Where our sources come from
We look first for systematic reviews and meta-analyses, for guidelines from professional societies and for recent randomised trials. We only use older work when nothing newer answers the question, which is why every reference shows the year. We check each statement against the study's abstract or the wording of the guideline, not against reports about it.
Rhythmus is free, shows no ads and links to no products. No manufacturer pays us to make a statement come out one way or another.
The three levels in the app
For every cycle phase and life stage, Rhythmus describes what is known. Each paragraph carries one of three levels:
- established: guidelines or several studies consistently reach the same result. Example: iron is lost with menstruation, so the recommended intake for women is higher.
- small effect: there is a difference, but it is small, doesn't show up in every study or is hard to separate from normal fluctuation. Example: the slightly higher resting metabolic rate in the luteal phase.
- not supported: a popular rule doesn't hold up against the research. Example: “train heavy in the follicular phase”.
Paragraphs without a label contain nothing that could be rated, such as “no particular nutrition rules are supported for this phase”. The level rates only the state of the research, never you or your body. Under each level you can expand why it was given, and each source shows what kind of work it is and how many studies or women it covers.
The four verdicts in the myth check
In the myth check we test popular claims. “True” and “False” mean the sources clearly support or contradict the claim. “Partly”: there's something to it, but the claim goes further. “Not supported”: the studies don't show the claimed benefit. Every verdict rests on an article in this knowledge section that uses the same source.
What we look at in a study
- Study type. A randomised trial can separate cause and effect; an observational study only shows associations. Reviews pool many studies and therefore usually carry more weight than a single one.
- Number of participants. Small studies are more likely to miss real effects. When they do find one, it is often overestimated, and other groups are less often able to reproduce it.
- How the cycle phase was determined. Estimating the phase from the calendar instead of measuring hormones assigns some measurements to the wrong phase. A 2025 review therefore only included studies with measured hormone levels: 19 studies with 279 women in total.
- Consistency. A single study isn't enough for “established”. We check whether other work points the same way.
Why small studies aren't automatically worthless
Many cycle studies have only two dozen participants. That alone doesn't make them useless; the design matters. The 2026 randomised trial on strength training had 24 women, and each leg was randomly assigned its own training variant. Each woman was her own comparison group, so differences between people drop out.
Even so, we never base a core statement on a single small study. What we say about cycle-based training also rests on reviews that point the same way.
Why we don't give impact factors
The impact factor measures how often a journal's articles are cited on average. It says little about any single article: respected journals also print opinion pieces without data of their own. The international declaration DORA explicitly advises against using it as a measure of the quality of individual articles. Instead we give the journal and year and link every source so you can read it yourself.
Your own data: Your pattern
Studies describe what holds on average across many women. What holds for you shows in your entries. “Your pattern” works like a small study with you as the only participant: it compares you with yourself, looks only at two points in the cycle fixed in advance and factors out weekdays. It only reports what holds up in two checks in a row. That's why it takes several cycles, and some real patterns stay undetected for a long time.
When the evidence changes
When a study supports or refutes a statement, we adjust the text and the level. Every change to the content is listed with date, version and reason under What we changed. If you spot a mistake, write to us at contact@rhythmus.app.
Where our rules reach their limits
- We follow fixed questions, but not a formal grading system of the kind guideline groups use.
- We don't check the authors' conflicts of interest for every study individually.
- An independent expert review of our texts is still pending.