Home / Blog / Stopping tirzepatide

What happens when you stop tirzepatide — and how to plan for it before you start

The least-discussed number in obesity medicine: what the scale does after the prescription ends. SURMOUNT-4 measured it directly, and the answer reshapes how you should think about cost, commitment, and exit strategy on day one.

The withdrawal experiment

SURMOUNT-4 gave everyone tirzepatide for 36 weeks — average loss, 20.9% of body weight — then randomized: half continued the drug, half switched to placebo without knowing. By week 88, the continuers had deepened their loss to about 25.3% below baseline. The switchers regained roughly 14 percentage points of what they'd lost, landing near 10% below baseline — still meaningfully lighter than the start, but on a clear upward trajectory. The biology is unsentimental: tirzepatide suppresses appetite and slows gastric emptying only while it's in your system, and its five-day half-life means the effect is functionally gone within a few weeks of the last dose. The hormonal environment that produced the original weight returns, and it is patient.

What this means for the price of the drug

The honest unit of cost isn't a month — it's a maintenance plan. A $169-a-month program is $2,028 a year, likely for multiple years; a program that's cheap for six months and unsustainable afterward buys a loop, not a result. This is why our dataset leads with 12-month totals and maintenance-dose pricing rather than starting prices, and why flat-rate structures with no dose-based increases matter more than introductory offers. Run the arithmetic on year two before you're impressed by month one.

If you do need to stop

People stop for real reasons — cost, pregnancy plans, surgery, side effects, or simply reaching a goal. The evidence-informed playbook: don't stop cold at peak dose without a plan; many clinicians step down through the titration ladder in reverse, watching appetite return in a controlled way, though tapering itself hasn't been trialed. Fortify the non-drug scaffolding first — the months before stopping are when protein targets, resistance training, and sleep discipline earn their keep, because they're what remains. Consider partial rather than full exit: a reduced ongoing dose as maintenance is unstudied but pharmacologically coherent (see the microdosing analysis), and stretching injections at a clinician's direction can cut cost without full withdrawal. Set a re-entry threshold in advance — a specific regain number that triggers restarting — because SURMOUNT-4's switchers didn't fail; their biology did exactly what untreated biology does. Plan for it the way you'd plan for any chronic condition, because that's what the data says obesity is.

SURMOUNT-4, read closely: what both curves actually show

Because this trial carries the whole stopping conversation, read it at full resolution. Design: everyone received tirzepatide for 36 weeks (reaching an average of about −20.9%), then randomization split the room — continue drug, or switch to placebo — for 52 more weeks, with lifestyle support continuing in both arms. The continuation curve kept falling, to roughly −25.3% at week 88: maintenance isn't a plateau on this drug, it's often continued slow progress. The withdrawal curve reversed within weeks and regained about 14 percentage points of the loss — ending near −9.9% from original baseline. Three readings matter more than the headline. First, the regain wasn't a discipline failure: both arms had identical lifestyle scaffolding, so the difference isolates the pharmacology — appetite signaling and energy defense snap back when the signal stops, which is physiology reasserting a set point, not character. Second, the withdrawal arm did not return to baseline within the year — partial retention is real, and it's the honest foundation for every tapering conversation. Third, the trial tested cold-turkey withdrawal, not tapering, not intermittent dosing, not diet-transition protocols — so "stopping means regaining everything" overstates the data exactly as much as "you can just stop" understates it. The clinical translation the trial supports: obesity behaves like the chronic condition it is, the medication behaves like maintenance therapy for one, and any exit plan should be built against a 14-point headwind, on purpose, with your clinician.

Tapering versus stopping cold: what clinicians actually do in 2026

No published trial has compared discontinuation strategies head-to-head, so current practice is physiology-guided consensus — worth knowing precisely because programs rarely explain it. The common architecture: stabilize first (eight-plus weeks at goal weight on your current dose — never exit from a moving position); step down the ladder you climbed, one rung at a time — 10 to 7.5 to 5 to 2.5 — holding each for six-to-twelve weeks with seven-day-average tracking, which converts one cliff into several small experiments; stretch the interval as an alternative or adjunct lever — the ~5-day half-life means moving injections to every 10–14 days lowers average drug exposure gradually, a pattern some clinicians prefer to dose cuts; and predefine the reversal trigger — a sustained 3–5 lb drift on the weekly average means step back up, as protocol rather than defeat. During every phase, the non-drug scaffolding stops being optional: the protein floor and resistance training that defended muscle on the way down are the entire metabolic defense on the way off, and the appetite-monitoring honesty ("food noise" returning is data, not weakness) determines whether a step-down holds. Two exit paths deserve naming alongside full discontinuation: maintenance at the lowest effective dose — often 2.5–5 mg, cutting cost and side-effect load while keeping the signal — which many clinicians now treat as the default destination rather than a waypoint; and molecule step-down to cheaper semaglutide for the holding pattern, a budget-driven variant with its own switching mechanics. The unifying principle: the question isn't "how do I get off," it's "what's the minimum ongoing signal that holds my result" — and that question has answers between fifteen milligrams and zero.

