Do GLP-1 medicines work the same for everyone?
No. Response and tolerability vary from person to person. In the STEP 1 trial, average weight loss with semaglutide 2.4 mg was 14.9% over 68 weeks, yet not everyone reached the same result: 86.4% lost at least 5%, and 50.5% lost at least 15%. A trial average does not predict what one person will experience.
Sources checked against Europe PMC abstracts and FDA/DailyMed label text: Nov 17, 2026; full text not read for abstract-only claims (recheck on publication date)
Quick answer
No. Response and tolerability vary from person to person. In the STEP 1 trial, average weight loss with semaglutide 2.4 mg was 14.9% over 68 weeks, yet not everyone reached the same result: 86.4% lost at least 5%, and 50.5% lost at least 15%. A trial average does not predict what one person will experience.
What “response” means
“Response” can mean different things, and a person can respond on one measure and not another:
- Weight. Percent change in body weight, or the share of people reaching a threshold such as 5% or 10%.
- Blood sugar. In diabetes trials, A1C is the usual measure. This page does not cover those results.
- Heart and blood vessel events. The SELECT trial followed adults 45 and older with existing cardiovascular disease and overweight or obesity, without diabetes. A major cardiovascular event occurred in 6.5% of those on semaglutide 2.4 mg and 8.0% on placebo over an average of about 40 months (hazard ratio 0.80). Source: Lincoff et al., N Engl J Med 2023.
- How people feel and function. Trials also record patient-reported outcomes. This page does not summarize them.
What trials show

STEP 1 response thresholds
STEP 1 enrolled 1,961 adults with obesity (BMI 30 or higher, or 27 or higher with a weight-related condition) and no diabetes, who received semaglutide 2.4 mg or placebo for 68 weeks, plus lifestyle intervention. Mean weight change was -14.9% with semaglutide and -2.4% with placebo. Among participants with a week-68 measurement, weight loss of 5% or more, 10% or more and 15% or more was seen in 86.4%, 69.1% and 50.5% on semaglutide versus 31.5%, 12.0% and 4.9% on placebo. Source: Wilding et al., N Engl J Med 2021.
The same trial shows the spread: some people on placebo lost 10% or more, and some on semaglutide lost less than 5%.
Tolerability varies too. In SELECT, adverse events led 16.6% of people on semaglutide and 8.2% on placebo to stop the study drug permanently. In the Wegovy label trials, 6.8% on Wegovy and 3.2% on placebo stopped because of adverse reactions. Sources: Lincoff et al. 2023; Wegovy label, section 6.1. These come from different trials of different people, so they cannot be compared with each other.
How to interpret your own progress
- Look at the full picture with your prescriber: weight, waist or measurements, blood pressure, blood sugar and lipids if relevant, strength and energy, side effects and goals.
- Anecdotes online come from a few people, usually with different health conditions, medicines and diets. Trial averages come from groups under study conditions. Neither tells you how your treatment is going.
- Bring your own numbers and questions to follow-up visits. Changes to a plan are your prescriber’s decision.
What this does not establish
The trials show group averages and proportions. They do not identify who will respond, and they do not say why people differ. They also do not tell you how a person’s results compare across different medicines. STEP 1 was a 68-week trial in adults without diabetes, so results can differ in other groups or over longer periods.
FAQs
What does research show about whether GLP-1 medicines work the same for everyone? Average results are clear in trials, but individual results spread widely. Some people lose much more than the average and some much less.
What kind of study supports this answer? Randomized, double-blind, placebo-controlled trials (STEP 1 for weight; SELECT for cardiovascular events).
What does the available evidence not establish? Who will respond, why, and how any one person’s results compare with the trial averages.