Simple AI models on wearables predict prolonged sitting in women with chronic pelvic pain

Key points

• A team at the Icahn School of Medicine at Mount Sinai created an AI forecasting model to predict 15-minute periods of sedentary behavior in women with chronic pelvic pain disorders.
• The study analyzed data from 134 women with chronic pelvic pain and 61 healthy controls who wore Fitbit devices for up to 90 days.
• Simple, interpretable models achieved accuracy comparable to more computationally intensive deep-learning methods, allowing for on-device processing.
• The approach aims to create 'just-in-time' interventions that prompt users to take short movement breaks before prolonged inactivity begins, reducing unnecessary alerts.
• The findings were published in the September 30, 2026 issue of npj Women's Health, with funding from the National Institutes of Health and the National Center for Advancing Translational Sciences.

Background

This development follows a broader trend in AI-powered wearables, as seen at IFA 2026 where devices expanded into pendants and keychains, and Meta's recent launch of the Muse Charm. It also aligns with ongoing efforts to address chronic pelvic pain, such as the case of a Fort Wayne server diagnosed with pelvic congestion syndrome after years of undiagnosed pain. The current study builds on the need for personalized digital health tools that account for the realities of living with chronic conditions, moving beyond generic advice to sit less and move more.

Why it matters

The ability to forecast sedentary behavior using lightweight AI on personal devices offers a privacy-preserving solution for managing chronic pelvic pain. By delivering timely, personalized prompts for movement, these tools could improve symptom management and quality of life for women affected by conditions like endometriosis, adenomyosis, and uterine fibroids. The success of simple models suggests that practical, real-world deployment is feasible without requiring complex infrastructure or remote data transmission.

What to watch

The research team is now integrating the forecasting framework into a just-in-time adaptive intervention to test whether personalized movement prompts reduce sedentary time and improve symptoms. Prospective clinical trials will be required to determine the effectiveness of these interventions in real-world settings. The approach may also be applied to other chronic conditions where prolonged sitting contributes to poorer health outcomes.