Last post, I explained what GLP-1 drugs actually do and where they came from. Short version: they mimic a gut hormone, turn down hunger signals in your brain at a neurological level, and were originally built for type 2 diabetes. The science is real and the cardiovascular data is legitimately impressive.
Now for the part of the research that isn’t on the Wegovy website.
I want to be clear: this isn’t a takedown piece. I’m not a doctor. I’m a guy who’s been doing bedtime yoga for 40 days and lost 18 pounds the slow way, and I got curious. What I found is that the clinical trial data is real and the benefits are real, but so is the fine print. And nobody’s reading the fine print.
The Muscle Problem #
Let’s start with the number that stopped me cold.
In the STEP 1 trial, the landmark semaglutide weight-loss study, participants lost an average of 15.2 kilograms (33.5 lbs) over 68 weeks. That sounds great. Here’s the part that got buried: 6.92 kg of that was lean mass.
Do the math: 45.5% of the weight lost was muscle, not fat.
Now, to be fair, every calorie restriction method causes some lean mass loss. Traditional diet-only approaches typically lose 20-30% of total weight as muscle. So semaglutide is somewhat worse on this measure, not dramatically different, but worse.
Tirzepatide (Mounjaro, Zepbound) is meaningfully better on this metric. The SURMOUNT-1 body composition substudy showed approximately 26% of weight lost was lean mass, which is closer to what you’d see with diet alone.
A 2024 network meta-analysis covering 22 randomized controlled trials and 2,258 participants found that GLP-1 drugs reduced lean mass by an average of 0.86 kg, and that semaglutide and tirzepatide at their highest doses were “among the least effective in preserving lean mass” despite being most effective for fat loss overall.
The mitigation strategy is clear in the research: 1.2 to 1.6 grams of protein per kilogram of body weight per day, plus resistance training 2-3 times per week. Which, great, but that’s the same advice I’m following right now without the drug. The difference is that on GLP-1s, it’s not a bonus recommendation, it’s a damage mitigation requirement.
You’re Probably On It Forever #
This is the part I think deserves the most attention, because it changes the whole calculus.
The STEP 1 trial didn’t just measure weight loss on the drug. Researchers followed participants for one year after they stopped. Here’s what happened: they regained two-thirds of their lost weight within 12 months.
Not a little regain. Not some expected bounce. Two thirds.
A 2025 meta-analysis pooled data across multiple GLP-1 withdrawal studies and found a mean weight regain of 5.63 kg after stopping treatment, along with reversal of HbA1c improvements, blood pressure gains, and BMI reduction. Everything came back.
The SURMOUNT-4 trial (tirzepatide) found more than 50% of weight loss rebounded within 52 weeks of stopping. The STEP-10 trial found similar results with semaglutide at 28 weeks.
The reason is straightforward once you understand the mechanism. The drug suppresses hunger signals. When you stop taking the drug, the hunger signals come back. The underlying biology of obesity, the dysregulated hormones, the elevated set points, none of that changed. The drug treated symptoms, not causes.
The researchers who ran the STEP 1 extension said it plainly: “Findings confirm the chronicity of obesity and suggest ongoing treatment is required.”
Ongoing treatment. As in, indefinitely. As in, this is a chronic medication, not a course of treatment.
That’s not a disqualifier on its own. Plenty of chronic conditions require long-term medication. But it is information that I think a lot of people signing up for Ozempic through a telehealth app don’t have when they sign up.
The Bone Density Problem #
This one is newer and less discussed.
Multiple studies now show that significant weight loss on semaglutide is associated with declining bone mineral density (BMD). A 2025 study found patients on semaglutide and tirzepatide showed:
- Lumbar spine BMD down 1.6%
- Femoral neck BMD down 1.8%
- Total hip BMD down 2.8%
And the hip decline was significantly correlated with total weight lost. The more weight you lost, the more bone density you lost.
Here’s the nuance: researchers believe this is driven primarily by the weight loss itself, not a direct drug effect on bone. When you lose a lot of weight, your skeleton carries less load, and bone adapts by becoming less dense. This happens with any significant weight loss method.
