GLP-1 side effects: what is expected, what is serious, what to do
Nausea, constipation and reflux hit roughly 40 to 70 percent of patients starting a GLP-1 and resolve in most within two to three months. The serious risks (pancreatitis, gallbladder events, thyroid C-cell carcinoma signal, NAION, aspiration risk under anesthesia) are real but uncommon, and most are mitigable with titration and pre-procedure planning. We walk through each category, what the SELECT, STEP and SURMOUNT trial data shows, and when to call the prescriber.
The short version
The expected side effects are gastrointestinal: nausea (roughly 40 to 70 percent of patients), constipation (25 to 40 percent), reflux and burping (15 to 25 percent), occasional vomiting. They peak in the first two to four weeks of each dose step and resolve for most patients within two to three months of stable dosing. The serious risks (pancreatitis, gallbladder events, NAION, aspiration risk under general anesthesia, the thyroid C-cell carcinoma boxed warning) are uncommon. SELECT, STEP, SURMOUNT, FLOW and SURPASS pooled safety data is the best evidence base; cumulative patient-years on semaglutide and tirzepatide now exceeds tens of millions. The most important practical point: slow titration is the single biggest determinant of tolerability.
Key facts
| Side effect | Frequency (semaglutide) | Frequency (tirzepatide) | Source |
|---|---|---|---|
| Nausea | 44% | 33% to 39% | STEP-1 NEJM / SURMOUNT-1 NEJM |
| Diarrhea | 30% | 21% to 26% | STEP-1, SURMOUNT-1 |
| Constipation | 24% | 17% to 19% | STEP-1, SURMOUNT-1 |
| Vomiting | 24% | 9% to 13% | STEP-1, SURMOUNT-1 |
| Gallbladder events | 2.6% | 0.6% to 1.6% | STEP-1, SURMOUNT-1, SELECT |
| Acute pancreatitis | 0.2 to 0.3% | 0.2% | SELECT NEJM 2023 |
| Discontinuation due to adverse events | 7% | 6% to 7% | STEP-1, SURMOUNT-1 |
| Boxed warning, medullary thyroid carcinoma | Theoretical (rodent data) | Theoretical (rodent data) | FDA Wegovy label |
Frequencies above are intent-to-treat at full maintenance dose. Per-dose frequencies are lower in the first three months and higher in the dose-escalation weeks. Both molecules are well-tolerated at the population level; individual tolerability varies.
Gastrointestinal effects: the common ones
Nausea is the most common adverse event on every GLP-1. STEP-1 reported nausea in 44 percent of semaglutide patients versus 16 percent of placebo. SURMOUNT-1 reported nausea in 33 to 39 percent of tirzepatide patients depending on dose. The mechanism is dual: direct vagal-afferent stimulation by the GLP-1 receptor and delayed gastric emptying. Both improve with continued exposure as the receptor downregulates. Most patients describe the first 48 to 72 hours after each dose escalation as the worst. By week three on a given dose, the nausea is usually manageable. By week six, it is typically a non-issue. Persistent nausea past month four on a stable dose is unusual and warrants a clinician conversation.
Constipation is the second-most-common GI effect, in roughly 17 to 24 percent of patients. The mechanism is the same delayed-motility effect that produces the early satiety patients use to lose weight. Hydration, fiber and gentle laxatives (polyethylene glycol, magnesium) manage it for most patients. Vomiting hits about 9 to 24 percent and is more common on semaglutide than tirzepatide. Reflux and burping are less commonly trial-reported but are widely reported in patient communities; the patient-reported pattern is covered in GLP-1 side effects ranked. Mitigation through titration timing is in titration guide.
Gallbladder events
Cholelithiasis (gallstones) and cholecystitis (gallbladder inflammation) occur at elevated rates on GLP-1. STEP-1 reported gallbladder-related events in 2.6 percent of semaglutide patients versus 1.2 percent of placebo. SURMOUNT-1 reported 0.6 to 1.6 percent on tirzepatide depending on dose. The mechanism is partly the medication itself (slowed gallbladder motility) and partly the rapid weight loss the medication produces; rapid weight loss is a known independent risk factor for gallstones. Most events are managed conservatively; cholecystectomy is occasionally required. Patients with prior gallstones or symptomatic gallbladder disease are not absolutely contraindicated but should be counseled about elevated risk.
Pancreatitis
Acute pancreatitis is the most-watched serious risk. The SELECT trial, published in NEJM in November 2023, enrolled 17,604 adults and reported 16 cases of acute pancreatitis on semaglutide versus 8 on placebo over a median 39.8 months of follow-up. Absolute incidence is low (roughly 0.2 to 0.3 percent over four years); relative risk roughly doubles. The signal is consistent across STEP-1, SURMOUNT-1 and the diabetes-indication trial program (SUSTAIN, SURPASS). The labeled warning is to discontinue the medication if pancreatitis is suspected and not to restart unless an alternative cause is identified. Patients with prior pancreatitis, symptomatic gallstone disease or significant alcohol use are typically not started on GLP-1. See GLP-1 honest risks for the broader risk-benefit framing.
