Switching From Semaglutide to Tirzepatide: The Evidence Guide

The switch question exists because of one trial. SURMOUNT-5 (Aronne et al., NEJM 2025; PMID 40353578) was the first head-to-head Phase 3 randomized comparison of tirzepatide against semaglutide at its full weight-management dose in adults with obesity without diabetes. Over 72 weeks, tirzepatide produced 20.2% mean weight loss (22.8 kg) versus 13.7% (15.0 kg) for semaglutide 2.4 mg — roughly a 6.5 percentage-point separation, with tirzepatide superior at every threshold measured: ≥10%, ≥15%, ≥20%, and ≥25% body-weight loss. Waist circumference fell more on tirzepatide as well.

The plateau is the other half of the why. The only published switch study enrolled people with type 2 diabetes who had lost under 3% of their weight on semaglutide — inadequate responders, not average ones (Kurinami et al., Endocrinol Metab 2025). That is the typical switch profile the telehealth literature describes too: not "semaglutide failed," but "response at an adequate dose was below expectations, and a mechanically different drug with a larger average effect is available." An inadequate response on one agent does not predict an inadequate response on the other, because the two drugs do not engage the same receptors.

The mechanistic reason tirzepatide averages better is receptor coverage: semaglutide is a GLP-1 receptor agonist only, while tirzepatide adds GIP receptor agonism, engaging adipose-tissue metabolism, additional insulin-secretion pathways, and further appetite circuits that GLP-1-only drugs do not touch. For an individual, SURMOUNT-5 settles the population-level question — tirzepatide wins on average — but not the personal one, which is a matter for the prescribing clinician weighing the full record.

The single most important fact on this page: 1 mg of semaglutide does not equal any number of milligrams of tirzepatide. The molecules are unrelated in structure (a 31-amino-acid GLP-1 analog vs a 39-amino-acid dual agonist), act on overlapping but different receptor sets (GLP-1-only vs GIP plus GLP-1), and have different half-lives (~7 days vs ~5 days). Pharmacological equivalence requires comparable molecules at comparable exposures; none exists here. As pharmacist-facing CPD guidance puts it, switching between weight-loss medications means restarting the titration ladder, not translating a dose.

The conversion charts that circulate on social media and some clinic pages are editorial conventions — someone's idea of a "comfortable equivalent" — with no basis in pharmacology or any regulatory label. The practical risk they create is dose-stacking: treating a high semaglutide dose as justifying a higher-than-starter tirzepatide dose ignores that the new drug's own titration schedule exists to let the gut adapt to a different molecule at different exposures.

The half-life difference does have one legitimate role in the switch conversation: it frames the gap week (covered next). Semaglutide levels decline slowly after the last injection, and tirzepatide builds to steady state over its first weeks — which is why the two drugs are sequenced with a washout-style gap rather than overlapped. That is the entire extent of what the pharmacokinetics supports; it supports no numeric conversion whatsoever.

Four elements appear consistently across every credible source. First, stop semaglutide — the last weekly injection simply runs out rather than being tapered. Second, wait a gap of roughly three to ten days, about one week in most descriptions (Affinity Whole Health, Fay) — long enough for semaglutide levels to begin falling, short enough to limit regain and keep the treatment continuum. Third, start tirzepatide at the 2.5 mg starter dose. Fourth, titrate in the label's standard four-week steps, exactly as a tirzepatide-naive patient would.

The 2.5 mg start deserves emphasis because it is the point where internet advice diverges most sharply from every serious source. In the only published switch study, all 15 patients started at 2.5 mg regardless of their prior semaglutide dose (Kurinami et al., 2025). Telehealth protocols describe the same default, with one clinician-judgment carve-out: Ro describes starting at 2.5 mg or 5 mg even for patients on high semaglutide doses, and Fay notes that a 5 mg start may be considered when semaglutide was well tolerated — a prescriber decision, not a patient-side rule. The 2.5 mg dose is sub-therapeutic for weight by design; it exists to establish tolerability before efficacy dosing begins.

