Do Probiotics Help Long COVID Fatigue? What a New Trial Found, and Why I Still Start Elsewhere
A 2026 randomized trial found a high-dose probiotic beat placebo for long COVID fatigue. What the numbers show, and why it is not the first step.
Photo: Emma-Jane Hobden / Unsplash
A multi-strain probiotic beat placebo for long COVID fatigue in a randomized trial published this July. Across four weeks, fatigue scores in the probiotic group fell by about a quarter, against roughly six percent in the placebo group. So yes, a probiotic outperformed a sugar pill here. And no, it is still not the first thing I reach for with post-viral patients. Those two answers sit together more comfortably than they sound.
How DELong#3 was built, and who paid for it
DELong#3 ran out of a single hospital in Milan. Forty-eight people completed it, all with a prior COVID infection and fatigue scoring at least 4 out of 11 on the Chalder Fatigue Scale, a standard fatigue questionnaire. Median illness duration was 30 months, so they had been exhausted for two and a half years.
For four weeks each participant took either VSL#3 or a matched placebo, two sachets a day, then four more weeks on nothing.
The headline numbers: a median fatigue-score improvement of 24.24 percent on the probiotic against 6.06 percent on placebo, p = 0.037. And 68 percent of the probiotic group counted as responders against 35.7 percent of placebo, p = 0.019.
Two caveats belong right next to those numbers. The spread overlaps heavily, since a quarter of the placebo group did at least as well as the median person on the probiotic. Beyond that, neither the paper's main text nor the registered protocol defines what a "responder" is, so you can't check the claim yourself.
VSL#3 is a commercial eight-strain formula, and the amount used was enormous. Each sachet held 450 billion colony-forming units, meaning 450 billion viable bacteria, and participants took two daily. This is the trial protocol, not a dose I am suggesting to anyone. The study drew public funding plus a non-conditioning grant from Actial Farmaceutica, which makes VSL#3 and supplied the product. Its conflict-of-interest statement nonetheless declares no commercial relationship that could be construed as a potential conflict, which sits awkwardly against the funding paragraph above it.
The bigger trial, and the review that still says no
I have to argue against my own instinct here. DELong#3 is not the largest evidence on this. The RECOVERY trial in Lancet Infectious Diseases randomized 463 people with post-acute COVID to a synbiotic or placebo for six months. Fatigue improved with odds roughly 2.3 times those of placebo. Side effects matched placebo at about one in ten. That is ten times the size of DELong#3 and six times the duration.
Set against that, a 2024 BMJ living review of 24 long COVID trials in 3,695 people found no compelling evidence for combined probiotics and prebiotics. So the field has one large positive trial and a review that declines to call the category proven.
Who these trials were not run on
DELong#3 excluded anyone with pre-existing ME/CFS, fibromyalgia, or irritable bowel syndrome, plus neurological and psychiatric diagnoses, and anyone on antidepressants, opioids, immunomodulators, or any antibiotic or probiotic in the prior month. Enrollment ran ages 18 to 65.
Read that list against the people who write to me about probiotics. Many meet at least one exclusion. Histamine was never measured, never mentioned, and never stratified in either trial, so neither says anything about histamine-sensitive people.
The placebo problem in fatigue research
About 36 percent of DELong#3's placebo group counted as responders. More than one in three people improved while taking a sachet of nothing, and that number belongs beside the two-in-three headline.
STOP-FATIGUE makes the point better. It tested a synbiotic in 26 people with post-COVID chronic fatigue over three months, and both arms improved significantly on general fatigue compared with their own baselines. On the symptom outcomes, the synbiotic only clearly separated from placebo on post-exertional malaise, the crash that follows activity. Without a placebo arm, everyone would have concluded the synbiotic fixed general fatigue. It didn't.
Which is why a testimonial, or four weeks of feeling better, cannot tell you whether the thing worked.
Why probiotics still come last in the order I teach
My practice and these trials part ways at the sequence, which has nothing to do with whether the trials are any good.
When someone's system is unstable, particularly when histamine reactivity is part of the picture, I don't open with a high-dose multi-strain product. I work people up to it. The order I use runs postbiotics first, then support for the gut lining, then a gentle prebiotic, and probiotics at the end. Butyrate is usually the opening move, seeing as it speaks to the immune system without introducing new organisms.