The thyroid C-cell carcinoma boxed warning
All GLP-1 receptor agonists carry a boxed warning for medullary thyroid carcinoma based on rodent studies in which sustained GLP-1 receptor activation produced C-cell tumors in rats. The human relevance is contested. Large observational studies have reported inconsistent signals; some find a slight increase in thyroid cancer incidence, others find none. The FDA's most recent review concluded the evidence does not establish a clear human risk but maintains the boxed warning. The labeled contraindication is personal or family history of medullary thyroid carcinoma, or multiple endocrine neoplasia syndrome type 2 (MEN 2). Outside the contraindication, GLP-1 is widely prescribed without thyroid monitoring. Reference: the Wegovy label carries the standard boxed warning text.
NAION and the eye signal
Non-arteritic anterior ischemic optic neuropathy (NAION) is sudden vision loss in one eye from disrupted blood flow to the optic nerve. A 2024 Mass Eye and Ear retrospective cohort study reported a roughly fourfold higher risk of NAION in semaglutide-treated patients compared to matched controls without semaglutide exposure. The study was a single-center retrospective cohort, not a randomized trial, and the absolute event count was small. Subsequent larger database studies have produced more mixed signals. The FDA is monitoring; no labeled contraindication has been added as of 2026. Patients with NAION risk factors (diabetes, hypertension, sleep apnea, "disc at risk" anatomy) should be counseled. Sudden vision loss in one eye on a GLP-1 is a same-day call. The full picture is in our risk overview at honest risks.
Aspiration risk and surgery
GLP-1 delays gastric emptying. Under general anesthesia, that delayed emptying creates aspiration risk even after the standard NPO fasting window. The American Society of Anesthesiologists issued guidance in June 2023 recommending GLP-1 be held for one week before elective surgery for once-weekly injectable agents (and longer for higher doses). The guidance was updated in late 2024 to allow more individualized timing based on dose, indication, and the surgical-fasting protocol; for diabetes-indication GLP-1, where holding the medication creates glycemic risk, the updated guidance favors enhanced fasting protocols (clear liquids only for 24 hours) over discontinuation. Always coordinate with the surgeon and anesthesiologist; do not stop the medication unilaterally before non-elective procedures or for diabetes indication without prescriber input. Reference: ASA 2023 guidance and the 2024 update.
Depression, anxiety and suicidality
Reports of suicidal ideation in GLP-1-treated patients in 2023 prompted FDA and EMA reviews. The FDA published its conclusion in the FDA Drug Safety Communication, finding the evidence does not establish a causal link. Subsequent cohort studies of millions of patient-years (including a Nature Medicine paper analyzing electronic health records, and a National Institutes of Health-funded analysis at PubMed) have not found a consistent increase in incident depression or suicidality on semaglutide; some have suggested a small protective effect. Patients with active major depressive disorder or recent suicidal ideation should be monitored and the medication is not contraindicated in well-managed mood disorders. SSRI interactions are covered in GLP-1 and SSRI safety. The bottom-line framing: take new-onset mood changes on GLP-1 seriously, but the population-level evidence does not support a strong causal signal.
Alcohol use disorder, the counterintuitive effect
One of the more unexpected GLP-1 effects is a reduction in alcohol craving and consumption. Patients consistently report needing or wanting less alcohol on the medication. The mechanism is reward-pathway modulation in the ventral tegmental area and nucleus accumbens, the same circuitry GLP-1 modulates in food reward. Multiple randomized trials of GLP-1 in alcohol use disorder are ongoing in 2026, with the Mass General Brigham AUD trial program the largest. There is no formal AUD indication yet. Heavy drinking on a GLP-1 worsens nausea and may increase pancreatitis risk; most clinicians advise moderation. Deeper treatment in semaglutide and AUD.
Muscle loss and lean mass
All caloric-deficit weight loss involves some lean mass loss, and GLP-1-mediated weight loss is no exception. STEP-1 sub-study DXA analysis found roughly 25 to 40 percent of total weight lost was lean mass, which is within the expected range for diet-driven weight loss of similar magnitude. SURMOUNT-1 and -3 produced similar proportions for tirzepatide. Whether the absolute lean mass loss is clinically meaningful depends on the patient's baseline (older, lower-baseline-muscle patients are at higher risk of sarcopenia), activity pattern, and protein intake. Resistance training two to three times a week and protein intake of 1.2 to 1.6 grams per kg of ideal body weight per day is the standard mitigation. Deeper analysis at muscle loss what we know and seniors and sarcopenia.
