The results of a randomized and controlled feeding trial were recently published in Cell, detailing the effects of a probiotic and a “restore” diet, designed to restore the microbiome to a non-industrialized state and document the associated cardiometabolic changes. The hypothesis underlying the study’s design was that the symbiosis between the microbiome and human host has evolved over millions of years in a setting quite unlike our industrialized environment, and an increase in many non-communicable diseases may be a consequence of the disruption of the pre-industrial microbiome.
The probiotic used in the study, Limosilactobacillus reuteri (L. reuteri), was chosen because it is rarely found in Western microbiome samples and is one of 39 OTUs (operational taxonomic units) previously found to be completely undetectable in U.S. samples, but highly prevalent among adults from non-industrialized regions in Papua New Guinea. The restore diet used in the study emphasized foods commonly consumed in rural Papua New Guinea (but also available in Canada, where the study took place), such as beans, sweet potato, rice, cucumber, and cabbage, as well as foods that provide growth substrates for L. reuteri, specifically, foods with high amounts of raffinose and stachyose, including Jerusalem artichokes, peas, and onions. The diet was also high in fiber, primarily plant-based, devoid of dairy and wheat, and with limited processed foods.
This study was designed to evaluate several questions, including the ability of supplementation to re-establish L. reuteri colonization and its effects, the effects of the diet, and any interactions between the diet and the probiotic. Thirty healthy Canadian adults completed the trial, with exclusion criteria including many gastrointestinal and chronic health conditions, as well as a vegan/vegetarian diet, recent antibiotic use, and detection of L. reuteri (>104 cells/gram feces) before the study onset. To evaluate the diet, participants were randomized to either their usual diets or the restore diet for 3 weeks, followed by a 3-week washout, and then a crossover to the other diet. Using a parallel-arm design, on the 4th day of each diet, participants also received one of three interventions; a single oral dose of L. reuteri (∼1010 viable cells) isolated from Papua New Guinea (strain PB-W1), a single dose of an alternate strain of L. reuteri (DSM 20016), or a placebo (maltodextrin).
Regarding colonization of L. reuteri, the single dose did not appear to reliably persist with either strain, becoming undetectable by day 12-17 in all but one participant (who had received PB-W1), with some evidence that the restore diet may have enhanced the survival of the PB-W1 strain but not the DSM 20016 strain. Additionally, the probiotic did not influence the risk markers of chronic disease utilized in this study.
However, the restore diet did modulate the overall gut microbiome community and had a quite favorable effect on chronic disease markers. This included significant reductions in LDL cholesterol (-16.8%), BMI (-1.4%), plasma glucose (-6.3%), C-reactive protein (-14%), and fecal calprotectin (-21%), an indicator of gastrointestinal inflammation. Insulin sensitivity and resistance improved, as marked by changes in the quantitative insulin sensitivity check index (QUICKI) and the homeostatic model assessment of insulin resistance (HOMA-IR), driven by reductions in glucose (but not insulin). Although they did not reach statistical significance, there were also reductions in lipopolysaccharide-binding protein (LPS) and fecal zonulin levels, suggestive of improved intestinal barrier function.
A number of favorable changes to the relative abundances of specific microbiota also occurred with the restore diet, in part driven by changes to the intestinal environment, such as a greater abundance of short-chain fatty acids and a decreased pH. This included increases in species generally associated with health benefits, including three Bifidobacterium species, F. prausnitzii, Roseburia hominis, and others, while decreasing the abundance of pro-inflammatory species, such as Bilophila wadsworthia and Alistipes putredinis. These changes varied substantially on an individual basis. For example, the change in the relative abundance of Bifidobacterium ranged from −12% to +760%. The changes that occurred also appeared to be transient and reversible, i.e., during the washout periods, there was a shift toward the baseline population levels.
Despite promising changes in cardiometabolic markers, there remain many questions. It is unclear if repeated dosing of L. reuteri would have more effectively increased colonization, and if the restore diet played a large role in determining this (as well as the importance of the two probiotic strains). The small sample size may have also lacked the power to detect more subtle changes, which a larger trial would have revealed (perhaps finding significant improvements in the markers of gut barrier function). Additionally, many of the benefits of this diet could be attributed to the adoption of a plant-based high-fiber diet (fiber intake doubled during the restore diet period), and it’s unclear which, if any, of the specific dietary interventions were essential to the observed benefits. Yet previously published high-fiber intervention trials have not consistently shown improvements in glycemic and lipid outcomes (despite anti-inflammatory effects), suggesting that the restore diet may have something additional to offer.