Answer first: GLP-1 side effects do not follow the calendar. They follow the dose ladder. Both flagship injectables, Wegovy (semaglutide 2.4 mg) and Zepbound/Mounjaro (tirzepatide), change the dose every four weeks by design, and the trials consistently report that gastrointestinal events cluster around those dose escalations, not around arbitrary week numbers. A page that tells you what happens in "week 5" without telling you which dose step week 5 is has skipped the only fact that makes the timeline predictable.
This guide maps the side effects to the actual titration ladders, row by row, and attaches the trial and label numbers to each step: STEP 1 for semaglutide (nausea in roughly 44% of participants versus placebo), SURMOUNT-1 for tirzepatide (nausea rising from roughly a quarter of participants at 5 mg to roughly a third at the highest doses), the 2026 Wegovy label's dose-escalation breakout, and the 2025 pooled analysis across the SURMOUNT program that settled when these events occur and how much they actually contribute to weight loss (less than most people assume).
This page is educational only. It does not provide medical advice, does not tell any individual what dose to take or when to escalate, and cannot substitute for the prescribing clinician who can see the full medical record. It reports what the trials, the FDA labels, and published clinical studies say, with the evidence grade attached to every claim.
One framing note before the table: the numbers below are group averages from trials of thousands of people. They describe how often an event occurred in a trial arm, not what any one person will experience on any one dose step.
Side-by-side comparison
| Compare | What trials and labels report | Typical pattern at this step | Evidence grade |
|---|---|---|---|
| Wegovy 0.25 mg (weeks 1–4) | The introductory semaglutide step, held four weeks before the first escalation. The 2026 Wegovy label pools the entire escalation period (weeks 1–16): nausea 42% vs 18% on placebo, diarrhea 22% vs 19%. | First exposure to the drug. STEP 1 describes its GI events as predominantly mild-to-moderate and transient (Wilding et al., NEJM 2021). | FDA label |
| Wegovy 0.5 mg (weeks 5–8) | First escalation. In STEP 1 (68 weeks, n=1,961), GI events typically occurred during dose escalation: nausea ~44%, diarrhea ~32%, vomiting ~24%, constipation ~24%, all versus placebo (Wilding et al., NEJM 2021). | First escalation injection. The clinically described pattern: peak 24–72 hours after the new dose, settling within 1–2 weeks. | Trial (STEP 1) |
| Wegovy 1.0 mg (weeks 9–12) | Mid-ladder escalation. The 2026 label's escalation-period breakout puts vomiting at 36% vs 10% on placebo across the escalation window — the widest drug-versus-placebo gap of the GI effects. | Same post-escalation shape as every step change: worst in the first days after the injection, improving across the four-week block. | FDA label |
| Wegovy 1.7 mg (weeks 13–16) | Final escalation step before maintenance. STEP 1 reported GI events as the leading cause of discontinuation, concentrated in the escalation phase rather than stable dosing (Wilding et al., NEJM 2021). | Last climb step; STEP 1's events were transient, so the described trajectory is improvement within the block, not accumulation. | Trial (STEP 1) |
| Wegovy 2.4 mg (week 17+, maintenance) | Stable maintenance dose; the label's escalation-period breakout no longer applies. STEP 1's GI events were transient, and tirzepatide labeling across the class describes events clustering in escalation rather than stable dosing. | New persistent symptoms at week 17+ are not the expected escalation pattern and belong with a clinician. | FDA label / Pooled trials |
| Tirzepatide 2.5 mg (weeks 1–4, starter) | Everyone starts here per the label — a tolerability starter that is sub-therapeutic for weight, not an efficacy dose. Tirzepatide labeling states the majority of nausea, vomiting, and diarrhea reports occurred during dose escalation and decreased over time. | Lowest exposure on the ladder; the step exists to establish tolerability before efficacy dosing begins. | FDA label |
