Seen in 1%)
Enterococcus faecium	High	Not reported (seen in 2.2%)
Methanobacteriaceae (family)	High	Not reported (seen in 27%)

As is seen clearly, there are differences — some very striking (Firmicutes). I will leave it to the test providers to offer explanations on differences of techniques, etc. See Below One of the key issues is how labs handle not detected for determining ranges: some will exclude it from the calculation (thus range IF detected) and others will give it a zero value and include it.

I was curious about Bacillus distribution, which is shown below

Explanation of differences from GI-MAP

COMMENT FROM DIAGNOSTIC SOLUTIONS LAB

Thank you for contacting us and providing an opportunity for us to comment.  Metagenomic sequencing has been popular for untargeted characterization of microbiome composition, whereas qPCR is used primarily for accurate quantitation of selected targets (microbes and genes of interest) especially those with particular clinical implications, such as common pathogens and opportunists. Metagenomic sequencing and qPCR are not considered competing methodologies for the same purpose, so it’s not an apples-to-apples comparison. 

A major difference between the two methods is the type of quantitation that they provide. Metagenomic sequencing results are expressed as relative abundance (usually as a percentage), whereas qPCR results are expressed as absolute abundance (usually microbial cells per gram of stool). With relative abundance, it’s not possible to directly determine how much of a given taxon (species, genus, etc) is actually present, since its relative abundance is dependent upon the relative abundance of all of the other detected taxa. A couple of example references are included below. 

The differences between relative and absolute abundance are likely key factors contributing to any differences in results between GI-MAP and metagenomic tests. From a clinical standpoint, there are demonstrated advantages of using methods, such as qPCR, that provide absolute quantitation, since absolute quantitation has been shown in research studies to be more effective in identifying true correlations with quantitative clinical markers. Examples of clinical markers that are included on GI-MAP include calprotectin, secretory IgA, pancreatic elastase and occult blood.

A quantitative sequencing framework for absolute abundance measurements of mucosal and lumenal microbial communities

https://pubmed.ncbi.nlm.nih.gov/32444602/

“Thus, an inherent limitation of methods that use relative abundance is that they cannot determine whether an individual taxon is more abundant or less abundant (the direction of the change) or by how much (the magnitude of the change) between two experimental conditions or samples.”

Benchmarking microbiome transformations favors experimental quantitative approaches to address compositionality and sampling depth biases

https://pubmed.ncbi.nlm.nih.gov/34117246/

“Our results demonstrated that quantitative approaches performed better than their relative and compositional counterparts when it comes to identifying taxon–metadata associations or studying taxon–taxon interactions. The observed performance gap among the methods profiled widened with increasing unevenness of taxa distributions, as exemplified by our analyses of the blooming scenario.”

Explanation from BiomeSight

As many of your readers know, it is not possible to compare Biomesight and GI MAP directly due to the differences in what and how it is measured. The strength of a 16s test is in the ability to see the overall ecosystem and this technology is widely used today in clinical and academic research, often alongside shotgun or other WGS techniques. It allows for coverage of a range of bacteria that’s not included in more focused PCR tests.

Additionally, Biomesight already has around 400 samples from the long covid community allowing for better profiling of the condition. We have shared around half of these samples’ results with MicrobiomePrescription enabling a strong comparison base.

Here’s some of our blog articles covering our findings from the study:

https://biomesight.com/blog/long-covid-study-update-1

https://biomesight.com/blog/the-role-of-lps-in-long-covid

Suggestions from GIMAP

Without applying PUBMED filtering

All of the above bacteria were used.

With PubMed Filtering

Only the following bacteria was selected. This filtering is not available on the site because the quality of select drops too low (as seen here).

Bacteria Name	Analysis	Taxonomy Hierarchy Level
Bacteroidetes	Too High	phylum
Firmicutes	Too High	phylum

Suggestions from BiomeSight

Without Filtering – using Kaltoft-Moldrup for bacteria selection

We have 70 bacteria selected

Using Special Studies

We have 93 bacteria selected.

Getting Worst with Long COVID

Taking ME/CFS as a typical pattern, there tend to be three main paths:

• Slow or Spontaneous Remission
• Steady state with waxing and waning
• Slow progressive deterioration, sometimes ending in complete system failue

There are two others, suicide from hopelessness and Remission cause by specific treatment (which was my personal case).

