Key facts
- What it measures: the share of participants losing at least a defined percentage of baseline body weight.
- Common thresholds: ≥5%, ≥10%, ≥15% and ≥20% weight loss.
- Why 5% matters: the long-standing minimum bar generally linked to improved metabolic risk factors.
- Contrast with the mean: an average hides variation; responder rates expose the spread.
- Where you see it: reported alongside the mean in modern obesity trials such as STEP 1 and SURMOUNT-1.
- Related concept: pair it with number needed to treat to judge how much extra benefit a drug adds over placebo.
What a responder rate is
A responder rate is the proportion of participants in a trial who achieve at least a predefined amount of weight loss by the end of the study. If a trial reports a 50% responder rate at the 15% threshold, it means half of the participants lost at least 15% of their starting body weight.
The word "responder" simply labels anyone who cleared the bar. Everyone else — people who lost less, stayed the same, or gained — counts as a non-responder for that threshold. Because the bar is fixed in advance and the same for every participant, a responder rate turns a messy distribution of individual results into a single, interpretable percentage: how many people got at least this much benefit.
The standard weight-loss thresholds
Obesity trials usually report responder rates at a ladder of thresholds — most often ≥5%, ≥10%, ≥15% and ≥20% of baseline body weight. Each rung answers a slightly different question, and the rate naturally falls as the bar rises: more people can reach 5% than can reach 20%.
- ≥5%: the classic minimum for a clinically meaningful response. Losing this much is generally associated with improvements in metabolic risk factors such as blood pressure, blood glucose and blood lipids.
- ≥10%: a more substantial response, often linked in the literature to broader metabolic benefit.
- ≥15%: a threshold that older weight-loss drugs rarely reached at a population level, so it became a marker of the newer, more effective agents.
- ≥20%: a high bar that approaches the range once associated mainly with bariatric surgery, now reported for the most effective medications at their top doses.
Reporting the whole ladder, rather than a single cut-off, lets a reader see how the benefit is distributed — how quickly the responder rate drops as the threshold climbs.
Why it beats the average for an individual
An average is a single summary of a whole group, and it can be produced by very different underlying patterns. Imagine two drugs that both report a mean weight loss of 15%. In one, almost everyone lands somewhere near 15%. In the other, a subset of strong responders lose 30% while a large group barely moves. The mean is identical, but the experience of a typical person is completely different.
Responder rates pull that difference into the open. By showing what share of people cleared 5%, 10%, 15% and 20%, they describe the spread of outcomes, not just the centre. For someone weighing whether to pursue a treatment, "half of participants lost at least 15%" is a more grounded expectation than "the average was 15%," because it acknowledges that results vary and puts a number on how often the better outcomes actually occurred.
This is also why responder rates pair naturally with the number needed to treat: one tells you how many people responded, the other tells you how many extra responders a drug produced compared with placebo.
Responder rates in STEP 1
The STEP 1 trial tested once-weekly semaglutide 2.4 mg against placebo in adults with overweight or obesity over 68 weeks, and it is a good example of responder-rate reporting done well. Alongside the headline mean weight change, the publication reported the share of participants clearing each threshold (PMID 33567185).
On semaglutide, the large majority of participants reached the 5% and 10% thresholds, and roughly half reached 15% — far above the corresponding placebo rates, where only a small fraction cleared the higher bars. The gap between the drug and placebo responder rates at each threshold is what makes the effect legible: it is not just that the average moved, but that a much larger proportion of people crossed each meaningful line. For the exact figures at every threshold, see the trial publication linked in the sources.
Responder rates in SURMOUNT-1
The SURMOUNT-1 trial applied the same logic to tirzepatide, a dual GIP/GLP-1 receptor agonist, versus placebo over 72 weeks (PMID 35658024). Because tirzepatide produced larger average weight loss than earlier agents, its responder rates were correspondingly higher, especially at the top dose.