| Tirzepatide 5 mg | First efficacy dose. SURMOUNT-1 (72 weeks): nausea ~24–27% vs ~9% on placebo; discontinuation due to adverse events 4.3% vs 2.6% on placebo (Jastreboff et al., NEJM 2022; per-dose figures approximate). | First real escalation test of tolerance; the same described 24–72-hour peak window applies after the step-up injection. | Trial (SURMOUNT-1) |
| Tirzepatide 7.5–10 mg | The range where most users settle per the label's maintenance options (5/10/15 mg). Nausea ~32–33% at 10 mg vs ~9% on placebo in SURMOUNT-1. | Mid-ladder steps; each step change reopens the described peak window, with the pooled data showing events diminish across the titration period. | Trial (SURMOUNT-1) |
| Tirzepatide 12.5–15 mg | Highest steps. SURMOUNT-1: nausea ~31–36% at 15 mg; discontinuation due to adverse events 6.2% at 15 mg vs 2.6% on placebo — dose-related but modest in absolute terms. | Top of the ladder; rates rise with dose, yet the gap versus placebo on discontinuation stays within a few percentage points. | Trial (SURMOUNT-1) |
| Tirzepatide program: the pooled escalation window | Pooled participant-level analysis across SURMOUNT-1 to -4 (Rubino et al., Diabetes Obes Metab 2025) — tirzepatide trials only: GI events occurred primarily during the 20-week titration period, diminished over time, and rarely led to discontinuation (1.0–10.5% across arms); rescue medication use clustered in the first 24 weeks. | The tirzepatide population-level timing curve: worst during the climb, quieter once the maintenance dose holds. The 2026 Wegovy label's escalation-period breakout shows the same shape for semaglutide. | Pooled trials (Rubino 2025) |
| Timing at any single escalation step | No Phase 3 publication reports hour-level symptom timing. The 24–72-hour peak and 1–2-week settle is a clinically described telehealth pattern, directionally consistent with the labels' escalation-window language. | Treat as descriptive, not proven — the trial-backed claim is the coarser one: events cluster in escalation weeks and diminish on stable doses. | Clinically described pattern |
GLP-1 side effects mapped to the actual titration ladders: what the trials and labels report at each dose step, with an evidence grade per row. Trial percentages are group rates versus placebo, not individual predictions.
Why generic week-by-week guides mislead: dose weeks are not calendar weeks
Most pages ranking for "GLP-1 side effects week by week" divide the first few months into generic calendar blocks: week 1, week 2, weeks 3–4, and so on. That structure reads neatly and matches nothing about how these drugs are actually taken. Wegovy changes the dose at weeks 5, 9, 13, and 17. Zepbound and Mounjaro change the dose after every four-week step once escalation begins. A side-effect timeline that ignores the ladder is describing a drug that does not exist.
The practical failure of calendar-week framing shows up immediately. "Week 5" for a Wegovy patient is always the first week at 0.5 mg — an escalation week, when the trials say GI events are most likely. "Week 6" on the same dose is pharmacologically identical to week 5's dose but is usually a settling week. Two patients both in "week 9" can be on different drugs at different steps entirely. The only unit that makes the timeline comparable between people is the dose step, not the week number.
The ladder also explains something generic pages treat as mysterious: why symptoms seem to return every month. If nausea faded by the end of a four-week step and came back days after the next injection, that is not a random recurrence — it is the escalation pattern the labels and trials describe. Conversely, someone whose clinician has held them at a stable dose for months is on a different curve entirely, one the pooled data describes as diminishing over time (Rubino et al., Diabetes Obes Metab 2025).
The two titration ladders, exactly as the labels define them
Wegovy (semaglutide 2.4 mg for weight management): 0.25 mg for weeks 1–4, then 0.5 mg for weeks 5–8, then 1.0 mg for weeks 9–12, then 1.7 mg for weeks 13–16, then 2.4 mg from week 17 onward as the maintenance dose. Every step is four weeks, and the schedule exists specifically to improve gastrointestinal tolerability — the label says so outright. The final dose is 1.7 mg on the way up, not 1.74.
