Your cycle and the weather: what your own record shows, and what research hasn't measured

Weather sensitivity usually gets described as a fixed trait — some people react to a shift in air pressure or a hot, humid day, and others don't. But if you keep a symptom diary for more than a few weeks, a different question tends to surface: the same weather, in the same body, doesn't always land the same way. A muggy afternoon that barely registers one week can feel noticeably heavier a couple of weeks later. People who track their symptoms often notice this long before they can explain it — the calendar seems to matter about as much as the forecast. A note before we start: this is informational material, not medical advice, and nothing here diagnoses anything.

When it's a doctor's question, not a diary's

Before anything else: severe pain, heavy bleeding, or a marked change in the character of your cycle are not things to observe and wait on. They're a reason to contact a doctor. NHS guidance is specific about it — see a GP if "your periods become more painful, heavier or irregular", if "your period pain is stopping you doing your usual daily activities", if you have "pain during sex, or when peeing or pooing", or if "you're bleeding between periods"; and get urgent advice if "your pelvic pain or period pain is severe or worse than usual, and painkillers have not helped" (NHS).

This matters more than it may sound. Conditions such as endometriosis sit behind exactly these symptoms, and they are already found late: a review of seventeen studies published from 2018 onward reports times ranging from a few months to twelve years, with typical waits of five to twelve years, and concludes that the diagnostic delay "is still present" (De Corte et al., 2024). A diary won't find something like that. At best it describes a pattern, and it can't rule anything in or out — but it can absorb months of attention that a symptom like this doesn't have to spare. If the pain is severe, or something about your cycle changed abruptly, that's a conversation with a doctor, not a data point to keep watching.

What's actually established

A few things about the cycle are documented reasonably well. Menstrual migraine affects about 6% of women of reproductive age, and diary-based studies have found that perimenstrual attacks "are often more disabling and can persist up to 35% longer than those unrelated to the menstrual cycle"; the response to triptans appears to differ too, with perimenstrual attacks showing a higher recurrence rate after treatment (Raffaelli et al., 2023).

The usual explanation offered is the premenstrual drop in estrogen. It's worth quoting the same review rather than smoothing that into certainty: the authors write that "the current body of evidence supporting this hypothesis remains limited, warranting further validation", and that existing studies are "limited by methodologic inconsistencies, small sample sizes, and variable case definitions, precluding definitive conclusions" (Raffaelli et al., 2023). A widely repeated mechanism and a well-supported one are not the same thing.

The most solid physiological finding here is about temperature, and it's worth being precise about how narrow it is. A 28-day study that followed 19 healthy women closely found that skin temperatures and basal body temperature all varied across the cycle in a similar pattern, "with highest and lowest values during the luteal and follicular phases, respectively" — that is, higher in the second half of the cycle, lower in the first (Kräuchi et al., 2014). A separate meta-analysis, looking specifically at exercising in heat, found the same direction: internal body temperature was significantly higher in the luteal phase both at the start of exercise and after it, while mean skin temperature did not differ significantly between phases (Giersch et al., 2020).

That's a real, replicated result, and it stops exactly where you'd want it to keep going. It says your baseline temperature shifts a little across the month. It doesn't say how heat will feel to you on a given day, whether a humid afternoon will bother you more in one week than another, or that a headache or a flat mood in a particular week comes from that shift. The distance between "temperature runs slightly higher" and "here's why this week feels harder" is where the research runs out.

How little has actually been measured

Given how often the cycle and the weather come up in the same sentence, the underlying research is thin — and frequently absent altogether. A scoping review of studies measuring pressure pain sensitivity in headache found that cycle phase was recorded in just 1 of 30 tension-type headache studies and 1 of 18 migraine studies. The authors' conclusion is blunt: "the phase of the menstrual cycle has been rarely considered in studies investigating sensitivity to pressure pain in primary headaches, such as TTH or migraine, although there is evidence showing the relevance of the phase of the menstrual cycle in pain perception" (Curiel-Montero et al., 2021). The heat meta-analysis above screened 367 articles to find 9 that reported the relevant data at all, across 83 participants in total (Giersch et al., 2020).

Part of this is a wider gap in who gets studied. An analysis of exercise thermoregulation research published between 2010 and 2019 — 1,407 articles, 28,030 participants — found women made up between 11.6% and 17.8% of participants, an underrepresentation the authors say "belies the growth in women active in sport and industry, and also the pressing need to expand the scientific evidence relevant to women" (Hutchins et al., 2021).

Where questions have been asked more directly, some come back flat. A meta-analysis of perceived exertion in athletes found no statistically significant difference between cycle days 1–5 and days 19–24 (Paludo et al., 2022). And when researchers surveyed 484 women in southern Saudi Arabia specifically about weather and menstrual symptom severity, their own conclusion was that "the scientific evidence in support of this connection is still limited and inconclusive" and that "further research is needed to better understand the underlying mechanisms" (Alqahtani et al., 2024).

None of that means the connection you notice is imaginary. It means it hasn't been measured well enough, in enough people, for long enough, for anyone — researcher or app — to say something general about it.

Your own record is the only data about you

Which is exactly where a personal record does something those studies can't: it describes you, not an average drawn from a few dozen participants. If you write down your symptoms, the weather that day, and where you were in your cycle, and after a few months something that looks like a pattern shows up, that pattern is real in the sense that it happened. It isn't evidence of a mechanism. It isn't something anyone else's record would necessarily show. It doesn't mean "the second half of the cycle causes this" any more than it means "humidity causes this" — it means that in your own record, these things tended to line up. That's worth noticing. It isn't worth explaining, and it isn't a rule you can extend to next month, let alone to anyone else.

Where your entries stay

Cycle entries you log in MeteoHealth stay on your device. The app doesn't sync them to its iCloud database — there's no server on our side holding this data, no account tied to it, and no third-party analytics reading it. Like other app data on your phone, they can be included in your device's own iCloud backup. If you allow it, recorded flow is additionally written into Apple Health. Machine learning runs on the device, the same as everywhere else in the app.

What the app does with it

To be precise about what's being described here: MeteoHealth doesn't measure anything about your cycle. Everything it holds about it is what you typed in. You record your own entries, and the app shows them next to that day's weather and your other symptom entries, arranged by phase — a way to see your own data side by side, not a way to learn something about your cycle you didn't already tell it.

It is not a contraceptive method, and it does not replace a consultation with your doctor. Cycle predictions are an estimate built from your own entries — not a diagnosis and not a medical conclusion.