Research literacy
How to read a botanical study without overpromising
A practical evidence guide for Kenyan buyers: study design, formulae, endpoints, effect sizes, safety signals and product comparability.
A transparent reading habit protects buyers from both exaggerated promises and reflexive dismissal of useful research.
Relative risk = risk in treatment group ÷ risk in control group
Relative measures can sound large when baseline risk is small; absolute differences and confidence intervals give needed context.
95% CI = a range reflecting statistical uncertainty around an estimate
A confidence interval that crosses a clinically important null value often signals imprecision or no clear difference, depending on the prespecified analysis.
Start with the product, not the headline
Write down the botanical species, plant part, extraction solvent, marker compounds, dose, route, duration and comparator. A study of 182 mg enteric-coated peppermint oil is not a study of a diffuser. A trial of 600 mg purified oral CBD is not a trial of an unknown 50 ml blend. A fruiting-body mushroom powder is not automatically equivalent to a mycelial erinacine extract. Comparability is a scientific requirement, not a marketing detail.
Classify the evidence level
In-vitro studies measure molecules, receptors or cells. Animal studies add metabolism and behaviour but still do not predict human benefit reliably. Observational studies find associations and can be confounded. Randomised controlled trials compare interventions, but sample size, blinding, missing data and endpoint choice matter. Systematic reviews are useful when their search, inclusion and bias assessment are clear. A single positive paper is a signal to inspect, not a license to promise.
Look at endpoints and uncertainty
A p-value is not a measure of clinical importance. Ask for the between-group difference, confidence interval, absolute response rate, adverse events and whether the primary endpoint was prespecified. In the peppermint RCT, secondary pain measures improved while the primary abdominal-pain response did not. In CBD arthritis trials, both placebo and CBD groups improved but the difference was negligible. Mixed results are part of the answer.
Keep safety in the same frame
Natural products can interact with medicines, irritate skin, change liver enzymes or be toxic when swallowed. Dose, route and manufacturing quality drive risk. Avoid personal dosing advice from a chatbot or blog; use the label and professional guidance. Report a persistent or serious symptom to a clinician. Enki’s commercial interest is disclosed: this library is educational product content, not independent medical advice or a clinical guideline.
Sources & credits
- CONSORT 2010 statement: updated guidelines for reporting parallel-group randomised trials — Trial-reporting standard
- Cochrane Handbook for Systematic Reviews of Interventions — Evidence-synthesis methods
- NCCIH: Using dietary supplements wisely — Authoritative consumer safety guidance