The reader had a early-COVID sample which allows comparisons. It is especially nice that the Lab Read Quality are similar, which makes interpretation more robust.

Criteria	Current Sample	Old Sample
Lab Read Quality	10.3	10
Bacteria Reported By Lab	639	507
Bacteria Over 99%ile	6	3
Bacteria Over 95%ile	26	13
Bacteria Over 90%ile	40	26
Bacteria Under 10%ile	198	284
Bacteria Under 5%ile	173	229
Bacteria Under 1%ile	155	161
Lab: BiomeSight
Rarely Seen 1%	3	7
Rarely Seen 5%	16	25
Pathogens	34	36
Outside Range from JasonH	10	10
Outside Range from Medivere	19	19
Outside Range from Metagenomics	9	9
Outside Range from MyBioma	7	7
Outside Range from Nirvana/CosmosId	19	19
Outside Range from XenoGene	6	6
Outside Lab Range (+/- 1.96SD)	14	8
Outside Box-Plot-Whiskers	67	42
Outside Kaltoft-Moldrup	181	186
Condition Est. Over 99%ile	0	0
Condition Est. Over 95%ile	0	0
Condition Est. Over 90%ile	3	0
Enzymes Over 99%ile	8	0
Enzymes Over 95%ile	28	14
Enzymes Over 90%ile	52	30
Enzymes Under 10%ile	150	240
Enzymes Under 5%ile	120	161
Enzymes Under 1%ile	90	88
Compounds Over 99%ile	7	8
Compounds Over 95%ile	37	101
Compounds Over 90%ile	146	360
Compounds Under 10%ile	94	443
Compounds Under 5%ile	60	162
Compounds Under 1%ile	35	101

A summary of the above would be:

• More bacteria are in play over time, post COVID
  • More bacteria at extreme high values
  • Less bacteria at extreme low values
• No difference seen for canned selection of bacteria ( JasonH thru XenoGene)
• More extreme values seen from Outside Lab Range (+/- 1.96SD), and Outside Box-Plot-Whiskers
• Enzyme production became more extreme over time, after COVID
• Compound production extreme values dropped a lot

This agrees with Special Studies on a large population (n > 150) of long COVID, especially for Compounds. There were no statistically significant compounds detected. There were statistically significant bacteria and enzymes detected.

Looking at the From Special Studies, we see that the degree of match for all of the top items have increased significantly (with the lab quality being similar), which is in agreement with “getting worse”

	Older Sample	Latest Sample
Inflammatory bowel disease	52% match	63% match
Small intestinal bacterial overgrowth (SIBO)	45%	54%
Depression	43%	50%
ME/CFS without IBS	42%	50%
COVID19 (Long Hauler)	41%	46%
Unrefreshed sleep	41%	51%

Which set of Suggestions is likely best?

Using BiomeSight data with Long COVID with Special Studies is by far the best. It detects 93 different bacteria at a detail level (low hierarchy – species, z-scores for each above 2.6) versus GI-MP with PubMed Long COVID studies which detects just 2 bacteria at the highest hierarchy level (phylum). While GI-MAP may be technically superior, there are two important factors:

• The number of bacteria reported is significantly less in GI-MAP. Ideally, they will, in time, provide a deep report with NCBI numbers available for download or automatic transfer.
  • Volume of data and volume of different bacteria has an exponential impact on the ability to detect what has statistical significance.
• We do not have a curated set of samples with Long COVID with GI-MAP. In other words, it does not point to the specific bacteria we should focus on.

Bottom Line

I am going to skip building a consensus here. The Special Studies filter for Long COVID feels right for a starting point. Doing the suggestions for 3 months and then doing a retest would be my way forward (after reviewing with your medical professional)

What we have not touched is probiotics.

From KEGG data, the top species were Azospirillum lipoferum, Streptomyces venezuelae, Azospirillum brasilense — all sitting in 110-120 range. These are from probiotics like Equilibrium and Prescript Assist (various versions) –  Prescript-Assist®/SBO Probiotic sits at the top of mixed impact.

