Leong KSW, Jayasinghe TN, Wilson BC, et al. Effects of fecal microbiome transfer in adolescents with obesity: the gut bugs randomized controlled trial. JAMA Netw Open. 2020;3(12):e2030415.
Randomized, double-blind, placebo-controlled trial, prospectively registered in the Australian New Zealand Clinical Trials Registry.1
Eighty-seven obese (body mass index [BMI] of 30 or more) adolescents (aged 14 to 18 years) in Auckland, New Zealand, without prediagnosed chronic diseases that could affect weight or metabolism. The fecal microbiome transplant, aka fecal microbiota transplantation (FMT), group had 42 participants, and the placebo group had 45 participants.
The primary outcome was difference in body mass index standard deviation score (BMI SDS, a scale used to tell how far someone’s body mass index deviates from the norm for their age and sex) at 6 weeks.
Secondary outcomes included insulin sensitivity, bone density, adiposity, anthropometry, cardiometabolic and metabolic biomarkers, general and gut-related quality-of-life surveys, and gut microbiome. There was a post-hoc analysis of before-and-after frequency of metabolic syndrome. Researchers advised the trial participants not to change their diet during the 6-week assessment period, and participants completed a 3-day diet diary at the end of the intervention.
Study Medication and Dosage
The intervention consisted of bacterial concentrate derived from screened donors’ stools. Each FMT capsule was double encapsulated in delayed-release capsules. Each participant swallowed 14 grams of bacterial concentrate (28 500-mg capsules). The 28 capsules were comprised of 7 capsules from each of the 4 screened, same-sex donors. Participants took these 28 capsules over the course of 2 days, with no food for 8 hours before or 2 hours after consuming them. Each participant also consumed 70 grams of oral macrogol bowel lavage on the day before the intervention began.
Safety, tolerability, and masking: There were no serious adverse events. Minor adverse events were uncommon and deemed unrelated to treatment. No participants reported difficulty ingesting the treatment capsules. The majority of participants in both the placebo and intervention groups could not identify which group they were in.
Changes in primary outcome: BMI SDS was almost unchanged between baseline and the 6-week primary outcome date, with a nonsignificant ~1% numerical increase in the FMT group, and a nonsignificant ~1% numerical decrease in the placebo group. There were also no significant differences between BMI SDS at 12- or 26-weeks postintervention.
Changes in secondary outcomes: The FMT group had reduced android-to-gynoid fat ratio compared with the placebo group at 6, 12, and 26 weeks, an effect which was more pronounced among female participants.
Metabolic parameters (insulin sensitivity, liver function, lipid profile, and inflammatory markers), total body fat percentage, blood pressure, and quality of life were not affected by FMT at 6, 12, or 26 weeks.
Post-hoc exploratory analysis: At baseline, none of the participants were diagnosed with any chronic diseases that could affect weight or metabolism, but data gathered during the trial revealed that 13 participants in the placebo group (29%) and 18 participants in the FMT group (44%) had metabolic syndrome at baseline.
Of the participants who had metabolic syndrome at baseline, 10 out of 13 in the placebo group (77%), but only 4 out of 18 (22%) in the FMT group, still qualified for that diagnosis by 26 weeks postintervention (P<0.01).
If we look only at the participants with metabolic syndrome, there were statistically significant drops in BMI SDS, fasting insulin, and fasting glucose in the FMT group at week 6, which did not persist in subsequent weeks.
Microbiome analysis: At baseline, comparison of lean-donor microbiomes to obese-participant microbiomes revealed that the donors had more Akkermansia muciniphila and some other bacterial species. Bacterial diversity and Firmicutes/Bacteroidetes ratio were not different between donors and participants (see “Note on microbiome” at the end of this article).
After FMT, the placebo group’s microbiome did not change substantially, whereas the FMT group’s microbiome changed significantly, demonstrating satisfactory engraftment. Female participants who received FMT had increased diversity at week 6, but not afterwards, and male participants did not have increased diversity at any time point.
