If I Stop, Will the Weight Come Back?

From fad diets to GLP-1 medicines: why weight returns, and what it takes to keep it off.

By Dr Abhijit Bhograj | Consultant Endocrinologist, Founder AiHealth.clinic

The 21-day reset. The 30-day detox. The ketogenic challenge. The 90-day transformation.

For decades, each new diet has promised to deliver what the previous one could not: lasting weight loss.

Now semaglutide and tirzepatide can help people lose substantially more weight than most traditional diets. Yet the question I hear most often is remarkably familiar:

"If I stop, will the weight come back?"

Often, yes. But that is only half the answer.

Weight regain did not begin with injections. These medicines have exposed a problem obesity researchers have understood for years: losing weight and maintaining weight loss are two different challenges.

We knew about regain long before GLP-1s

A fad diet often ends when the target weight is reached. The problem is that the body does not receive the message that the project is finished.

In a review of 29 structured weight-loss programmes, people had kept off, on average, only 17–29% of the weight they had initially lost after four to five years. [1]

In plain language: someone who had lost 10 kg might still be only 1.7–2.9 kg lighter several years later, based on those programme averages.

This does not mean lifestyle treatment fails. In the Look AHEAD trial, adults with type 2 diabetes received years of dietary, activity and behavioural support. Eight years later, their average weight remained 4.7% below the starting point; about half maintained at least 5% weight loss. [2]

The DiRECT trial went further: intensive dietary treatment helped some people achieve remission of type 2 diabetes. But among those in remission after two years, only about one in four remained in remission at five years. [3] Continued weight control was closely linked to continued benefit.

The television programme The Biggest Loser offered an extreme illustration. In a small follow-up of 14 contestants, much of a very large initial weight loss had returned six years later, while measured resting metabolism remained unusually low for their body size. This was an exceptional group, not a prediction for ordinary patients. [4]

The lesson is not that diets are useless. A sustainable diet and structured lifestyle treatment can be effective. The fad is the promise that a short intervention automatically produces a permanent result.

Why your body pushes back

We often describe weight loss as simple arithmetic: eat less, move more. Energy balance matters, but the body responds to the change.

As fat stores fall, leptin, a signal of stored energy, generally decreases. Hunger signals, including ghrelin, may increase. Energy expenditure can also fall more than expected from the smaller body alone. This is called adaptive thermogenesis.

In a landmark study, people lost an average of 13.5 kg on a very-low-energy diet. One year later, several hormonal changes persisted and hunger remained higher than before weight loss. [5]

You may have heard that the body has a fixed "set point" it tries to defend. That is a useful explanation, but not a settled scientific law. Researchers also describe settling points, shaped by food environments, behaviour and biology. [6]

The precise model is debated. The response to weight loss is not: your body can make keeping weight off harder than losing it.

What actually happens when injections stop?

Three major withdrawal trials tested what happened when people continued treatment or stopped it.

Semaglutide: STEP 1 extension. Participants had lost an average of 17.3% of their starting weight. One year after stopping medication and structured lifestyle support, they had regained about two-thirds of that loss. [7]

Imagine a starting weight of 100 kg:

• Before treatment: 100 kg
• After treatment: about 83 kg
• One year after stopping: about 94 kg

These are the study's average percentage changes translated into kilograms, not a prediction for any one patient. Some maintained much more weight loss; others regained more.

Semaglutide: STEP 4. People who continued treatment kept losing weight, while those switched to placebo regained weight. [8]

Tirzepatide: SURMOUNT-4. After participants initially lost about 21%, those switched to placebo regained substantial weight over the following year; those who continued treatment lost more. [9]

The pattern was consistent: continuing effective treatment protected weight loss better than withdrawal.

A 2026 BMJ review combined 37 studies involving more than 9,300 people. Weight returned at an estimated average of 0.4 kg per month after stopping weight-management medication overall, and around 0.8 kg per month after stopping newer medicines such as semaglutide or tirzepatide. [10]

That means roughly 800 grams a month on average for the newer-drug group, not a fixed rate for everyone. Most included studies had relatively short follow-up, so longer-term outcomes remain uncertain.

Regain after medication was faster than after behavioural programmes ended, even after accounting for initial weight loss. We should neither dismiss medication nor pretend that stopping it is the same as completing a short course of antibiotics.

Does everyone need the full dose forever?

Not necessarily. This is the most important recent development.

The 2026 SURMOUNT-MAINTAIN trial tested three approaches after people had already lost substantial weight on tirzepatide: continue the maximum tolerated dose, reduce to 5 mg weekly, or switch to placebo. [11]

To make the final results easier to picture, imagine a starting weight of 100 kg. At the end of the trial, the group averages would translate to approximately:

• 78 kg: continued maximum tolerated dose
• 83 kg: reduced to 5 mg weekly
• 90 kg: switched to placebo

An essential caveat: about two-thirds of those assigned placebo received rescue tirzepatide after substantial regain. So 90 kg does not represent what would happen if everyone stopped medication and remained untreated.

