This page lays the timing out in order: your own dose-escalation dates, when side effects were reported to begin, how common each one is at each dose, how long an episode lasts, which effects arrive later, and which symptoms are not part of the usual course and need a call.
When is the stretch in which side effects are most likely?
The label ties the common stomach side effects to one period: the weeks in which the dose is being raised. Enter your first dose, the dose you are heading for and how long you stay on each step, and the calendar shows that period by date.
Log side effects in the app next to your doses and see them against this schedule.
Open the appThe numbers along the bottom are weeks of treatment. The lighter band is the 4 weeks after the last increase, the time the label gives for the level to become steady. Your own dose changes may fall on other dates.
The sentence this calendar rests on is in the side effects section of the label: "The majority of nausea, vomiting, and/or diarrhea events occurred during dose escalation and decreased over time."1 The label does not break that down by week. What it does give is the schedule itself: 2.5 mg for the first 4 weeks, then an increase of 2.5 mg after at least 4 weeks on each dose, up to a maintenance dose of 5, 10 or 15 mg.
At the fastest pace the label allows, someone heading for 15 mg has five increases and reaches the top dose after 20 weeks. The level in the blood takes about 4 more weeks to settle, so the body is adjusting to something new for roughly the first 24 weeks. A person who stops at 5 mg has one increase and is through the same process in 8 weeks.
A worked example
Take a first dose on Monday, January 4, and the fastest pace. The increases fall on February 1 (5 mg), March 1 (7.5 mg), March 29 (10 mg), April 26 (12.5 mg) and May 24 (15 mg). The level is steady from about June 21. For this person the stretch the label points to runs from early January to late June, with a fresh climb starting on each of those five dates.
Now take the same start with 8 weeks on each dose and a target of 10 mg. The increases fall on February 1, March 29 and May 24, and the level is steady from about June 21 as well. The period is the same length, but it holds three climbs with eight weeks between them, where the first schedule packed in five with four weeks between them.
| Weeks | Dose at the fastest pace mg | Step up from the previous dose | What the sources say |
|---|---|---|---|
| 1 to 4 | 2.5 | the first dose | stomach emptying is slowed most after the first dose; half of reported events had begun by day 16 |
| 5 to 8 | 5 | doubled | the first increase and the lowest maintenance dose |
| 9 to 12 | 7.5 | +50% | dose escalation continues |
| 13 to 16 | 10 | +33% | over 80% of reported events had begun within 3 months |
| 17 to 20 | 12.5 | +25% | dose escalation continues |
| 21 to 24 | 15 | +20% | the last increase; the level becomes steady |
| 25 onwards | 15 | none | the label's "decreased over time" |
Schedule from section 2 of the label. The third column is our arithmetic.
The third column explains part of the pattern. Going from 2.5 to 5 mg doubles the dose. Going from 12.5 to 15 mg adds a fifth. Every step is the same 2.5 mg, but each one is a smaller change for the body than the one before it.
When do side effects usually start?
Early. The best timing data come from the FDA's adverse event reporting system, the database that collects reports of suspected side effects from patients, doctors and manufacturers. A 2026 analysis looked at the stomach and bowel events reported with tirzepatide and at how many days into treatment each one began.2
The numbers along the bottom are days of treatment. The dots are the published points: a quarter of events by day 2, half by day 16, three quarters by day 70, over 80% within 3 months. The line joins them. These are reports of events, not a count of all patients.
Stomach and bowel events reported with tirzepatide
A quarter of the reported events began within two days of the first injection. That fits how the drug is absorbed: the label puts the peak level in the blood at about 24 hours after a dose, with a range of 8 to 72 hours.1 The first days after the very first dose are when the body meets the drug with no adjustment behind it.
Half had begun by day 16, in the third week on 2.5 mg, and three quarters by day 70, which on the fastest schedule falls in the weeks on 7.5 mg. The analysis found no meaningful difference in timing between men and women (a median of 13.5 against 18 days) or between people over and under 65 (12 against 21 days).2
Two limits come with this kind of data. Reports are voluntary, so the database cannot say how many of all patients had a given effect: that is what the trial table further down is for. And a report records when an event started, not how bad it was. What the reports are good at is timing, and here they agree with the label: the events cluster at the beginning.