In the FMT group, improvements in android-to-gynoid fat ratio were associated with reductions in relative abundance of Escherichia coli and improvements in relative abundance of Faecalibacterium prausnitzii and other healthy bacteria, and weakly but nonsignificantly correlated with increases in diversity.
I became a licensed naturopathic doctor in 2011, the first year a paper was indexed in PubMed with the term “fecal microbiota transplantation.” Five papers were published using that term that year, and by 2016, there were 298.2
In May 2016, I spent a weekend in Hamilton, Ontario, with pioneers in the world of FMT research: Emma Allen-Vercoe, PhD (biologist and lead researcher in the RePOOPulate Project),3 Christine Lee, MD (infectious disease doctor and medical microbiologist),4 and Peter Kim, PhD (statistician).5
At that time, controlled trials of FMT for humans had only been conducted for infectious and autoimmune colitis, but there was 1 case report of a woman who had gained 34 pounds after receiving fecal transplant from her overweight daughter.6 This wasn’t a straightforward case, because both the woman and her daughter had a BMI of about 26 before FMT, and both gained about 30 pounds after FMT. In addition, the mother had received at least 7 rounds of antibiotics in the few months before FMT, for bacterial vaginosis, Clostridioides difficile (formerly known as Clostridium difficile, aka C diff) colitis, and Helicobacter pylori gastritis. We know that both taking antibiotics7-9 and treating H pylori[10 may be obesogenic.
Important corollaries from other trials are that the donor should have a more diverse microbiome and the recipient must have a less diverse microbiome.
Murine trials had shown a decade earlier that there was an obesogenic microbiome: We could reliably make skinny rats fat through transplanting fat rats’ microbiota.11,12 Transplanting human microbiomes into mice increases adiposity in the mice, with significantly more adiposity if the human donor is obese,13 but in trials like these, diet appears to be a bigger regulator of mouse weight gain than microbiome.14
Regardless of adiposity or exercise status, a transplanted microbiome from an obese donor can impair glucose regulation in a lean mouse recipient.15 The same group that published the first-ever randomized, controlled trial of FMT for C diff in 201316 published the first-ever prospective trial of FMT in humans the year before: They transplanted fecal microbiota from lean human donors to men with metabolic syndrome and observed a statistically significant increase in insulin sensitivity, which lasted for weeks before reverting to baseline status.17 However, they noted in a larger 2017 follow-up randomized, controlled trial (RCT) that the post-FMT insulin sensitivity is improved only if the donor has a diverse microbiome and the recipient has a nondiverse microbiome before FMT. They also noted that good responders had significant increases in Akkermansia muciniphila.18
Obese humans tend to have a more proinflammatory microbiome.19 Probiotics can favorably affect the inflammatory microbiome and lead to statistically significant weight loss (half a pound or more per month).20 Back in Hamilton, Kim asked me what he called the million-dollar question: Could FMT substantially reduce weight in overweight and obese humans? Other researchers were wondering the same.21
A 2018 retrospective looked at C diff patients who had received FMT from self-selected family-member donors. The donors ranged from lean to obese in their body weight.22 Although all C diff infection patients had lost weight during their diarrheal illness and appeared to generally regain that weight after a successful FMT, there was no additional weight gain noted in dozens of patients who received FMT from overweight or obese donors. That said, a case report in 2019 described a 26-year-old female with anorexia nervosa who increased her BMI to 19 in a treatment center, but saw it drop to and stabilize at 15 after discharge despite compliance with a hypercaloric, 2,500-calorie daily diet and a desire to achieve a healthy weight. After FMT from a donor with BMI 25, her body fat increased by 55%, and she gained and maintained almost 14 pounds over 36 weeks without any change in diet.23
In 2020 there were 620 papers published about FMT. Excitingly, 2 of them were randomized, controlled trials looking at FMT vs placebo in obese subjects. In the first, published in March 2020, researchers at Massachusetts General Hospital gave FMT or placebo capsules (30 capsules, then 15 capsules a week for 5 weeks) to 24 obese adults with mild to moderate insulin resistance. Despite apparently satisfactory microbiome engraftment, there were no significant differences in insulin sensitivity (primary outcome) or BMI (a secondary outcome).24
That brings us to the study that is the topic of this commentary, which intentionally enrolled enough participants to try to detect change in BMI as a primary outcome. Although it did not achieve this outcome in the group as a whole, the researchers did detect it in the subset of patients with metabolic syndrome. There is enough signal in this and previous studies to consider FMT for overweight people with metabolic syndrome, in whom it appears that FMT can influence insulin sensitivity and BMI. Important corollaries from other trials are that the donor should have a more diverse microbiome and the recipient must have a less diverse microbiome (perhaps Shannon diversity index of 5.9 or below).
