PeriClue uses an established scientific method called n-of-1 observation to find patterns between your daily habits and your symptoms. Instead of telling you what works for the average woman, we show you what works for you.
This is the same approach used in precision medicine and clinical research. The difference is that we automate the data collection and the analysis, so you get personalized insights without running a clinical trial on yourself.
Most health advice comes from population-level studies. Researchers take hundreds or thousands of women, measure an outcome, and report the average result. That is valuable science. But it has a fundamental limitation: the average result may not apply to you.
A landmark example from perimenopause research illustrates this perfectly. The Australian Women's Midlife Years (AMY) Study (Islam et al., The Lancet Diabetes & Endocrinology, 2025; n=5,509 analysis cohort from 8,096 recruited) found that women with regular cycles but changes in menstrual flow and vasomotor symptoms had symptom severity profiles similar to women already classified as early perimenopausal. Under the standard STRAW+10 staging criteria, these women would be classified as premenopausal and told “you are not there yet.” Their lived experience told a different story.
Similarly, the Women Living Better survey (Coslov, Richardson, Woods, Menopause, 2021; n=2,406) found that women in the Late Reproductive Stage reported nearly identical symptom profiles to women in the menopausal transition, with less than a 10% difference across 54 of 61 symptoms surveyed.
Population averages missed these women entirely. An individual-level approach would not.
The n-of-1 approach treats the individual as the sole unit of observation. The concept originates in n-of-1 clinical trials, formally described by Lillie et al. (2011) in Personalized Medicine as “the ultimate strategy for individualizing medicine.” Those trials use randomized crossover designs with blinding and washout periods to test treatments in a single patient.
PeriClue uses the observational variant of this approach, not the experimental trial design. We do not randomize interventions or use blinding. Instead, we track naturally occurring variation in your daily habits and symptoms and compute within-person correlations over time. This is sometimes called a “single-subject observational design” or “idiographic analysis” in the research literature (McDonald et al., 2017).
The core principle is the same: instead of asking “does this work on average?”, we ask “does this pattern hold for this specific person?”
The observational approach involves:
Because this is observational rather than experimental, our findings are associations, not causal claims. We are explicit about this throughout the app.
Our daily check-in system uses a daily diary methodology informed by the principles of Ecological Momentary Assessment (EMA), as described in the foundational review by Shiffman, Stone, and Hufford (2008) in the Annual Review of Clinical Psychology. True EMA captures data multiple times per day at or near the moment of experience. PeriClue's evening check-in is an end-of-day diary, which reduces recall bias compared to weekly or monthly retrospective surveys but does not achieve the real-time capture that defines strict EMA.
When you report your symptoms during your evening check-in rather than trying to remember them a week later, you are generating data with substantially less recall bias than retrospective surveys.
We supplement your self-reported data with passive signals from your Apple Watch: heart rate variability, resting heart rate, sleep stages, workout intensity, and skin temperature. These continuous, objective measurements capture data that self-report alone would miss, and they are collected in real time throughout the day, which is closer to the EMA ideal.
For each factor-symptom pair (for example: sleep efficiency and brain fog), PeriClue computes a correlation over a rolling window of your recent data. We use a 90-day observation window by default, which gives us enough data points to identify meaningful patterns while staying sensitive to changes in your life.
We account for:
Not all correlations carry the same weight of scientific evidence. We classify every factor-symptom pair into one of four tiers, and we tell you which tier you are looking at:
Most of our 85+ tracked factor-symptom pairs fall into the last category. This is not a weakness. It is an honest reflection of the state of the science. The daily-granularity, individual-level correlation methodology we use is new. No published study has validated it for perimenopause specifically because the technology to collect this data at scale did not exist until recently. We are transparent about that rather than overclaiming.
Three things distinguish PeriClue's approach from generic “wellness tracking”:
1. Established methodology, new application. Single-subject observational designs and daily diary methods are established research approaches with decades of use in behavioral and clinical science. We are applying them to a domain (daily perimenopause symptom tracking) where they have not been applied before, but the methods themselves are not novel or experimental.
2. Conservative claims. We never tell you “X causes Y.” We tell you “in your data, X is associated with Y.” We show you the strength of the correlation, the evidence tier, and the number of days of data behind it. You and your healthcare provider make the clinical decisions.
3. Symptom monitoring itself may be beneficial. A preliminary systematic review (Andrews et al., Frontiers in Global Women's Health, 2021) found promising early evidence that symptom monitoring may improve patient-doctor communication and treatment adherence in menopausal women, though the authors note the evidence base remains limited and included studies carry a high risk of bias. Even before our correlation engine finds a pattern, the act of consistent tracking appears to generate value by giving women and their providers a shared, objective record to work from.
Women with Polycystic Ovary Syndrome present a unique challenge. As women with PCOS age past 35, their cycles often become more regular, not less, due to declining androgen levels (Shah & Rasool, 2021). This means a PCOS woman arriving at perimenopause may have more regular cycles than she had at 25, masking the onset of transition.
Additionally, women with PCOS may experience fewer climacteric symptoms during perimenopause due to persistent androgenicity (Schmidt et al., 2011), but report more urogenital symptoms and vaginal dryness (Sharma & Mahajan, 2021). Our correlation engine accounts for this by using PCOS-adjusted thresholds: doubled cycle variability windows (14 days instead of 7) and extended amenorrhea thresholds (90 days instead of 60), tracking changes from your PCOS baseline rather than absolute cycle metrics.
Both the AMY Study (2025) and the Women Living Better survey (2021) confirm that significant perimenopause symptoms can emerge while cycles remain regular. Our multi-factor classifier can identify early perimenopause based on symptom burden even when cycle data alone would classify you as premenopausal. This is a deliberate and documented extension beyond the standard STRAW+10 staging framework, supported by the most recent research.
Last updated: March 2026