Do I have to stay on a GLP-1 forever?

For most people, sustained benefit requires sustained treatment, because the drug counteracts a biologically defended set point rather than resetting it. In the STEP 1 extension, participants regained about two thirds of their lost weight within a year of stopping. That said, full abrupt withdrawal is the harshest scenario studied, and lower maintenance dosing, extended intervals, and genuine behavioural consolidation are all being used in practice. The framing that helps most is that obesity behaves like a chronic relapsing condition with effective long-term treatments.

What is a maintenance dose of semaglutide?

Many people maintain their result on a dose below the one that produced the loss, because the dose needed to hold a weight is often not the dose needed to change it. Some do well moving from weekly to a longer interval, though the evidence there is thinner than for dose reduction and is clinical practice ahead of trial data. This is a decision to make with your prescriber, revisited annually against actual measurements rather than impressions.

What happens if I lose insurance coverage for my GLP-1?

It is the most common form of discontinuation and the least planned for. Decide in advance what your non-negotiable floor is — the resistance training and protein targets you hold regardless — and set a specific regain percentage at which you contact your prescriber rather than waiting until you feel bad enough about it. Ask now what your plan covers, what the cash price would be, and whether a lower maintenance dose would be affordable if the alternative were stopping entirely.

Thinking About the Rest of My Life

I did not escalate to the maximum dose. I want to start there, because I think the assumption that everybody titrates to the top is one of the more quietly harmful defaults in how this class of drug gets used.

Why I stopped where I stopped

At 1.7 mg the appetite effect is sufficient. The side effects are essentially absent. The trajectory is where my endocrinologist and I want it, which is to say slower than it was in month three and deliberately so.

There is no clinical reason to keep climbing to the maximum studied dose simply because the maximum studied dose exists. The trial protocols escalated to a target because that is how you generate a clean efficacy signal across a population. It is not a treatment goal for an individual who is doing well below it.

I mention this because I see the opposite constantly — patients who are miserable at a high dose, holding on because they believe the number on the pen is the assignment. It is not. The right dose is the lowest one that does the job.

What nine months looks like

Down 18.6% of starting weight. The rate of loss over the last three months is a fraction of what it was over the first three, which is exactly the expected shape and not, as I had to remind myself in July, a problem.

Third DXA next month. Strength is still climbing. Blood pressure is stable off any antihypertensive, which I was on the verge of starting last autumn.

I feel, and I say this with the appropriate suspicion of my own narration, essentially well.

The question I have been avoiding

Here it is: what is the plan for 2031?

Not next month. Not the maintenance dose. The actual long horizon, which is the horizon this condition operates on and the one almost nobody plans for.

I have written the analysis of the discontinuation data and I stand behind every word of it. Applying it to myself is a different exercise, and it forces three questions I have been letting sit.

1. Am I prepared for this to be indefinite?

Intellectually, yes, and I have been saying so publicly for a year. The framing I give patients — that we do not congratulate someone on their normalised blood pressure by stopping their lisinopril — is one I believe.

Emotionally there is still a small residue. A voice that treats an indefinite medication as an unfinished piece of work. I notice it most when I am tired, and I have learned to file it accordingly rather than argue with it.

What I have landed on: I would like to find the lowest dose and the longest interval that hold the result, and then hold there for as long as it is working and tolerated, and revisit annually with actual data rather than with feelings. That is a plan. “See how it goes” is not a plan, and it is what most people are running.

2. What is the floor if the drug disappears?

This is the one I think everybody should write down, and almost nobody has.

Coverage is not stable. My employer changes plans. Formularies revise in January. Prices move. The supply has already been disrupted once in the time I have been taking it. Any of these could end my access with about six weeks of notice.

So my floor, written explicitly, is:

Three lifting sessions a week. This is the non-negotiable, and it is non-negotiable precisely because it is the thing that protects what I have built if the pharmacologic support goes away.
The protein target holds, drug or no drug. It is now simply how I eat.
A quarterly weigh-in and waist measurement, so that a drift gets caught at 4% rather than at 15%.
A quarterly review of the actual numbers, not my impression of them. Everything from the last nine months is in Zenday App and will keep accumulating whether or not I am paying attention, which is precisely the property you want from a record that has to survive a bad quarter.
A named threshold for re-engaging — a specific regain percentage at which I go back to my endocrinologist rather than waiting to feel bad enough about it.

None of that prevents regain. The data is clear that habits alone do not fully counteract the physiology. But there is a large difference between a controlled partial regain caught early and an eighteen-month slide nobody was measuring.

3. What did I actually build?

This is the softer question and I think it is the important one.

Nine months of a quiet appetite is, functionally, nine months of unusually favourable conditions in which to install things. If I spent that window only losing weight, I will arrive at any future discontinuation with nothing but a smaller body and the same environment that produced the larger one.

What I think I built: a genuine lifting practice, now in its fifteenth month and no longer requiring willpower. A default way of eating that starts with protein. A sleep schedule I defend. A relationship with the tape measure rather than the scale.

What I did not build, and am still working on: any real change in how I handle a genuinely bad week. The drug has been carrying more of that than I like to admit. When work goes sideways, I am not reaching for food, but I have not replaced it with anything either — I have just stopped. That is not the same as having a coping strategy, and it is the thing I would work on next if I were my own patient.

To the person reading this at month two

A few things I would say, having got here.

The middle is boring, and the boring part is the part that works. Months four through eight had no revelations. They had 34 gym sessions and a lot of chicken.

Your dose is your dose. Not the maximum, not your friend’s, not the one on the box.

Get one measurement that isn’t the scale, whatever you can access — a tape measure costs three dollars.

Write your floor down now, while things are going well, because you will not write it in the week the pharmacy calls.

And the thing I most want to say, which I could not have said in November: this got a great deal less interesting as a story, and that is the good outcome. The first three months were dramatic. The last three have been a medication I take on Sunday mornings and mostly do not think about, sitting alongside a life that has more room in it than it used to.

That is what I was after. It turns out it does not photograph well.

The evidence behind the decision I am describing here — what regain after discontinuation actually looks like — is in what happens when you stop. And what is and is not known about staying on it for years is in what we know about long-term safety.