The most frequently reported stomach and bowel events in the same analysis were nausea, at 27.7% of the reports, and diarrhea, at 12.8%.2
Does the day after the injection matter?
Within each week the level is not flat. It peaks about a day after the injection, anywhere from 8 to 72 hours according to the label, and then drifts down. In our model, on a settled dose, the level just before the next injection is about 44% of that peak.1
None of the sources used here counted symptoms by day of the week, so there is no trial figure showing that day two is worse than day six. What can be said is where the level is highest, the first three days after each dose, and that a quarter of the reported events began within two days of the very first injection.
The label allows the injection day to be changed, as long as at least 3 days (72 hours) pass between two doses. That leaves some room to choose which days of the week the peak falls on.
How common is each side effect, and does the dose matter?
The label's main table counts every side effect reported by at least 2 in 100 people in the two large weight-loss trials and more often than on placebo. The chart shows the eight most common for the dose you pick. The second button subtracts the placebo figure, which leaves what the drug added.
People per 100 who reported the effect at any time in trials of up to 72 weeks. Placebo: 958 people. Zepbound: 630 on 5 mg, 948 on 10 mg, 941 on 15 mg.
| Side effect, people per 100 | Placebo | 5 mg | 10 mg | 15 mg |
|---|---|---|---|---|
| Nausea | 8 | 25 | 29 | 28 |
| Diarrhea | 8 | 19 | 21 | 23 |
| Vomiting | 2 | 8 | 11 | 13 |
| Constipation | 5 | 17 | 14 | 11 |
| Stomach pain | 5 | 9 | 9 | 10 |
| Indigestion | 4 | 9 | 9 | 10 |
| Injection site reactions | 2 | 6 | 8 | 8 |
| Feeling tired | 3 | 5 | 6 | 7 |
| Allergic reactions | 3 | 5 | 5 | 5 |
| Belching | 1 | 4 | 5 | 5 |
| Hair loss | 1 | 5 | 4 | 5 |
| Heartburn | 2 | 4 | 4 | 5 |
| Gas | 2 | 3 | 3 | 4 |
| Bloating | 2 | 3 | 3 | 4 |
| Dizziness | 2 | 4 | 5 | 4 |
| Low blood pressure | 0 | 1 | 1 | 2 |
Zepbound label, Table 1, Studies 1 and 2. The names are the plain ones from the patient section of the label.
Three things stand out. The first is that the dose matters less than people expect. Nausea was reported by 25 in 100 on 5 mg and 28 in 100 on 15 mg. Counting every stomach and bowel effect together, the label gives the same figure for all three doses: 56 in 100, against 30 in 100 on placebo.1 Everyone in those groups passed through the low doses first, and that early stretch is where most of the events happen.
The second is the placebo column. Eight in 100 people on placebo injections reported nausea and eight reported diarrhea. Some of what happens in the first months would have happened anyway, and the "above placebo" view takes that part out.
The third is constipation, which runs the other way: 17 in 100 on 5 mg, 14 on 10 mg and 11 on 15 mg. Diarrhea rises slightly over the same doses. The label reports the numbers and does not explain them.
What the drug adds, dose by dose
Subtracting the placebo column from each dose column leaves the part of each figure that belongs to the drug.
| More people per 100 than on placebo | 5 mg | 10 mg | 15 mg |
|---|---|---|---|
| Nausea | 17 | 21 | 20 |
| Diarrhea | 11 | 13 | 15 |
| Vomiting | 6 | 9 | 11 |
| Constipation | 12 | 9 | 6 |
Our subtraction from Table 1 of the label.
What do the numbers say about each symptom?
The figures for the main symptoms are scattered across several sources. Here they are side by side, one symptom at a time.
Nausea
The most common one. It was reported by 25 to 29 in 100 people on Zepbound and 8 in 100 on placebo, so the drug adds about 20. It is also the most reported stomach event in the FDA database, at 27.7% of reports.2 On Wegovy one episode lasted a median of 8 days. It is the symptom with the best evidence for slower dose steps: 24.1% in the small-step group of the schedule trial against 35.7% and 39.3% in the fast groups.
