How we work

Last updated 2026-07-23

1. Who writes this

Everything here is written by Sarah Lee, who founded Idunia. She holds no clinical, dietetics, or exercise-science credentials, and does not hold herself out as a dietitian or nutritionist. What she does is narrower and stated plainly: she selects published research, reads it in full, and explains what it says.

2. What this site is, and is not

This is informational writing about nutrition research. It is not individualized advice, not medical nutrition therapy, and not a substitute for a clinician who knows your history. It reports what studies found in the populations they studied. It never tells you what to do with your own body.

3. How topics are chosen

Topics are chosen because the research is strong enough to explain honestly and the question actually matters to female athletes — iron, energy availability, the menstrual cycle. Novelty alone is never a reason to cover something. A surprising single study is a reason to be careful, not a reason to publish.

4. The evidence hierarchy

Not all evidence is equal, and the writing says which kind it is leaning on. Roughly, from stronger to weaker: randomized controlled trials, then prospective cohorts, then cross-sectional studies, then animal work. Observational studies show association, not causation, and are described that way. A meta-analysis is only as good as the studies inside it.

5. How populations are handled

Every number carries the population it came from: how many people, their sex, age, training status, and country, and the instrument used to measure it. Findings are never quietly carried from one population to another. A result in elite cyclists is not a result in recreational lifters, and the sentence will say so. This single discipline is what most consumer nutrition writing gets wrong.

6. How uncertainty is flagged

Each research piece states how a finding sits against the wider literature, using four labels:

Consistent — it agrees with the weight of the evidence.
Extends — it adds to a consistent picture.
Contested — good studies disagree, and the piece says so rather than picking a side for you.
Outlier — it runs against the consensus. An outlier may be covered, but never as a headline that implies the advice has changed.

7. How AI is used, and how it is not

A human chooses the sources. A human reads the full papers. AI assists in drafting prose from material a person has already selected and verified. Every number is checked by a human against the full text before it is published. AI is never the origin of a fact, and is never credited as an author. If a claim exists only because a model produced it, it does not run.

8. Full text, not abstracts

This is the hard gate, and it is published here as a public commitment. No number is published if it was read only in an abstract. Abstracts strip exactly the two things that go wrong most often — the population a number applies to, and the instrument used to measure it. Full text, or the number does not appear.

9. Conflicts of interest

Idunia is a paid app for female athletes, and Sarah owns it. This site exists partly to introduce people to that app. That is a real commercial interest and it is stated openly. To be clear about what that does not buy: content is not gated on using the app, there are no supplement sponsorships, and there are no affiliate links. This site does not claim to be independent, because a publication attached to a product it sells cannot honestly make that claim.

10. Corrections

If something here is wrong, it gets fixed, visibly and with a date. To report an error, email hello@getmaxoutput.com. DOIs are checked against the retraction record before a paper is cited, and a citation to a retracted paper is removed when found.

11. Last updated

This page was last updated on 2026-07-23. It will change as the practice does; material changes are dated.