Most GLP-1 stomach side effects arrive in short episodes — a few days for diarrhoea or vomiting, about a week for nausea — and cluster in the months when the dose is being stepped up. Two complaints break that pattern: constipation, whose episodes last weeks, and hair loss, which tends to show up months in rather than at the start. And after the last dose, the drug itself lingers for weeks. This page puts the duration evidence for Wegovy, Ozempic, Zepbound and Mounjaro in one place, says which figures come from trials and which from a smaller real-world cohort, and names what nobody has measured.
Sources were read on 6 October 2026: the Wegovy (effective 18 June 2026) and Zepbound labels on DailyMed, the pooled STEP tolerability analysis, and a 12-month observational study. This page reports those documents; it is not a guide to managing symptoms, and severe or persistent symptoms are a matter for a clinician.
Three different questions hiding in "how long"
| Question | Best evidence | Short answer |
|---|---|---|
| How long does one episode last? | Pooled STEP 1-3 trials, semaglutide 2.4 mg | Days for nausea, diarrhoea and vomiting; weeks for constipation |
| For how many months do side effects keep coming? | Labels' timing language; STEP pooled timing; 12-month cohort | Peak during escalation, then a decline that does not reach placebo levels |
| How long after the last dose? | Labels' pharmacokinetics | Semaglutide ~5–7 weeks in circulation; tirzepatide half-life ~5 days |
One episode: days, except constipation
The only published median episode lengths for a GLP-1 at weight-loss doses come from the pooled analysis of the STEP 1, 2 and 3 trials of semaglutide 2.4 mg (Wharton et al., 2022):
| Symptom | Median episode on semaglutide | On placebo |
|---|---|---|
| Nausea | 8 days | similar |
| Diarrhoea | 3 days | similar |
| Vomiting | 2 days | similar |
| Constipation | 47 days | 35 days |
The detail behind these figures — rates, the STEP 4 run-in, and how often stomach effects led to a pause or a stop — is on GLP-1 nausea, and is not repeated here. The point for this question is the contrast: a person can have several short nausea episodes over a year, while one bout of constipation can outlast all of them.
No equivalent median exists for tirzepatide. Lilly's labels report rates and timing, not episode lengths.
Over the months: an escalation peak, then a slope
What the labels say. The Zepbound label: "The majority of nausea, vomiting, and/or diarrhea events occurred during dose escalation and decreased over time" — and most people who stopped because of adverse reactions "did so during the first few months of treatment due to gastrointestinal adverse reactions". The Mounjaro label uses almost the same words. The semaglutide labels put the same pattern in their own terms; the pooled STEP analysis places the peak around week 20, just after Wegovy's schedule reaches 2.4 mg at week 17. How each brand's schedule steps up is on GLP-1 titration schedules.
What a year looks like, symptom by symptom. The trials report the peak; they rarely publish a month-by-month table by symptom. One 12-month observational study does. Richards et al. (JMIR Formative Research, 2025) surveyed 339 people who completed a UK remote weight-management programme on tirzepatide (209) or semaglutide (130) at months 1, 3, 6 and 12:
| Reported at the survey | Tirzepatide m1 | m3 | m6 | m12 | Semaglutide m1 | m3 | m6 | m12 |
|---|---|---|---|---|---|---|---|---|
| Feeling sick | 27.8% | 28.2% | 20.1% | 10.5% | 31.5% | 35.4% | 10.8% | 3.1% |
| Constipation | 29.2% | 21.5% | 20.1% | 12.9% | 31.5% | 30.8% | 17.7% | 15.4% |
| More tired than usual | 21.5% | 20.1% | 13.9% | 8.1% | 24.6% | 20.0% | 17.7% | 3.1% |
| Heartburn or indigestion | 17.7% | 17.7% | 10.0% | 4.8% | 8.5% | 20.8% | 8.5% | 5.4% |
| Headaches | 7.7% | 4.3% | 2.4% | 2.4% | 15.4% | 13.8% | 2.3% | 2.3% |
| Hair loss | 1.4% | 2.9% | 10.5% | 9.6% | 1.5% | 0.8% | 3.1% | 5.4% |
| No side effects | 41.6% | 39.2% | 44.0% | 60.3% | 53.8% | 31.5% | 53.8% | 67.7% |
Table 5 of the paper; percentages of each cohort's completers. glp1ledger reading: m = month.
Three patterns are visible in it. Month 3 is often no better than month 1 — on semaglutide, feeling sick and indigestion were higher at month 3, which fits a dose still rising. Most complaints roughly halve or better by month 12, with constipation the slowest to fall. And hair loss runs the other way: rare at month 1, highest at month 6 on tirzepatide. The curve is consistent with shedding that begins months after rapid weight loss starts; what the labels and studies record about it is on GLP-1 hair loss.
Read this table with its limits. It counts only people who stayed for 12 months, so anyone who stopped because of side effects is missing from every column; it is self-reported, has no placebo arm, and the programme gave tirzepatide users closer monitoring than semaglutide users. It describes a pattern, not a rate a reader can expect.
After the last dose: weeks, not days
Neither label measures how long each symptom takes to fade after stopping. They do state how long the drug remains:
| Drug (brands) | What the label states | Implied clearance |
|---|---|---|
| Semaglutide (Wegovy, Ozempic) | Half-life about 1 week; present in circulation "for about 5 to 7 weeks after the last" dose of 2.4 mg, 7.2 mg or the 25 mg tablet | stated by the label |
| Tirzepatide (Zepbound, Mounjaro) | Half-life "approximately 5 days" | about 3–4 weeks at five half-lives (glp1ledger arithmetic, not a label figure) |
The same long half-life is why the Wegovy label says to stop at least 2 months before a planned pregnancy, and why the Zepbound label tells people on oral contraceptives to add a barrier method for 4 weeks after starting and after each dose step. What happens to appetite and weight after stopping is on stopping a GLP-1.
Fatigue, specifically
The Wegovy label lists fatigue in 11% on 2.4 mg against 5% on placebo, and gives no timing. In the cohort above, feeling more tired than usual fell from about a fifth of people at month 1 to 8.1% (tirzepatide) and 3.1% (semaglutide) at month 12. The label figures and the lack of a trial-measured duration are on GLP-1 fatigue.
What is not established
- No median episode length exists for tirzepatide, so durations for Zepbound and Mounjaro are inferred from timing language and observational data, not measured.
- No study reviewed here measured symptom resolution after the last dose. The drug's clearance time is known; how fast each side effect follows it is not.
- Trials and cohorts lose their least tolerant people early. A falling rate partly reflects symptoms fading and partly reflects who is left — a point the GLP-1 side effects hub sets out from the labels.
Sources
Wharton S et al., "Gastrointestinal tolerability of once-weekly semaglutide 2.4 mg in adults with overweight or obesity, and the relationship between gastrointestinal adverse events and weight loss", Diabetes Obes Metab 2022;24:94-105, PMC9293236 · Richards R, Lunt W, Whitman M, Spaltro G, Hall R, "Semaglutide and Tirzepatide in a Remote Weight Management Program: 12-Month Retrospective Observational Study", JMIR Form Res 2025, doi:10.2196/81912, PMC12475876 (Table 5) · Novo Nordisk, WEGOVY prescribing information, DailyMed setid ee06186f-2aa3-4990-a760-757579d8f77b (effective 2026-06-18), sections 6.1, 8.3 and 12.3 · Eli Lilly, ZEPBOUND prescribing information, DailyMed setid 487cd7e7-434c-4925-99fa-aa80b1cc776b, sections 6.1, 7 and 10. All read 6 October 2026.