A Bayesian meta-analysis pooled timepoint-level data from six studies, ten intervention arms and 1,776 participants who discontinued semaglutide or tirzepatide, with observed follow-up from 4 to 52 weeks. It modelled how much weight had been lost at the point of stopping, how fast it returned, and how long until half of it and then all of it was back.
Weight loss at cessation was 15.35 kg (95% credible interval 11.18 to 19.60). Regain then ran at 1.04 kg per month (0.80 to 1.29). On a linear projection, half the lost weight was back at 7.50 months (5.05 to 10.57) and baseline weight was reached at 15.00 months (10.10 to 21.13). Tirzepatide looked numerically faster in unadjusted analysis but no drug-specific difference survived adjustment. Greater initial weight loss pointed towards faster regain and behavioural support towards slower regain, but both estimates were imprecise.
The clinical point is not that regain happens — that was already known — but its pace. A kilogram a month is fast enough that a patient who stops because of cost, supply or side effects will notice within a season, and fast enough that a six-month gap in supply undoes most of a year's work. In Indian practice, where these drugs are self-funded and interruptions are usually financial rather than clinical, that trajectory should be part of the conversation before the first prescription, not after the last one.
Treat discontinuation as a planned event with follow-up attached. Weigh at three months rather than waiting for the annual review, agree in advance what regain would trigger restarting or switching, and make sure the alternatives — dietary support, resistance exercise, other agents — are in place before the drug stops rather than after.
- Discuss the regain trajectory before starting, not when the patient asks about stopping.
- Book a review three months after any discontinuation, whatever the reason for it.
- Agree in advance what degree of regain triggers restarting, dose reduction rather than stopping, or a switch.
- Where cost drives the decision, consider a lower maintenance dose over full cessation.
- Put structured dietary and resistance-exercise support in place before the drug is stopped.
The statistics, in plain English
Credible intervals are the Bayesian equivalent of confidence intervals and are read the same way: the range where the true value most plausibly sits. The regain rate of 1.04 kg per month with an interval of 0.80 to 1.29 is estimated tightly. The longer projections are weaker. Observed follow-up went to 52 weeks, so the 15-month return to baseline is an extrapolation that assumes regain continues in a straight line, which real weight trajectories rarely do — they usually flatten. Read the monthly rate as solid and the 15-month figure as a direction of travel.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for diabetes & endocrinology, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free