GLP-1 side effects: the actual rates, from the actual trials

Side-effect coverage of GLP-1s runs on anecdote in both directions: miracle-drug testimonials and horror-story headlines. Base rates are the antidote. Below are the numbers from the pivotal trials and the FDA labels — what's common, what's rare-but-serious, what's still being adjudicated, and the two under-discussed issues (muscle and regain) that deserve a plan on day one.

The common stuff: your gut, mostly, and mostly early

Gastrointestinal effects dominate. In STEP 1, semaglutide 2.4 mg produced nausea in about 44% of participants, diarrhea in ~32%, vomiting in ~25%, and constipation in ~24% — versus much lower placebo rates. In SURMOUNT-1, tirzepatide's rates were somewhat lower at comparable efficacy: nausea roughly 25–31% depending on dose, diarrhea ~19–23%, vomiting ~8–13%, constipation ~12–17%. Two patterns matter more than the exact percentages. These effects cluster during dose escalation and usually fade at a stable dose; and they're the reason titration schedules exist — climbing slower is a legitimate medical strategy, not failure. Despite all that GI noise, only about 4–7% of trial participants quit the drugs over adverse events, which tells you most people found the effects manageable. Practical mitigations with decent support: smaller meals, less fat and alcohol during escalation, hydration, and telling your prescriber early rather than toughing out vomiting (dehydration is how "annoying" becomes "kidney injury").

Rare but serious: the list to actually memorize

Boxed warning — thyroid C-cell tumors. Both drugs caused thyroid C-cell tumors in rodents; human relevance is unknown, but both are contraindicated with a personal or family history of medullary thyroid carcinoma or MEN 2. This is a hard stop, and any telehealth intake that doesn't ask about it is a red flag about the whole operation.

Pancreatitis. Reported in trials and post-marketing at low rates (well under 1% in the pivotal studies). Severe, persistent abdominal pain radiating to the back means stop the drug and seek care.

Gallbladder disease. Rapid weight loss by any method raises gallstone risk; in STEP 1, cholelithiasis occurred in ~2.6% on semaglutide, and gallbladder events appear across the class. Right-upper-quadrant pain after fatty meals is the classic signal.

Gastroparesis and bowel obstruction. These drugs slow gastric emptying by design; a small number of users develop severe, persistent stomach paralysis or ileus, now reflected in labeling and litigation. Persistent vomiting weeks into a stable dose is not "normal side effects."

Anesthesia and procedures. Because food lingers in the stomach, anesthesia societies issued guidance from 2023 onward on holding GLP-1s before elective procedures under sedation. Tell every proceduralist you're on one; aspiration is the risk being managed.

Eyes, mood, and pregnancy. For semaglutide, diabetic retinopathy complications worsened early in one diabetes trial (SUSTAIN-6), and 2024–2025 studies plus a European review flagged NAION — a rare optic-nerve stroke — as a possible very-rare risk; sudden vision change warrants urgent care. On mood: after case reports, both the FDA (2024) and EMA reviewed the evidence and did not find a causal link between GLP-1s and suicidal thoughts, though labels advise monitoring. Both drugs should be stopped well before a planned pregnancy (the Wegovy label says two months); and hormonal-pill users on tirzepatide are advised to add barrier contraception around initiation and dose increases because slowed gastric emptying can reduce pill absorption.

The two quiet issues worth planning for

Muscle. In body-composition substudies, roughly a quarter to forty percent of GLP-1 weight loss came from lean mass, not fat. That ratio isn't unique to these drugs — it tracks rapid weight loss generally — but at 15–20% total loss the absolute lean-mass number gets big enough to matter for strength and metabolic health, especially past age 50. The countermeasures are unglamorous and effective: resistance training twice or more weekly and protein intake around 1.2–1.6 g/kg/day. A provider who never mentions this is selling the scale number, not the health outcome.

