Modeling menstrual cycle phases from urinary hormone trajectories using an autoregressive hidden Markov model
4 affiliations
- 1School of Nursing, Duquesne University, Pittsburgh, Pennsylvania, United States of America
- 2Center for Fertility Awareness Education and Research, Nasuti College of Osteopathic Medicine, Duquesne University, Pennsylvania, United States of America
- 3Department of Family Medicine, University of Calgary, Calgary, Canada
- 4Department of Mathematics and Computer Science, Duquesne University, Pittsburgh, Pennsylvania, United States of America
Abstract
The recent availability of at-home menstrual cycle tracking technology has created opportunities for personalized assessment of reproductive health, alongside improved characterization of hormone patterns in women with and without reproductive disorders such as polyendocrine metabolic ovarian syndrome (PMOS), which affects approximately 10% of reproductive-age women. In this study, we leverage self-tracked urinary hormone data to develop an autoregressive Hidden Markov model (arHMM) that maps cycle days to physiologically meaningful phases based on hormone trajectories. By modeling day-to-day hormonal dynamics rather than absolute hormone levels, and allowing variable phase durations, this approach accommodates substantial variability in menstrual cycles, thereby enabling meaningful comparisons within and between individuals. Across more than 3800 cycles from over 1100 individuals, we find that arHMM-derived phases reproduce expected hormonal patterns within follicular, periovulatory, and luteal phases, and that phase-based timing for hormone testing outperforms conventional cycle day–based testing in capturing the luteinizing hormone surge and post-ovulatory progesterone rise, highlighting limitations of fixed-day clinical protocols. We identify phase-specific differences between healthy controls and individuals with self-reported PMOS, including lower luteinizing hormone in the periovulatory phase, and reduced luteal phase progesterone levels in PMOS. Furthermore, features derived from arHMM phase assignments show the potential for non-invasive PMOS screening using machine learning classification. Author summary: Tracking menstrual cycle information has become increasingly common, thanks in part to the availability of at-home urinary hormone data tracking technologies. In this study, we leveraged data from more than 3,800 menstrual cycles contributed by over 1,100 individuals in order to develop a model that predicts physiologically meaningful cycle phases based on day-to-day hormone patterns, exploring the potential of this model to provide ovulation proxies, improve clinical testing, and screen for reproductive disorders such as PMOS, a common endocrine disorder affecting approximately 1 in 10 reproductive age women. This study demonstrates the power of physiologically informed probabilistic modeling for leveraging data from at-home hormone monitoring towards a better characterization of menstrual cycle physiology and the development of noninvasive approaches for identifying reproductive health conditions such as PMOS.
Related
- Recombinant follicle‐stimulating hormone (rFSH) versus other recombinant or urinary gonadotropins for ovarian stimulation in assisted reproductive technology cycles
- Effects of occupational noise exposure on menstrual abnormalities among female workers: a path analysis based on XGBoost and structural equation modeling
- Timing luteal urinary pregnanediol glucuronide assessment using urinary luteinizing hormone: agreement with ultrasound-determined ovulation
- Preliminary laboratory analytical comparison of quantitative at-home urinary hormone monitoring systems using artificial urine
- First-morning urinary luteinizing hormone and bone age decomposition for diagnosis of central precocious puberty in girls: a two-step prediction model study.
BibTeX
@article{stujenske2026modeling,
title = {{Modeling menstrual cycle phases from urinary hormone trajectories using an autoregressive hidden Markov model}},
author = {Theresa M. Stujenske and Thomas P. Bouchard and Alexis Troy and Sara Kelemen and Brianna Folino and Tara Wills and Lauren A. Sugden},
journal = {PLOS Digital Health},
year = {2026},
month = nov,
doi = {10.1371/journal.pdig.0001502},
url = {https://journals.plos.org/digitalhealth/article?id=10.1371/journal.pdig.0001502},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Modeling menstrual cycle phases from urinary hormone trajectories using an autoregressive hidden Markov model
AU - Theresa M. Stujenske
AU - Thomas P. Bouchard
AU - Alexis Troy
AU - Sara Kelemen
AU - Brianna Folino
AU - Tara Wills
AU - Lauren A. Sugden
PY - 2026
DA - 2026/11/05
JO - PLOS Digital Health
DO - 10.1371/journal.pdig.0001502
UR - https://journals.plos.org/digitalhealth/article?id=10.1371/journal.pdig.0001502
AB - The recent availability of at-home menstrual cycle tracking technology has created opportunities for personalized assessment of reproductive health, alongside improved characterization of hormone patterns in women with and without reproductive disorders such as polyendocrine metabolic ovarian syndrome (PMOS), which affects approximately 10% of reproductive-age women. In this study, we leverage self-tracked urinary hormone data to develop an autoregressive Hidden Markov model (arHMM) that maps cycle days to physiologically meaningful phases based on hormone trajectories. By modeling day-to-day hormonal dynamics rather than absolute hormone levels, and allowing variable phase durations, this approach accommodates substantial variability in menstrual cycles, thereby enabling meaningful comparisons within and between individuals. Across more than 3800 cycles from over 1100 individuals, we find that arHMM-derived phases reproduce expected hormonal patterns within follicular, periovulatory, and luteal phases, and that phase-based timing for hormone testing outperforms conventional cycle day–based testing in capturing the luteinizing hormone surge and post-ovulatory progesterone rise, highlighting limitations of fixed-day clinical protocols. We identify phase-specific differences between healthy controls and individuals with self-reported PMOS, including lower luteinizing hormone in the periovulatory phase, and reduced luteal phase progesterone levels in PMOS. Furthermore, features derived from arHMM phase assignments show the potential for non-invasive PMOS screening using machine learning classification. Author summary: Tracking menstrual cycle information has become increasingly common, thanks in part to the availability of at-home urinary hormone data tracking technologies. In this study, we leveraged data from more than 3,800 menstrual cycles contributed by over 1,100 individuals in order to develop a model that predicts physiologically meaningful cycle phases based on day-to-day hormone patterns, exploring the potential of this model to provide ovulation proxies, improve clinical testing, and screen for reproductive disorders such as PMOS, a common endocrine disorder affecting approximately 1 in 10 reproductive age women. This study demonstrates the power of physiologically informed probabilistic modeling for leveraging data from at-home hormone monitoring towards a better characterization of menstrual cycle physiology and the development of noninvasive approaches for identifying reproductive health conditions such as PMOS.
SN - 2767-3170
ER -