GLP-1 Side Effects: What the Trials Report, What Patients Experience, and What Helps

About 44% of adults on semaglutide 2.4 mg in the STEP trials reported nausea; 30% diarrhoea, 24% vomiting, 24% constipation. Placebo arms reported 16%, 16%, 6%, 11%. That is the gap defining the side-effect conversation: GLP-1 drugs do produce real GI symptoms at meaningfully higher rates than placebo, concentrated in the dose-escalation window.

This article maps what the pivotal trials measured, what patients describe in clinic, what helps in the published management literature, and which symptoms warrant a same-day prescriber call. Every percentage links to its source label or trial publication.

TL;DR

GI side effects dominate the GLP-1 profile, but most are mild-to-moderate, transient, and concentrated at dose-escalation steps. The pooled STEP analysis classified 98.1% of GI events on semaglutide 2.4 mg as mild or moderate; 4.3% of patients permanently discontinued for GI reasons. Tirzepatide's profile is broadly similar, with somewhat lower nausea and higher injection-site reactions. Serious label-listed events are uncommon but real: thyroid C-cell tumour risk, acute pancreatitis, gallbladder disease, severe hypoglycaemia, AKI, and diabetic retinopathy progression.

The most common side effects, with incidence rates

Figures below come from the FDA Wegovy and Zepbound labels, which aggregate the STEP and SURMOUNT trial datasets. Percentages are the share of patients who reported the event at least once during the trial period (68 weeks Wegovy, 72 weeks Zepbound), not the share who experienced it continuously.

Sources: Wegovy prescribing information, Table 4, and Zepbound prescribing information, Section 6.1. Both labels report adverse reactions occurring in ≥2% of treated patients at a higher rate than placebo.

Three patterns are worth pulling out. Nausea on Wegovy at 44% is the single highest number in either label, against 28% on Zepbound 15 mg. Constipation on Zepbound decreases with higher doses (17% → 11%), the opposite of the nausea dose-response. Injection-site reactions are a Zepbound signal that does not appear in the Wegovy ≥2% table.

What "nausea" actually feels like in practice

The worst nausea is in the first 48 to 72 hours after a dose, particularly an escalation dose. For weekly injections, peak intensity lands on days 2 through 4; Sunday is a difficult day after a Friday injection.

Patients describe two distinct sensations. The first is low-grade, persistent stomach-fullness that makes food unappealing without an acute episode of feeling sick. The second is more conventional nausea triggered by smells, sights, or eating past fullness, typically worse with high-fat meals and large portions.

Intensity is dose-dependent. The first 0.25 mg semaglutide dose usually produces minor symptoms, the 0.5 mg step a sharper increase, and 1 mg is where many first consider whether to continue. Tirzepatide follows the same pattern at the 2.5 → 5 mg and 5 → 7.5 mg steps. At the maintenance dose, most patients describe symptoms as gone or substantially diminished. The pooled STEP analysis noted GI events were "transient and occurred most frequently during or shortly after dose escalation" (Wharton et al., 2022).

Evidence-grounded management of GI side effects

Management starts with titration. The Wegovy and Zepbound labels both specify a 16-week dose-escalation schedule and both explicitly permit a prescriber to delay the next dose increase by four weeks if a patient is not tolerating the current dose. Neither label encourages skipping doses; the recommendation is to slow the schedule, not interrupt it.

Beyond titration, the patient-side strategies that recur in published clinical experience:

• Smaller, more frequent meals. Patients who continue eating their pre-treatment portion sizes report worse symptoms than those who shift to four or five small meals.
• Lower-fat meals. Fat slows gastric emptying further on top of the drug effect. Higher-protein, moderate-fat meals are typically better tolerated than fried or high-cream foods.
• Hydration. GLP-1 drugs reduce thirst signalling along with hunger; patients commonly under-drink and develop fatigue, dizziness, or constipation that are downstream of dehydration rather than the drug itself.
• Antiemetic medication if the prescriber recommends. Ondansetron and similar agents are sometimes prescribed for the first one to two days after a dose increase. This is a clinical decision your prescriber makes based on your full medication picture.
• Slower escalation or a dose pause. Both labels allow a patient to remain at a lower dose longer than the standard schedule if tolerance is the limiting factor.

