What Happens When You Stop

There is a moment, usually somewhere between month eight and month fourteen, when a patient sits down and says: I think I’m ready to come off it.

I understand the impulse entirely. They feel well. The number is where they wanted it. There is a cost, monthly, that never stops arriving. And underneath all of that, usually unspoken, is the belief that needing the drug indefinitely means they never really fixed anything.

This is the conversation I most want to get right, so I try to have it at the beginning instead.

The data, stated plainly

STEP 1 extension. Participants who had completed 68 weeks of semaglutide came off the drug, and off the lifestyle intervention, and were followed for a year. They regained roughly two thirds of the weight they had lost. The cardiometabolic improvements — blood pressure, lipids, glycemia — reverted along with it.

SURMOUNT-4. A cleaner design. Everyone took tirzepatide openly for 36 weeks, then was randomised either to continue or to switch to placebo. Over the following year, those who continued lost a further several percent. Those switched to placebo regained a large share of what they had lost.

Different molecule, different design, same shape.

Why this is not a moral event

When people hear these numbers they hear an indictment. I hear a mechanism working exactly as described.

Body weight is not passively determined. It is defended. Lose a substantial amount of weight by any method and your body responds with a coordinated set of adaptations: circulating leptin falls, ghrelin rises, satiety signalling weakens, and resting energy expenditure drops somewhat below what your new size alone would predict. Sumithran’s work showed these hormonal changes were still measurable a full year after weight loss, in people who had done everything asked of them.

That is the headwind every person who has ever lost weight has been walking into. It is why the long-term results of diet-alone interventions have been so consistently disappointing for so many decades, and why the people in those studies were blamed for it anyway.

A GLP-1 counteracts part of that defence, pharmacologically, for as long as it is present. It is a long-acting hormone. Its effects last as long as the hormone does.

We do not tell a patient whose blood pressure normalised on lisinopril that they have cured their hypertension and can stop. We say the medication is working. This is the same sentence.

So is it forever?

Not necessarily, and I do not want to oversell the pessimism either.

What the trials measured was full withdrawal, abruptly, usually without a structured maintenance program. That is one specific scenario, and it is the harshest one. Several other paths exist and are being actively studied:

Lower maintenance dosing. Many patients maintain well on a dose below the one that produced the loss. The dose required to hold a weight is often not the dose required to change it. This is my most common approach.

Extended intervals. Some people do well moving from weekly to every 10 days or every 2 weeks. The evidence base here is thinner than I would like — this is clinical practice ahead of trial data, and I say so to patients.

Genuine behavioural consolidation. The window on the drug is the easiest time in a person’s life to build habits, because the appetite headwind is temporarily gone. Someone who spends that year establishing a real resistance training practice, a protein-forward way of eating, and a sleep schedule is in a materially different position at withdrawal than someone who spent the year only losing weight. I cannot show you a trial that quantifies this. I can tell you the difference between those two patients is obvious in clinic.

Deliberate, supervised tapering rather than a cliff, with a plan for what to do if the trajectory turns.

The scenario I actually worry about

Not the patient who chooses to stop. The patient who is stopped.

Coverage changes. A job changes. The employer switches plans in January and the drug is suddenly excluded. A shortage arrives. The out-of-pocket price becomes impossible in a month when something else broke.

This happens constantly, and it happens abruptly, and it is the version of discontinuation with no plan attached. So I ask everyone, before we start:

• What happens to this prescription if your insurance changes?
• Could you afford 3 months at cash price in an emergency?
• If you had to stop tomorrow, what is the floor you would hold — the protein target, the two training sessions, the non-negotiables?

Deciding those things while you feel well and in control is much easier than deciding them in the week the pharmacy tells you the price went up by an order of magnitude.

Catching drift early

One practical point that sits underneath all of the above.

Whether you taper deliberately or lose access abruptly, the difference between a manageable partial regain and an eighteen-month slide is almost entirely how early you notice. Four percent caught in month two is a dose conversation. Fifteen percent noticed a year later is starting over.

Noticing requires that somebody is still measuring after the interesting part is over, which is exactly when most people stop. My own version of this is a quarterly weigh-in and waist measurement that goes into the Zenday App alongside everything from the preceding 2 years, and a specific regain percentage written down in advance at which I go back to my endocrinologist rather than waiting until I feel bad enough about it. The tool matters less than the fact that the record survives the months when you are not paying attention — but a record that requires effort to maintain is a record that stops existing at precisely the moment it becomes useful.

What I’d want you to hold onto

If you have had bariatric surgery, this question has a longer history and a different shape — regain years after an operation is now one of the commonest reasons these drugs get started, and that is GLP-1s and bariatric surgery.

If a supply gap or a decision does take you off it, the restart is not simply picking up where you left off — tolerance fades, and resuming a high dose after several weeks can produce the side effects of a first injection. What to do if you miss a dose covers the rules, and I wrote about facing this question myself in thinking about the rest of my life.

If you regain weight after stopping, you have not proven anything about your discipline. You have demonstrated the pharmacokinetics of the drug you stopped taking.

The framing that helps most of my patients: obesity behaves like a chronic, relapsing, biologically defended condition. We now have effective long-term treatments for it. Long-term treatments are taken long-term. The alternative framing — that this is a temporary intervention to be graduated from — is the one that sets people up to feel like failures on a schedule.

Decide which of those two stories you are in before you start. It changes almost everything about how the next 2 years go.

If the reason you are considering stopping is that the scale has not moved, read why you stopped losing weight first — a plateau and a treatment failure look identical from the outside and are handled completely differently.