A new Swiss survey delivers exactly the numbers fat activists have been asking for. Almost two thirds of affected respondents say health professionals have discriminated against them. Nine out of ten people say fat people face contemptuous looks in everyday life. Half the population still believes it comes down to lifestyle and discipline.
The survey was commissioned by Novo Nordisk, the company that makes Wegovy.
That is not a scandal, and it is not a reason to throw the numbers away. It is a reason to read them properly. This piece is about how to do that.
What the Adipositas-Barometer 2026 actually found
gfs.bern, a Swiss polling institute, surveyed 1,539 Swiss residents aged 16 and over between April and May 2026, plus 116 physicians (general practitioners and specialists in endocrinology, surgery, gastroenterology, and obesity centres). The results are published as an open “cockpit” on gfsbern.ch.
The stigma findings are blunt:
- 82 percent of the general population say fat people in Switzerland are strongly or fairly strongly stigmatised. Among physicians, that figure is 97 percent.
- Only cardiovascular disease (26 percent) and type 2 diabetes (28 percent) score far lower. Mental illness (81 percent) and addiction (79 percent) sit at the same level as fatness.
- 92 percent agree that fat people frequently face contemptuous looks. 82 percent say they face prejudice from health professionals and at work.
- Among respondents who described themselves as affected, 64 percent report having been discriminated against by health professionals. 72 percent say shame keeps people from seeking medical help.
And then the part that makes the study genuinely interesting rather than merely useful:
- 79 percent of the population accept fatness as a disease requiring treatment.
- 76 percent say affected people carry a special personal responsibility.
- 49 percent agree that being fat comes down mainly to lifestyle and lack of discipline.
The disease framing has won, and the blame has not gone anywhere. Both statements are true at once, in the same population, at the same time. That is the finding worth carrying forward.
Who commissioned it, and why that matters
The client is named openly at the top of the cockpit page and again in the methods box: Novo Nordisk. No hiding, no shell foundation, no “unrestricted educational grant” phrasing. Credit where it is due, the disclosure is clean.
But disclosure is not neutrality. Novo Nordisk sells semaglutide. A study establishing that fatness is stigmatised, medically undertreated, and recognised as a disease is a study whose conclusions run in a commercially useful direction. That does not make the answers wrong. It shapes which questions were asked in the first place.
The brand questions sitting inside the stigma survey
Read the cockpit closely and you find something a purely academic stigma study would have no reason to include.
One section reports how respondents perceive Novo Nordisk specifically: among those who had noticed content about the company, 17 percent rated it positively, 13 percent neutrally, 14 percent negatively, and 42 percent said they had heard nothing specific about Novo Nordisk at all.
Another section ranks pharmaceutical companies by perceived engagement on the issue. Novo Nordisk leads at 43 percent, ahead of Roche (37 percent), AstraZeneca (28 percent) and Eli Lilly (21 percent).
That is brand tracking. It is a standard, legitimate market research instrument, and it is embedded in the same questionnaire that produced the stigma numbers now circulating in the press. Whoever cites the 64 percent is citing an instrument that also asked how well the sponsor is doing against its competitors.
The policy question that pays the sponsor
Physicians in the survey were asked to rate how effective various measures would be. Among the options offered: “relaxed reimbursement criteria for medications.” 76 percent rated that effective.
The number is real. The framing is worth noticing. A survey does not discover that respondents want looser reimbursement rules. It offers the option and counts the ticks. The measure that would most directly expand the sponsor’s market was one of the items on the list, sitting between “better representation of services in the tariff system” and a “national action plan.”
Again: this is how commissioned research works everywhere, in every industry. It only becomes a problem when the resulting number is later quoted as if physicians had raised the demand spontaneously.
What the sample can carry, and what it cannot
The methods box reports a sampling error of plus or minus 2.5 percentage points at a 50/50 split. That figure applies to the full population sample of 1,539.
It does not apply to the subgroup that produced the most quotable findings. Nine percent of respondents described themselves as affected. That is roughly 139 people, and the cockpit does not publish a separate n for this subgroup or a separate margin of error for it. The 64 percent discrimination figure, the 72 percent shame figure, the 43 percent who found treatment costs a heavy burden: all of these rest on that small subgroup.
They are still worth reporting. They are not worth reporting to the decimal point, and they are not worth treating as equivalent in precision to the headline figures. Anyone quoting them should say which base they come from. Most of the coverage does not.
This is not a one-off
The pattern is older and larger than one Swiss survey. The most widely cited international research on perceptions and barriers in obesity care carries the same sponsor.
ACTION-IO, the eleven-country survey of people with obesity and healthcare professionals published in Diabetes, Obesity and Metabolism in 2019, is registered at ClinicalTrials.gov as NCT03584191. Lead sponsor: Novo Nordisk A/S. Its US predecessor, the ACTION study (NCT03223493), is registered to the same company.
These studies are the source of the frequently repeated statistics about how few fat people are offered structured care and how rarely doctors raise the topic. They are, as far as we can tell, methodologically competent. They are also, without exception, funded by a company that sells the treatment the studies conclude is underprovided. The same asymmetry shapes who can actually get these drugs, which we looked at in Fat and Poor: The GLP-1 Access Gap Punishes Twice and in Covered in France, “Lifestyle” in Germany.
We have not found a comparable body of independently funded international research at that scale. If it exists, we would like to hear about it. Either way, the difficulty of finding one is itself a finding about who gets to fund knowledge about fat people.
Why we use the numbers anyway
The alternative to citing sponsored research is citing nothing, because for many of these questions nothing else at this scale exists. Refusing the data does not produce better data. It produces silence, and silence has never been on our side.
So we use them, with three rules:
- Name the funder in the same sentence as the number. Not in a footnote, not at the end.
- Separate the descriptive findings from the policy conclusions. “Two thirds of affected people report discrimination by health professionals” is a description. “Therefore reimbursement criteria should be relaxed” is a conclusion the sponsor benefits from, and it does not follow automatically from the description.
- Say what base the number rests on.
What nobody asked
The most revealing thing about a questionnaire is usually the question that is missing.
The Adipositas-Barometer asks at length about treatment: which therapies people used, why they stopped, whether reimbursement should be easier, whether interdisciplinary structures are missing. The list of reasons for discontinuing treatment includes “reached target weight” as an endpoint.
It does not ask whether care could be organised so that fat people receive competent treatment for the thing they actually came in for. It does not ask whether stigma might be reduced by changing how practices work rather than by expanding what they prescribe. It does not ask affected people whether they want their bodies treated at all. What competent care would look like instead is the subject of our guide to finding a weight-inclusive therapist.
Those questions are answerable. They are just not the questions a pharmaceutical company has a reason to pay for. Until somebody else pays, they will keep not being asked, and the evidence base will keep pointing in one direction because that is the only direction anyone bought a map for.
Sources, all read at the primary source on 2 August 2026: gfs.bern, “Adipositas Barometer 2026”, cockpit.gfsbern.ch, commissioned by Novo Nordisk, fieldwork April to May 2026, N=1,539 population and N=116 physicians, project code CH26OB00079_06/2026. ClinicalTrials.gov entries NCT03584191 (ACTION-IO) and NCT03223493 (ACTION), lead sponsor Novo Nordisk A/S.
Fatosphere does not publish diet or weight loss content. This article reports on how research about fat people is funded and framed. It is not a recommendation for or against any treatment.