The trial does show that a lower dose preserved meaningful weight loss in selected patients, although not as much as continuing the higher dose. It does not prove that 5 mg is sufficient for everyone, or that everyone should reduce treatment.

This changes the question from "full dose forever or stop completely?" to "What is the lowest effective treatment intensity for this person?"

Can tapering prevent regain?

Reducing to an effective maintenance dose is not the same as tapering to zero.

A gradual withdrawal sounds sensible: reduce the dose, watch appetite and weight, then decide what to do next. But we do not yet have a validated taper-to-zero strategy proven to prevent long-term regain. [12]

That gap matters. A plan that sounds biologically plausible is not necessarily a plan supported by trials.

If treatment must stop, the early months deserve close monitoring, not a final prescription and goodbye.

What about exercise and diet?

Medication is not a substitute for a sustainable lifestyle. In trials of exercise combined with liraglutide, structured activity helped maintain weight loss and improve body composition. A follow-up suggested better maintenance after treatment ended among those who had exercised, particularly with medication. [13,14]

A small 2026 pilot study involving 39 people also suggested that digital support or medically tailored meals might reduce early regain after stopping GLP-1 therapy. Encouraging, but too small and short to establish a long-term solution. [15]

The practical message: maintaining weight requires a system, not another 30-day challenge.

What I discuss with patients from day one

1. Plan maintenance before starting. Discuss duration, affordability, side effects, pregnancy plans, and what happens if treatment changes.
2. Protect strength. Include adequate protein, appropriate to medical needs, and resistance training. Muscle function matters beyond the scale.
3. Choose an eating pattern you can live with. A six-week transformation is of limited value if the diet cannot be maintained for years.
4. Track more than kilograms. Depending on the patient, follow waist circumference, strength, hunger, food preoccupation, glucose, blood pressure and other health markers.
5. Agree on review triggers. Decide together what amount of regain, returning hunger or metabolic deterioration should prompt reassessment. Some patients will need continued treatment, some may do well on a lower dose, and others may maintain without medication.

A 2025 Lancet Commission distinguished clinical obesity, a chronic systemic illness, from preclinical obesity. Not everyone with excess body fat has the same disease severity or needs the same treatment. [16]

The real measure of success

We have spent decades celebrating kilograms lost at three months. Modern medicines have transformed what is possible. But the harder question comes at three years, or ten.

How much weight stayed off? Did diabetes risk, blood pressure, mobility and quality of life improve? Can the person afford and tolerate the plan?

So, if you stop, will the weight come back?

Often, yes. Not always, and not necessarily all of it. Regain varies, and it is not evidence of a personal failure.

What matters is having a plan before it happens.

The goal is not the lowest weight for the shortest time. It is better health for the longest time.

The diet may end. The injection may stop. The maintenance plan must continue.

Dr Abhijit Bhograj
Consultant Endocrinologist | Founder, AiHealth.clinic

Educational information only. Decisions to start, reduce or stop medication should be made with your treating clinician.

References

1. Anderson JW, et al. Am J Clin Nutr. 2001. doi:10.1093/ajcn/74.5.579
2. Look AHEAD Research Group. Obesity. 2014. doi:10.1002/oby.20662
3. Lean ME, et al. Lancet Diabetes Endocrinol. 2024. doi:10.1016/S2213-8587(23)00385-6
4. Fothergill E, et al. Obesity. 2016. doi:10.1002/oby.21538
5. Sumithran P, et al. N Engl J Med. 2011. doi:10.1056/NEJMoa1105816
6. Speakman JR, et al. Dis Model Mech. 2011. doi:10.1242/dmm.008698
7. Wilding JPH, et al. Diabetes Obes Metab. 2022. doi:10.1111/dom.14725
8. Rubino D, et al. JAMA. 2021. doi:10.1001/jama.2021.3224
9. Aronne LJ, et al. JAMA. 2024. doi:10.1001/jama.2023.24945
10. West S, et al. BMJ. 2026. doi:10.1136/bmj-2025-085304
11. Horn DB, et al. Lancet. 2026. doi:10.1016/S0140-6736(26)00656-2
12. Shah E, et al. Diabetes Obes Metab. 2026. doi:10.1111/dom.70713
13. Lundgren JR, et al. N Engl J Med. 2021. doi:10.1056/NEJMoa2028198
14. Jensen SBK, et al. eClinicalMedicine. 2024. doi:10.1016/j.eclinm.2024.102475
15. Hollis-Hansen K, et al. Obes Sci Pract. 2026. doi:10.1002/osp4.70194
16. Rubino F, et al. Lancet Diabetes Endocrinol. 2025. doi:10.1016/S2213-8587(24)00316-4