Diarrhea
Reported by 19 to 23 in 100 against 8 on placebo, rising slightly with the dose. It made up 12.8% of the FDA reports. On Wegovy an episode lasted a median of 3 days. In the Zepbound trials 5.2% to 9.3% of people took a medicine for it at some point. The patient section of the label adds a practical line for this symptom and for vomiting: drink fluids to lower the chance of dehydration.
Vomiting
Less common, and the one that rises most clearly with the dose: 8, 11 and 13 in 100 on 5, 10 and 15 mg, against 2 on placebo. On Wegovy an episode lasted a median of 2 days. Vomiting that does not stop is on the label's list of things to report right away, because of fluid loss.
Constipation
Reported by 17, 14 and 11 in 100 on 5, 10 and 15 mg, against 5 on placebo. It is the long one: a median of 47 days per episode on Wegovy, with the share of people affected levelling off at about week 10 and staying there.
Stomach pain, indigestion, heartburn, belching, gas and bloating
Each of these was reported by fewer than 10 in 100 people. Stomach pain and indigestion: 9 to 10 in 100 against 4 to 5 on placebo. Heartburn and belching: 4 to 5 against 1 to 2. Gas and bloating: 3 to 4 against 2. Belching has the largest gap in relative terms, four to five times the placebo figure. The FDA report analysis noted the same: belching and slowed stomach emptying were the events reported most out of proportion for tirzepatide compared with other drugs.2
Tiredness and dizziness
Feeling tired, a heading under which the label groups fatigue, weakness, lethargy and malaise, was reported by 5 to 7 in 100 against 3 on placebo. Dizziness was reported by 4 to 5 against 2. The label gives no timing for either.
How long does one episode last?
For Zepbound the label does not say. It states that the events "decreased over time" and gives no number of days. The pooled analysis of the diabetes trials of tirzepatide uses the word directly: nausea, diarrhea and vomiting "were transient and of mild-to-moderate severity".3
The clearest figures in days come from Wegovy, a drug of the same family taken on a similar step-up schedule. An analysis of 2,117 people on semaglutide 2.4 mg measured how long each single event lasted from start to finish.4
| Side effect | Median length of one episode on Wegovy days | On placebo |
|---|---|---|
| Nausea | 8 | similar |
| Diarrhea | 3 | similar |
| Vomiting | 2 | similar |
| Constipation | 47 | 35 days |
Wharton and colleagues, Diabetes, Obesity and Metabolism 2022, pooled STEP 1 to 3 trials. Semaglutide, not tirzepatide.
A median means half of the episodes were shorter and half were longer. So a typical bout of nausea ran for about a week, vomiting and diarrhea for two or three days, and constipation for more than six weeks. The authors describe the first three as "relatively short-lived".
Two cautions apply. These are figures for a different drug, and no equivalent table has been published for tirzepatide among the sources used here. And an episode is not the same as the whole course: a person can have one short bout after the first dose and another after an increase two months later. The numbers describe each bout, not the total.
Constipation behaves differently from the other three. It lasts much longer per episode, and on Zepbound it was reported less often at the higher doses. For someone planning around side effects, that makes it the one to expect to manage for weeks, not to wait out for days.
When do they ease?
After the dose stops rising. That is the plain reading of the label's sentence, and three other observations line up with it.
In the Wegovy analysis, the share of people who had nausea, vomiting or diarrhea at any given moment rose through the step-up period, "peaked at about week 20 and decreased thereafter", with the steepest fall for nausea. Constipation levelled off at about week 10.4 Wegovy takes 16 weeks to reach its full dose, so the peak came shortly after the last increase.
In the Zepbound trials, researchers recorded when people first took a medicine for nausea or diarrhea. First use was most common in the first 24 weeks, which is the 20 weeks of dose increases plus the weeks just after.5
And the label contains a small experiment that shows the body adapting. Researchers used a common painkiller, acetaminophen, as a tracer for how fast the stomach empties. After the first 5 mg dose of tirzepatide, the tracer's peak level was 55% lower and came an hour later, a sign of a slowed stomach. At week 6, on 15 mg, there was "no meaningful impact".1 In the label's words, the effect on stomach emptying "is largest after the first dose and this effect diminishes over time".
