You’re caught up.

Home/Issues/Issue 09/Study

12 week randomised, double blind, placebo controlled, multicentre parallel trial Online publication date confirmed against the PubMed publication date field and the PMC record.

A killed bacterial supplement did nothing overall, and moved weight in the half of people whose guts were least diverse

AI narration, generated on first listen
Journal
Gut Microbes 18(1):2722835
Authors
Lim A, Kim S, Yoon S
Institution
the LOTTE R and D Center
Published
25 August 2026
Source
PMID 42640624 · DOI 10.1080/19490976.2026.2722835
Design
Twelve week randomised, double blind, placebo controlled, multicentre parallel trial with a pre-specified primary outcome and a prospective power calculation. Registered with the Clinical Research Information Service as KCT0008119 with no protocol amendments after registration. Subgroup analysis by baseline gut microbial diversity was not pre-specified.
Sample
120 overweight Korean adults aged 19 to 65 with a body mass index between 25 and 30, randomised 60 per group. Per protocol set 101, being 49 on the supplement and 52 on placebo, with diversity subgroups of 24 to 26 each.

What a paraprobiotic is

A paraprobiotic is a bacterial product that has been deliberately killed before it is swallowed. The cells in this one were heated to 90 degrees for half an hour and then freeze dried, so, as the authors put it, it is not expected to replicate or stably colonise the gut. The idea is that the dead cells and their components can still be sensed by the host even though nothing takes up residence.

Drawn from the paper's introduction.

Why they ran it

The authors write that baseline gut microbial diversity may represent a reproducible stratification marker for paraprobiotic responsiveness, but that this proposition has not been comprehensively evaluated in controlled trial settings.

Drawn from the paper's introduction.

One hundred and twenty overweight Korean adults were randomised 60 to each arm and took either a 1,600 mg daily sachet containing 1,500 mg of the heat inactivated preparation or a matching placebo for twelve weeks. The trial was registered in advance and powered for its primary outcome: 50 participants per group were required to detect a 1.5 kg between group difference in body fat mass at two sided alpha 0.05 with 90 percent power.

The primary outcome came back null. Change in total body fat mass by DXA showed no significant between group difference across the per protocol population, with fat mass falling in both arms. Body weight fell by 1.33 plus or minus 0.28 kg on the supplement against 0.94 plus or minus 0.26 kg on placebo, which was not statistically significant.

Splitting participants by how diverse their gut community had been at the start produced a different picture. In the low diversity subgroup, body weight fell by 1.98 plus or minus 0.39 kg against 0.95 plus or minus 0.33 kg on placebo at p 0.0483, with BMI at p 0.0412, circulating leptin at p 0.0423 and HDL cholesterol at p 0.0342. The high diversity subgroup showed no consistent response across any outcome domain. Each subgroup contained between 24 and 26 people.

The clinical changes in the low diversity subgroup were accompanied by compositional shifts in the gut community, elevated faecal acetate and butyrate and altered bile acid composition, with changes in the relative abundance of particular taxa inversely correlated with changes in body weight, body fat mass and leptin.

All six authors are affiliated with the LOTTE R and D Center, and the sponsor is LOTTE Chilsung Beverage.

The numbers

Primary outcome, change in total body fat mass by DXA over 12 weeksno significant between group difference; fat mass fell in both arms
Body weight, whole per protocol populationminus 1.33±0.28 kg supplement vs minus 0.94±0.26 kg placebo, not significant
Low diversity subgroup, body weightminus 1.98±0.39 kg vs minus 0.95±0.33 kg, p=0.0483
Low diversity subgroup, other outcomesBMI p=0.0412; leptin p=0.0423; HDL cholesterol p=0.0342
Randomisation and analysis sets120 randomised (60 per group); per protocol 101 (49 vs 52); subgroups 24 to 26 each
Doseone 1,600 mg sachet daily containing 1,500 mg heat inactivated preparation
Powering50 per group for a 1.5 kg between group fat mass difference at alpha 0.05, 90% power
RegistrationCRIS KCT0008119, no protocol amendments after registration
SponsorLOTTE Chilsung Beverage; all six authors affiliated with LOTTE R and D Center

Why this might happen

Proposed by the authors This is the explanation the authors offer in their discussion. This study did not test it.

The authors argue that communities with low diversity harbour a less densely occupied niche, a condition which permits broader and more coherent compositional reorganization in response to external stimuli, whereas species rich communities are less susceptible to reorganization because the constituent taxa use limited resources more efficiently and collectively occupy a denser network of nutritional niches.

They are explicit that their own evidence for this is indirect. They write that the observed cross domain coherence provides indirect support for the possibility that the supplement contributed to reshaping the resident microbial community, rather than evidence of a strain specific causal mechanism, and that the correlations do not establish temporal sequence, mediation, or causality.

Drawn from Discussion, PMC13523931.

What this does not show

  • This does not show that the supplement causes weight loss in people with low gut diversity. The subgroup split was analysed after the fact within a trial that missed its primary outcome, and the authors themselves say the result requires prospective validation in independently designed trials with pre-specified diversity based stratification.
  • This does not show that the supplement reached or acted on the gut community. The authors state that no strain specific measure of engagement was included, that their sequencing approach could not track the strain, and that no strain associated metabolites or proximal host response markers were assessed.
  • This does not establish the barrier and inflammation pathway the discussion describes. Plasma markers of leakiness, intestinal permeability and the relevant gut hormones were not measured, so that chain is imported from other work rather than shown here.
  • This does not tell you whether anything lasts. The authors note that durability beyond the twelve week intervention was not evaluated, and quantitative dietary intake was not measured at all.

Where this leaves us

A properly powered trial of a killed bacterial supplement found nothing on its pre-specified primary outcome, and the only signals appeared in a subgroup defined by how diverse someone's gut was to begin with.

Korean adults aged 19 to 65 with a body mass index between 25 and 30, over twelve weeks.

A trial that measures baseline diversity first and randomises within those strata, designed by a group with no commercial stake, is what would turn this from a hypothesis into a finding.

Caveats worth holding

  • The trial is run and staffed by the manufacturer of the product tested. This is disclosed in the paper and is recorded here so it can be weighed; it was not treated as grounds for exclusion.
  • The headline subgroup finding was not pre-specified and sits inside a trial that missed its primary outcome.
  • Analysis is per protocol rather than intention to treat, and 19 of 120 randomised participants are absent from it.
  • Subgroups of 24 to 26 people are small for the comparisons drawn from them.
  • Dietary intake was not measured quantitatively over the twelve weeks.

Newsletter

Each issue by email, when the newsletter launches. Leaving your address puts you on the list, nothing is sent yet.