The GLP-1 receptor agonist class has changed obesity medicine. The marketing has caught up faster than the nuance. This piece is about the nuance.
How much weight do GLP-1 trials actually show?
The headline numbers are accurate. STEP 1 (semaglutide 2.4 mg/week, 68 weeks, n=1,961) delivered mean weight loss of 14.9% vs 2.4% for placebo (p=0.001) ( Wilding et al. (STEP 1) 2021, n=1961 ). SURMOUNT-1 (tirzepatide 15 mg/week) went further: ~22.5% mean weight loss at 72 weeks. The cardiovascular arm of the semaglutide program (SELECT Lincoff et al. (SELECT) 2023, n=17604 ) found a 20% relative risk reduction in MACE among overweight/obese patients with cardiovascular disease.
The less-discussed findings:
- Lean mass loss. DXA substudies in STEP and SURMOUNT show roughly 30-40% of the lost body mass is lean tissue when no resistance-training arm is added. This is comparable to what happens with aggressive caloric restriction alone; GLP-1s don't make it worse, but they don't fix it either.
- Regain on cessation. STEP 4 extension Rubino et al. (STEP 4) 2021, n=803 : stopping semaglutide produced regain of approximately two-thirds of the lost weight within ~52 weeks. GLP-1s are closer to statins than to antibiotics in duration-of-use logic.
- Side effects. Nausea, constipation, diarrhea, occasional pancreatitis. Most GI issues peak during titration and remit.
GLP-1 and dual-agonist trials: weight, maintenance, and cardiovascular outcomes
| Study | N | Duration | Design | Outcome | Finding |
|---|---|---|---|---|---|
| STEP 1 (Wilding 2021) cite | 1,961 | 68 wk | double-blind RCT | Mean body weight change | Semaglutide 2.4 mg: 14.9% loss vs 2.4% placebo (p=0.001) |
| STEP 4 (Rubino 2021) cite | 803 | 68 wk total (20-wk run-in then 48-wk randomized) | double-blind withdrawal RCT | Weight maintenance after run-in | Continuing semaglutide kept losing; switching to placebo regained ~two-thirds of the lost weight |
| SURMOUNT-1 (Jastreboff 2022) cite | 2,539 | 72 wk | double-blind RCT | Mean body weight change | Tirzepatide 15 mg: ~22.5% loss vs ~2.4% placebo |
| SELECT (Lincoff 2023) cite | 17,604 | median 39.8 mo | double-blind RCT (CV outcomes) | MACE composite | Semaglutide 2.4 mg: HR 0.80 (p<0.001) in overweight/obese without diabetes |
Synthesis Across four pivotal trials, semaglutide and tirzepatide produce 15-22% mean weight loss at 68-72 weeks, the magnitude scales with dose and dual-agonism, and the cardiovascular benefit in SELECT confirms outcomes follow weight on a multi-year horizon. STEP 4 anchors the durability question: the loss persists only while the drug continues.
How do you protect lean mass on a GLP-1?
Three levers, all non-negotiable if body composition matters to you:
| Phase | Dose | Frequency | Notes |
|---|---|---|---|
| Protein | 1.6-2.2 g/kg goal bodyweight | daily, split across 3-4 meals | Goal bodyweight, not current. Morton 2018 shows diminishing returns above ~1.6. |
| Resistance training | 8-12 working sets per muscle group | 2-3 sessions/week | Full-body splits easier to adhere to; emphasize compounds. |
| Creatine | 5 g/day | daily, any timing | Cheap lean-mass insurance. See the creatine deep-dive. |
| Step count | 7-10k/day floor | daily | Cheapest NEAT lever. GLP-1 appetite suppression often masks spontaneous activity drop. |
The protein target deserves a number, not a vibe: Morton 2018 meta-analysis (n=1,863, 49 studies) found resistance training + protein supplementation plateaued at ~1.6 g/kg/day, with diminishing returns thereafter ( Morton et al. 2017, n=1863 ). Stuart Phillips argues the real-world target should be closer to 2.2 g/kg to account for ageing and anabolic resistance ( Phillips et al. 2016 ).
How should you titrate GLP-1 dosing in practice?
Standard titration (semaglutide, adult): 0.25 mg/week for 4 weeks, 0.5 mg/week for 4 weeks, 1 mg/week for 4 weeks, 1.7 mg/week for 4 weeks, then 2.4 mg/week maintenance. Slow down if GI side effects are severe. See the semaglutide compound entry for half-life and source links.
The counter-view
Some clinicians (David Ludwig, Gary Taubes camps) argue the whole obesity-pharma approach misses the point: the underlying metabolic dysregulation isn't fixed, just masked by appetite suppression. They have a fair critique about the regain-on-cessation data. The empirical counter: many people can't achieve or sustain large weight loss through diet alone, and for them the comparison isn't "GLP-1 vs fixing metabolism", it's "GLP-1 vs continued obesity".
What we'd do differently if we were you
If you're starting semaglutide or tirzepatide, wire the resistance-training and protein protocol before your first injection, not after. The lean-mass loss happens regardless; the difference is how much of it returns when you stop the drug.