The stomach-emptying test in the label
Put together: the body reacts most to the first dose, reacts again to each increase, and adapts in between. Once there are no more increases, there is nothing new to adapt to.
Why each increase sets them off again
With a weekly injection and a half-life of about 5 days, the amount in the body builds up over several doses. The label says the level becomes steady after 4 weeks on the same dose.1 Our model of that build-up puts the peak after the very first injection at about 61% of the peak reached once the dose has settled, and at 85%, 94% and 98% after the second, third and fourth.
So the level is still climbing through the whole first month, and then the dose doubles. Each later increase starts another four-week climb to a new plateau. The calendar at the top of the page marks those climbs, and the lighter band at the end marks the last one.
Which side effects come later?
Not everything belongs to the first weeks. Four effects in the label are tied to something other than the dose going up.
Hair loss. The label states that hair loss "adverse reactions in ZEPBOUND-treated patients were associated with weight reduction". It was reported by 7.1% of women and 0.5% of men on Zepbound, against 1.3% and 0% on placebo. Nobody taking Zepbound left the trials because of it.1 The label ties it to the weight loss and not to the injection, and weight loss takes months.
Gallbladder problems. The label says the same about these: "Acute gallbladder events were associated with weight reduction." Gallstones were reported in 1.1% of people on Zepbound and 1% on placebo, inflammation of the gallbladder in 0.7% against 0.2%, and gallbladder removal in 0.2% against none.1
Injection site reactions. These were reported by 6 to 8 in 100 people. The label connects them with antibodies that some people's immune systems make against the drug over time. Among people who developed such antibodies, 11.3% had injection site reactions. Among those who did not, 1% did.1
Constipation. As above, it is the slow one. In the Wegovy data a single episode ran for a median of 47 days, and the share of people affected stayed level from about week 10 instead of falling.4
| Effect | What it follows | Zepbound | Placebo |
|---|---|---|---|
| Hair loss, women | weight reduction | 7.1% | 1.3% |
| Hair loss, men | weight reduction | 0.5% | 0% |
| Gallstones | weight reduction | 1.1% | 1% |
| Gallbladder inflammation | weight reduction | 0.7% | 0.2% |
| Injection site reactions, with antibodies | antibodies to the drug | 11.3% | — |
| Injection site reactions, without antibodies | 1% | — |
Zepbound label, sections 5.4 and 6.1, pooled Studies 1 and 2.
Which symptoms are not part of the usual course?
The patient section of the label separates the common side effects from a short list of serious ones, and for each serious one it says what to do. The wording below follows it closely.1
| What you notice | What it may be | What the label says to do |
|---|---|---|
| Severe pain in the stomach area that will not go away, with or without vomiting, sometimes felt through to the back | inflammation of the pancreas | stop using Zepbound and call your healthcare provider right away |
| Swelling of the face, lips, tongue or throat; trouble breathing or swallowing; fainting; very rapid heartbeat; severe rash or itching | serious allergic reaction | stop using Zepbound and get medical help right away |
| Pain in the upper stomach, fever, yellowing of skin or eyes, clay-colored stools | gallbladder problems | tell your healthcare provider right away |
| Nausea, vomiting or diarrhea that does not go away | fluid loss that can lead to kidney problems | drink fluids; tell your healthcare provider right away |
| Stomach problems that are severe or will not go away | severe stomach problems | tell your healthcare provider |
| Dizziness, sweating, shakiness, confusion, fast heartbeat, hunger | low blood sugar, mainly with insulin or a sulfonylurea | talk to your healthcare provider about how to manage it |
| Changes in vision, in people with type 2 diabetes | changes in the eye | tell your healthcare provider |
Zepbound Medication Guide, revised April 2026, "What are the possible side effects of ZEPBOUND?"
