Our editorial policy

An article appears on GesundeFakten only if its key claims can be backed up by studies in humans. This page describes how we find those studies, how we weigh them, and what happens when the research changes. The rules apply to every piece and to each of the eleven language versions.

Which topics we take on

It almost always starts with a question people are actually asking. Can you eat beetroot raw? Does psyllium husk help with constipation? Before a question becomes an article, we check whether there is enough solid research on it. If there isn’t, the topic stays on the shelf, even if it’s being searched for a lot right now. We would rather write about fewer topics than about one where we have nothing reliable to say.

Where we search

We search the two major medical databases, PubMed, run by the US National Library of Medicine, and Europe PMC. Where guidelines or position statements exist on a topic, for example from the German Nutrition Society (DGE), the World Health Organization or Cochrane reviews, we include them.

How we weigh studies

Studies differ in how much they can tell us, and we take that into account every time we assess one.

Systematic reviews and meta-analyses carry the most weight. They pool all the good studies on a question and analyse them together according to fixed rules, which is why they are considered the gold standard. Where they exist, our key claims are based on them.

Next come randomised controlled trials. In these, chance decides who receives a treatment and who ends up in the comparison group. This is the only way to reliably show whether something causes an effect. When the number of participants or the length of the study matters for the assessment, we say so.

Observational studies often follow thousands of people over many years, which makes them valuable. But they only show that two things occur together, not that one causes the other. That is why the text says, with appropriate caution, that something is associated with a lower risk.

Animal experiments and laboratory studies offer clues as to how something might work in the body. They don’t count as evidence of an effect in humans. We mention them only when they help with understanding, and we always label them for what they are.

What a topic needs at minimum

Every article requires at least five relevant studies in humans, including at least one systematic review or meta-analysis. We prefer work from the past 15 years. We draw on older studies if they are foundational and more recent research hasn’t contradicted them.

Every source is opened and read before we cite it, and every link is checked to make sure it works before publication. A study must also genuinely investigate the question the article asks. A paper on high-dose nitrate powder, for instance, says little about a beetroot salad.

How we present results

What matters to us is how large an effect is and who it applies to. A simple "works" or "doesn’t work" rarely does justice to the research. When good studies reach different results, we show both sides and explain what might account for the differences. We are open about a study’s weaknesses, such as a small number of participants, a short duration or funding by the manufacturer.

We report amounts as they were used in studies and guidelines. We don’t derive a personal dosage from them, because what suits an individual belongs in a conversation with a doctor or pharmacist. For illnesses and medication, we word things especially cautiously and advise seeking medical advice.

Sources

At the bottom of every article, the "Scientific sources" section lists all the studies used, with authors, title, journal, year of publication and a link to PubMed or the DOI. Key statements in the text point directly to the relevant study.

Updates and corrections

Research keeps moving, and our articles should reflect that. When new reviews or guidelines appear, we revise the affected pieces and note the date of the last content review.

We correct factual errors openly. The end of the article then states what was changed and when. We welcome tips about errors or new studies through the contact page.

Tools, translations and images

Digital tools help us with research, translation and drafting, including AI-assisted ones. The German version is only published once it has been editorially reviewed, checked against the sources and revised. The translations into the ten other languages are produced from that version and each one is checked individually against the original. Cover images created with AI carry a visible label reading "AI-generated". They illustrate the topic and do not represent study results.

Responsibility

Simon Gütter is responsible for the content; full details can be found in the Legal Notice. Our articles are not a substitute for medical advice, diagnosis or treatment.

Policy last updated: November 2026