Skip to main content
Do the weight-loss effects seen with GLP-1 medicines in the general population also show up in people on peritoneal dialysis? Until now, that question had almost no direct evidence behind it, because dialysis patients are routinely left out of the large trials that establish how these medicines work. A new retrospective study presented at ASN Kidney Week 2026 in Denver takes a first serious look, tracking nearly 5,900 adults with type 2 diabetes who started peritoneal dialysis and comparing those who used GLP-1 medicines with those who did not. The headline finding is that greater cumulative exposure to the medicines was associated with progressive weight reduction, and with a lower modelled burden of hospital stays.
That sentence needs careful unpacking, because this kind of study can show patterns but cannot prove cause and effect the way a randomized trial can. This article explains who was studied, what changed and what did not, why this patient group has been missing from the evidence until now, and what would need to happen before these findings could change clinical practice. It is research news, not medical guidance: anyone on dialysis has nutritional and treatment needs that are managed with their own care team, and nothing here is a suggestion about any individual treatment.
The Study at a Glance
| Detail | What the researchers did |
|---|---|
| Study type | Retrospective cohort study using prescription and health records |
| Patients | 5,897 adults with type 2 diabetes who started peritoneal dialysis |
| Study window | Patients starting dialysis between April 2021 and September 2024 |
| GLP-1 users | 516 patients, or 8.7 per cent of the group, initiated a GLP-1 medicine |
| Follow-up | From day 90 after the start of peritoneal dialysis |
| Main outcome | Relative change in body weight |
| Other outcomes | Hospital stays, ultrafiltration, urine output, plus exploratory looks at blood sugar control and mortality |
| Where presented | ASN Kidney Week 2026, Denver, poster FR-PO1115 |
The design matters for how much weight the findings can carry. Rather than assigning patients at random to a medicine or a comparison group, the researchers looked back at records of people who had already started dialysis and sorted their follow-up time according to prescription data. They used statistical models designed to account for the fact that treatment decisions change over time and that patients leave follow-up for different reasons. That is a respectable approach for this kind of question, but it cannot fully remove the possibility that the people who received the medicines differed from those who did not in ways the records could not capture.
Why Dialysis Patients Were Missing From the Evidence
The gap this study addresses is not an accident. As the study’s presenter, Steven M. Brunelli of DaVita Clinical Research, put it, dialysis patients, and peritoneal dialysis patients in particular, generally do not make it into the landmark clinical trials. When a major breakthrough arrives, clinicians are left guessing whether the findings apply to a group whose biology and daily treatment differ sharply from the trial populations.
Peritoneal dialysis is a distinct world from the general population in which GLP-1 medicines were first studied. Patients face cardiovascular risk, a substantial metabolic burden from the glucose in dialysis fluid itself, and a progressive loss of whatever kidney function remains. Those features could plausibly change how a medicine behaves, how much weight changes, and whether benefits outweigh risks. The FLOW trial had already shown that semaglutide reduced the risk of death from cardiovascular causes in people with type 2 diabetes and chronic kidney disease, but evidence specifically in peritoneal dialysis populations remained limited. That is the gap the new analysis tries to narrow.
There is a human side to this evidence gap that is easy to overlook. When a group is excluded from trials, every later decision about that group rests on extrapolation — assuming that what worked elsewhere will work here too. Extrapolation is sometimes the only option, but it is a weaker basis for decisions than direct evidence, and it leaves patients and clinicians in an uncomfortable position: either deny a potentially helpful treatment for lack of proof, or offer it on the basis of borrowed proof. Studies like this one are valuable precisely because they start replacing extrapolation with observation, even if observation alone cannot settle the question.
What Changed: Weight and Hospital Stays
The central result concerns weight. Greater cumulative exposure to GLP-1 medicines was associated with progressive weight reduction of 1.19 percentage points, with a 95 per cent confidence interval from 1.97 to 0.42 percentage points in the negative direction and a p-value of 0.003. In plain terms, the more exposure patients accumulated, the more their weight tended to fall relative to the comparison, and the statistical signal was strong enough that it is unlikely to be mere chance in this dataset.
The pattern also showed a dose-response shape, which strengthens the impression that the association is real rather than a quirk of one subgroup. Each additional quarter of exposure was associated with a further 0.73 percentage-point reduction, a relationship with a p-value below 0.001. When the analysis was restricted to people newly starting the medicines, the association grew stronger, at 1.69 percentage points. A graded pattern like this — more exposure, more change — is one of the features researchers look for when judging whether an observational finding deserves to be taken seriously.
The second notable result was a lower modelled hospitalization burden. Greater cumulative exposure was associated with an incidence rate ratio of 0.935 per exposure quarter, with a p-value of 0.004. In everyday language, each additional quarter of exposure was linked to a lower rate of hospital stays in the model. Hospital stays matter to patients far beyond the numbers: they interrupt dialysis routines, disrupt family life and often signal that something in the overall health picture has worsened. A finding that points toward fewer of them is therefore worth attention, even while remembering that this is a modelled association, not a proven effect.
For readers trying to place these numbers in context, it helps to remember what weight change means in a dialysis setting. Body weight in dialysis patients is not a simple measure of fat. Fluid shifts, dialysis fluid itself and changes in muscle can all move the scale. That complexity does not invalidate the finding, but it does mean the headline weight change should not be read as a pure measure of fat loss. It is a signal that something changed in the direction researchers expected, and it invites the more detailed, prospective research that can pin down what exactly changed.
What Did Not Change
Just as important as the positive findings are the places where the study found no significant association. Greater GLP-1 exposure showed no significant relationship with two dialysis-specific measures: body surface area-indexed ultrafiltration, which reflects how well excess fluid is removed, and urine output, which reflects remaining kidney function. For patients and clinicians, those null results are informative. They suggest that, within this analysis, the medicines were not associated with measurable changes in those particular aspects of dialysis function, for better or worse.
