Why GLP-1 Weight Loss Varies

Weight loss on GLP-1 receptor agonists and dual agonists varies widely between people. A 2026 meta-analysis of 32 trials involving 40,408 participants reported a pooled mean weight reduction of 9.36%, but also found very high variation across trials. Its message is not that one person can be predicted from an average. It is that drug class, treatment duration, participant mix and study design can all change the result readers see reported.^[1]

The analysis specifically included GLP-1 receptor agonists and dual agonists, including tirzepatide. That makes it useful context for headlines about these medicines, but it cannot tell an individual reader what result to expect. Weight change in a trial is a group average, not a personal forecast.^[1]

The pooled result hides a wide spread

Lu and colleagues searched four databases through February 1, 2026 and included 32 eligible trials. Their pooled estimate was 9.36% mean weight reduction, with a 95% confidence interval from 7.92% to 10.80%. The reported I² statistic was 99.6%, which indicates that results differed substantially across the studies rather than clustering around one shared estimate.^[1]

That spread is not surprising when a review places several medicines, trial lengths and participant groups into one analysis. A short study of one GLP-1 medicine is not directly interchangeable with a longer study of a dual agonist. The authors found that GLP-1/GIP dual agonists, represented by tirzepatide in their subgroup analysis, and treatment lasting at least 72 weeks were associated with larger weight reductions across studies.^[1]

A single landmark trial illustrates the difference between a clear trial estimate and a universal promise. In SURMOUNT-1, adults with obesity or overweight without diabetes were randomly assigned to tirzepatide or placebo alongside lifestyle intervention. At week 72, the mean change in body weight was -15.0%, -19.5% and -20.9% in the three tirzepatide groups, compared with -3.1% with placebo. Those are results from that population and protocol, not a result every patient will have.^[2]

Associations are not personal predictions

The meta-analysis also used meta-regression to examine factors linked with different study-level results. Baseline anxiety or depression was positively associated with weight-loss efficacy in that analysis. Type 2 diabetes, obstructive sleep apnea and metabolic dysfunction-associated fatty liver disease were negatively correlated with treatment response. These findings describe relationships across the included studies. They do not establish that any one diagnosis determines how a particular person will respond.^[1]

That distinction matters. Meta-regression can generate useful questions, but it is vulnerable to differences between trials that are not fully measured or reported. The analysis does not show that anxiety or depression causes greater weight loss, nor that the listed metabolic conditions make a medicine ineffective. It reports correlations in a mixed body of evidence. A clinician considering treatment has information that a pooled analysis does not, including medical history, other medicines, adverse effects and follow-up needs.^[1]

The practical implication is not to turn a review into a scorecard. A trial can show that a medicine changed average weight in a defined group, while still leaving a broad range of individual experiences inside that average. Readers can use the tirzepatide evidence summary, the tirzepatide safety section and the tirzepatide regulation section to distinguish trial evidence, safety information and legal status. Those are separate questions, and a percentage alone answers none of them completely.^[1]

What the study does and does not compare

The review combines evidence across GLP-1 receptor agonists and dual agonists. It is therefore a broad map of heterogeneity, not a head-to-head trial designed to rank every medicine. The finding that the tirzepatide subgroup was associated with greater loss is compatible with the trial evidence, but it does not replace direct comparisons made under the same conditions.^[1]

Readers comparing medicines should separate three questions: what a specific randomized trial found, whether the trial population resembles their situation, and what is authorised where they live. Our tirzepatide overview explains the molecule and its evidence base, while the regulation section covers its prescription status by country. Neither page substitutes for individual medical advice.

What we do not yet know

The review cannot settle why different people lose different amounts of weight. Its authors identify emotional state and metabolic comorbidities as possible contributors, but the included trials were not built to test a personal response formula. The high heterogeneity itself is a warning against treating a pooled percentage as a guaranteed outcome.^[1]

There are also practical unknowns beyond weight change. Trial results depend on who enrolled, how long treatment continued, which outcomes were measured and how missing data were handled. Longer follow-up and more direct comparisons may clarify which patterns are reproducible. Until then, it is more accurate to say that response varies than to claim that a simple profile can identify the right medicine or result for everyone.^[1]

Why the variation matters

The useful takeaway from this review is modest. GLP-1 medicines and dual agonists have produced meaningful average weight reductions in trials, while the size of that reduction differs across studies and people. Readers should be wary of before-and-after stories or a single percentage presented as a promise. If you are considering a prescription medicine for weight management, discuss the evidence, expected monitoring and possible adverse effects with a qualified healthcare professional.^[3]