Which weight loss jab or pill actually works best? Here’s what the real evidence shows.

Six things this study found that nobody else is telling you

1. Semaglutide currently has the strongest mortality and heart-attack evidence

In this BMJ analysis, injectable semaglutide, the active ingredient in Wegovy, was the only drug associated with a statistically supported reduction in all-cause mortality, with a risk ratio of 0.81, and myocardial infarction, with a risk ratio of 0.72. These estimates were largely informed by cardiovascular outcome trials in higher-risk populations. Both injectable semaglutide and tirzepatide were also associated with lower heart-failure risk in the network analysis.

That does not prove other treatments lack cardiovascular benefit. It reflects the evidence available to this review and the fact that semaglutide has been tested in large cardiovascular outcome trials. It is also important not to apply results from high-risk cardiovascular populations automatically to every person using a weight-management medicine. For someone with established cardiovascular disease, treatment choice should be discussed with a clinician using the medicine’s current licensed indication and the person’s individual risk profile.

2. No treatment reached the study’s quality-of-life threshold

The researchers analysed quality-of-life data from 43 trials involving 45,663 participants and used a 10-point change as the minimally important difference. None of the drugs supported by moderate or high-certainty evidence reached that threshold at one year. The largest point estimate was 4.3 points for phentermine-topiramate; oral semaglutide was 4.1, tirzepatide 3.9, CagriSema 3.5 and injectable semaglutide 2.9.

That finding needs careful interpretation. It does not mean nobody feels better after losing weight, and different trials used different quality-of-life instruments that were converted to a common scale for the network analysis. It means the average effects in the evidence reviewed did not cross the threshold the researchers had chosen for a clinically important change. Weight loss and measured quality of life are related, but they are not interchangeable outcomes.

3. The weight comes back faster than most people expect

This figure comes from a separate BMJ systematic review published on 7 January 2026, not from the July network meta-analysis above. That review included 37 studies and 9,341 participants. Across all weight-management medicines, average regain after stopping was about 0.4kg a month, with body weight projected to return to baseline at around 1.7 years. Cardiometabolic markers were projected to return to baseline within about 1.4 years. These are modelled averages, not a timetable that will apply to every individual.

The July BMJ paper also notes that real-world persistence with weight-management medicines can be much lower than in clinical trials, with previous observational evidence suggesting substantial discontinuation during the first year. Reasons can include adverse effects, cost, access and individual circumstances. The practical message is not that everyone will regain weight on the same schedule, but that stopping treatment commonly leads to regain and should be planned with the prescribing team rather than treated as an afterthought.

4. Lean mass loss varies between treatments

Tirzepatide produced the largest reduction in fat mass in the moderate/high-certainty evidence, but it was also associated with an 8.3 percent greater reduction in lean mass than lifestyle modification alone in the single trial informing that outcome. Injectable semaglutide was associated with a 5.8 percent greater reduction in lean mass. Oral semaglutide and liraglutide showed little or no effect on lean mass compared with lifestyle modification in this analysis.
Lean mass is not the same thing as skeletal muscle: it includes muscle but also other non-fat tissue and body water. The finding therefore should not be reported as though the study measured pure muscle loss. The BMJ authors conclude that pharmacotherapy should be integrated with lifestyle measures and strategies to preserve lean mass; individual exercise and nutrition advice should take account of age, health and physical ability.

5. Retatrutide has moved on significantly since the BMJ evidence search

Retatrutide needs special context on this page because the BMJ literature search stopped on 12 November 2025. In that evidence set, emerging agents including retatrutide, mazdutide and ecnoglutide were estimated to reduce body weight by around 13.1 to 14.6 percent versus lifestyle modification at one year, but the certainty was rated low to very low. Those figures are network estimates from the evidence available at the time; they are not the same thing as retatrutide’s subsequently reported dedicated Phase 3 results.

Since that search cut-off, Lilly has reported several Phase 3 retatrutide trials. TRIUMPH-1 reported 28.3 percent average weight loss at 80 weeks with 12mg under Lilly’s efficacy estimand. A pre-specified extension in participants with a baseline BMI of at least 35 reported 30.3 percent average weight loss at 104 weeks for the 12mg-to-maximum-tolerated-dose group. TRIUMPH-2 subsequently reported up to 20.8 percent average weight loss in adults with obesity or overweight and type 2 diabetes, while TRIUMPH-3 reported up to 22.6 percent in adults with severe obesity and established cardiovascular disease, with or without type 2 diabetes.

The important lesson is not to place the BMJ 13.1 to 14.6 percent estimate beside the newer 28.3 percent TRIUMPH-1 figure as though they are competing measurements from the same study. They answer different questions, use different evidence and were produced at different stages of retatrutide’s development. Retatrutide remains investigational and is not approved for routine use in the UK. We track its current Phase 3 results and regulatory position in our dedicated Retatrutide UK guide.

6. What about Saxenda, orlistat and the older options?

The older options are worth keeping in context. In this network meta-analysis, liraglutide, orlistat, naltrexone-bupropion, exenatide, SGLT-2 inhibitors, dulaglutide and metformin were not convincingly different from lifestyle modification alone for percentage body-weight change at one year. That is a statement about this pooled comparison and its decision thresholds, not proof that an individual medicine never produces weight loss in an individual patient.

The drugs also differ substantially in mechanism, licensed indication, contraindications, side effects, cost and availability. Orlistat, for example, reduces dietary fat absorption rather than acting on GLP-1 pathways. The BMJ analysis is useful for comparing average trial effects, but it should not replace the individual assessment needed to decide whether any particular treatment is appropriate.