Regain after stopping. In STEP 1's extension, participants regained roughly two-thirds of lost weight within a year of stopping semaglutide; in SURMOUNT-4, people switched from tirzepatide to placebo regained about 14% of body weight while those who continued kept losing. Obesity behaves like the chronic condition it is. The honest framing: starting a GLP-1 is a long-horizon decision, and the exit plan (maintenance dosing, taper strategies, intensified lifestyle support) should be discussed at the start — including its budget, which is why our cost guide prices the year, not the month.

Compounded products: same molecule, wider error bars

Everything above describes FDA-approved products. Compounded versions add a separate risk layer — dosing errors from vial-and-syringe administration (the FDA logged hundreds of adverse-event reports during the compounding boom, many involving overdoses), incorrect salt forms, and variable sterility assurance — covered in our regulatory explainer. If you use a compounded product, dosing education and a pharmacy you can verify are not extras; they're the safety system.

Managing the big three, practically

Since gastrointestinal effects dominate every trial table, the management playbook deserves equal prominence. Nausea, the most common, is heavily behavior-responsive: smaller meals eaten slower, fat and grease dialed down during titration weeks, cold and bland options on rough days, hydration in sips rather than volumes, and — the highest-yield move — telling your prescriber rather than toughing it out, because holding a titration step longer is an explicitly sanctioned strategy and antiemetics exist for the stubborn cases. Constipation, the quiet second, responds to a deliberate fluid floor, fiber built up gradually, movement, and early conversation about over-the-counter options rather than week-three misery; it tends to be most prominent in the early months. Reflux and burping (the "sulfur burps" of forum fame) improve with smaller evening meals, upright time after eating, and trigger-food awareness; persistent or severe reflux belongs in the clinician conversation, particularly with the anesthesia considerations our surgery guide covers. Across all three, the pattern from the trials is consistent and genuinely reassuring: effects cluster around dose escalations and fade with time at stable doses for most people — which is why the label's flexible titration exists, and why the difference between a tolerable and an intolerable experience is so often just pace.

The symptoms that skip the tips and go to a clinician

A short, serious list, in prose and without drama. Severe, persistent abdominal pain — especially radiating to the back with vomiting — is the pancreatitis-pattern symptom the labels flag for immediate evaluation. Right-upper-abdomen pain, particularly after meals, raises the gallbladder question that rapid weight loss earns independent of drug. Inability to keep fluids down risks dehydration and kidney strain and warrants same-day contact. Vision changes in diabetics, symptoms of low blood sugar in anyone combining these drugs with insulin or sulfonylureas, signs of allergic reaction, and a personal or family history of medullary thyroid carcinoma or MEN2 (a contraindication to starting at all) round out the list. None of these is common; all of them are why "24/7 medical support" is a feature worth weighting in any program comparison, and why the cheapest program without reachable clinicians is not cheap.

The timeline most patients actually experience

Trial tables report totals; patients live sequences, and the sequence is more reassuring than the totals suggest. Weeks one through four, at starting doses, are commonly quiet or mildly queasy — the starting tiers exist precisely to be gentle. Each titration step then brings the highest-probability window for nausea and appetite-adjacent effects, typically peaking in the days after the first one or two injections at the new dose and settling as the body adapts; this is the window where the eating tactics above earn their keep and where a clinician call about holding a step longer is most valuable. By the time a stable maintenance dose has been held for a month or two, most people's side-effect burden has fallen to occasional and manageable — the trials' discontinuation-for-adverse-events figures, in the mid-single digits to high teens depending on molecule and dose, describe the minority for whom it doesn't. Constipation runs on a longer, flatter arc; injection-site reactions are usually minor and technique-responsive; and fatigue, when it appears, often tracks eating too little rather than the drug directly — worth auditing against the protein floor in our composition guide before attributing it to pharmacology. Knowing the shape of the curve changes behavior: patients who expect the escalation-week wobble ride it; patients surprised by it quit in week six of what would have been a fine year.