About Live Well News

Health coverage has a structural problem: the incentive is to make every study sound like a breakthrough, and almost none of them are. A single trial of forty people becomes a headline, the caveats live in paragraph nine, and the reader is left with an impression the evidence never supported.

Live Well News is built the other way round. We publish fewer studies than we read, we say plainly when a finding is preliminary or mixed or measured in mice, and we put the number that matters in the headline rather than the payoff at the bottom. Publishing since July 2026, we cover Longevity & Aging, Nutrition & Diet, Supplements, Fitness & Exercise, Sleep, Brain & Mental Health and Heart & Metabolic.

What we publish

News (333 articles) covers a single new study: what was measured, what was found, how much to trust it, and what would have to be true for it to be wrong. Every news story is tied to one paper, and one paper gets one story, permanently.

Verdicts (23 pages) are the opposite shape: standing reference pages on a single compound, updated rather than replaced. A verdict pools everything worth citing on creatine or magnesium or turmeric into one answer to the only question a reader actually has, which is whether to take it. Several conclude that the answer is no.

Everything is filed under seven sections, and cross-cutting topics that no single section owns - menopause, children's health, medication safety - are collected on the topic index.

How a study becomes an article

The largest quality lever on a site like this is not the writing, it is refusing to write. A beautifully reported article about a thirty-six-person supplement trial is still a bad article. So a paper has to clear a written bar before anyone drafts a word, and that call is made on the paper itself rather than on the press release about it:

• People, not mice. Animal-only and cell-only work is rejected. Where a paper mixes the two, the human part has to carry the finding.
• Primary research or a systematic review. No editorials, commentaries, viewpoints, letters or conference abstracts - if nobody collected data, there is no finding to report.
• Enough participants to mean something, on a floor that changes with the design: a randomised trial and a cohort study are not held to the same number, because they do not earn their power the same way. A meta-analysis needs at least ten pooled studies as well as the headcount.
• A journal we have not deny-listed, and not a retracted paper or one under an expression of concern.
• Recent. News covers work published in the last eight weeks. Older research is still worth writing about, but as an explainer, not as news.

Papers that fail are not covered, however interesting the result. That is the part most health coverage skips, and it is the reason a study you saw everywhere else may be absent here.

What gets checked before publishing

Standards decay when nothing measures them. These are the rules an article has to meet before it goes live, and one that misses a rule goes back rather than out.

• Every claim is read back against its source. Each cited sentence is checked against the paper it points at, and anything that has drifted from what the source actually says is rewritten or cut. This is the rule that earns its keep: it catches claims that are plausible, well written and simply not in the paper.
• The writing has to be readable. Every article is held to a reading level a general reader can follow. Long, clause-stacked sentences get broken up. If a finding cannot be said clearly, we do not understand it well enough to publish it.
• The summary is a summary. The key points at the top are the cliff notes: what was found, what it means, and how far to trust it. Read only those and you should be able to tell someone else what the study showed. They are not a results table.
• Statistics are translated, not transcribed. P-values, confidence intervals and risk ratios are not much use to most people, so we say what they mean - "about a third lower", "small enough to be chance" - and keep the raw figures in the questions at the foot of each article, for anyone who wants to check them. A paragraph that needs statistical training to parse is transcription, not reporting.
• Jargon gets introduced. An acronym or a technical term is explained where it first appears - every time, however obvious it seems to whoever wrote it.
• Headlines name the thing. A title has to say what was actually found. No withheld objects, no curiosity gaps, nothing that makes you click to learn what the article already knows.
• An association is never written as a cause. If a study can only show that two things travel together, the article says so, in the headline as well as the caveats.

None of this is left to memory. Each rule is written down and applied the same way to every article, including the ones we would rather wave through.

How to check our work

You should not have to take any of this on trust, so every article is built to be audited:

• Every non-obvious claim carries a numbered citation to the exact source it came from, with a link and the date we retrieved it.
• Every article lists its sources in full, with DOIs, so you can go to the paper rather than to our summary of it.
• The questions at the foot of each article carry the precise figures - sample sizes, effect sizes, intervals - in the researchers' own terms, so you can check the strength of a finding against the paper rather than against our description of it.
• Each article states what kind of study it was and what that design can and cannot show, so you can weigh the finding yourself rather than taking our word for its strength.

Across the site that comes to 5,727 citations against 1,113 listed sources. If one of them does not support what we wrote, we want to know.

What we do not do

• We do not give medical advice. We report research. We do not diagnose, prescribe, or tell you what to take. Research describes averages across groups; decisions about your health belong with a clinician who knows your history.
• We do not sell supplements or take commission. No affiliate links, no sponsored posts, no paid placements. We publish verdicts concluding that popular supplements do not work, which would not survive a commission model.
• We do not run advertising yet. No ads are currently served. When they are, they will be labelled and kept away from editorial judgement.
• We do not cover a study twice to get a second bite at a trending topic.

Who is accountable

Articles are published by the Live Well News editorial team, and every article carries a byline linking to the author page responsible for it. Our Editorial standards set out how we source, review and disclose, including how we use AI tools and where a person is required in the loop.

When we get something wrong we correct it and say what changed - the policy and the log are on our Corrections page. To report an error, or to argue with a conclusion, email [email protected]. Every report is read.

Get in touch

General enquiries: [email protected]. Story tips and corrections: [email protected]. Advertising and partnerships: [email protected], and our Advertise page. Everything is also on our Contact page. You can follow the work by email or RSS.