Every dossier in this field answers whether to start. Almost none answer what happens when you stop — which is the question with the consequences, because the average operator does stop. SURMOUNT-4 is the trial built to ask it: 36 weeks of tirzepatide produced a 20.9% reduction, then half the participants were switched to placebo. Over the next 52 weeks they regained 14.0% of body weight, while those who continued lost a further 5.5%. Only 16.6% of the withdrawal group held even 80% of what they had lost — against 89.5% of those who stayed on. A meta-analysis of 8 RCTs found the regain arrives 'regardless of lifestyle interventions', a phrase we will not soften. But the post-hoc analysis carries the finding that matters: the metabolic damage is not all-or-nothing. It scales with how much comes back — and operators who held regain under 25% finished the year with waist circumference, non-HDL cholesterol and fasting insulin statistically indistinguishable from their treated best.
- ›Treat the exit as a planned phase with a physician, not an event that happens to you. The trial that defines this question is a withdrawal trial — the regain is the expected outcome, not a personal failure.
- ›Defend the band, not the number. The target supported by the data is holding regain under ~25% of what was lost — the threshold at which the cardiometabolic gains survived the year.
- ›Build the resistance substrate BEFORE the taper, not after. The only arm in the body-composition literature that meaningfully changes the lean-mass fraction is lifestyle plus resistance training (17.5% vs ~26%).
- ›Re-measure on a schedule you set in advance — waist, fasting insulin, non-HDL-C, HbA1c. These are the parameters the post-hoc shows moving, and they move quietly.
The operator who plans the exit converts a cliff into a controlled descent. Not because willpower beats pharmacology — the meta-analysis is explicit that it does not — but because the outcome is graded, not binary, and the difference between the top and bottom regain tiers is 14 cm of waist and a 46% swing in fasting insulin. The operator who does not plan it discovers that the metabolic improvements he paid for are leased, not owned, and that the lease expires on a schedule nobody told him about.
Certainty ends at the regain itself, which is settled, and at the association between regain tier and cardiometabolic reversal, which is a post-hoc analysis and therefore hypothesis-generating rather than confirmatory. What is NOT established is that any behavioural intervention moves an operator between regain tiers: the post-hoc identifies the target, not the method. The pooled evidence explicitly reports regain 'regardless of lifestyle interventions'. The resistance-training figure is imported from body-composition trials during treatment, not from withdrawal trials, and is therefore inference, not proof. Every trial cited is industry-sponsored. Tirzepatide is a prescription pharmaceutical: initiation, titration and discontinuation are clinical decisions, and nothing here is a schedule to self-administer.