Even though the study from Mass General dosed 4 times higher than this study and didn’t see an impact on insulin sensitivity, as someone who has used FMT for over a decade with patients with a wide variety of indications, I wonder if this and other trials of FMT for people with obesity and/or metabolic syndrome have been underdosing a little. I don’t personally have experience using FMT for a patient with a chief complaint of obesity or metabolic impairment, but I usually use a minimum of 10 doses of FMT for chronic hyperinflammatory conditions I’ve treated. It is also possible that adolescents with metabolic syndrome may be more likely than adults (the Mass General trial enrolled participants aged 25-60 years) to respond favorably and/or rapidly to microbiome changes.
I can’t remember who in the weight-loss world quipped, “Exercise for ounces, and diet for pounds.” The literature in some ways supports that quip; it may take as long as 6 to 18 months of intense exercise, 30 to 60 minutes daily to achieve a 10- to 15-pound (average 3%) weight loss.25 In contrast, following a very-low-calorie ketogenic diet for just 4 weeks can reduce overweight by an average of 20 pounds.26 Other diets achieve reductions similar to those achieved by exercise: The Mediterranean diet can reduce overweight by 8 to 20 pounds after a year or longer, and vegetarian diets reduce overweight by an average of 7 to 10 pounds.27
Obesity, of course, is not an entity; it is a state of being. It is the human in front of us who is obese, with their unique history and metabolism. That human may or may not be sedentary, may or may not have endocrine disruption, may or may not have a whole-foods diet or a diverse microbiome. Certainly, among other factors, we should take exercise, diet, and microbiome into account, tailoring our interventions to the individual patient. Ways to beneficially alter the microbiome to avoid cardiovascular and other risks in overweight and obese patients can include diet (especially Mediterranean diet),28,29 avoidance of superfluous antibiotics, and the use of probiotics; and in some overweight patients, especially those with impaired metabolism and low gut diversity (perhaps below 6 on the Shannon diversity index) and lacking Akkermansia, it may include FMT.
Note on microbiome: Early research indicated that the ratio between the 2 phyla that make up 90% of gut microbiota, Bacteroidetes and Firmicutes, represented lean (Bacteroidetes dominant) and overweight (Firmicutes dominant) microbiomes.30 However, other research suggests that the Bacteroidetes/Firmicutes (BF) ratio can be flipped from what that earlier research indicated in nonobese diabetic mice,31 and the data on the BF ratio in humans have been contradictory,32 with larger and more recent studies not supporting a BF-ratio connection with overweight phenotype,33 so I no longer recommend using that ratio as a marker.
Note on regulation of FMT: As of this writing, FDA still allows clinicians to use FMT outside of clinical trials only for patients with C diff infections not responding to standard therapies. This exception, allowing FMT for C diff, is likely to be removed when a commercial product is approved that substitutes for FMT. This is likely to happen this year. That being said, FDA regulates clinicians but does not regulate health choices that individuals make in their own homes, and it is almost certain that individual patients will continue to try home FMT to self-treat a variety of conditions unrelated to C diff. It behooves us to understand the ongoing data on FMT for various conditions and inform patients accordingly.