Digital biomarkers for women's health
The newest area in this library, and the one where the gap between consumer tracking apps and clinical research measurement is widest.
Women's health is the newest area in this library and the one where the distance between consumer products and clinical research measurement is widest. Cycle tracking is among the most used categories in consumer health apps, yet the measures that reach registered clinical trials are far fewer and far more narrowly defined.
The measures collected here are physiological rather than self reported: cycle length and its variability derived from logged menses, the small nocturnal wrist temperature shift that follows ovulation, and the vasomotor events of the menopause transition detected from skin conductance and temperature. What they share is that a sensor observes something a diary can only estimate in retrospect.
Two boundaries define this category. It is written for research use, not for contraception or conception decisions, where the regulatory bar is entirely different and consumer apps have repeatedly overstated what their algorithms support. And it excludes population slices dressed as measures: sleep disruption during menopause is measured by total sleep time and wake after sleep onset, which already have their own pages, so adding a menopause specific sleep entry would duplicate rather than add.
Women's health measures in this library
How these measures are used
The most defensible research use is methodological. Any wearable study enrolling premenopausal participants is measuring people whose resting heart rate, heart rate variability, skin temperature and sleep shift with cycle phase. Treating that variation as noise makes a study less sensitive; recording cycle phase as a covariate makes it more so. That argument holds regardless of whether the study is about women's health at all.
The second use is as an outcome in its own right, in trials of treatments for menstrual disorders, endometriosis, polycystic ovary syndrome and menopausal symptoms, where the burden is episodic and poorly captured by a clinic visit. Objective vasomotor event counting is the clearest example: it separates how many events occurred from how distressing they were, which a symptom diary conflates.
These measures have no questionnaire counterparts in this library, which is unusual and is stated on each page rather than glossed over.
What the evidence supports today
Evidence here is genuinely early, and this category says so. Very large app based datasets have characterised real world cycle length and variability at a scale clinical studies never reached, which is a real contribution, but those data are self reported and skew towards users motivated enough to log consistently.
The nocturnal wrist temperature shift after ovulation is physiologically well grounded and has been demonstrated against reference methods in research cohorts. Accuracy at the level of an individual cycle is lower than the population level signal suggests, and it is affected by alcohol, illness, room temperature and shift work.
Objective hot flash measurement using skin conductance has a research literature going back decades and a documented divergence from self report, but no wearable vasomotor measure holds a regulatory qualification as a trial endpoint.
No measure in this category is cleared for contraception, conception or diagnosis, and none should be presented as if it were.
Common questions
Are these measures for fertility tracking?
No. They are listed for research use: as covariates in wearable studies, and as outcomes in trials of menstrual and menopausal conditions. Using wearable temperature to make contraception or conception decisions is a different claim with a different regulatory bar, and this library does not support it.
Why does cycle phase matter in a study that is not about women's health?
Because it moves the other measurements. Resting heart rate, heart rate variability, skin temperature and sleep all shift across the cycle in premenopausal participants. A study that ignores that treats real physiological variation as noise, which reduces its power to detect whatever it is actually looking for.
How accurate is wrist temperature for detecting ovulation?
The population level pattern is clear and has been shown against reference methods, but single cycle accuracy in an individual is lower, and the signal is disturbed by alcohol, illness, ambient temperature and disrupted sleep. It is a research signal, not a decision tool.
Why are there no questionnaires linked to these measures?
Because this library does not currently contain the relevant instruments. Menopause symptom scales and cycle specific instruments are a recognised gap, recorded as a separate work item. Where a measure has no counterpart, the page says so rather than substituting a loosely related scale.
Run a study on these measures
WeGuide captures wearable data and patient reported outcomes in one platform, from screening through to analysis.
