Reliability criteria

We score each strain from 1 to 5 based on the strength of human clinical data. A score of 1 means no direct health studies exist for that strain. The strain may be well characterized genetically or used in food production, but nobody has tested it in people. A score of 2 means the evidence comes from animal models or lab dishes. These studies show biological plausibility but cannot confirm effects in humans. A score of 3 means at least one human trial exists, but the evidence is limited. This could be a small sample size, a single study, or a trial where the strain was mixed with others so its individual contribution is unclear. A score of 4 means multiple randomized controlled trials have tested the strain in humans with consistent results, but no one has pooled those trials in a meta-analysis yet. A score of 5 means the strain has both multiple randomized controlled trials and at least one meta-analysis or systematic review confirming the effect. This is the strongest evidence available in medical research. We do not count marketing claims, company-funded white papers, or studies on different strains of the same species. Only peer-reviewed PubMed-indexed studies on the specific strain count toward the score.