Zepbound and Mounjaro (tirzepatide): every patient starts at 2.5 mg for four weeks — a starter dose that is sub-therapeutic for weight and exists for tolerability — then escalates 5, 7.5, 10, 12.5, and 15 mg with a minimum of four weeks per step. Approved maintenance doses are 5, 10, and 15 mg, and most users settle at 7.5–10 mg. Not everyone climbs the whole ladder; the label treats each escalation as a decision, not an obligation.
Two structural differences matter for side effects. First, tirzepatide's ladder is longer (up to 20 weeks to reach the top vs Wegovy's 16), meaning more escalation events, each of which is a chance for the post-escalation symptom pattern to recur. Second, semaglutide's ladder tops out at one fixed maintenance dose while tirzepatide's has three approved maintenance levels — so "typical" tirzepatide exposure varies more between patients. Both ladders share the design logic: slow enough for the gut to adapt at each dose, fast enough to reach an effective dose within a few months.
What the trials actually report at each step
For semaglutide, the anchor is STEP 1 (Wilding et al., NEJM 2021; 1,961 adults with obesity, 68 weeks): nausea in roughly 44% of participants, diarrhea roughly 32%, vomiting roughly 24%, and constipation roughly 24%, each versus placebo, with events predominantly mild-to-moderate and transient. The trial's key timing finding is structural: GI events typically occurred during dose escalation, and they were the leading cause of discontinuation. The 2026 Wegovy label adds the escalation-period breakout: nausea 42% vs 18% on placebo and vomiting 36% vs 10% during escalation, with diarrhea closer between groups (22% vs 19%).
For tirzepatide, the anchor is SURMOUNT-1 (Jastreboff et al., NEJM 2022; 2,539 adults with obesity, 72 weeks): nausea rose with dose — roughly 24–27% at 5 mg, 32–33% at 10 mg, and 31–36% at 15 mg, versus roughly 9% on placebo (per-dose figures approximate). Discontinuation due to adverse events stayed modest across arms: 4.3% at 5 mg, 7.1% at 10 mg, and 6.2% at 15 mg, versus 2.6% on placebo. Tirzepatide labeling summarizes the same pattern in plain language: the majority of nausea, vomiting, and diarrhea reports occurred during dose escalation and decreased over time.
The strongest synthesis comes from the pooled participant-level analysis across SURMOUNT-1 through -4 (Rubino et al., Diabetes Obes Metab 2025): any GI adverse event occurred in 27.8–72.8% of tirzepatide arms versus 12.2–32.5% on placebo, severe or serious GI events were rare (0.3–5.6%), events occurred primarily during the 20-week titration period and diminished afterward, and rescue antiemetic or antidiarrheal use clustered in the first 24 weeks. Read together, the three evidence layers agree on the core claim of this page: the side-effect timeline of a GLP-1 medication is, to a first approximation, the shape of its titration ladder.
When symptoms peak and when they settle
Telehealth sources that treat patients on these drugs converge on a consistent description: symptoms tend to peak 24–72 hours after an escalation injection and settle within one to two weeks at the new dose. PeptaHub labels this a clinically described pattern, not a trial finding — no Phase 3 publication reports hour-level symptom timing. But it is directionally consistent with what the trials and labels do say: events cluster in escalation windows and diminish on stable dosing.
The settling half of the pattern is better evidenced than the peaking half. The tirzepatide label states that reports decreased over time; the pooled SURMOUNT analysis found events diminished after the titration period and rarely led to discontinuation; STEP 1 described its GI events as transient. For someone on a stable maintenance dose who has a rough few days every month, the pattern to watch for is whether those days align with the injection and whether they shorten over successive months — the described trajectory is improvement, not accumulation.
What this framework should not become is a reason to dismiss symptoms. A pattern that peaks and settles is the expected course; a pattern that intensifies across weeks on a stable dose, or that crosses into the red-flag territory in the next section, is not on the described trajectory at all and belongs with a clinician regardless of where it falls on the ladder.
