The Research Landscape: How to Read Migraine Studies Critically

Hierarchy of evidence, briefly

Not all studies carry equal weight. Roughly, in ascending order of reliability:

• 01Case report: one person. Interesting, never conclusive.
• 02Observational or cohort study: shows association, not cause.
• 03Randomised controlled trial: participants allocated by chance, ideally blinded.
• 04Systematic review and meta-analysis: pools multiple trials, the strongest single source.
• 05Clinical guideline: for the UK, NICE guidance and SIGN, which synthesise the above.

Placebo response is unusually high in headache trials

Headache and migraine trials show substantial placebo responses, sometimes affecting a third or more of participants. That is precisely why an uncontrolled study, or a page of testimonials, cannot establish that something works. Any claim about a treatment effect needs a controlled comparison.

Questions to ask of any study

Five that resolve most doubt:

• 01How many participants, and over how long?
• 02Was there a control group, and was allocation random?
• 03Was the outcome pre-specified, or found afterwards?
• 04Who funded it, and who ran the analysis?
• 05Has it been replicated by an independent group?

Where to look instead of a headline

Start with NHS guidance, NICE clinical guidance, The Migraine Trust and the International Headache Society's classification. PubMed is free to search if you want the primary literature, and Cochrane reviews are the most reliable summaries of intervention evidence.

How this applies to cosmetic products

Cosmetic aromatic products, including ours, are not licensed medicines and are not required to demonstrate clinical efficacy. That cuts both ways: it means they can be enjoyed for how they feel, and it means nobody, including us, should tell you they treat or prevent migraine. If you want a treatment with an evidence base, that conversation belongs with a clinician.