The order is always the same. Someone calls your name, walks you to a scale, writes down a number, then sits you down and wraps a cuff around your arm. Weight first, blood pressure second. Nobody picked that sequence for a clinical reason. It is just how the room is laid out.
A new experiment suggests the sequence has a cost, and that the cost lands on the measurement that actually matters.
What the study did
“Please step on the scale”: The experience of being weighed promotes sustained systolic blood pressure levels, by Angela C. Incollingo Rodriguez, L. S. Nunes and M. S. Kirschner, was published as an advance online article in the APA journal Stigma and Health in 2026 (DOI 10.1037/sah0000690).
190 undergraduates went through a mock health care visit. Everyone gave a baseline blood pressure reading and a saliva sample. Then they were randomised three ways. One group was weighed the standard way, by a researcher, immediately after the baseline. One group weighed themselves alone in a room and typed the number into a form. One group was not weighed at all until every other measurement was finished. Blood pressure and saliva were taken again twenty minutes later.
Both weighed groups held their systolic blood pressure at the level it started at. The group that was not weighed saw theirs fall over the same twenty minutes. The effect size was Cohen’s f = .30.
That is the finding, and it is worth being exact about what it is not. There was no difference in diastolic blood pressure. None in cortisol. None in self-reported stress. Baseline systolic readings were already slightly elevated in all three groups, above 120 mmHg, which may say something about walking into a lab or a clinic in the first place.
The authors’ own summary: being weighed may promote sustained stress physiology outputs via systolic blood pressure, regardless of how weight is measured. Self-weighing did not help. Whatever is happening here does not depend on who holds the clipboard.
The obvious objection, and why it cuts the other way
The mean BMI in the sample was 24.29. Participants were young, predominantly white, and from relatively comfortable backgrounds. Most of them were not fat. So the study did not measure what happens to fat patients on a scale.
The authors say this themselves, and add a point worth sitting with: if the sample was at low risk of having been stigmatised on a scale before, then finding an effect anyway is a more stringent test of the hypothesis, not a weaker one. For people who have been weighed with unsolicited commentary attached, year after year, the prediction is not a smaller effect.
They also explain why they did not simply recruit higher-weight participants. Deliberately exposing people to something already known to distress them raises an ethical problem they were not willing to wave through. Their proposal is replication with community participants in real clinical settings.
Until that happens, this is one experiment on 190 students. A reason to look, not a settled fact. And the full text sits behind the publisher’s paywall. What is quoted here comes from the published abstract and impact statement, and from the paper’s conclusions as reproduced at length by Ragen Chastain, who heard the lead author present the work.
The measurement error that is already documented
The weighing effect is new and unreplicated. There is a second error in the same room that is neither.
In 2023, a randomised crossover trial at Johns Hopkins put 195 adults through four rounds of blood pressure measurement and varied only one thing: the cuff. The Cuff(SZ) trial (JAMA Internal Medicine, DOI 10.1001/jamainternmed.2023.3264) found that measuring someone who needs a large cuff with a regular cuff produced a systolic reading 4.8 mmHg too high. For people who need an extra-large cuff, the regular cuff produced a reading 19.5 mmHg too high.
The second number is not a rounding error. In that group, the average reading taken with the standard cuff was 144/87 mmHg. The same people, measured with the cuff that fit their arm, averaged 125/79 mmHg. Under US guidelines the first number is stage 2 hypertension. The second is not hypertension.
The authors also show how big the problem is. A national US survey found a regular cuff is the right size for 51 percent of adults; 40 percent need a large one. Most home monitors ship with a single cuff, and it is the regular one.
Put the two findings in sequence and you get the standard appointment: a fat patient is weighed, then measured with a cuff that does not fit, and the resulting number goes into the file. From there it is read as a fact about the body.
What follows, and what does not
None of this is an argument against measuring blood pressure. It is the opposite. Blood pressure is one of the few cheap and genuinely useful things a clinic does. A wrong reading is worse than no reading, because it gets treated as real: more tests, medication decisions, a line in the record that travels with you and shapes what the next doctor assumes before you speak.
Three things follow from that.
The order is not fixed. The study authors suggest measuring blood pressure before weighing, or moving the weigh-in to the end of the visit. That costs a practice nothing.
The cuff has to fit. Arm circumference is measured with a tape at the midpoint between shoulder and elbow, and the cuff is chosen from that number. Guessing by eye is where the 19.5 mmHg comes from.
A weigh-in is not automatic. Where weight is not needed for a dose, a specific clinical question or a monitoring task, there is no measurement reason to do it. In the US, More-Love.org sells “Don’t Weigh Me” cards for handing over at the desk, precisely because saying it out loud in a waiting room is hard.
What applies in Germany
Section 630d of the German Civil Code requires the treating side to obtain the patient’s consent before carrying out a medical measure, and paragraph 3 states that consent may be withdrawn at any time, without giving reasons, and informally. The provision’s core case is an intervention in body or health, and no statute settles whether a routine weigh-in counts as one. But the structure of the German treatment contract is consent, not compliance. Patients agree to measures; they are not subjected to them. Asking to skip the scale, to stand on it facing away, or to have the cuff size checked is a request inside that structure, not obstruction of it.
Section 630g gives patients the right to inspect their own file. Useful to know, although which cuff was used is almost never recorded.
What to ask for
- Blood pressure first, scale afterwards.
- Upper arm circumference measured with a tape, cuff selected from that number. Above roughly 32 cm, a regular cuff is the wrong one.
- A second reading after five minutes seated, if the first one is high.
- If the weight is not needed for a specific clinical purpose, say you would rather skip it. If it is needed, ask to face away from the display and not to be told the number.
If a practice reacts badly to any of that, our guide to what to do when a doctor blames everything on your weight covers the next steps, and our guide to finding a weight-neutral doctor covers the longer-term one.
None of this is about whether anyone should weigh less. It is about whether the number in the chart describes the patient or the procedure. Right now, sometimes, it describes the procedure.
This article does not give medical advice and does not recommend for or against any treatment. Decisions about blood pressure, medication and monitoring belong to you and the people treating you.