Characterizing physiological and symptomatic variation in menstrual cycles using self-tracked mobile-health data

Area of Science:

• Women's health
• Digital health
• Reproductive health

Background:

• Menstrual cycle tracking traditionally relied on surveys.
• Menstrual tracking mobile apps offer rich, longitudinal data on women's health.
• User-tracked data from apps like Clue provide insights into menstrual health experiences.

Purpose of the Study:

• To analyze a large dataset of self-tracked menstrual cycle data.
• To investigate the relationship between menstrual cycle length variability and reported symptoms.
• To explore the potential of digital health data for understanding women's health.

Main Methods:

• Utilized a database of over 378,000 users and 4.9 million cycles from the Clue app.
• Developed a procedure to exclude cycles with low user engagement to mitigate tracking artifacts.
• Analyzed statistical relationships between cycle length variability and self-reported symptoms.

Main Results:

• Self-reported menstrual tracker data show statistically significant links between cycle length variability and symptoms.
• Women with different menstrual variability patterns exhibit distinct cycle characteristics and symptom tracking.
• Cycle and period length statistics remained stable over time across the variability spectrum.

Conclusions:

• Longitudinal, high-resolution self-tracked data can enhance understanding of menstruation and women's health.
• Identified symptoms associated with timing data may help predict cycle variability and serve as health indicators.
• Findings highlight the potential of digital health tools for clinical insights and user awareness.