The useful distinction is between a symptom that fades and one that does not. Ordinary nausea comes around a dose change and eases. Pain that is severe and does not let up, vomiting that goes on until you cannot keep fluids down, or any sign of an allergic reaction is outside that pattern at any week of treatment.
The numbers show how uncommon the serious ones were in the trials. Pancreatitis was confirmed in 0.2% of people on Zepbound and 0.2% on placebo. Acute kidney injury was reported in 0.5% against 0.2%. Allergic reactions within a day of an injection occurred in 2.1% against 0.4%, and the label notes that most allergic reactions in the trials were skin reactions such as rash and itching.1
The kidney warning has a timing note of its own. The label asks prescribers to watch kidney function in patients whose side effects could cause fluid loss, "especially during dosage initiation and escalation". That is the same period the calendar shades. Low blood sugar belongs mostly to people who also take diabetes medicines: in the trial of people with type 2 diabetes it was reported in 4.2% on Zepbound against 1.3% on placebo, and in the trial without diabetes in 0.3% against none.
Does going up more slowly help?
The evidence says it does for nausea. Before the large studies, researchers gave 111 people with type 2 diabetes tirzepatide for 12 weeks on three different schedules. One group started low and rose in small steps to 12 mg. Two groups reached 15 mg in 8 weeks with bigger jumps.6
People who reported nausea in the 12-week schedule trial
The authors concluded that "lower starting doses and smaller dose increments are associated with a more favourable side effect profile".6 Today's label starts at 2.5 mg and moves in steps of 2.5 mg, the kind of schedule that did better in that trial.
The label leaves room to go slower still. Its wording is that the dose may be increased after "at least" 4 weeks on the current one, so 4 weeks is a minimum.1 The calendar above lets you set 8 or 12 weeks per step and see what that does to the dates: the escalation period gets longer, and each climb has more time to settle before the next begins. Whether to take that time is a decision for you and your prescriber. The dosing schedule has every step by date.
What did the trials do when someone felt sick?
The four main Zepbound trials followed a fixed order of steps for stomach symptoms a participant could not tolerate, and a pooled analysis of those trials describes it.5
- Food first. Participants were counselled on three changes: eat smaller meals, split three daily meals into four or more smaller ones, and stop eating when you feel full.
- A medicine for the symptom. The trial doctor could prescribe something for nausea or for diarrhea.
- Skip one dose. One weekly injection was left out and treatment resumed at the same dose.
- Step down. The dose was dropped to the next lower one.
- Stop. Only if nothing above had worked.
The same analysis counted how often the second step was used. Across three of the trials, 13.8% to 27.5% of people on tirzepatide took an anti-nausea medicine at some point, against 5.8% to 6.7% on placebo. For anti-diarrhea medicine the figures were 5.2% to 9.3% against 1.2% to 2.2%.5
| Used at some point in the trial | Tirzepatide | Placebo |
|---|---|---|
| A medicine for nausea | 13.8% to 27.5% | 5.8% to 6.7% |
| A medicine for diarrhea | 5.2% to 9.3% | 1.2% to 2.2% |
Rubino and colleagues, Diabetes, Obesity and Metabolism 2025. Ranges across three trials.
So at most about one participant in four needed a medicine for nausea, and most needed none. Severity points the same way: most stomach and bowel events were mild to moderate, and severe or serious ones occurred in 0.3% to 5.6% of people, depending on the trial.
Outside a trial each step is the prescriber's decision. The order is still worth knowing before the first bad week. It shows that stopping was the last of five options, and that a pause or a step down are ordinary tools and no sign that the treatment has failed.
Do side effects mean the drug is working?
No, and their absence does not mean it is failing. The pooled analysis compared weight loss in people who had nausea, vomiting or diarrhea with weight loss in people who had none.5
| Weight lost | 5 mg | 10 mg | 15 mg |
|---|---|---|---|
| Had at least one of the three symptoms | 17.2% | 22.2% | 22.8% |
| Had none of them | 15.3% | 20.6% | 22.2% |
The gap is between half a point and two points. People who never felt sick lost almost as much as people who did. The authors estimated that these side effects accounted for at most 3.1% of the total weight reduction and wrote that they "appeared to contribute slightly".
