How Many People Actually Quit Because of the Nausea — and What the Placebo Group Did

Everyone starting a GLP-1 has read that nausea is common. Almost nobody has read how many people it actually stops, which is a different and more answerable question — and the labels report it.

Here is what they say, with revision dates.

The share who stop

Wegovy, revised 30 June 2026:

"The most common adverse reactions leading to discontinuation were nausea (1.8% versus 0.2%), vomiting (1.2% versus 0%), and diarrhea (0.7% versus 0.1%) for WEGOVY and placebo, respectively."

Saxenda, revised 15 June 2026:

"The most common adverse reactions leading to discontinuation were nausea (2.9% versus 0.2% for SAXENDA and placebo, respectively), vomiting (1.7% versus less than 0.1%), and diarrhea (1.4% v..."

Mounjaro, revised 6 May 2026, reports it by dose:

"More patients receiving MOUNJARO 5 mg (3.0%), MOUNJARO 10 mg (5.4%), and MOUNJARO 15 mg (6.6%) discontinued treatment due to gastrointestinal adverse reactions than patients receiving placebo (0.4%)."

That last one is the most informative sentence in the set, because it shows the thing the others do not: the rate more than doubles from the lowest dose to the highest. Three percent at 5 mg, six and a half at 15 mg. Titration is not just about comfort in the first weeks — the destination dose changes the odds of staying on the drug at all.

The same tirzepatide dosing schedule applies to Zepbound.

The column almost nobody quotes

From the Saxenda label:

"Gastrointestinal Adverse Reactions In the adult clinical trials, approximately 68% of SAXENDA-treated patients and 39% of placebo-treated patients reported gastrointestinal disorders"

Thirty-nine percent of people who got a placebo injection reported gastrointestinal symptoms.

That number reframes the whole conversation. It does not mean two-thirds of GLP-1 nausea is imaginary — 68 against 39 is a real and large difference. It means that in a group of people paying attention to their digestion, a substantial fraction will have something to report regardless, and that a single bad week is weak evidence about the drug.

Mounjaro's label makes the same point about severity:

"In the pool of placebo-controlled trials in adults, severe gastrointestinal adverse reactions occurred more frequently among patients receiving MOUNJARO (5 mg 1.3%, 10 mg 0.4%, 15 mg 1.2%) than placeb..."

Severe reactions run between 0.4% and 1.3% — and notice they do not climb neatly with dose the way discontinuations do.

The older drugs were harder to stay on

Byetta, revised 8 September 2025, across different trials:

"The most common adverse reactions leading to withdrawal for BYETTA-treated patients were nausea (9%) and vomiting (5%)."

and elsewhere in the same label:

"The most frequently reported adverse reactions leading to withdrawal for BYETTA-treated patients were nausea (5.1%) and vomiting (2.9%)."

Nine percent quitting over nausea, against Wegovy's 1.8%. Byetta is twice-daily exenatide and was discontinued in the US in October 2024; Bydureon BCise, its weekly successor, reports 0.7% for both nausea and diarrhea. Same molecule, different release profile, an order of magnitude apart on the thing that makes people stop.

What to do with this

Ask what the destination dose is, not just the starting one. On tirzepatide the discontinuation rate roughly doubles between 5 mg and 15 mg.

Give it more than one bad week. With 39% of a placebo group reporting GI symptoms, a single rough day is not a signal about the drug.

Know which number is which. "Nausea is common" and "1.8% stopped because of nausea" are both true and describe completely different experiences. The first is about how many feel it; the second is about how many it defeats.

Severe is rare and it is not the same as common. Between 0.4% and 1.3% on tirzepatide, and severity does not track dose the way discontinuation does.

Check the revision date. Wegovy's label is June 2026, Mounjaro's May 2026, Byetta's September 2025.

Where this comes from

Every figure above is quoted from the manufacturer's own Structured Product Label as published on DailyMed, the National Library of Medicine's repository. We use the manufacturer's label rather than a repackager's copy, which can lag by several revisions.

Nothing here is our arithmetic. Where two numbers are compared — 9% against 1.8%, or 68% against 39% — both sit in quotations above.

This is information, not medical advice. Whether a side effect is worth stopping over is a decision for you and your prescriber; the labels give you the base rate to hold it against.