Benefits
Raises fecal Bifidobacterium counts at 1.4 to 2.8 g/day
In an 8-week randomized placebo-controlled trial in 32 healthy adults, XOS at 1.4 g/day and 2.8 g/day increased fecal Bifidobacterium counts, with the higher dose working better. Counts of Lactobacillus, Enterobacteriaceae and Clostridium did not change, and total anaerobes and the Bacteroides fragilis group also rose at 2.8 g/day, so the effect is not confined to bifidobacteria. Raising a stool bacterial count is a laboratory measurement: no symptom, digestive, immune or metabolic outcome was recorded in that trial.
Microbiome composition shifts, with limited measured health outcomes
Randomized human trials of XOS agree on one thing: it raises fecal Bifidobacterium counts. They do not support the wider claims usually attached to that result. The 32-person Finegold trial reported no notable shift in bacterial diversity, no change in stool pH and no change in stool short-chain fatty acids or lactic acid, and no difference in Enterobacteriaceae or Clostridium counts. Bowel movements per day did rise in the Childs trial, but at 8 g/day, four times the usual iXOS® serving. A separate 4-week study gave 4.2 g/day to severely constipated pregnant women and reported stools rising from about 1 to nearly 7 over the study, but it had no placebo group, so how much of that came from XOS is unknown. Whether the composition changes themselves make a person feel or function better has not been tested.
Short-chain fatty acids, and what the metabolic trials actually showed
Gut bacteria ferment XOS to butyrate, propionate and acetate, short-chain fatty acids that feed colon cells and act on immune and metabolic signalling in laboratory work. In people the results are mixed and the trials are small. In an 8-week randomized placebo-controlled trial in 26 adults with type 2 diabetes, 4 g/day of XOS lowered blood glucose, HbA1c, total and LDL cholesterol and oxidized LDL. In an 8-week pilot in 13 prediabetic and 16 healthy adults, 2 g/day changed none of those things: no effect on fasting glucose, HOMA-IR, triglycerides, TNF-alpha or body composition. Nobody has repeated the diabetes result, and XOS is a food ingredient, not a treatment for diabetes or high cholesterol. On the immune side the human measurements have moved in mixed directions rather than showing enhanced defence, and no infection or illness endpoint has been measured.
Mechanism of action
Selective Bifidobacterium fermentation and SCFA production
XOS passes undigested to the colon where Bifidobacterium species — possessing specific beta-xylosidase enzymes — selectively ferment it as a preferred substrate. This selective fermentation increases Bifidobacterium abundance and produces butyrate (colonocyte fuel, anti-inflammatory, histone deacetylase inhibitor), propionate (hepatic gluconeogenesis inhibitor, satiety signal), and acetate. These short-chain fatty acids have systemic metabolic and immune roles in laboratory work, but the human trials have not reliably reproduced that: the largest of them measured no change in stool short-chain fatty acids at all, and the metabolic results have gone in different directions in different trials.
Clinical trials
Randomized, double-blind, placebo-controlled trial in 32 healthy adults given 1.4 g/day XOS, 2.8 g/day XOS or placebo for 8 weeks, with stool culture, sequencing, pH and short-chain fatty acid analysis. (Finegold SM et al. 2014, Food Funct 5(3):436-45)
Healthy adults.
Bifidobacterium counts rose in both XOS groups compared with placebo, and more at 2.8 g/day than at 1.4 g/day. Bacterial diversity, stool pH, short-chain fatty acids and lactic acid did not change, and counts of Lactobacillus, Enterobacteriaceae and Clostridium were no different between groups. Blood glucose was not measured in this trial. XOS was well tolerated, with no significant gastrointestinal symptoms at either dose.