None of these is a guaranteed fix. The decision tree typically runs: try meal-size and composition changes first; if symptoms persist, talk to the prescriber about slowing titration or pausing the next increase; if symptoms become severe or fail to resolve within a week or two, the conversation shifts to whether to continue the drug.

Less common but worth knowing

Gallbladder disease. The Wegovy label reports cholelithiasis at 1.6% on drug versus 0.7% on placebo and cholecystitis at 0.6% versus 0.2%. In the larger SELECT trial of 17,604 patients, gallbladder-related disorders ran 2.8% on semaglutide versus 2.3% on placebo, driven mainly by cholelithiasis. The Zepbound label reports cholelithiasis at 1.1% versus 1.0% and cholecystitis at 0.7% versus 0.2%. Rapid weight loss raises gallstone risk on its own; the GLP-1 effect appears to be on top of that.

Pancreatitis. Both labels carry a Warnings and Precautions section on acute pancreatitis. In Wegovy trials, four adjudicated cases (0.2 per 100 patient-years) versus one in placebo. In Zepbound trials, 0.2% on drug and 0.2% on placebo. The label instruction is to discontinue promptly if suspected and not to restart if confirmed.

Injection-site reactions. More frequent with Zepbound (6–8%) than Wegovy. Usually mild redness or itching; rotate sites between abdomen, thigh, and upper arm.

Mood changes. Both labels include a Suicidal Behavior and Ideation precaution. The trial data do not show a clear causal signal (depression rates in STEP 1 were similar between groups), but FDA included the language given the regulatory history of weight-loss drugs. Flag a depression history to your prescriber before starting.

Hair loss. Real and label-listed. Mechanism is most likely the rapid weight loss itself (telogen effluvium), not a direct drug effect. Begins two to four months after starting and resolves over six to twelve months once weight stabilises.

Serious adverse events from the trials

The Warnings and Precautions sections of both labels list the same set of serious events:

• Thyroid C-cell tumours (boxed warning). In rats, both molecules produce dose-dependent C-cell tumours at clinically relevant exposures. Human relevance is not established. Both drugs are contraindicated in patients with personal or family history of medullary thyroid carcinoma (MTC) or Multiple Endocrine Neoplasia syndrome type 2 (MEN 2).
• Acute pancreatitis. Low incidence; the label instruction is unequivocal: discontinue if suspected, do not restart if confirmed.
• Acute gallbladder disease. Discussed above; higher rate than placebo even after adjusting for weight loss.
• Severe hypoglycaemia. Primarily a concern in T2D patients also on insulin or a sulfonylurea. GLP-1s do not cause hypoglycaemia on their own.
• Acute kidney injury. Generally dehydration-mediated; the label notes most reported events followed severe nausea, vomiting, or diarrhoea. In Wegovy adult trials, AKI ran 0.4 per 100 patient-years on drug versus 0.2 on placebo.
• Diabetic retinopathy complications in T2D. In a T2D trial with BMI ≥27, diabetic retinopathy ran 4.0% on Wegovy versus 2.7% on placebo. The mechanism is thought to be rapid glycaemic improvement rather than direct drug effect.

The Wegovy label also notes heart-rate increases (average +1 to +4 bpm) and a hypersensitivity section covering post-marketing anaphylaxis and angioedema.

What's a red flag: call your prescriber if

Signs that warrant same-day contact with the prescribing clinician:

• Persistent vomiting for more than 24 hours. The volume-depletion mechanism behind reported AKI cases.
• Severe upper-abdominal pain, particularly radiating to the back. Classic pancreatitis presentation.
• Right-upper-quadrant pain, fever, or yellowing of the skin or eyes. Possible cholecystitis or gallstone obstruction.
• Sudden vision changes in patients with type 2 diabetes. Possible retinopathy progression.
• Signs of dehydration: light-headedness on standing, very dark urine, no urination for many hours.
• Severe injection-site reaction beyond mild redness: spreading rash, fever, painful swelling.
• Chest pain, palpitations, or new shortness of breath. Not GLP-1-specific but a general contact threshold.
• Suicidal thoughts or significant new depression. Per label monitoring guidance.