At the highest dose, a large majority of participants cleared 5%, and — notably — a substantial share reached the demanding 20% threshold, a level of response that used to be discussed mainly in the context of surgery. Reading STEP 1 and SURMOUNT-1 side by side shows why responder rates are useful for comparison: the two trials used different populations and durations, so the means are not perfectly interchangeable, but the responder ladders make it easy to see that a higher fraction of people reached the steeper thresholds with tirzepatide. The precise percentages at each cut-off are in the SURMOUNT-1 publication.
Reading a responder rate carefully
A responder rate is a powerful summary, but it has limits worth keeping in mind:
- It is a group statistic, not a personal prediction. A 90% responder rate at 5% does not mean you personally have a 90% chance — trial populations are selected, supervised and adherent in ways that may not match any individual situation.
- The threshold is arbitrary by design. Someone who lost 14.9% is a non-responder at the 15% bar and a responder at the 10% bar. Small movements around a cut-off can shift the label without meaning much clinically.
- Estimands and analysis choices matter. Trials can calculate responder rates in more than one way — for example, counting only people still on treatment, or including everyone regardless of whether they stopped. Two figures for the "same" threshold can differ for this reason, so it is worth checking which estimate a source quotes.
- Always compare against placebo. A responder rate alone can look impressive until you see that a meaningful share of the placebo group also cleared the bar. The difference between the arms is the real signal.
Used with those caveats, responder rates are one of the most honest ways to communicate a weight-loss result, because they refuse to let a single average stand in for a range of very different outcomes. For the bigger picture of which weight-loss peptides have this quality of evidence, see our pillar guide on peptides for weight loss.
Frequently asked questions
What is a responder rate in a weight-loss trial?
A responder rate is the proportion of participants who reach a predefined weight-loss threshold — usually at least 5%, 10%, 15% or 20% of their starting body weight — by the end of the trial. Instead of averaging everyone together, it counts how many people cleared a specific, clinically meaningful bar, so a 50% responder rate at the 15% threshold means half the participants lost at least 15% of their body weight.
Why does a responder rate matter more than the average weight loss?
An average can hide wide variation. A mean of 15% might come from a group where nearly everyone lost around 15%, or from a group where some lost 30% and others barely responded. Responder rates unpack that spread by showing what share of people hit each threshold, which is usually closer to the question an individual actually cares about: what are the odds this works well for me?
Why is 5% weight loss used as a threshold?
Losing at least 5% of body weight is a long-standing benchmark in obesity medicine because that amount is generally associated with measurable improvements in metabolic risk factors such as blood pressure, blood glucose and lipids. Regulators and researchers adopted it as a minimum bar for a clinically meaningful response, which is why the 5% responder rate appears so consistently across weight-loss trials.
What responder rates were reported in STEP 1 and SURMOUNT-1?
In STEP 1, which tested semaglutide 2.4 mg, most participants on the drug reached the 5% and 10% thresholds and about half reached 15%, far above placebo. In SURMOUNT-1, which tested tirzepatide, responder rates were higher still at the top dose, with a large majority clearing 5% and a substantial share reaching 20% or more. Exact figures are in each trial's publication, linked in the sources below.
Does a high responder rate guarantee I will lose weight?
No. A responder rate describes a trial population under trial conditions, not any single person. Even in trials with strong responder rates, some participants did not reach the threshold, and real-world results depend on adherence, individual biology, other conditions and clinical supervision. This page is educational and not medical advice; treatment decisions belong with a licensed clinician.
Sources
This glossary entry draws on the responder-rate reporting in two peer-reviewed obesity trials:
- Once-Weekly Semaglutide in Adults with Overweight or Obesity (STEP 1) — New England Journal of Medicine, 2021; PubMed 33567185.
- Tirzepatide Once Weekly for the Treatment of Obesity (SURMOUNT-1) — New England Journal of Medicine, 2022; PubMed 35658024.