Semaglutide Side Effects: What the STEP Trials and Real-World Data Reported

Semaglutide is the single-receptor GLP-1 agonist behind Ozempic and Wegovy, and STEP 1 is the trial that established the modern GLP-1 weight-management category, reporting 14.9% mean body-weight loss at 68 weeks on the 2.4 mg dose. ^[1] Efficacy, though, is only half of what a research buyer or prospective patient needs. This article covers the other half: what side effects the STEP trial program and subsequent real-world research actually reported, the mechanism behind why they happen, when they peak, how they differ by sex, and the less common but more serious warnings worth knowing.

A note on how to read this: side-effect incidence figures are trial-specific and depend on dose, population, and titration schedule. Rather than treat any single percentage as a personal prediction, this piece describes the reported patterns (which effects, how frequent relative to each other, and when in the course they cluster) and attributes each to its source.

What are semaglutide's side effects?

The anchor trial for semaglutide's obesity indication is STEP 1, a 68-week Phase 3 trial published in the New England Journal of Medicine in 2021, which enrolled adults with obesity or overweight and at least one weight-related condition. ^[1] Its safety top-line was that the most frequently reported adverse events were gastrointestinal (nausea, diarrhea, constipation, and vomiting), and these were dose-dependent, mostly mild to moderate in severity, and occurred primarily during the dose-escalation period. ^[1] STEP 3, which paired semaglutide with intensive behavioral therapy and reached 16.0% mean weight loss, reported the same broad GI pattern, as did the earlier SUSTAIN program in type 2 diabetes that supported semaglutide's original approval as Ozempic. ^[3]

Two framing points matter before the detail. First, the dominant effects are on-target: they are the digestible consequence of the same receptor activity that produces the weight loss and the glycemic benefit, not an unrelated toxicity. Second, the trial reported the GI cluster as concentrated in the dose-escalation window rather than sustained at high rates across the full 68 weeks. Both points come directly from how the trial characterized its adverse events. ^[1]

GI side effects: nausea, diarrhea, constipation, and vomiting, and why they happen

The gastrointestinal effects are the headline of semaglutide's tolerability profile, and they are not incidental: they are a direct extension of the mechanism. Semaglutide activates the GLP-1 receptor, which slows gastric emptying, the rate at which the stomach passes its contents to the small intestine. That slowing is part of how the compound reduces appetite and food intake, since food stays in the stomach longer, satiety signals persist, and intake falls. ^[2] The same slowing, when exposure rises faster than the gut adapts, is what produces nausea, and often the diarrhea or constipation that come with a shifting digestive tempo.

This is why the effects are reported as dose-dependent rather than all-or-nothing: a larger dose means more receptor activation, more slowing, and a greater chance the gut is pushed past its current tolerance. It also explains the specific complaints people search for, including semaglutide nausea, semaglutide diarrhea, semaglutide constipation, and reports of yellow diarrhea, which likely reflects altered transit time and bile mixing rather than a distinct toxicity. Delayed gastric emptying can push in either direction on stool consistency, all from the same root cause: a digestive system running at a deliberately slowed pace while it recalibrates. Constipation in particular was reported by some participants as more persistent than nausea, likely because slowed transit does not resolve as quickly as the acute nausea response does once a dose step is tolerated.

The practical implication in the trial data is that GI effects were worst when exposure was changing fastest, during the escalation weeks, which is the reason the program built the dose up in steps over 16 to 20 weeks rather than starting at the 2.4 mg target dose. ^[1] None of that is dosing advice; it is the trial-reported rationale for why the escalation schedule is shaped the way it is.

Semaglutide side effects the first week, and how long nausea lasts

A frequent search is what the first week on semaglutide feels like, and the trial pharmacokinetics explain why it is typically the mildest week from an exposure standpoint even though it is often when people notice the most novelty in how their stomach feels. The starting dose in STEP 1 was 0.25 mg, a dose too low to have a measurable weight-loss effect on its own, used purely to begin the gut's adaptation before the dose climbs. ^[1] Some people notice mild nausea or appetite change even at this low starting dose, which the trial data attributes to the same gastric-emptying mechanism operating at a smaller scale.

Semaglutide has an elimination half-life of approximately 7 days, so plasma levels build gradually and reach steady state in roughly 4 to 5 weeks after any dose change. ^[4] During that ramp, exposure is rising, and rising exposure is exactly the condition that provokes GI effects. Once a dose is held and plasma levels plateau, exposure stops climbing, the gut finishes adapting at that level, and the effects tend to settle, which is why nausea reported early in a dose step is generally described in the literature as easing within days to roughly a week or two as the body adjusts to that specific level.

