The dopamine.co.uk Evidence Standard

Every dopamine.co.uk page that answers a question about evidence carries a grade from A to E. The grade tells you how strong the human evidence is for the page’s main question or claim — not whether we like the answer.

• A
  Multiple high-quality human studies

  Several well-conducted human studies — typically consistent systematic reviews, meta-analyses, multiple randomised trials, or repeated human imaging studies that measure dopamine directly — point the same way.

• B
  Good human evidence with limitations

  Good human evidence that has real limitations: trials that are small, short or few; results that are consistent but not yet replicated at scale; or strong evidence for part of the claim only.

• C
  Limited or observational human evidence

  Limited human evidence: observational or cross-sectional studies, small pilots, or brain scans that measure a proxy (such as blood flow) rather than dopamine itself.

• D
  Primarily animal or mechanistic evidence

  Mainly animal or mechanistic evidence. Findings in rodents or cells can be important, but they do not tell us what happens in people.

• E
  Unsupported or contradicted

  Unsupported or contradicted. The human evidence is absent where it should exist, or points the other way.

• ?
  Evidence currently insufficient

  Too little evidence of any kind to grade. We publish this openly: knowing that a popular claim has not been tested is useful information.

How a grade is assigned

1. We identify the page’s main question or claim.
2. We search PubMed and prefer, in order: systematic reviews and meta-analyses; randomised controlled trials; human studies that measure dopamine directly (for example PET imaging, pharmacological challenge or patient studies); cohort and observational studies; animal and mechanistic studies. NICE, NHS and public-health guidance inform safety advice.
3. Every reference is verified against its PubMed record, and every number is checked against the source abstract.
4. The grade is assigned by editorial judgement against the definitions above, and the reasoning is shown on the page with the studies that justify it. No grade is final without human validation, and no grade is published without supporting references.

How studies are assessed and grades validated

For every assessment we consider: study design; sample size; human versus animal evidence; randomisation, blinding and control groups where relevant; replication and consistency between studies; effect sizes and confidence intervals; risk of bias and attrition; the population studied and the duration; clinical versus surrogate outcomes; systematic reviews and meta-analyses; publication date; how directly the evidence addresses the claim; and conflicts of interest or funding. We distinguish statistical significance (an effect is unlikely to be chance) from practical significance (an effect is big enough to matter in real life).

In plain language, the grades are: A — Strong, B — Good, C — Limited, D — Early (mainly mechanistic, animal, small or preliminary studies) and E — Unsupported.

Validation. AI tools may help draft a proposed grade and its reasons, but no grade is final until a human editor has checked it against the cited studies. Until then the page shows the grade as provisional. Validated grades show the validation date. Every grade is auditable: the page lists the studies and the reasoning behind it.

How sure are we?

Each page also gives a plain-language confidence rating — very confident, reasonably confident, uncertain or very uncertain — with the reasons. Where researchers genuinely disagree, the page says so and explains why; where evidence overwhelmingly points one way, we don’t invent balance.

Why a brain scan is not a dopamine measurement

Most headlines about phones, likes and “dopamine hits” come from functional MRI, which tracks changes in blood flow. fMRI can show that a reward-related brain region became more active; it cannot show that dopamine was released, or how much. Studies that measure dopamine more directly — such as PET imaging with dopamine-receptor tracers — are rarer, smaller and more expensive. Our grades reflect that difference.

Verdicts and grades are different things

A verdict (true, partly true, false…) says whether a claim matches the evidence. A grade says how strong that evidence is. A claim can be confidently false (verdict: false, grade: A) or plausibly true on thin evidence (verdict: unproven, grade: C).

Review cycle

Every graded page shows when it was last reviewed and when its next review is due (at most twelve months). Grades can go up or down as new research is published. Changes that alter a verdict or grade are logged on the corrections page.