Two other analyses found the same. In the diabetes trials of tirzepatide, weight reduction was consistent between people who reported these symptoms and people who did not.3 For Wegovy, a formal calculation found that less than 1 percentage point of the extra weight loss ran through the stomach side effects.4 The drug lowers weight by lowering appetite. Feeling sick is a separate effect that some people get on top.
How many people stop because of side effects?
Few, in the trials. The label gives the share who gave up treatment because of a side effect of any kind, and the share who did so because of a stomach or bowel effect.1
| Stopped treatment because of | Placebo | 5 mg | 10 mg | 15 mg |
|---|---|---|---|---|
| Any side effect | 3.4% | 4.8% | 6.3% | 6.7% |
| A stomach or bowel side effect | 0.5% | 1.9% | 3.3% | 4.3% |
Zepbound label, section 6.1, pooled Studies 1 and 2.
On 15 mg, about 4 in 100 people stopped because of their stomach, and 96 did not. Here the dose does show: the share roughly doubles from 5 mg to 15 mg, from 1.9% to 4.3%, while the share who report nausea barely changes, from 25 to 28 in 100.
Two further trials in the label give similar figures. In the trial that began with an intensive diet and exercise programme, 10% of people on Zepbound stopped because of side effects, against 2% on placebo. In the withdrawal trial, 7% stopped for that reason during the first 36 weeks.
The timing of stopping follows the timing of the side effects. In the Wegovy trials, where it was recorded, most of the people who gave up treatment over stomach and bowel effects did so during the dose-escalation period.4 A person who gets through the step-up is past the stretch in which most stopping happens.
Outside trials, side effects weigh more. In a study of US health records covering 125,474 adults, people who had moderate or severe stomach and bowel side effects were 19% more likely to stop the drug, among those without diabetes.8 The article on what happens after stopping covers what follows.
Do side effects come back after a break?
They can, because a break undoes the adaptation. With a half-life of about 5 days, our model puts the level at 17% of the usual peak two weeks after the last injection, at 6% after three weeks and under 1% after five. A body that has been without the drug for a month meets the next injection almost as it met the first one.
The Zepbound label covers a late dose of up to 4 days and gives no restart dose after a longer gap.1 The Wegovy label does address it for its own drug, and its reason is side effects: after 2 or more missed doses in a row, restart the step-up at a lower dose "to reduce the risk of gastrointestinal adverse reactions".7 If you have been off Zepbound for several weeks, ask your prescriber which dose to resume at before injecting. The article on missed doses has a checker for any gap.
How long do side effects last after stopping?
The label does not give a figure, so the drug level is the best guide to when an effect caused by the drug should fade.
| Days after the last injection | Level, share of the usual peak |
|---|---|
| 7 | 44% |
| 14 | 17% |
| 21 | 6% |
| 28 | 2.4% |
| 35 | 0.9% |
Our one-compartment model with the label's half-life of 5 days.
For the first week after the last dose the level is no lower than in any ordinary week. From the second week it falls below anything the body has seen during treatment, and after about a month very little is left. The label uses the same half-life when it talks about overdose: symptoms may need a period of observation, "taking into account the half-life of tirzepatide of approximately 5 days".1
Effects that follow the weight loss, such as hair loss, do not run on this clock, because they were never tied to the level in the first place. The article on how long Zepbound stays in the body has the full curve by date.
What about the less common effects?
Below the main table the label lists several effects that are too rare for it, each with a number from the same two trials.1
| Effect | Zepbound | Placebo |
|---|---|---|
| Low blood pressure | 1.6% | 0.1% |
| Low blood pressure, in people on blood pressure medicines | 2.2% | — |
| Dry mouth or throat | 1% | 0.1% |
| Change in taste | 0.4% | 0% |
| Altered skin sensation | 0.2% to 0.4% | 0.1% |
| Heart rate | up 1 to 3 beats a minute | no increase |
Low blood pressure is the one with a timing story. The label notes that it was more frequent in people already taking blood pressure medicines, 2.2% against 1.2% in those who were not, and that it "also occurred in association with gastrointestinal adverse events and dehydration". The label does not give its timing. Dizziness on standing is the symptom to report, and people on blood pressure tablets are the group the label singles out.