But with GLP-1s producing 15-20% body weight reductions, “any weight loss method” isn’t doing the same work as it was when the guideline was “lose 1-2 lbs per week.” A tirzepatide study found it was associated with higher osteoporosis risk than other GLP-1 drugs, with a hazard ratio of 1.44.
Resistance training, again, helps. The bones respond to load. Which is, again, great advice, but it’s starting to feel like the solution to every GLP-1 side effect is “do the thing you would have needed to do anyway.”
The Risks That Are Real But Small #
Pancreatitis: A meta-analysis of 62 studies covering over 66,000 patients found a significantly elevated risk of pancreatitis with GLP-1 use (relative risk 1.44). That’s real. It’s also a small absolute risk increase. The standard protocol is to stop immediately if pancreatitis is suspected.
Thyroid cancer: This is the contested one. Rodent studies found thyroid C-cell tumors dose-dependently. That’s why there’s a black box warning for anyone with a personal or family history of medullary thyroid carcinoma. In humans, the data conflicts: a French study found increased thyroid cancer risk, a 2024 multinational study of 145,000 patients found no significant increase, and a 2025 Danish study found no significant change. No causal relationship has been established in humans, but it’s unresolved enough that the contraindication for high-risk patients is appropriate.
Pancreatic cancer: No significant association found in the meta-analysis data.
What’s genuinely unknown: There is no long-term safety data past 10 years for semaglutide or tirzepatide. The drugs are too new. That’s not a condemnation, it’s a fact. Any drug that’s been widely used for less than a decade has a gap in its long-term safety record.
The 50% Dropout Rate #
Something that got lost in the conversation: roughly half of people who start GLP-1 drugs quit within the first year. The primary drivers are gastrointestinal side effects, cost, and insurance coverage.
More than 30% drop out during the first four weeks of dose titration alone, when nausea and vomiting are at their worst.
A 2025 survey of over 2,000 primary care physicians found that 56% cited overprescribing as their top concern with telehealth-driven GLP-1 prescriptions. Fifty percent cited lack of continuity of care. Fifty-seven percent said they caution patients against accessing GLP-1 treatment through third-party telehealth platforms.
That’s not a fringe physician position. That’s a majority of primary care doctors saying the way these drugs are being distributed right now doesn’t match how they should be used.
So What Do I Think? #
I think these are real drugs with real benefits and real risks, and the conversation around them is being conducted almost entirely in bad faith from both sides. The “it’s a miracle” camp skips the muscle loss and the permanence. The “it’s cheating” camp skips the cardiovascular data and the biology of obesity.
The honest picture is in the middle. These drugs work, sometimes dramatically, for the people they were designed for. They also come with trade-offs that deserve to be understood before someone hands over a prescription through a website.
I’m doing this without them. That’s a choice I made, and it’s working. Forty days on the streak. Eighteen pounds down. The boring daily way.
But I’m not going to pretend that choice makes me morally superior to someone dealing with clinical obesity and cardiovascular risk who needs pharmacological help. I’m just going to tell you what the data says.
Part three is about who the data says actually needs these drugs, and what the research says about doing it without them.
Sources #
- STEP 1 body composition substudy (Journal of the Endocrine Society, 2021)
- SURMOUNT-1 body composition analysis (Diabetes, Obesity and Metabolism, 2025)
- GLP-1 effect on body composition: network meta-analysis of 22 RCTs (Metabolism, 2024)
- STEP 1 extension: weight regain after stopping semaglutide (Diabetes, Obesity and Metabolism, 2022)
- Metabolic rebound after GLP-1RA discontinuation: meta-analysis (eClinicalMedicine/Lancet, 2025)
- SURMOUNT-4: weight regain after tirzepatide withdrawal (JAMA, 2024)
- Bone mineral density changes on semaglutide/tirzepatide (Endocrine Society, 2025)
- Tirzepatide and osteoporosis risk (Diabetes Research and Clinical Practice, 2025)
- Pancreatitis risk with GLP-1 agonists: meta-analysis of 62 studies (Endocrinology, Diabetes & Metabolism, 2025)
- GLP-1 agonists and thyroid cancer risk (Diabetes Care, 2023)
- GLP-1 agonists and thyroid cancer: Scandinavian cohort of 145,000 patients (BMJ, 2024)
- PCP survey on telehealth GLP-1 prescribing concerns (Omada Health, 2025)