The broader GLP-1 research in PCOS includes studies with different comparison treatments. Replacing metformin and adding to it answer different questions.
Direct comparisons with metformin
A 2019 review of eight randomized trials reported better BMI and insulin resistance results with GLP-1 drugs than with metformin [2]. The authors rated the evidence moderate to low quality. The reported BMI effect was a standardized measure.
The review supporting the international guideline could pool only exenatide against metformin [3]. It found no clear difference in weight or fasting insulin. Fasting glucose results slightly favored metformin, with very low certainty.
Adding a GLP-1 to metformin
In a 2025 review, adding a GLP-1 drug was associated with 1.37 kg greater weight loss than metformin alone (95% CI 0.67 to 2.07 kg) [4]. The trials also found a smaller waist circumference and lower fasting glucose with the combination.
A 2026 review included 11 trials and several different comparators [1]. Adding a GLP-1 drug lowered BMI by an estimated 1.38 kg/m², with low certainty. Evidence was insufficient for conclusions about menstrual regularity or excess hair growth.
If pregnancy is a near-term goal, the specific drug's pregnancy guidance can affect the treatment plan. The 2026 review found no included trials assessing quality of life, mental health or cost-effectiveness.
Which treatment fits depends on the symptom or health risk being treated. The weight data alone do not establish which medication is better for PCOS overall.