How common is semaglutide nausea? The trial incidence

Nausea led the adverse-event list across semaglutide's dose groups in STEP 1, reported in a dose-dependent way. ^[1] Diarrhea and constipation were each common and similarly dose-related, while vomiting was reported less frequently than nausea. ^[1] The large majority of these events were characterized in the trial as mild to moderate in severity. ^[1]

Discontinuation due to adverse events is the figure that best captures real-world tolerability burden, because it reflects effects severe enough that a participant stopped the drug entirely. In STEP 1, treatment discontinuation attributable to adverse events was reported in a low single-digit to high single-digit percentage range across the semaglutide arm over 68 weeks, compared with a lower rate on placebo, and gastrointestinal events were the leading cause. ^[1] That figure is useful context: even though a majority of participants experienced some GI symptom during the trial, the overwhelming majority did not find it severe enough to stop.

How semaglutide's side effects were managed in the trials

The trial's answer to tolerability was structural, not reactive: gradual dose titration. The clinical program escalated the dose stepwise over roughly 16 to 20 weeks, from 0.25 mg up through intermediate steps to the 2.4 mg target dose, rather than starting at the target dose. ^[1] Titration works because of the multi-day half-life: each step's exposure layers onto the last, so the body adapts to a slowly rising level rather than a sudden jump. ^[4]

It is worth stating plainly what this section is and is not. Describing the trial's titration schedule is a description of study design, not an instruction to titrate a particular way, on a particular schedule, at a particular dose. Those are clinical decisions made under a prescriber's supervision, outside the scope of this article. The value here is understanding why the escalation exists: it is the trade the program made to buy tolerability while exposure climbed to a level where the efficacy showed up.

Semaglutide side effects in women versus men

A commonly searched question is whether semaglutide's side-effect profile differs by sex. The honest answer is that the STEP trial program, whose populations skewed heavily female, did not report a formal sex-stratified breakdown of GI adverse-event incidence, so a rigorous head-to-head comparison within the semaglutide literature is not well established. What broader pharmacology and adjacent literature suggest is that women report nausea somewhat more frequently than men across many drug classes generally, a pattern related to differences in gastric emptying rate and other physiological factors rather than something specific to semaglutide. Menstrual cycle timing and hormonal status have also been raised as possible modifiers of GI tolerability in general clinical experience, though this is not something the semaglutide trials specifically measured. The GI mechanism itself, slowed gastric emptying from GLP-1 receptor activation, applies the same way regardless of sex, which is the more clearly established point. ^[2]

Thyroid and cancer risk: what the boxed warning means

Semaglutide carries a boxed warning, the FDA's most serious labeling category, for thyroid C-cell tumors. The basis for the warning is preclinical: in rodent studies, GLP-1 receptor agonists caused a dose-dependent and treatment-duration-dependent increase in thyroid C-cell tumors, including medullary thyroid carcinoma. ^[6] Whether this rodent finding translates into increased medullary thyroid carcinoma risk in humans has not been determined, since rodent thyroid C-cells are more sensitive to this signaling pathway than human C-cells appear to be, and long-term human epidemiological data remain limited. ^[6] Because of this uncertainty, semaglutide is contraindicated in people with a personal or family history of medullary thyroid carcinoma or Multiple Endocrine Neoplasia syndrome type 2 (MEN 2), and the labeling advises attention to symptoms like a neck mass, hoarseness, difficulty swallowing, or shortness of breath.

Eye side effects: retinopathy and NAION

Two distinct eye-related signals appear in the semaglutide literature, and they are mechanistically different from each other. The first comes from SUSTAIN-6, semaglutide's cardiovascular outcomes trial in people with type 2 diabetes, which reported a higher rate of diabetic retinopathy complications in the semaglutide group than placebo. ^[5] The leading explanation is not a direct toxic effect on the eye but rather the speed of glucose lowering itself: rapid improvement in blood glucose control is a recognized, if paradoxical, risk factor for temporary worsening of existing diabetic retinopathy, a pattern also seen with rapid insulin intensification, independent of the specific drug used to achieve it.

The second, more recently reported signal is an observational association between semaglutide and nonarteritic anterior ischemic optic neuropathy (NAION), a rare condition involving sudden, typically painless vision loss in one eye due to reduced blood flow to the optic nerve. ^[8] This finding comes from retrospective observational research rather than a randomized trial, which means it can identify an association but cannot establish that semaglutide causes the condition; confounding by the underlying conditions semaglutide is prescribed for (obesity, diabetes, cardiovascular risk factors that independently raise NAION risk) has not been fully ruled out. The practical takeaway is a symptom to know: sudden vision loss or blurring in one eye is a reason for urgent ophthalmologic evaluation, not something to wait out.

