Stopping a GLP-1 costs most of the weight back, and more after tirzepatide

Design

Systematic review and meta-analysis of randomised and controlled trials

Population

Six studies, 8,993 adults with overweight or obesity; 5,553 discontinued, 3,440 continued

Primary outcome

Weight change after discontinuation versus continued treatment

Effect

Between-group mean difference 17.90% (95% CI 14.11–21.69); within-group change before to after withdrawal 9.11% (7.91–10.30)

Six trials covering 8,993 adults with overweight or obesity — 5,553 who discontinued a GLP-1 receptor agonist and 3,440 who continued — were pooled to quantify what withdrawal does. The weight difference between discontinuation and continuation was 17.90% (95% CI 14.11–21.69, P < 0.0001), and the within-group change from before to after withdrawal was 9.11% (7.91–10.30).

Subgroup analysis found the rebound after tirzepatide larger than after semaglutide, which the authors attribute to differing mechanisms. That comparison is between trials rather than within one, so it reflects differences in trial populations and in how much weight was lost in the first place as much as in pharmacology — a drug that produces more weight loss has more weight available to regain.

On the other side, adverse events fell after withdrawal: overall and gastrointestinal adverse reactions both decreased, and cardiovascular event rates were unaffected. Statistical heterogeneity was present (P = 0.0082). None of this is surprising, and that is rather the point: these drugs treat obesity while they are being taken, and the trials have been read as though they treated it permanently.

• Frame GLP-1 therapy as long-term treatment at the first prescription, not as a course.
• When a patient wants to stop, plan the stop: agree a monitoring interval and what happens if weight returns.
• Continue to address diet, activity and sleep throughout — they are what remains when the drug does not.
• Do not read the tirzepatide-versus-semaglutide difference as pharmacology; more weight lost means more available to regain.
• In Indian practice, where these agents are self-funded, an interrupted course is the common pattern rather than the exception — plan for it.

Why it matters

The question at the first prescription is not whether the drug works but whether the patient can stay on it.

Don't overread it

The tirzepatide versus semaglutide difference is an unadjusted between-trial subgroup comparison, not a head-to-head result.

The statistics, in plain English

A mean difference of 17.90% between the discontinuation and continuation groups combines two things: weight regained by those who stopped and weight still being lost by those who continued. The within-group figure of 9.11% is the cleaner estimate of actual regain. Heterogeneity was statistically significant, and with six trials of different drugs, durations and populations the pooled number is best read as confirming a direction everyone already suspected rather than as a precise quantity.

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.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.