Educational framing, stated plainly: the above is a description of what studies, labels, and clinical protocols report — not instructions for any individual. Timing the gap, choosing a start dose, and filing the new prescription are prescribing-clinician decisions involving contraindications, coverage, and history that no web page can see.

The entire dedicated literature on this switch is one 2025 retrospective case series: Kurinami et al. (Endocrinol Metab Seoul 2025;40(6):1012-15; PMC12765867), 15 patients with type 2 diabetes at a single Japanese center who switched from semaglutide 1.0 mg after losing under 3% of body weight. Every patient started tirzepatide at 2.5 mg with standard four-week titration steps. Outcomes split by where titration landed: the group that escalated early to 10 mg (n=5) achieved a significant HbA1c reduction (−0.7%, from 6.6% to 5.9%, P<0.01) and lost 6.6 kg on average (not statistically significant), while the 7.5 mg group (n=10) showed no significant changes. Across all 15 switchers there were zero GI adverse events and zero discontinuations — which the authors plausibly attribute to the fact that every participant had already tolerated semaglutide.

The limitations are the findings' shadow, and stating them is the difference between evidence grading and hype: 15 patients, retrospective design, single center, a type 2 diabetes cohort (not the obesity-without-diabetes population of SURMOUNT-5), and subgroup analyses of 5 and 10 people. The authors' suggestion that early escalation to 10 mg may benefit inadequate responders is explicitly hypothesis-generating. It is a real signal from real switchers — the only one published — and it should carry exactly the weight a 15-patient retrospective signal carries.

Why this study matters anyway: Google ranks it on page one for the consumer switching query, because it is the closest thing to a primary source that exists. A reader who finds it deserves the context around it — which is the job of the table above and this page's grading.

The transition re-runs the titration clock. Starting tirzepatide at 2.5 mg after semaglutide means beginning a new ladder from its bottom rung, and the side-effect pattern of the first months on tirzepatide is the same escalation-linked GI pattern described for any new starter: events cluster at dose changes, peak in the days after an escalation injection per the clinically described pattern, and settle as dosing stabilizes. PeptaHub's GLP-1 side effects week-by-week guide maps that pattern to every step of both ladders, with trial rates per dose.

The switch-specific twist is that prior semaglutide tolerance may soften the re-titration. The zero-GI-event finding in the 15-patient series suggests — weakly, at n=15 — that a gut already adapted to a GLP-1 agonist may handle the first tirzepatide steps more easily than a treatment-naive one. Whether that carries to any individual is unknowable in advance; it is a reason for realistic optimism, not an expectation.

Weight in the gap and early weeks: telehealth guidance describes minimal regain in the one-week gap, with loss pausing rather than reversing, and then a re-acceleration once tirzepatide reaches efficacy dosing (5 mg and up, after the sub-therapeutic starter step). No trial has formally measured gap-week dynamics, so this row of the table carries a telehealth guidance grade rather than a trial grade — the honest floor of what is known.

Two legitimate sources pull in opposite directions on titration speed after a switch, and this page presents both rather than picking a winner. The conservative position: pharmacist-facing CPD guidance (Pharmaceutical Journal) frames restart-at-the-lowest-dose-and-titrate as the safest default regardless of prior therapy — the standard against which deviations need justification. The accelerative position: the switch study's 10 mg subgroup showed better HbA1c outcomes with early escalation in inadequate responders (Kurinami et al., 2025), suggesting that for some switchers, a deliberate climb is slower than the evidence tolerates.

The asymmetry between the two positions matters. The conservative default is backed by professional standards, applies to everyone, and costs time. The accelerated path is backed by a 15-patient retrospective subgroup, applies to an ill-defined subset, and saves time. When a universal default and a small-sample signal conflict, the resolution is not a web page's call — it is the prescribing clinician weighing the individual's response history, tolerance, and goals. That is not a dodge; it is what the evidence quality actually supports.

What both positions agree on is everything else on this page: no conversion, a gap week, a 2.5 mg-class start, four-week steps as the frame, and red-flag symptoms that override all scheduling logic. The disagreement is confined to pace, within a structure both sides share — which is why the switch conversation with a clinician is short once the patient understands the structure. This guide's job is that understanding; the pace decision stays where the responsibility sits.