Strongest evidence
The most robust research area is reduced upper respiratory infections (moderate evidence strength). Across 6 studies, 5 reported beneficial effects, with effect sizes ranging from small to large. A meta-analysis found a moderate effect (RR=0.79), and one high-quality RCT in healthy infants and young children used a dose of 10 billion CFU/day. The evidence is considered preliminary due to potential publication bias and a small study base.
Mixed or weaker evidence
All other outcomes have low or very low evidence strength. Increased short-chain fatty acid production (4 studies, all beneficial, small to moderate effects) is limited to older adults, but only one study reported statistical significance. Improved intestinal barrier function (4 studies, all beneficial, moderate effects) relies on systematic reviews, in vitro, and animal models—no human clinical trials. Improved gut microbiota composition (3 studies, all beneficial, small effects) also lacks human clinical data. Improved gastrointestinal health (3 reviews, mixed effect sizes) and improved gut health (3 reviews, moderate effects) are based entirely on review articles without primary trial data. Reduced inflammation (4 studies, small effects) and improved immune function (4 reviews, mixed effects) show no statistically significant findings and no human clinical trials. Reduced antibiotic-associated diarrhea (3 studies, 2 beneficial, 1 neutral) is mixed; the only high-quality RCT found no effect.
Effective dose patterns
Only one outcome—upper respiratory infections—has a reported effective dose: 10 billion CFU/day. No other outcomes provided consistent dose information, limiting the ability to generalize dosing across conditions.
Population insights
Population data are sparse. The strongest evidence applies to healthy infants and young children (0–3 years) for upper respiratory infections. One outcome (short-chain fatty acid production) was studied exclusively in older adults (aged ≥60 years). Most other syntheses did not specify populations, making it unclear which groups might benefit.
Notable caveats
A recurring caveat across all outcomes is the small evidence base (3–6 studies) and the risk of publication bias—null-result studies are less likely to be published or indexed. Many syntheses rely on review articles, animal models, or in vitro experiments rather than human clinical trials, limiting causal inference. Statistical significance was often not reported, and effect sizes were frequently described qualitatively. The lack of consistent dosing, duration, and population data further weakens the overall evidence.