My reasoning is practical rather than theoretical. Say you react to most of what you eat, take a capsule holding eight strains at once, and feel worse. You have learned nothing about which strain did it. One thing at a time is slower, yet it is the only way to get an answer you can act on. That is a sequencing argument, not a safety claim about VSL#3.
I want to be careful here, because the science underneath the usual "probiotics make histamine" story is thinner than the internet suggests. Histamine production in bacteria traces to a gene called hdc, which builds an enzyme that converts the amino acid histidine into histamine. That gene sits at the strain level, often on a plasmid that strains can lose in culture. The best-characterised producers come out of cheese and wine. Even so, at least one common probiotic species, Lactobacillus reuteri, makes histamine too, and in mice that histamine acts as an anti-inflammatory signal rather than a harmful one. No one seems to have published a screen of the eight VSL#3 strains either way, and in one rat model VSL#3 reduced mast cell degranulation rather than increasing it.
So the accurate statement is that we do not know, in either direction. I could not find a clinical trial of probiotics in histamine intolerance or in mast cell activation syndrome, a condition where mast cells release their chemical contents far too easily. I hold my sequence because it works in my practice and because it fails gently. I do not hold it because a trial supports it. There isn't one.
Two different starting points
If you are relatively stable, eating a broad diet, and not reacting to new foods, a multi-strain probiotic is reasonable to raise with your provider. RECOVERY and DELong#3 give that conversation some evidence behind it.
If you are in the unstable phase, where the safe-food list keeps shrinking and every new supplement feels like a coin flip, this is not your starting gun. Build the terrain underneath first, starting with why your food sensitivities keep widening. Probiotics keep.
Either way, take it to the person who knows your history. Neither trial enrolled anyone selected for histamine reactivity, and neither included ME/CFS, so for those groups the position is simply that nobody has tested this yet.
Gut work sits inside a longer order of operations, and where it belongs depends on what else is running. The EBV Reactivation Treatment Algorithm lays that order out for post-viral fatigue and EBV reactivation, including the points where the answer is to leave the gut alone for now.
FAQ
Do probiotics help long COVID fatigue?
Two randomized trials say yes. The 463-person RECOVERY trial found a synbiotic improved fatigue with odds about 2.3 times placebo over six months, and the 48-person DELong#3 trial found VSL#3 improved fatigue scores by a median of 24.24 percent against 6.06 percent on placebo. A 2024 BMJ living systematic review of 24 long COVID trials still concluded the probiotic evidence was not compelling. Promising, not settled.
Should I take a probiotic if I have MCAS or histamine intolerance?
I could not find a clinical trial answering it. Probiotics have not been tested in mast cell activation syndrome or histamine intolerance in any clinical study I could locate, so any claim in either direction is extrapolation. In my own practice I introduce probiotics late and one strain at a time, so that a reaction tells you something usable. Discuss it with a practitioner you trust.
Do probiotics produce histamine?
Some bacteria do, through an enzyme built by the hdc gene that converts histidine into histamine. This is a strain-level property, often carried on a plasmid that can be lost, rather than a fixed feature of a species. The best-characterised producers come from cheese and wine, though Lactobacillus reuteri, a common probiotic species, also produces histamine. In mouse studies that bacterial histamine acted as an anti-inflammatory signal.
Does this research apply to people with ME/CFS?
No. Pre-existing ME/CFS was an explicit exclusion criterion in DELong#3, alongside fibromyalgia and irritable bowel syndrome. The findings cannot be transferred to an ME/CFS population without a trial that actually enrolled them.
Why would probiotics come last in a post-viral gut plan?
Because sequencing protects information. Postbiotics such as butyrate signal to the immune system without adding live organisms, gut-lining support and a gentle prebiotic prepare the terrain, and probiotics introduce new bacteria into a system that can then be assessed one change at a time. Starting with eight strains at once means a bad reaction tells you nothing about its cause.
How much of the improvement could just be placebo?
In DELong#3, 35.7 percent of the placebo group met the responder mark, so more than a third improved on nothing at all. In the STOP-FATIGUE synbiotic trial, the placebo arm improved significantly on general fatigue too. Fatigue outcomes in post-viral illness respond substantially to placebo, which is why uncontrolled results carry very little weight here.
References
- 1.
- 2.
- 3.
- 4.
- 5.
- 6.
- 7.
- 8.
- 9.