The economics of staying: maintenance math that changes the decision

Half of stopping decisions are budget decisions wearing clinical clothes, so run the budget honestly. The maintenance scenario most people never price: a successful step-down to 5 mg on a flat-rate program costs exactly what titration did — $139–169 listed at the operator floor, $1,668–2,028 a year — while on dose-priced programs the step-down is where costs finally fall, and on LillyDirect the 5 mg tier runs $399 against the $449 maintenance band. Against that, price the stopping scenario with SURMOUNT-4's numbers: a 14-point regain on a 200-lb starting weight is roughly 28 pounds returning over a year — with the health costs, the re-treatment costs (re-titration from 2.5 mg, months of sub-therapeutic dosing, the side-effect gauntlet again), and the demoralization tax that shows up in every real-world adherence dataset. Annualized, the gap between "maintain at $1,700" and "stop, regain, restart" is far smaller than it looks in the month you cancel — and that's before the cheaper maintenance paths: the semaglutide floor at $1,320–1,668, a future negotiated-price brand lane, or the Bridge's $600 year for eligible beneficiaries. None of this argues that everyone stays forever; pregnancy planning, adverse effects, and genuine informed preference all end regimens legitimately. It argues that the exit should be priced like the medical decision it is — against the trial's 14-point headwind and the real cost of the round trip — rather than decided by a renewal email. The commitment calculator runs your numbers; the conversation that matters runs with your clinician; and the version of you twelve months out is the client both should be serving.

The behavioral handoff: what has to be running before the dose comes down

The medication has been doing measurable work you'll need to replace or release deliberately, so audit the handoff like an engineer. Appetite regulation: on-drug, satiety arrived early and "food noise" stayed quiet — off-drug or on less, both revert toward baseline, and the practical countermeasures are structural rather than willpower-based: protein-anchored meals on a schedule (structure substitutes for the missing satiety signal), pre-decided portions plated before eating, and the honest reframe that returning hunger is expected physiology to be managed, not evidence of failure to be ashamed of. Energy expenditure: you now maintain a smaller body with a lower calorie budget, and the muscle preserved through the resistance-and-protein rules is that budget's biggest line item — training doesn't become optional at goal weight, it becomes the load-bearing wall. Monitoring cadence: the seven-day average, a weekly waist or clothing checkpoint, and the predefined reversal trigger from the tapering section turn drift into data within two weeks instead of two sizes. Support architecture: the months after a step-down are precisely when a clinician check-in schedule, a household that knows the plan, and — where it helps — structured programs earn their keep; going quiet is the classic failure mode. None of this is a lecture about lifestyle; it's the sober engineering point that SURMOUNT-4's placebo arm had lifestyle support and still regained 14 points — which means the behavioral layer is necessary, insufficient alone, and exactly why the maintenance-dose middle path from the sections above is the strategy most exits should be built around rather than the cliff.

The pre-exit checklist, in one paragraph

Before any dose comes down, be able to check every box in this paragraph. Eight-plus weeks stable at goal on your current dose — exits from moving positions fail. A written taper architecture with your clinician — rungs, hold durations, and the reversal trigger number chosen in advance. The behavioral layer running, not planned: protein floor hit routinely, resistance training scheduled, the seven-day average tracked without drama. The economics decided consciously — maintenance-dose cost priced against the round-trip cost of regain-and-restart, with the calculator's numbers in front of you rather than the renewal email. The special cases cleared — pregnancy timing coordinated properly (these drugs stop before conception attempts, with obstetric guidance), surgery holds calendared, and any adverse-effect exit routed through the prescriber who can document and manage it. And one honest sentence said out loud about which exit this is: a planned step-down experiment with a safety net, a medically necessary stop, or a budget decision that the maintenance-floor options in this article haven't actually been tested against yet. Five boxes and a sentence — the difference, in the SURMOUNT-4 era, between discontinuing a medication and abandoning a result.

Practical companions: Half-life & washout · Results timeline · Protein defense

Sources: SURMOUNT-4 (JAMA 2024) — 36-week lead-in, randomized withdrawal, week-88 outcomes; tirzepatide pharmacokinetics from FDA labeling (~5-day half-life); SURMOUNT-1 (NEJM 2022). Catalog: /sources/.