The Side-Effect Curve: Why Most Quit in Month One

Two claims get made about GLP-1 side effects, and only one of them holds up. The first — that gastrointestinal symptoms cluster around dose escalation and are mostly mild and time-limited — is well documented in the pivotal trials. The second, that most people therefore quit in the first month, is not what the published persistence data shows. It is worth separating them, because the shape of the curve and the shape of the drop-off are different curves.

What the trials observed about symptom timing

The clearest picture comes from a pooled tolerability analysis of STEP 1–3, published in Diabetes, Obesity and Metabolism in 2022, covering 2,117 participants randomized to 68 weeks of semaglutide 2.4 mg and 1,262 to placebo.

Gastrointestinal event	Semaglutide 2.4 mg	Placebo
Nausea	43.9%	16.1%
Diarrhoea	29.7%	15.9%
Vomiting	24.5%	6.3%
Constipation	24.2%	11.1%

The rates are high, and the framing matters more than the rates. Of all gastrointestinal adverse events recorded in the semaglutide group, 99.5% were non-serious and 98.1% were mild-to-moderate. The authors describe them as transient, and as occurring “most frequently during/shortly after dose escalation.” STEP 1’s own report says the same of nausea and diarrhoea: typically transient, mild-to-moderate, and subsiding with time. SURMOUNT-1 reported the same pattern for tirzepatide — most gastrointestinal events mild to moderate, “occurring primarily during dose escalation.”

Across pooled STEP 1–3 data, 99.5% of gastrointestinal adverse events on semaglutide were non-serious and 98.1% mild-to-moderate, concentrated around dose escalation — and only 4.3% of participants stopped treatment because of them.

That last figure is the one most often missed. In the pooled STEP trials, 4.3% of semaglutide-treated participants permanently discontinued for gastrointestinal events. STEP 1 alone put it at 4.5%, against 0.8% on placebo. In SURMOUNT-1, adverse events of any kind caused discontinuation in 4.3%, 7.1% and 6.2% of the 5 mg, 10 mg and 15 mg tirzepatide groups, versus 2.6% on placebo. Over the much longer SELECT trial — a mean 34.2 months of exposure — adverse events led to permanent discontinuation in 16.6% of the semaglutide group versus 8.2% on placebo.

Where the “month one” story breaks down

Real-world stopping rates are much higher than trial discontinuation rates, but they are not concentrated in the first four weeks. A single-centre retrospective cohort of 2,306 patients treated with semaglutide or tirzepatide in an academic obesity clinic between January 2022 and December 2024, published in Diabetes, Obesity and Metabolism, tracked persistence directly. Median persistence was 10.7 months (IQR 5.4–16.3).

Time point	Cumulative discontinuation (real-world cohort)
3 months	14%
6 months	24%
9 months	35%
12 months	50%

Half of patients had stopped by one year — a far worse retention picture than any trial. But 86% were still on treatment at three months. Discontinuation is front-loaded relative to a flat curve, and it is much heavier than the trials imply, and it is still not a month-one phenomenon. Among those who did persist, median weight loss was 9.4% at six months and 14.4% at twelve.

What the data does and does not tell you

• Symptom timing is well established. Gastrointestinal events cluster during and shortly after dose escalation, and are overwhelmingly non-serious and mild-to-moderate.
• Trial discontinuation for those symptoms is low — roughly 4–7% depending on drug and dose, against 0.8–2.6% on placebo.
• Real-world stopping is a different phenomenon. At 50% by twelve months it dwarfs trial discontinuation, and the cohort authors note that the reasons were not directly captured — intolerance, supply shortages, cost, insurance, and people reaching their own goal are all candidates.
• “Most quit in month one” is unsupported. The best available persistence data puts three-month discontinuation at 14%.

The takeaway

The side-effect curve is real: trials consistently show gastrointestinal symptoms peaking around dose increases and settling with continued exposure, and they rarely reach the severity that ends treatment inside a trial. The discontinuation curve is a separate thing and is driven by more than symptoms — half of a real-world clinic cohort was off treatment within a year, but only 14% within three months. Attributing all of that to early nausea overstates what the evidence shows, and understates the access, cost and expectation factors sitting behind it. This is educational content, not medical advice, and nothing here is a dosing or titration recommendation — those decisions belong with a prescriber. For what the trials recorded in more detail, see GLP-1 side effects: what the trials report and managing GLP-1 nausea.

Sources

• Gastrointestinal tolerability of once-weekly semaglutide 2.4 mg in adults with overweight or obesity (Diabetes, Obesity and Metabolism, 2022)
• Once-Weekly Semaglutide in Adults with Overweight or Obesity (STEP 1), N Engl J Med 2021
• Tirzepatide Once Weekly for the Treatment of Obesity (SURMOUNT-1), N Engl J Med 2022
• Real-world titration, persistence and weight loss of semaglutide and tirzepatide in an academic obesity clinic (Diabetes, Obesity and Metabolism, via PMC)
• Semaglutide and Cardiovascular Outcomes in Obesity without Diabetes (SELECT), N Engl J Med 2023

This is sample content created during site scaffolding. Sources have been attached but each claim still needs verification against them before launch.