This study adds observational evidence to the broader research on GLP-1 treatment for PCOS by comparing women prescribed the same drug. Participants reported their PCOS status and logged their own weights. The study did not assess whether tirzepatide improved ovulation, hormone levels, or other PCOS symptoms.[1]
Weight change among women who reached 10 months
At month 10, the 40 women with self-reported PCOS who reached that point on treatment lost a mean 19.40% of their starting weight (95% confidence interval, 16.42% to 22.38%). The 397 women without self-reported PCOS lost a mean 18.74% (95% confidence interval, 17.81% to 19.67%). The difference was not statistically significant (p = 0.68). That comparison leaves room for a true difference between the groups.[1]
Why the follow-up group was smaller
The study included women who began treatment between February 2024 and January 2025, and the average follow-up was 2.28 months. Later starters could run out of observation time. Others stopped treatment. Table 2 reports who contributed at each visit, and supplemental figures show follow-up and censoring in both groups.[1]
What the 57.66% figure measures
The authors used Kaplan–Meier analysis to estimate the chance of reaching at least 20% weight loss by month 10. For the self-reported PCOS group, the estimate was 57.66% (95% confidence interval, 47.44% to 68.29%). The analysis used the timing of recorded weight loss milestones and accounted for when follow-up ended. It drew from participants as they contributed follow-up, beyond the 40 women in the month-10 mean. This estimate depends on the assumptions used to account for incomplete follow-up.[1]
The digital engagement finding
Among women with self-reported PCOS, digital engagement was associated with 3.42 percentage points more weight loss by month 10 in an adjusted analysis (95% confidence interval, 1.71 to 5.12 percentage points, p < 0.001). Engagement included a coaching session, weekly weight tracking on average, and use of the app’s information materials. When weighing evidence on GLP-1 coaching, consider that participants chose whether to engage and that motivation or treatment adherence may partly explain the difference.[1]
Funding and author disclosures
The paper reported no relevant funding. Four authors worked for Menwell Ltd, which operates Voy, the UK service that supplied the data. The fifth disclosed consulting or speaking fees from several drugmakers, including Eli Lilly, and grants paid to his university outside this study.[1]