The label also records average rises in two pancreas enzymes in blood tests, amylase by 20% to 25% and lipase by 28% to 35%. A rise in a test is not pancreatitis, which the trials confirmed in 0.2% of people on Zepbound and 0.2% on placebo.
Who reports stomach side effects more often
The FDA report analysis looked for patterns in who was affected. Stomach and bowel events were reported more often in older adults, in men, and in people taking other medicines at the same time. They were also reported more often when tirzepatide was used for type 2 diabetes than when it was used for weight loss.2 These are patterns in voluntary reports, so they show where to look and do not prove a cause.
Is the timeline different on Mounjaro or with diabetes?
Mounjaro is the same drug on the same step-up schedule, approved for type 2 diabetes, so the timing is the same. Across its five main trials, with 6,263 participants, nausea was reported by 12% to 24% of people on tirzepatide, diarrhea by 12% to 22% and vomiting by 2% to 13%, and the analysis describes them as transient and mild to moderate.3
Two things are different with diabetes. In the FDA report analysis, stomach and bowel events were reported more often when tirzepatide was used for diabetes than for weight loss.2 And low blood sugar becomes a real possibility for people who also take insulin or a sulfonylurea: 4.2% on Zepbound against 1.3% on placebo in the trial of people with diabetes. The label asks those patients to talk through low blood sugar with their prescriber before starting, and to report any change in vision.1
How does this compare with Wegovy?
The Wegovy label has its own table from its own trials.7
| People per 100 | Wegovy 2.4 mg | Its placebo | Zepbound 15 mg | Its placebo |
|---|---|---|---|---|
| Nausea | 44 | 16 | 28 | 8 |
| Diarrhea | 30 | 16 | 23 | 8 |
| Vomiting | 24 | 6 | 13 | 2 |
| Constipation | 24 | 11 | 11 | 5 |
| Stopped treatment because of side effects | 6.8 | 3.2 | 6.7 | 3.4 |
Wegovy label, Table 3, and Zepbound label, Table 1. Different trials, different people.
The Wegovy figures are higher, and so are the figures in Wegovy's placebo group: twice as many people on placebo reported nausea in the Wegovy trials as in the Zepbound trials. The placebo groups show that the two sets of trials are not directly comparable, and it is the reason one column cannot simply be set against the other. The share who stopped because of side effects, a harder measure, is almost the same for the two drugs.
The Wegovy analysis also recorded what people did about their symptoms. Stomach and bowel effects led to a lower dose or a pause in 12.5% of people on Wegovy, against 1.7% on placebo, and to stopping for good in 4.3% against 0.7%. Of the people on Wegovy who had such effects, 95.5% stayed in their trial to the end. Nearly all of the events, 98.1%, were rated mild or moderate.4
The timing is alike. For both drugs the events cluster in the step-up period and decline afterwards, and for Wegovy the peak was at about week 20.4 The head-to-head comparison covers the trial that tested the two drugs against each other.
Short answers to common questions
How long does Zepbound nausea last?
The label gives no number of days. For Wegovy, a similar drug, one episode of nausea lasted a median of 8 days. Episodes cluster after the first dose and after each increase, and the label says they decreased over time.
Do side effects get worse with every dose increase?
The trial table does not show that. Nausea was reported by 25, 29 and 28 in 100 people on 5, 10 and 15 mg. Each step of 2.5 mg is also a smaller change than the one before: the first increase doubles the dose and the last adds a fifth.
When do side effects peak?
For reported stomach and bowel events, at the very start: a quarter had begun within 2 days of the first dose and half within 16 days. Across a whole course they cluster in the weeks of dose increases. The Wegovy data, which tracked the share of people affected week by week, put the peak at about week 20, shortly after the last dose increase.
Is constipation different from the others?
Yes. It lasts longer per episode, a median of 47 days in the Wegovy data, and on Zepbound it was reported less often at higher doses: 17, 14 and 11 in 100.
I have no side effects. Is the drug working?