Persistent mild nausea, occasional vomiting after a dose increase, soft stools, constipation, and titration fatigue do not by themselves warrant a same-day call. Bring them up at the next scheduled visit, or sooner if you are reconsidering the drug.

The "Ozempic face" question

The term entered the press in late 2022 and refers to a gaunt or hollowed facial appearance some patients develop after substantial weight loss on a GLP-1. The dermatology consensus is that this is not a direct drug effect. Subcutaneous fat in the cheeks, temples, and periorbital areas decreases along with fat elsewhere on the body; the face is simply where the change shows first.

Patients who lose 15% or more of starting weight, who are over 40, and who had more facial volume to begin with are likeliest to notice the shift. The same pattern appears after equivalent weight loss from any cause: bariatric surgery, very-low-calorie diet, illness. Dermatology responses range from skincare adjustments to hyaluronic-acid fillers and fat grafting; none is GLP-1-specific. Rapid weight loss exposes facial volume loss that gradual weight loss tends to mask.

Side-effect-by-dose-stage map

Side effects cluster at titration steps and ease at the maintenance dose. The stages where most patients report the worst symptoms:

• Semaglutide 0.25 → 0.5 mg (week 5). First meaningful jump in receptor activation.
• Semaglutide 0.5 → 1.0 mg (week 9). Often the hardest step.
• Tirzepatide 2.5 → 5 mg and 5 → 7.5 mg (weeks 5 and 9). Frequently the worst steps in the SURMOUNT-1 titration pattern.

Patients who plateau at the maintenance dose (semaglutide 2.4 mg or tirzepatide 10 or 15 mg) typically describe a tolerance pattern: by the third or fourth week at the same dose, GI symptoms have diminished or resolved. Patients who do not tolerate the next step can remain at the current dose for an additional four weeks before re-attempting escalation.

What helps before each dose

Strategies that recur in published clinical experience and patient communities:

• Lower-fat meal the evening before the injection. Reduces the load on the already-slowed stomach.
• Smaller portions for the 48 hours after the dose. Front-loading lighter eating around the injection reduces symptom intensity.
• Steady hydration. Two to three litres of fluid per day across the dose window, more in warm weather. Coffee and alcohol do not count.
• Injection time of day. Many prefer morning injections so the peak GI effect (12–48 hours later) lands during the day rather than overnight.
• Site rotation. Abdomen, thigh, and upper arm in rotation, particularly for tirzepatide given the higher injection-site reaction rate.
• A simple anti-nausea plan, if the prescriber agrees. Ginger products or low-dose ondansetron in the first 48 hours after a dose increase. A clinical decision that belongs to the prescriber.

These do not eliminate side effects. They tend to move a patient from "intolerable" to "manageable" during the difficult titration weeks.

How we keep this article current

We refresh this page when an FDA label update, a major trial readout, or a published safety analysis changes the picture. As the evidence base grows, a few figures here tend to drift faster than the rest:

• Low-incidence figures. SELECT extended the semaglutide safety dataset to 17,604 patients over 40 months, and the SURMOUNT extension trials are doing similar work for tirzepatide. As these cohorts grow, the precision of rare-but-serious signals (gallbladder disease, pancreatitis) will improve.
• The thyroid C-cell tumour question. This is the most consequential open one; the boxed warning is based on rat data, and human relevance is unestablished. FDA will revisit the language as post-marketing data matures.

If you spot an error or a missing source on this page, please email [email protected]. We acknowledge corrections within five business days and publish the resolution within fifteen.

Where to read next

If you are deciding whether a GLP-1 is right for you, start with the mechanism guide. For the head-to-head, see Wegovy vs Zepbound. For the contraindications list, see who shouldn't take a GLP-1. Our methodology page explains how we source these figures and the disclaimer sets out what this content is and is not.

Methodology and status: this article aggregates adverse-event data from the FDA labels of Wegovy and Zepbound, the STEP and SURMOUNT trial publications, SELECT, the Wharton pooled GI tolerability analysis, and the ASA perioperative guidance documents. Every incidence figure is linked to the source label or publication. This article is awaiting medical review; no individual clinical recommendation here should override your prescriber's specific advice for your situation.