The myth buster: GI events explain at most a sliver of the weight loss
The most persistent folk explanation for GLP-1 weight loss is that the drug "only works because it makes you sick" — that nausea and reduced intake from feeling unwell do the real work. The trial data refutes this directly. In the pooled participant-level analysis across SURMOUNT-1 to -4, a formal mediation analysis attributed at most approximately 3.1% of weight loss to gastrointestinal events (Rubino et al., Diabetes Obes Metab 2025). The other ~97% of the effect is attributable to the medication's pharmacology — central appetite regulation, slowed gastric emptying contributing to satiety, and, for tirzepatide, the added GIP receptor pathway — not to illness.
The mediation number also settles a related worry from the other direction: patients with minimal side effects sometimes fear the drug is "not working" on them. If GI events mediated most of the effect, that fear would be rational; since they mediate at most a small fraction, the absence of nausea carries no information about response. Trial participants who lost substantial weight without GI events are the rule, not an anomaly.
This is also why "do side effects mean it's working?" is the wrong mental model entirely. The side-effect curve tracks the dose ladder; the weight-loss curve tracks the pharmacology. They move through the same calendar but are not the same signal — which is exactly why a titration-mapped page like this one separates them row by row.
Red-flag symptoms that are never a normal part of titration
The escalation pattern described above covers nausea, vomiting, diarrhea, and constipation in their mild-to-moderate, transient forms. The prescribing labels define a separate category of symptoms that are not tolerability events at any dose step and warrant prompt medical attention. This section reports what the labels say; it is not a diagnostic guide, and any concerning symptom belongs with a clinician or urgent care, not with a wait-and-see approach.
Per the labels: severe, persistent abdominal pain, especially pain that may radiate to the back, is the hallmark warning sign for acute pancreatitis and instructions are to stop the medication and contact a clinician. Gallbladder warnings cover cholelithiasis and cholelithiasis-related complications — trials across the class report gallbladder events at higher rates than placebo, plausibly linked to rapid weight loss itself. Persistent vomiting or diarrhea that prevents eating and drinking, with signs of dehydration, is a label-level concern during escalation periods in particular. Both drugs also carry the class boxed warning regarding medullary thyroid carcinoma and MEN2 history, based on rodent findings.
The framing that separates this section from the rest of the page: everything else in the titration table describes an expected, monitored, generally self-limiting pattern. The items in this paragraph do not sit anywhere on that spectrum — they are orthogonal to the ladder. Escalation step zero or maintenance week sixty, the answer is the same: medical evaluation, not patience.
The discontinuation reality: what stopping actually looks like in the trials
GI events during escalation are the leading cause of discontinuation in STEP 1 (Wilding et al., NEJM 2021) — the single most common exit ramp from semaglutide treatment. The absolute numbers are smaller than the folk reputation suggests: SURMOUNT-1 discontinuation due to adverse events ran 4.3–7.1% across tirzepatide arms versus 2.6% on placebo (Jastreboff et al., NEJM 2022), and the pooled SURMOUNT analysis found GI events led to treatment discontinuation in 1.0–10.5% of participants across arms (Rubino et al., Diabetes Obes Metab 2025). Most people who start these medications in trials stay on them; the ones who stop mostly stop early, during the climb.
That structure has a practical implication the labels support: the ladder is adjustable. Delaying an escalation to let a current dose settle is standard, label-consistent clinical practice — the four-week steps are minimums, not a fixed countdown, and titration decisions belong to the prescribing clinician. A patient struggling at a step is not failing the schedule; the schedule exists to be tuned. What the trials do not support is pushing through red-flag symptoms to stay on schedule, since nothing in the ladder's design requires tolerating anything from the previous section.
