Research Methodology

The question we ask of every claim

Before publishing any health claim, our team asks: what level of evidence supports this, and what would a working clinician make of it? The strength of the evidence determines how confidently we present the recommendation.

Evidence hierarchy we use

Not all sources are equal. We classify the evidence behind every recommendation into one of five tiers:

• Tier 1 — Strongest evidence: Cochrane systematic reviews, large randomised controlled trials (RCTs), meta-analyses of multiple RCTs.
• Tier 2 — Strong evidence: Individual RCTs of good quality, large prospective cohort studies (e.g., Nurses' Health Study, EPIC).
• Tier 3 — Moderate evidence: Smaller cohort studies, case-control studies, official guideline recommendations (NICE, NIH, WHO).
• Tier 4 — Limited evidence: Cross-sectional studies, animal models, expert opinion from peer-reviewed publications.
• Tier 5 — Insufficient evidence: Case reports, anecdotal observations, traditional use without clinical validation.

For any condition page, we aim to anchor every recommendation in Tier 1–3 evidence. Where only Tier 4 or 5 evidence exists, we say so explicitly with phrases like "limited evidence suggests" or "traditionally used, but not clinically validated."

Primary sources we draw from

• Clinical guidelines: NICE (UK), CDC and NIH (US), WHO, NHMRC (Australia), ESC (cardiology), ADA (diabetes), AAP (paediatrics).
• Nutrition authorities: Academy of Nutrition and Dietetics, British Dietetic Association, EFSA, USDA Dietary Guidelines.
• Research databases: PubMed, Cochrane Library, NIH PubChem, Examine.com (cross-referenced with primary sources).
• Peer-reviewed journals: NEJM, BMJ, Lancet, JAMA, American Journal of Clinical Nutrition, Diabetes Care, European Heart Journal, Gut.

How we present evidence to readers

We translate clinical evidence into practical guidance using these conventions:

• "Strong evidence shows..." — reflects Tier 1–2 evidence with consistent results across multiple studies.
• "Research suggests..." — reflects Tier 2–3 evidence with some inconsistency or smaller sample sizes.
• "Limited evidence indicates..." — reflects Tier 3–4 evidence; results may not generalise.
• "Traditionally used..." — reflects Tier 5; mentioned for completeness but not recommended as primary intervention.

What we don't do

• We don't cherry-pick. If a claim is supported by some studies but contradicted by larger ones, we note both and explain the consensus position.
• We don't conflate correlation with causation. Observational studies (e.g., "people who eat X have lower risk of Y") are flagged as such and not presented as proof of cause.
• We don't recommend supplements without dosage and risk context. If we mention a supplement, we cite the dose used in clinical trials and note any documented risks or contraindications.
• We don't replace clinical judgement. Our content informs but does not diagnose, treat, or prescribe. Every condition page links to our Medical Disclaimer.

Recipe and meal-plan methodology

For recipes and meal plans, we follow these standards:

• Nutrition data from USDA FoodData Central, McCance & Widdowson (UK), or peer-reviewed analyses.
• Portion guidance aligned with NHS, USDA, or EFSA serving recommendations.
• Allergen and intolerance flags on every recipe (gluten, dairy, nuts, soy, eggs).
• Realistic timing based on tested kitchen workflows, not aspirational ideal-condition timing.

Healing Stories methodology

The healing stories on our site are composite, dramatised narratives based on common patterns observed in clinical practice and published patient outcomes. They are not interviews with named individuals, and they should not be read as testimonials. Each story page carries a clear notice. They illustrate realistic recovery patterns to help readers set appropriate expectations — not to suggest any specific outcome can be guaranteed.

Continuous improvement

This methodology is reviewed annually and updated when better practice emerges. The most recent review was in April 2026.

Questions about our methods

If you'd like to know which sources we used for a specific claim, or you have suggestions for improving our process, email [email protected] with subject "Methodology."