Titration: the single biggest determinant of tolerability
The labeled titration schedule for semaglutide injectable is 0.25 mg weekly for four weeks, then 0.5 mg for four weeks, then 1.0 mg for four weeks, then 1.7 mg for four weeks, then 2.4 mg maintenance. For tirzepatide it is 2.5 mg weekly for four weeks, then escalation in 2.5 mg increments every four weeks to a maximum of 15 mg. The labels permit slower titration when tolerability is poor, and most clinicians use that flexibility. Common patterns we see: hold at the current dose for an additional two to four weeks when GI side effects are intense; drop back to the prior dose if escalation triggered persistent vomiting; and skip the final dose increment if the patient is meeting weight-loss targets on the lower dose. None of these are contraindicated and all reduce side effects. Full titration walkthrough at titration guide and injection technique.
Storage and travel
Semaglutide and tirzepatide require refrigeration before first use (2 to 8 degrees Celsius). Once in use, the labels permit room temperature storage (up to 30 degrees Celsius) for 28 to 56 days depending on the specific product. Travel logistics are covered in storage and travel cold chain. The relevance to side effects: a pen left in a hot car or stored at high temperature can degrade, which produces both reduced efficacy and unpredictable injection-site reactions. If a dose was stored outside the labeled window, replace rather than inject.
Absolute and relative contraindications
Absolute contraindications, present on the labels of all GLP-1 receptor agonists, are personal or family history of medullary thyroid carcinoma, multiple endocrine neoplasia syndrome type 2, and known hypersensitivity to the active ingredient. Pregnancy and breastfeeding are also labeled contraindications, covered specifically in pregnancy and GLP-1 and the related condition page at breastfeeding postpartum. Type 1 diabetes is not an absolute contraindication but is off-label and carries its own dosing considerations, covered in GLP-1 and type 1 diabetes. Relative contraindications include prior pancreatitis, symptomatic gallbladder disease, severe gastroparesis at baseline, and end-stage renal disease, the latter covered in dialysis and ESRD dosing.
When to call the prescriber
Same-day call: severe abdominal pain, especially radiating to the back (pancreatitis signal); persistent vomiting that prevents fluid intake past 24 hours; yellowing of skin or eyes (biliary obstruction); sudden vision loss in one eye (NAION signal); new-onset suicidal thoughts; severe headache with vision changes. Routine call within a week: persistent nausea past the typical four-to-six-week window on a stable dose; new gallbladder-area pain; injection-site reactions that do not resolve in 48 hours; significant mood changes; new constipation that does not respond to standard measures. Expected and managed at home without a call: mild nausea in the first two weeks of each dose step, occasional reflux, mild constipation, burping, fatigue in the first month. The triage framing is in honest risks.
Long-term safety data
The longest randomized follow-up on semaglutide is SELECT (median 39.8 months) and on tirzepatide is SURMOUNT-4 (88 weeks total). Real-world postmarket data on semaglutide now exceeds 15 million patient-years globally. The major emerging signals from postmarket are pancreatitis (confirmed but rare), gallbladder events (confirmed at elevated rates), and the NAION question (open). The signals notably not borne out by postmarket include medullary thyroid carcinoma (not confirmed in humans), heart-failure exacerbation (the SELECT and SURPASS programs actually showed cardiovascular benefit), and renal harm (the FLOW trial published in NEJM showed renal benefit, not harm, in T2D plus CKD). Real-world postmarket adverse-event reports are tracked in the FDA Adverse Event Reporting System (FAERS public dashboard). The European Medicines Agency runs a parallel surveillance and published its most recent semaglutide safety summary at the EMA Wegovy assessment. The renal data is in FLOW trial and the CV data in SELECT deep dive.
Population-specific considerations
Older adults: sarcopenia risk is the dominant consideration, covered in seniors and sarcopenia. Pediatric patients: liraglutide and semaglutide carry pediatric indications for adolescents; tirzepatide does not as of 2026. Pregnancy and preconception: GLP-1 should be discontinued at least two months before planned conception per the labels. Bariatric surgery patients regaining weight: GLP-1 is increasingly used as adjunct therapy; see post-bariatric regain. Specific high-risk groups have their own deep-dives at HFpEF with obesity, moderate to severe OSA and T2D with CKD.
What to do next
If you are starting a GLP-1, the most useful next read is the titration guide and the injection technique guide. If you are months in and hitting unexpected side effects, match the symptom to the relevant section above and the linked deep-dive. If you are weighing GLP-1 against alternatives, the safety profile of GLP-1 stacks favorably against most historical anti-obesity agents; we cover the comparative view in GLP-1 vs cheap alternatives honest. For program selection that prioritizes onboarding quality and clinical oversight, see the chart sort on the homepage chart and the methodology at our scoring methodology.