Scanning down the list for familiar names, we see

• Bacillus subtilis subsp. natto, Bacillus subtilis, and other bacillus probiotics in the 75 to 83 range
• Escherichia coli is at 68
• Clostridium beijerinckii is at 63
• Clostridium butyricum is at 55
• Enterococcus faecalis at 40
• Lactococcus lactis at 39
• Lactobacillus plantarum at 36
• Lactobacillus gasseri at 23
• reg’activ / immune & vitality – contains lactobacillus brevis (Levilactobacillus brevis) at 28 and lactobacillus fermentum (Limosilactobacillus fermentum) at 24.3
• Pädia GmbH (DE)/Mambiotic Kapseln – contains lactobacillus fermentum (Limosilactobacillus fermentum) at 24.3
• spain (es) / lactanza hereditum – contains lactobacillus fermentum (Limosilactobacillus fermentum) at 24.3
  

Remember: The above are specific species, you may wish to view the Probiotic Mixtures page to see which species are in each mixture. My take for simplicity would be: Prescript Assist, Symbioflo-2 (E.Coli probiotic),  miyarisan (Clostridium butyricum) and one of those listed above containing lactobacillus fermentum (Limosilactobacillus fermentum).

Reminder: One set of probiotic suggestions is obtained from the (very few) studies using the probiotics species and seeing what is affected. The other set of suggestions is obtained from the genomics of the bacteria that you have and the genomics of probiotic species. Two very different approaches which lead to the same probiotics.

Questions

This person is a new person to microbiome intrepretation (with a dose of brain fog)

Q: Are the items on the green add or increase lists the things I should be adding to my diet to help lower the levels of the bad bacteria I have.?

A: Correct, Some items have a ruler beside it, 📏, This is a link to dosages used in clinical studies (i.e. likely both safe dosages, and likely sufficient dosage to cause change. They are slowly being added for more and more modifiers.

I have shown suggestions from BOTH of your samples below, you will note that they are very similar.

For example for walnuts we see that 40-80 gram/day (1.5 to 4 ounce) or 1/3 – 2/3 cups/day.

A: The Red ones are ones to reduce or eliminate from your diet when possible. These are substance that encourages the bad shifts. Often there can be surprises when the items to avoid are those often suggested. Best example is Vitamin B12 — I have a hypothesis (which some people agree with from their experience) that there are greedy bacteria in the gut that thrives on B-12 (thus accounting for low B-12 levels in the blood). It is suggested that B-Vitamins be taken by injection or transdermal (via skin patches).

Q: What should be the priorities in terms of diet and supplements to help reduce the levels of bad bacteria? 

A: After reducing as many red items slowly (don’t go cold turkey!) as practical, start adding the green items with the highest confidence items according to practically (cost, ease of preparation, other issues). Again, small steps, adding one every two days. For probiotics, do one for two weeks and then change to a different one (See this post for background)

Q: Also the top three probiotics that the AI program came up with are Symbiopharm/ symbioflo, Regactiv immune and vitality and LGasseri probiotic. Is that correct? I have started taking Visbiome? 

A: Yes, Visbiome has mixed benefits, some of it’s component may shift items in the wrong way. It’s estimated confidence is 11 versus another mixed item Global Healing Center / FloraTrex with a value of 95. I would suggest taking it for two weeks (or until one of the other arrives, which ever occurs last).

Q: Also I see that Enterobacter, Bacteroides, Firmicites and Bacillus are too high. Which of these is the biggest problem and what particular bacteria in these bacteria families are over grown? 

A: Unfortunately, there is no way to detects which is the biggest problem. People will often speculate. With specials studies, we get a measure of how strongly bacteria are associated (called a z-score), For you, the most statistically significant for Long Covid that you have are

• Catonella
• Catonella morbi
• much lower: Clostridia

For Chronic Fatigue Syndrome the top ones for significance that you have are:

• Sporolactobacillus
• Sporolactobacillus putidus
• Prevotella copri
• Shuttleworthia

Q: I see from watching your other videos that I fit the pattern of a CFS patients pattern with a high number of rare bacteria. Have you seen people being able to reverse this to a more healthy normal microbiome and recover? 

A: Yes, take a look at this recent review A History of Several 16s Tests and Suggestions where we see measurable objective improvement as well as the person reporting improvement of some symptoms. There are several people that are doing test-> suggestions –> retest repeat. They have been forwarding their new samples to do an update as they are done (usually 3 months between tests). The pattern has been improvement both subjectively and objectively — unfortunately, it is not an overnight recovery, rather a slow methodological journey.

Different combination of bacteria cause different symptoms so the progress is from waypoint to waypoint .