Mortality estimates pointed in a protective direction but were imprecise, with a hazard ratio of 0.706 and a p-value of 0.33. That combination — a favourable-looking estimate with a wide margin of uncertainty — is the statistical equivalent of a shrug. It does not show harm, and it does not show benefit. It shows that this dataset, with its follow-up and event numbers, could not answer the mortality question, and that anyone claiming otherwise from these numbers would be overreaching.
Blood sugar analyses ran into a different problem. The researchers reported that those analyses violated the assumptions of their statistical approach and were therefore interpreted descriptively, because of confounding by indication — the tendency for the reasons a medicine is prescribed to be entangled with the outcomes being measured. In plain terms, people prescribed a diabetes medicine often differ from those not prescribed one precisely because of their blood sugar history, which makes it hard to separate the medicine’s effect from the reason it was given. The honest response, which the researchers adopted, is to describe rather than to claim.
Taken together, the pattern of results is coherent rather than sweeping. Weight moved in the expected direction with a dose-response shape. Hospital stays moved in a favourable direction. Dialysis-specific measures did not move significantly. Mortality could not be judged. Blood sugar could not be cleanly assessed. That is a realistic profile for an early observational study, and it is more credible for its restraint than a story claiming everything improved would have been.
A Timeline of the Evidence
The story of this evidence stretches back further than the conference presentation. The cohort covers patients who started peritoneal dialysis between April 2021 and September 2024, a period in which GLP-1 medicines were becoming widely used in the general population while dialysis patients remained largely outside the trial evidence. Follow-up began at day 90 after dialysis started, a design choice that avoids the unstable early period when treatment is still being established.
The analysis itself was then brought to ASN Kidney Week 2026, held from October 21 to 25 in Denver, Colorado, as poster FR-PO1115. Conference presentations are a normal first step for this kind of research: they put findings before specialists for scrutiny before the fuller, peer-reviewed publication that typically follows. The HCPLive report on the presentation provides the detailed account of the study’s design and results, and readers who want the primary record can trace it through the conference abstract.
What happens next, according to the study’s presenter, is the step that matters most. Brunelli said he hopes the work builds momentum for a dedicated trial in peritoneal dialysis patients. That is the right ambition, because only a prospective, randomized trial can answer the questions this study raises but cannot settle: whether the weight change reflects fat loss or fluid shifts, whether hospital stays genuinely fall, and whether the balance of benefits and risks favours treatment in this specific population. Until such a trial exists, the reasonable clinical stance is the one Brunelli himself described — that these data are consistent with not withholding a medicine from patients who would otherwise be good candidates merely because they are on peritoneal dialysis, while recognising that the evidence remains observational.
What This Means Beyond Dialysis
For the general reader, the most transferable lesson of this study is not about any medicine. It is about how weight and energy balance are studied. Weight change is the visible outcome of a long chain: energy taken in, energy used by the body at rest and during activity, and the body’s own regulation of appetite and storage. A guide to what TDEE means explains total daily energy expenditure, the total calories the body uses in a day, and why it is the starting point for understanding any weight change. A comparison of TDEE versus BMR clarifies the difference between the energy the body needs at complete rest and the energy it uses across a full day of living.
That general framework is also why observational weight studies need such careful reading. When a study reports a percentage-point change in weight, the number alone does not say whether the change came from eating less, moving more, losing fluid or losing muscle. Each of those has different implications, and only detailed follow-up can separate them. The dialysis study is a good illustration: a real, statistically strong signal that still leaves the mechanism open, and that is exactly why its authors call for a dedicated trial rather than declaring the question closed.
For readers curious about their own numbers, a step-by-step explanation of how to calculate TDEE walks through the standard method for estimating daily energy needs from basic personal factors. That kind of estimate is a general educational tool, not a clinical instrument, and it is not designed for people with complex medical conditions such as kidney failure, whose energy and nutrition needs are managed individually with their care teams.
Frequently Asked Questions
What did the new study find about GLP-1 medicines in peritoneal dialysis?
In a retrospective cohort of 5,897 adults with type 2 diabetes who started peritoneal dialysis, greater cumulative exposure to GLP-1 medicines was associated with progressive weight reduction of 1.19 percentage points, with a stronger association of 1.69 percentage points among new users. Greater exposure was also associated with a lower modelled hospitalization burden.
Was this a randomized clinical trial?
No. It was a retrospective cohort study based on prescription and health records, presented as a conference poster at ASN Kidney Week 2026. Its presenter stressed that it is observational research, not a clinical trial, and that prospective studies are needed to confirm the long-term effects in this population.
Why were dialysis patients missing from earlier GLP-1 evidence?
Dialysis patients, and peritoneal dialysis patients in particular, generally do not take part in the landmark clinical trials of new medicines. That left clinicians with little direct evidence about whether findings from the general population applied to people on dialysis, whose treatment involves glucose-containing dialysis fluid and progressive loss of remaining kidney function.
Did the study find changes in dialysis function or survival?
No significant associations were found with ultrafiltration or urine output. Mortality estimates pointed in a protective direction but were imprecise and could not support a conclusion. Blood sugar analyses could not be interpreted as causal because they were affected by confounding related to why the medicines were prescribed.
What research should come next?
The study’s presenter hopes the findings build momentum for a dedicated prospective trial in peritoneal dialysis patients. Such a trial could establish whether the weight change reflects fat loss, whether hospital stays genuinely fall, and what the balance of benefits and risks is for this specific group.
For general readers, the study is a reminder that weight is an outcome of daily energy balance, not a single number on a scale. A free TDEE calculator gives a general estimate of daily energy needs, which is the starting point for understanding how weight changes over time.