For context on the other side of the stopping decision: STEP 4 showed that withdrawing semaglutide after weight loss leads to substantial regain, so discontinuation for tolerability is a genuine cost, not a neutral off-ramp. That is exactly why the escalation-tuning path — slower steps, held doses, clinician-managed symptom support — is the common middle route between toughing it out and quitting, and why conversations about a difficult step belong with the prescriber rather than at the pharmacy window alone.
Frequently asked questions
Because the dose changes every four weeks. Both Wegovy (semaglutide) and Zepbound/Mounjaro (tirzepatide) use titration ladders that escalate the dose after each four-week step, and the trials consistently report that GI events cluster around those escalations: STEP 1 found events typically occurred during dose escalation, the tirzepatide label states the majority of reports occurred during escalation and decreased over time, and the pooled SURMOUNT-1 to -4 analysis found events concentrated in the 20-week titration period (Wilding et al., NEJM 2021; Rubino et al., Diabetes Obes Metab 2025). Symptoms that return on schedule after each escalation injection are the described pattern; a calendar-week guide that never mentions the ladder cannot explain them.
There is no clean head-to-head answer, and any page offering one is comparing across trials. The reference points: STEP 1 reported nausea in roughly 44% of semaglutide 2.4 mg participants versus placebo (Wilding et al., NEJM 2021), while SURMOUNT-1 reported roughly 31–36% at tirzepatide's highest doses versus ~9% on placebo (Jastreboff et al., NEJM 2022) — similar magnitudes from different designs, populations, and dose exposures. Tirzepatide's design includes receptor-bias features thought to contribute to tolerability, and head-to-head data in type 2 diabetes (SURPASS-2) did not show a dramatic tolerability separation. The honest summary: both drugs share the same GI-dominated, escalation-timed side-effect profile; individual tolerance varies more than the average difference between the drugs.
No. This is one of the most testable myths in the category, and the data refutes it: a formal mediation analysis pooling SURMOUNT-1 through -4 attributed at most approximately 3.1% of weight loss to gastrointestinal events (Rubino et al., Diabetes Obes Metab 2025). The weight loss comes from the pharmacology — central appetite regulation and, for tirzepatide, dual GIP/GLP-1 receptor activity — not from being sick. Feeling nothing at a dose step carries no information about whether the medication is working, and feeling awful is not evidence that it is working harder.
The labels define symptoms that are never part of normal titration at any dose step: severe, persistent abdominal pain (especially radiating to the back), which is the labeled warning sign for acute pancreatitis; gallbladder symptoms such as pain in the upper abdomen after meals; and persistent vomiting or diarrhea with signs of dehydration. These warrant prompt medical contact — the label language for suspected pancreatitis is to stop the medication and contact a clinician. Beyond the red flags, symptoms that intensify week over week on a stable dose, or any tolerability question about an escalation step, are prescribing-clinician conversations; the four-week schedule is a minimum, and delaying an escalation is standard label-consistent practice.
The ladders are designed to be adjustable, and nothing in either label requires climbing on a fixed countdown. The four-week steps are minimums, holding a dose longer to let symptoms settle is standard clinical practice, and dose selection at every step is a decision for the prescribing clinician. The trial context: GI events during escalation were the leading cause of discontinuation in STEP 1 (Wilding et al., NEJM 2021), yet absolute discontinuation rates stayed modest — 4.3–7.1% across tirzepatide arms in SURMOUNT-1 versus 2.6% on placebo — meaning most participants who hit a rough step continued. Educational framing only: this describes what labels and trials report, not what any individual should do.
On the clinically described pattern reported by telehealth sources, symptoms peak 24–72 hours after an escalation injection and settle within one to two weeks at the new dose — a pattern, not a trial finding, though it aligns with label language. The better-evidenced version is step-level: the pooled SURMOUNT analysis found GI events occurred primarily during the 20-week titration period and diminished over time on stable dosing (Rubino et al., Diabetes Obes Metab 2025), and STEP 1 described its GI events as transient (Wilding et al., NEJM 2021). The trajectory to expect across the ladder is recurring-but-shortening episodes at each escalation, fading toward background once the maintenance dose is reached.