Our Standards

Every summary should make the evidence easier to understand.

Health headlines often stretch what a study found. Our job is to report research accurately, show its limits, and send you to the original source.

How we review studies

Most of our articles follow one format: one thing from one study. Each one gives you:

  1. The one thing. What the study tested, in one plain sentence, at the top of the page.
  2. Do it today. Practical steps using the amounts and schedule the study actually used.
  3. What the study found. The real numbers, who was studied, and what kind of study it was.
  4. Keep in mind. The study’s limits: size, length, design, and funding.
  5. Check first. Who should talk to a doctor before trying it, when that applies.
  6. The source. A link to the paper or the researchers’ own published summary.

Every number we publish is checked against the original paper or the research institution’s own release.

How we choose research

We prioritize peer-reviewed studies in established journals, systematic reviews and meta-analyses, and research that is getting public attention and needs context. We label preprints (studies not yet peer reviewed) clearly, and we label animal and cell studies so they aren’t mistaken for findings in people.

How we write about findings

  • We say “linked to” or “associated with” for observational studies, and reserve causal language for well-designed trials that support it.
  • We report absolute numbers alongside relative ones when the paper provides them.
  • We don’t call anything a cure, breakthrough, or miracle.
  • We only suggest steps that match what a study tested, and we note who should check with a doctor first. We don’t diagnose, and nothing here replaces care from a qualified clinician.

Sources and links

Every summary links to the original study. Where possible we link to free full-text versions. When we mention later research in a Research Revisited article, we cite it too.

Independence

We don’t accept payment to cover a study, and sponsors never review or approve our summaries. Any relationship that could look like a conflict of interest is listed on our Disclosures page.

Use of AI tools

We use software tools, including AI, to find studies and draft summaries. Every number is checked against the original source before it is published, and the editor is responsible for everything on the site.

Images

Some images on this site are conceptual illustrations or AI-generated photography used for visual context. They are labeled where they could be mistaken for scientific imagery and are not data.

Corrections

If we get something wrong, we fix it and say so. See our Corrections policy, or email team@herhearth.com.