How Long Do Semaglutide Side Effects Last? The Answer, in Days

Almost nobody asks this question before the first injection. They ask it somewhere around day four, sitting on the edge of the bed at seven in the morning, deciding whether to eat anything. The question is not "is nausea a known side effect." Everyone knows that. The question is how long do semaglutide side effects last, and the answer people get back from most sources is a shrug dressed up as reassurance: they are usually temporary, they tend to improve, talk to your prescriber. Temporary is not a number. You cannot plan a work week around temporary.

The original work in this piece: the trial literature does report durations in days, and it is buried in a supplementary tolerability analysis that almost no patient-facing article quotes. We pulled those numbers out and laid them alongside the dose escalation calendar, so you can see not just how long an individual episode lasts but which week of treatment you are statistically most likely to feel worst. The two things together are the actual answer. Neither one alone is.

How long do semaglutide side effects last, measured in days rather than adjectives? The best available answer comes from a pooled tolerability analysis of the STEP 1 through 3 trials published in Diabetes, Obesity and Metabolism in 2022, covering 2,117 people on semaglutide 2.4 mg once weekly for 68 weeks against 1,262 on placebo. Using a Kaplan-Meier estimate of individual events, the median duration of a nausea event was 8 days. Diarrhea was 3 days. Vomiting was 2 days. Constipation was 47 days. Those are medians, so half of all episodes ran longer, but they are real numbers from more than three thousand people rather than a clinical impression.

The number that should reframe the whole conversation is the constipation figure. Every consult, every pamphlet, every forum thread leads with nausea. Nausea is the loud one. But in this dataset the median nausea episode resolved inside of a week and a half, while the median constipation episode ran nearly seven weeks, and constipation prevalence plateaued at around week 10 and stayed elevated for the entire 68-week treatment period rather than fading. In other words, the side effect patients are warned about most is the one that ends fastest, and the side effect treated as a footnote is the one most likely to still be there next season. If your gut has slowed and stayed slow, you are not an outlier, and the fix is a protocol rather than patience, which is what we cover in constipation on a GLP-1.

The second surprise is the calendar. Every STEP trial used the same ladder: 0.25 mg once weekly for four weeks, then a step up every four weeks to 0.5, then 1.0, then 1.7, reaching the full 2.4 mg dose at week 16. The intuitive assumption is that the first four weeks are the hardest and it improves from there. The prevalence curves say the opposite. Nausea, diarrhea and vomiting prevalence in the semaglutide arm peaked at roughly week 20 and declined afterward, with nausea falling most sharply. Week 20 is four weeks after arriving at the top dose, not four weeks after starting. So the person who tells you "the first month is rough and then you are fine" is describing something the pooled data does not show. Each step up is a fresh provocation, and the biggest one is the last one. That is the practical reason your prescriber refuses to move faster, which we unpack in why your GLP-1 dose starts so low.

How common is any of this, in context. In the same pooled analysis, nausea affected 43.9 percent of the semaglutide group against 16.1 percent on placebo, diarrhea 29.7 against 15.9, vomiting 24.5 against 6.3, and constipation 24.2 against 11.1. The severity picture matters as much as the frequency: 99.5 percent of gastrointestinal adverse events were non-serious and 98.1 percent were mild to moderate, and only 4.3 percent of semaglutide-treated participants stopped treatment permanently because of them. The placebo columns are worth staring at for a second, because roughly one in six people reported nausea while injecting nothing at all. Some of what gets attributed to the drug is the ordinary background noise of being a human with a digestive system, which is a useful thing to know before you blame every off day on the pen. The broader safety picture, including the rare events that are not about tolerability at all, sits in the full picture of GLP-1 side effects.

An original four-part log you can start this week. This is not a diagnostic tool and it does not replace your prescriber. It exists because "how long does this last" is unanswerable without knowing what your own pattern looks like, and almost nobody records it. First, write your injection day at the top of a page and number the seven days beneath it. Second, for each day, note only two things: the symptom and a severity from one to three, where one means you noticed it, two means it changed what you ate or did, and three means you could not work or keep fluids down. Third, mark the day each episode started and the day it fully stopped, and count the days between. Fourth, keep the same page running through your next dose increase. Now read it. An episode that starts one to three days after the injection and clears within about a week and a half is sitting on top of the published median and needs management, not alarm. An episode that keeps running past two weeks, or a constipation pattern with no bowel movement for three or more days, has left the typical range and is a call to the clinic. A severity of three at any point, or any symptom that arrives for the first time weeks after a dose has been stable, is not a titration story and should be evaluated rather than logged. And if the same eight-day nausea episode recurs at every single step up, that is the pattern that argues for a slower ladder rather than a stronger anti-nausea plan, which is a different conversation from the one in managing GLP-1 nausea.

What the studies do not tell you. Three specific gaps, stated plainly. The first is that these are durations of individual events, not of a person's experience. Someone who gets a fresh 8-day nausea episode at each of four dose steps has not had 8 bad days, they have had roughly a month of them spread across four months, and no published figure describes that person. The second is that every number here comes from 2.4 mg branded semaglutide administered on a rigid four-week ladder inside a trial. Compounded semaglutide, microdosing schedules and faster telehealth escalations have no equivalent published duration data, so anyone quoting these numbers at you for a non-trial protocol is extrapolating. The third is survivorship: these durations were measured in people who stayed enrolled. The 4.3 percent who quit for gastrointestinal reasons took their worst episodes out of the dataset with them, which means the published medians are, if anything, slightly optimistic. None of that makes the numbers useless. It makes them a floor rather than a promise.

For general orientation on what these medications are and where they sit among the options, the NIDDK overview of prescription obesity medications and the Mayo Clinic drug reference for subcutaneous semaglutide are both worth a read. The pivotal efficacy trial behind the 2.4 mg dose, for anyone who wants the primary source, is Wilding and colleagues in the New England Journal of Medicine, PMID 33567185.

The honest summary is this. Most semaglutide side effects are short at the level of the episode and long at the level of the treatment course, they peak later than anyone expects, and constipation is the one that overstays. If your log shows episodes that fit the published pattern, the reasonable move is to manage them and hold the ladder steady. If it shows escalating severity, a symptom that will not resolve, or a life you cannot actually run, that is worth raising early rather than quietly quitting, and it is also the point at which some people start weighing the other routes, which is the ground covered in when weight-loss surgery makes sense.