Fact-check: UCSF grand rounds on anti-obesity medications, beyond the scale

What the video covers

Dr. Diana Thiara, medical director of the UCSF Weight Management Clinic and program director of its obesity medicine fellowship, gives Department of Medicine grand rounds. She sets goals in percentages rather than BMI targets, reviews lifestyle components including protein and resistance training, then goes through the FDA-approved medications in order of approval with one-year and two-year figures: orlistat, naltrexone-bupropion, liraglutide 3 mg, phentermine-topiramate, semaglutide 2.4 mg and tirzepatide. She covers GLP-1 pharmacology, side-effect rates, contraception, the proportion of large responders, and subgroup analyses including the stronger response in women in the SURMOUNT trials.

What it gets right

The percentage thresholds are the standard clinical framing: 5 to 10 percent for diabetes prevention and early comorbidity improvement, 10 to 15 percent for cardiovascular and sleep-apnoea benefit, above 15 percent for remission of some conditions. The medication history and figures are consistent with the labels and trials: orlistat about 3 percent, naltrexone-bupropion 5 to 8 percent, liraglutide 5 to 10 percent, phentermine-topiramate 9 to 14 percent. The large-responder figures are correct: about a third of the STEP 1 semaglutide group lost 20 percent or more (32 percent; PubMed 33567185), and 36 percent of the SURMOUNT-1 15 mg group lost 25 percent or more (PubMed 35658024). The side-effect account, near-universal early nausea that should settle before each escalation, pancreatitis and cholecystitis under 1 percent, and suicidal ideation not seen in trials but monitored after post-marketing reports, matches the label wording. The oral contraceptive caution for tirzepatide is in the label.

The sex-difference observation is real: across SURMOUNT-1 to 4, women lost more than men at each dose, with the largest gap in SURMOUNT-3 after a lifestyle run-in. She correctly calls it a subgroup finding.

What it leaves out or overstates

Two numbers need their denominators. "Wegovy 12.5 percent at one year" is close to the placebo-subtracted difference in STEP 1 (12.4 points); the semaglutide arm itself lost 14.9 percent. "Zepbound 17.8 percent" is the placebo-subtracted figure for 15 mg in SURMOUNT-1; the arm lost 20.9 percent. Both ways of stating results are legitimate, but a listener comparing her figures with other videos in this library will see a mismatch unless they know which is which. "About 85 percent of participants" reporting nausea is higher than the label's 44 percent for Wegovy and 25 to 29 percent for Zepbound; the higher number may include any GI event (74 percent in STEP 1).

The sex-difference analysis is post hoc and does not change dosing or indication. Nothing here is a reason to expect a particular result as an individual.

Where to go next

The Stanford pharmacist's webinar earlier in this playlist is the diabetes-side companion. For the two-year data in one table, see the agent comparison on GLP-1s Explained, and for trial digests with confidence intervals, FormBlends Research. Protein and resistance training during loss are covered on Weight Loss on GLP-1s. FormBlends' compounded products at formblends.com are not FDA approved and were not studied in any of these trials.