The 30-second summary
- Only 30.7% of adults were still persistently filling a prescription at twelve months. A fifth stopped almost at once.
- The reasons are mostly practical: side effects, cost, less weight loss than expected, and fear of rare harms.
- The first four to eight weeks decide most of it. That is the dose escalation window and where side effects peak.
- Cycling on and off may carry its own cost, through swings in weight and blood sugar.
- Stopping deliberately is fine. Drifting off is the version to avoid, because restarting at the old dose goes badly.
The numbers
A retrospective cohort study used United States commercial claims data from 2021 to 2023, covering everyone aged 10 or over who started a GLP-1 for obesity without diabetes: 201,229 adults and 1,139 paediatric patients. Adherence was measured as the proportion of days covered by a prescription, each month, for a year, and the analysis let distinct patterns emerge rather than imposing categories.
Five trajectories appeared in adults.
| Pattern | Share |
|---|---|
| Immediate discontinuers | 19.6% |
| Early discontinuers | 23.5% |
| Late discontinuers | 20.7% |
| Intermittent | 5.5% |
| Persistent users | 30.7% |
Mean annual proportion of days covered across those groups ranged from 0.10 to 0.87. Among paediatric patients the five groups were more evenly distributed, each around a fifth of the cohort, with persistent users at 20.3%.
One finding is difficult to explain in terms of individual behaviour. Region predicted persistence. Adults in the Northeast had 1.32 times the odds of being persistent users compared with those in the South, and in paediatric patients the gap was wider at 1.65. Where you live is a proxy for what your insurance covers and what the medicine costs you, not for how much you want it to work.
Why people stop
A 2026 review in Nature Reviews Endocrinology set out the common reasons: gastrointestinal adverse effects, less-than-desired efficacy, high cost, and fear about uncommon or rare adverse effects.
Each of those has a different answer.
Gastrointestinal side effects peak during dose escalation and settle for most people. They are also the most modifiable item on the list, because a large share of nausea, constipation and reflux responds to what and how you eat rather than to the dose alone.
Less weight loss than hoped for is often a comparison problem. Trial averages are averages, they follow people who stayed in the trial, and they attach to specific doses at specific durations. Someone at week 12 on a middle dose comparing themselves to a 68-week headline figure is not failing.
Cost is the one that cannot be fixed with information, and it is the likeliest explanation for the regional pattern above. It is worth asking directly about coverage, patient support schemes and whether a different medicine in the class is cheaper for you.
Fear of rare adverse effects deserves respect rather than dismissal. The answer is usually proportion: knowing which symptoms genuinely mean act today, and knowing that most of the alarming headlines describe rare events or unsettled evidence.
What repeated stopping and starting may cost
This is the part that is newer and less widely known.
When treatment stops, weight regain is common and multiple cardiometabolic risk factors deteriorate. That much is established. The 2026 review goes further and raises a concern about cycling: repeated rounds of initiation, interruption and re-initiation might induce fluctuations in body weight and HbA1c, and both of those fluctuations are themselves recognised risk factors for cardiovascular and microvascular events.
The review also notes that incretin-based medicines lack direct anti-atherosclerotic and plaque-stabilising effects, which may contribute to elevated cardiovascular risk particularly in the period straight after stopping.
Be careful with how much weight to put on this. The review is explicit that few data are currently available on hard outcomes in people who discontinue, so this is a well-reasoned concern rather than a demonstrated harm. What it supports is a practical conclusion: if you are going to stop, a planned stop is better than an unplanned one, and a stop is better than a pattern of stopping and restarting every few months.
The first eight weeks
Roughly four in ten adults in the trajectory study were gone by the end of the early period. That is where the leverage is.
Expect the dose steps to reset things. Each increase is a small reset of the gut, and the few days afterwards are the hardest. Knowing that in advance turns a reason to stop into a week to get through.
Eat smaller and more often. Three full plates is often three chances to feel overwhelmed. Four or five smaller occasions is easier to finish and delivers more in total.
Deal with constipation before it arrives. It is the side effect people treat in week three rather than prevent in week one. Fibre and fluid together, from the first dose.
Drop the fat and the fried food on the days after a dose. Fat is the slowest thing to leave a stomach that is already slow.
Ask about pausing a step rather than stopping. Staying at the current dose for an extra few weeks is a normal, sanctioned option in the prescribing information for these medicines, and many people do not know it exists. A held dose is not a failure and it is not the same as stopping.
Ask about cost before you run out, not after. A gap created by a pharmacy problem is the version that becomes permanent.
When stopping is the right call
Nothing here is an argument that everyone should stay on a GLP-1 forever.
Reaching your goal and moving to a maintenance plan is a legitimate outcome. So is an adverse effect that warrants stopping, a change in circumstances, or simply deciding with your prescriber that this is not for you. Trial evidence on stepping down to a lower maintenance dose after weight loss is now emerging, and that is a conversation worth having.
The distinction that matters is between a decision and a drift. A decision has a plan attached: what happens to your eating, whether you step down rather than stop, what you watch for, and when you review it. A drift is a missed refill that becomes a month.
What to do with this
Most of what makes the first eight weeks survivable is food: protein you can actually finish, fibre and fluid together, and portions sized to the appetite you have rather than the one you used to have.
Sources
- Huang W, et al. Real-World Adherence Trajectories of GLP-1 Receptor Agonists Among Adults and Paediatric Patients With Obesity. Diabetes Obes Metab 2026. PubMed
- Ceriello A, et al. Causes and consequences of discontinuation of GLP1RAs or tirzepatide. Nat Rev Endocrinol 2026. PubMed
- Characterisation of real-world patients who discontinued a glucagon-like peptide-1 agonist. Diabetes Obes Metab 2026. PubMed
- Sievenpiper JL, et al. Nutritional and lifestyle supportive care recommendations for management of obesity with GLP-1 based therapies. Obes Pillars 2026. PubMed
Medical disclaimer: Articles in the foodose research library are educational, not medical advice. Follow the instructions from your prescriber and the leaflet in your pack. See our full medical disclaimer.