Side effects are not the measure. People with none of the three main stomach symptoms lost 15.3%, 20.6% and 22.2% of their weight on 5, 10 and 15 mg, almost the same as people who had them.
Side effects started months in, on a steady dose. Is that normal?
It is outside the usual pattern. Over 80% of reported stomach and bowel events began within the first 3 months. A new symptom on a steady dose is worth raising with your prescriber, and pain in the upper stomach with fever or yellowing of the skin is on the label's list of gallbladder signs to report right away.
Will side effects return if I restart after a break?
They can. After five weeks off, under 1% of the usual peak level remains, so the body has lost its adaptation. Ask which dose to restart at.
Which symptoms mean I should stop and get help?
The label names two for which it says to stop: severe stomach pain that will not go away, which can mean pancreatitis, and signs of a serious allergic reaction such as swelling of the face or throat or trouble breathing.
How the calendar and charts work
The calendar builds the schedule from section 2 of the label: 4 weeks on 2.5 mg, then an increase of 2.5 mg after each period you choose, up to your maintenance dose. It shades the time from the first dose to the last increase, and adds a lighter band of 4 weeks, the time the label gives for the level to become steady on a dose. It does not predict that you will have side effects in that period. It marks the period in which the label says most of them occurred.
The onset chart joins four published points from the analysis of FDA reports: a quarter of events by day 2, half by day 16, three quarters by day 70 and, by the paper's own wording, over 80% within the first 3 months, which we plot at day 90. The line between the points is drawn straight, and the sentence under the chart only states which two points your day lies between. The dose chart shows the eight largest figures from Table 1 of the label for the dose you select. The level figures come from our one-compartment model with the label's half-life of 5 days.
Sources
- ZEPBOUND (tirzepatide) prescribing information and Medication Guide, NDA 217806/S-041, revised August 2026. U.S. Food and Drug Administration; sections 2.1, 2.3, 5.3 to 5.6, 6.1, 10, 12.2 and 12.3. accessdata.fda.gov
- Shen P, Peng MS, Kim SJ, Joung KI, Kim KJ. Gastrointestinal adverse events associated with tirzepatide: a bibliometric and pharmacovigilance analysis. PLoS One 2026;21:e0344289. doi:10.1371/journal.pone.0344289
- Patel H, Khunti K, Rodbard HW, et al. Gastrointestinal adverse events and weight reduction in people with type 2 diabetes treated with tirzepatide in the SURPASS clinical trials. Diabetes Obes Metab 2024;26:473-481. doi:10.1111/dom.15333
- Wharton S, Calanna S, Davies M, 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. doi:10.1111/dom.14551
- Rubino DM, Pedersen SD, Connery L, et al. Gastrointestinal tolerability and weight reduction associated with tirzepatide in adults with obesity or overweight with and without type 2 diabetes in the SURMOUNT-1 to -4 trials. Diabetes Obes Metab 2025;27:1826-1835. doi:10.1111/dom.16176
- Frias JP, Nauck MA, Van J, et al. Efficacy and tolerability of tirzepatide in patients with type 2 diabetes: a 12-week, randomized, double-blind, placebo-controlled study to evaluate different dose-escalation regimens. Diabetes Obes Metab 2020;22:938-946. doi:10.1111/dom.13979
- WEGOVY (semaglutide) prescribing information, NDA 215256/S-025, revised June 2026. U.S. Food and Drug Administration; sections 2.4 and 6.1. accessdata.fda.gov
- Rodriguez PJ, Zhang V, Gratzl S, et al. Discontinuation and reinitiation of dual-labeled GLP-1 receptor agonists among US adults with overweight or obesity. JAMA Netw Open 2025;8:e2457349. doi:10.1001/jamanetworkopen.2024.57349
Related analyses
Wegovy side effects: how common each one is, when it peaks and how long it lasts
Zepbound hair loss: how common it is, when it starts, and whether it grows back
Mounjaro side effects: how common each one is by dose, and when they ease
Ozempic side effects: how common each one is by dose, and when they ease
Log side effects next to your doses and see them against your schedule.
Open Dose Insider