This site usually explains stopping with money. That explanation has a problem, and this study is the problem.
A program with no price
A multidisciplinary obesity clinic at an academic medical center ran a Medical Weight Loss Bundle at no cost to the patient, and looked back at everyone aged 18 to 75 who enrolled between January 2022 and December 2024. [1] Of 2,855 enrollees, 2,306 — 81% — filled at least one GLP-1 prescription. Median age was 46, and 85% were women. Persistence was defined mechanically: continuous fills with no gap of 84 days or more.
Median persistence came to 10.7 months, with an interquartile range of 5.4 to 16.3. A quarter of this group was gone inside six months, in a setting where nobody was billed anything and a team of clinicians was actively managing them.
Where the dose stopped
Titration is the other half. Of semaglutide users, 81% escalated to at least 1 mg — but only 23% ever reached 2.4 mg, the dose the weight-loss trials ran at. Of the small tirzepatide group, 75% reached 10 mg or more and 28% reached 15 mg.
Note what that means for a price comparison. Most people in this cohort were taking less drug than the trials used, in a program with no financial reason to hold back. A seller that charges by the dose is selling most buyers a mid-rung, and a seller that charges a flat rate is charging a mid-rung buyer for a top-rung supply. That question is worked through separately and it is where the money actually sits.
Tirzepatide was the sole agent for only 117 people here, about 5% of the cohort. Its titration numbers come from a small group and should be read that way.
The weight figures have a condition attached
Among patients persistent for at least six months, median weight loss was 9.4%, IQR 6.0 to 13.4. Among those persistent for at least twelve, it was 14.4%, IQR 9.5 to 20.5. The second figure lands close to the 15% to 21% range the trials reported, which is the study’s headline and is fair.
It is also conditional. Both numbers describe people who were still filling prescriptions at that mark, and the people who stopped at month four are not in either one. That is not a flaw in the study — it is what persistence-stratified reporting means — but it makes the figures useless as an expectation for somebody about to start. The honest version is that if you keep taking it for a year, results like the trials are plausible, and roughly half of this cohort did not keep taking it for a year.
What to do with it before buying
Price the outcome you are likely to have, not the one on the label. If median persistence in a free, clinically supported program is under eleven months, an annual prepay bought from a website is a bet against a base rate, and the term checker shows which sellers make that bet a condition of their headline figure.
Sellers here publish a median of $179 a month for semaglutide by injection, read September 2026. At 10.7 months that is what the typical course costs, which is a different and more useful number than twelve times the monthly rate. Reasons people give for stopping are worth reading alongside it — cost leads that list, and it does not finish it.