TOPLINE

Although online GLP-1 users' reports of their medication largely matched photos of the drugs they uploaded, more than 65% of those using compounded semaglutide or tirzepatide said they took a brand name version, suggesting that self-reports alone may not reliably differentiate between compounded products and FDA-approved products.

METHODOLOGY

• Compounded semaglutide and tirzepatide are easy to miss in research because many users get them through telehealth platforms and spas, outside the health records and claims data that studies usually rely on.
• To test whether medication photos can verify GLP-1 use in online surveys, researchers recruited US adults through an online research panel between October 2025 and January 2026. All reported at least 3 months of once-weekly semaglutide or tirzepatide use. The final analysis included 322 individuals (mean age, 45.06 years).
• Participants first completed a screening survey that required a photo of their medication or its label or prescription records. Those who qualified then completed a main survey on the GLP-1 they used and their demographic and health characteristics.
• Two trained coders, blinded to what participants reported, classified each medication type based on its photo into one of four brand name or compounded products; the compounded version was classified as so when the photos showed semaglutide or tirzepatide as active ingredients and no brand name was included. An independent adjudicator, also blinded, resolved disagreements.
• Photo classifications were compared with each participant's reported drug. Because everyone had already reported taking semaglutide or tirzepatide during screening, a "none of the above" answer on the main survey was counted as the use of compounded medication. The drugs were then regrouped as brand names and compounded to test how well self-report captured the use of compounded medication.

TAKEAWAY

• Before adjudication, the two coders' classifications were nearly identical, differing in terms of only five photos.
• Based on photos, 83 participants (25.8%) were deemed to be using compounded semaglutide or tirzepatide; the rest had documentation showing a brand name.
• Self-reports matched photos for 266 of 322 participants (82.6%), but 54 of the 56 mismatches involved compounded medication users. Of the 83 photo-coded compounded medication users, 54 (65%) reported a brand name drug, and the other 29 chose "none of the above."
• When drugs were grouped simply as brand names or compounded, self-reports and photos agreed only moderately. Self-reports identified just 34.9% of compounded medication users (29 of 83). It never labelled a brand name user as compounded, and 81.6% of those who reported a brand name drug were confirmed by photo (239 of 293).

IN PRACTICE

"Review procedures are increasingly important where medication formulation is clinically meaningful, particularly in the United States, where compounded semaglutide‐ or tirzepatide‐containing products remain widely accessible outside traditional health systems. Notably, photo‐coded compounded products represented more than one quarter of the present sample despite data collection occurring after semaglutide and tirzepatide shortages were resolved, emphasizing the continued importance of medication review beyond the shortage period. In this context, reliance on self‐report alone may not reliably classify medication type, potentially biasing inferences about medication effects and user experiences," the authors wrote.

SOURCE

The study was led by Jacqueline M. Katz, University of Michigan, Ann Arbor, Michigan. It was published online in Obesity Science & Practice.

LIMITATIONS

Because the sample came from a narrow, highly screened pool of experienced online research panel participants, the findings may not apply to all online samples or all GLP-1 users. The survey did not offer an explicit answer choice for "compounded" medications; therefore, some mismatches between the reported and photo-coded drugs may reflect the question rather than inaccurate reporting. As photo-based classification was not checked against dispensing records or chemical tests, it showed what the documentation displayed but could not confirm what a product actually contained.

DISCLOSURES

The study was supported by the University of Michigan Eisenberg Family Depression Center and University of Michigan Department of Psychology. Two of the study's authors disclosed having financial relationships, including receiving research grants, consulting fees, speaking honoraria, and royalties from several organizations, with one author also receiving honoraria from Medscape/WebMD.

This article was created using several editorial tools, including AI, as part of the process. Human editors reviewed this content before publication.