Keeping weight off: what the long-term data show

The common belief is that almost nobody keeps weight off. The research is more encouraging than that, and more specific about what it takes. There are two bodies of evidence worth knowing: the National Weight Control Registry, which has followed people who succeeded for decades, and the GLP-1 trials that stopped the drug and watched what happened.

The National Weight Control Registry

The NWCR enrols adults who have lost at least 30 lb (13.6 kg) and kept it off for at least a year. Wing and Phelan's 2005 review reported that members had lost an average of 33 kg and kept it off for more than five years, and estimated from wider research that about 20 percent of people with overweight succeed at long-term weight loss, defined as losing 10 percent and keeping it off for a year.

What maintainers do

From Wing and Phelan's review and the follow-up studies, the habits that recur:

• High physical activity, about an hour a day on average.
• A consistent eating pattern across weekdays and weekends, rather than strict weekdays and free weekends.
• Breakfast most days (78 percent ate it every day).
• Regular self-weighing, so small regains are caught early.
• Less TV. Most watched 10 hours a week or less.

The 10-year study also found what predicted regain: less leisure-time activity, less dietary restraint, less frequent self-weighing, more fat in the diet, and more disinhibited eating. Larger initial losses and longer time already maintained predicted better outcomes, which matches Wing and Phelan's observation that once people have kept weight off for two to five years, their chance of longer-term success rises substantially.

The caution: the NWCR is a self-selected group of people who succeeded and volunteered. It tells you what successful maintainers do, not that those habits caused the success or that they will work for everyone. The breakfast finding is the clearest example: eating breakfast is common among maintainers, but a 2019 BMJ meta-analysis of 13 short randomised trials found no weight benefit from eating breakfast (if anything, a small difference of 0.44 kg favoured skipping it).

What the GLP-1 withdrawal trials show

The STEP 1 extension and SURMOUNT-4 make the same point: for most people, the effect of a GLP-1 receptor agonist on weight lasts as long as the drug does. In STEP 1, lifestyle support was also stopped at week 68, so the extension tests stopping everything, not the drug alone. S-LITE is the hopeful result: after diet-induced weight loss, exercise on its own kept off 4.1 kg more than placebo over a year, and exercise plus the drug did best. Our pages on stopping a GLP-1 and the maintenance dose cover the medication side.

A maintenance plan

• Keep moving more than you did before. The registry members averaged about an hour a day. The 2025 joint GLP-1 advisory lists 60 minutes or more a day of activity among the behaviours associated with maintenance.
• Keep lifting. Muscle kept during the loss is easier to keep than to rebuild. See exercise to keep muscle.
• Weigh regularly, and set a regain threshold (for example 2 to 3 kg) that triggers a deliberate response rather than drift.
• Keep the eating pattern you lost weight on, with more food, rather than returning to the old one. Protein and fibre help with fullness; see the sample week.
• Sleep. Short sleep raises intake by hundreds of calories a day in trials. See sleep and health.
• If you plan to stop a GLP-1, do it with your prescriber, with a plan for the months after, not by running out.

Logging it

Maintenance is a long game with small signals: a creeping weekly average, fewer steps, shorter nights. Peptide AI puts weight, steps and sleep from wearable integrations, meals from nutrition tracking and sessions from the training log on one timeline, with the GLP-1 dose log if you are on one, so a regain shows up as a trend early rather than as a surprise.