Fact-check: Harvard's teaching video on GLP-1 agonists for weight loss

What the video covers

Dr. Lee-Shing Chang, an endocrinologist at Brigham and Women's Hospital, sets out the trials that established GLP-1 receptor agonists for weight management. He covers GLP-1 physiology briefly, the five drugs with a long-term weight indication as of 2021, then SCALE for liraglutide, STEP 1 and STEP 8 for semaglutide, and STEP 4 for what happens when semaglutide is stopped.

What it gets right

Every number we could check is correct. SCALE randomised 3,731 adults without diabetes to liraglutide 3.0 mg daily or placebo; at 56 weeks the liraglutide group lost a mean 8.0 percent (8.4 kg) against 2.6 percent (2.8 kg) on placebo (PubMed 26132939). STEP 1 randomised 1,961 adults to semaglutide 2.4 mg weekly or placebo; at 68 weeks the change was 14.9 percent against 2.4 percent, an estimated treatment difference of 12.4 percentage points, with 86 percent versus 32 percent losing at least 5 percent and 51 percent versus 5 percent losing at least 15 percent (PubMed 33567185). STEP 8 randomised 338 participants and found 15.8 percent with semaglutide against 6.4 percent with liraglutide, a 9.4-point difference (PubMed 35015037).

The STEP 4 description is also faithful. After a 20-week run-in on semaglutide, those switched to placebo regained weight while those who continued kept losing; the published figures are a further 7.9 percent loss on continued semaglutide against a 6.9 percent regain on placebo between weeks 20 and 68 (PubMed 33755728). The video draws the correct conclusion: the drug has to be continued to keep the effect.

The pharmacology is right too. Liraglutide and semaglutide share high homology with native GLP-1 and are modified to bind albumin, which extends the half-life; both started as glucose-lowering drugs and were later studied in people without diabetes.

What it leaves out or overstates

Nothing is overstated. What is missing is what did not exist yet. Tirzepatide's SURMOUNT-1 result, a mean 20.9 percent reduction at 15 mg over 72 weeks against 3.1 percent on placebo (PubMed 35658024), is absent, as is its November 2023 approval for weight management. SELECT, the cardiovascular outcomes trial, had not reported. Adverse events are mentioned only in passing; the STEP 1 paper reports gastrointestinal events in 74.2 percent of the semaglutide group versus 47.9 percent on placebo, and discontinuation for adverse events in 7.0 percent versus 3.1 percent. A viewer weighing the decision needs those figures as much as the efficacy ones.

Where to go next

The companion Harvard video on prescribing, contraindications and titration is next in the semaglutide playlist. For the trial digests with confidence intervals, see FormBlends Research. For a dated titration plan from the label, the Semaglutide Hub has a planner. FormBlends itself offers compounded semaglutide at formblends.com/products/semaglutide; compounded products are not FDA approved and were not the products studied in these trials.