Kidney side effects: what's reported

Acute kidney injury has been reported rarely in association with semaglutide and other GLP-1 receptor agonists, and it is a labeled consideration for the drug class. The mechanism generally invoked is not a direct toxic effect on the kidney but rather dehydration secondary to severe GI effects, particularly vomiting and diarrhea, which can reduce kidney perfusion and, in susceptible individuals, precipitate acute kidney injury. This is the same indirect pathway described for other drugs that cause significant GI fluid loss. Adequate hydration during periods of GI upset, especially during dose escalation, is the general clinical guidance associated with this risk, and persistent vomiting or diarrhea alongside reduced urination is a reason for prompt medical evaluation rather than something to manage by waiting it out.

Hair loss, fatigue, and headache: other commonly searched effects

Hair shedding is a commonly searched concern across the entire weight-loss field, including semaglutide, tirzepatide, and rapid weight loss by any means. The leading explanation invoked across the field is telogen effluvium, a temporary, stress-related shift in the hair growth cycle triggered by rapid weight loss and the nutritional changes that accompany a sharp drop in intake, rather than a direct toxic effect of the drug on hair follicles. This framing should be confirmed by a clinician, since it is not characterized in the semaglutide trial literature as a primary, direct adverse effect in the way the GI events are.

Fatigue is also commonly reported, and in the trial framing it is generally attributed to the combination of reduced caloric intake, the metabolic adjustment to a lower body weight, and, for some, the GI effects themselves rather than to a specific pharmacologic mechanism unique to semaglutide. Headache was reported in the trial program as well, though it was not characterized as clearly dose-dependent the way the GI effects were, and it is a common nonspecific adverse event across many trials regardless of arm.

Semaglutide vs tirzepatide side effects: is one gentler?

Because tirzepatide activates both GIP and GLP-1 receptors while semaglutide activates GLP-1 alone, a natural question is whether the added receptor changes the tolerability profile. The honest answer starts with a caveat: the two compounds have not been extensively compared head-to-head for tolerability at matched doses within the same trial, so most of what follows is drawn from separate trials with different populations and titration schedules.

What those separate trials show is broad class similarity. In semaglutide's STEP program and tirzepatide's SURMOUNT program, the most common adverse events were gastrointestinal, with nausea leading the list in both, and were dose-dependent, mostly mild to moderate, and concentrated early. ^{[1] [7]} Some analyses have suggested tirzepatide's added GIP activity may soften the GI burden somewhat relative to GLP-1 alone, since GIP signaling is theorized to counteract some GLP-1-driven nausea in animal models, but this remains an active area of research rather than a settled comparative finding, and cross-trial rate comparisons are confounded by different doses and populations. The defensible statement is narrower: the GI profile is a class property, driven substantially by the shared GLP-1 mechanism, and it is managed the same way across both compounds, gradual titration timed to each drug's pharmacokinetics.

How the side-effect timeline relates to when semaglutide starts working

A related troubleshooting search, semaglutide not working or why am I not losing weight, usually describes the same early weeks that produce the GI effects, and the pharmacology offers a clean explanation that ties the two together. Weight-loss and glycemic effects in the trials were dose-dependent, and plasma levels take roughly 4 to 5 weeks to reach steady state after any dose change. ^{[1] [4]} That means the early, low-dose weeks, the same weeks when GI effects are most likely, are also the weeks when the metabolic effect is weakest, because exposure has not yet built to the level where the trial measured its results. Reading the first few weeks as the verdict is a common misread; STEP 1's primary endpoint was measured at 68 weeks for exactly this reason.

Putting the side-effect profile in context

Semaglutide's tolerability profile reads as a coherent whole once the mechanism is in view. The dominant effects are gastrointestinal and on-target: the digestible cost of slowed gastric emptying, the same process that drives appetite reduction. They are dose-dependent and concentrated in the 16 to 20 week escalation window, easing as exposure plateaus, and only a modest share of participants discontinued because of them. ^[1] Less common but more serious signals, the thyroid C-cell tumor warning, diabetic retinopathy complications, the observational NAION association, and rare kidney injury tied to dehydration, are recognized concerns worth knowing even though they affect a small minority. The whole GI profile was managed in the trials through gradual titration timed to the compound's roughly 7-day half-life.

For a research or clinical decision, the useful takeaway is that the side-effect timeline and the efficacy timeline are the same timeline, both governed by how exposure builds toward steady state. Understanding that one curve explains both why the early weeks feel bumpy and why they understate the compound's eventual effect.

For research and educational purposes only. Not medical advice. Side-effect patterns and incidence figures describe findings from published clinical trials, observational research, and FDA labeling information; they are not predictions, instructions, or management advice for any individual. FDA-approved semaglutide (Ozempic, Wegovy) is available only by prescription and under clinician supervision. Research-labeled semaglutide is not approved for human use.