When a study says people who lost more weight on a GLP-1 medication developed fewer chronic conditions, it is tempting to read it as a promise: lose more, get protected more. This analysis, presented at the European Congress on Obesity in May 2026, is a good case study in why that reading goes further than the data.
What the researchers did
A team led by Professor John Wilding of the University of Liverpool used Optum Market Clarity, a large US database of electronic health records and insurance claims. They identified 89,718 adults who started a GLP-1-based medication between January 2021 and June 2024, then followed them through June 2025.
About 76% started semaglutide, 17.5% tirzepatide, and 7% liraglutide. The average participant was 57.5 years old with a BMI of 34.7, and roughly 61% had type 2 diabetes.
People were sorted by how much their BMI changed in the first year of treatment: under 5% reduction (the reference group), 5–10%, 10–15%, 15% or more, and weight gain. The researchers then tracked, on average, about 11 more months for four outcomes: osteoarthritis, chronic kidney disease, obstructive sleep apnea, and heart failure.
What they found
Compared with the under-5% group, people whose BMI fell 15% or more had lower rates of:
- Osteoarthritis: 37% lower
- Chronic kidney disease: 30% lower
- Obstructive sleep apnea: 69% lower
- Heart failure: 32% lower, though this one was not statistically significant
People whose weight went up during the first year had higher rates of the same conditions, with the sleep apnea and heart failure differences reaching significance.
Those are the numbers behind the headlines. Now the part the headlines skip.
What the analysis does not establish
It is an association, not a demonstration of cause. The researchers themselves used associational language, noting that larger reductions were associated with lower risk. People who lose more weight may differ from those who lose less in ways a database cannot see: how closely they were followed, their diet and activity, their baseline health, their access to care. Any of those could account for part of the gap.
The groups were formed after the fact. Treatment was not randomly assigned and people did not get randomly assigned to a weight-loss amount. Someone who responds strongly to a medication may already be different, biologically or behaviorally, from someone who does not.
Half the cohort stopped treatment within a year. The analysis grouped people by weight change regardless of whether they stayed on the drug. So this is not a clean picture of what happens while taking a medication, and it cannot tell you what any particular dose, duration, or product does.
Follow-up was short. About 11 months after the first year is not long for conditions like kidney disease and osteoarthritis that develop over years. Rates over a longer window could look different.
It is conference data from a press release. The sources available describe the findings secondhand. Funding and conflict-of-interest details were not stated in the coverage we could find, and conference results are often preliminary until they pass peer review. Read this as an early signal.
Averages do not predict you. A group-level association says nothing about what will happen for any one person. Response to these medications varies widely, and a larger or smaller change on the scale is not a verdict on anyone’s health or on whether treatment is working.
Why the sleep apnea number deserves extra care
The 69% figure for sleep apnea is the largest, and the most likely to be quoted out of context. Sleep apnea is underdiagnosed, so a database counts only the people whose condition was found and recorded. People engaged in closer follow-up, such as those having more visits during a medication course, may simply be more likely to be tested. That could inflate or distort differences between groups. We covered the broader evidence in GLP-1s and sleep apnea, and the same caution applies here: a number from a population is not your own measurement.
The part that applies to compounded medications
The people in this database were prescribed manufactured, branded products (semaglutide, tirzepatide, and liraglutide). The evidence covers those products and does not transfer to a compounded preparation. Compounded medications are not FDA-approved, are prepared by state-licensed pharmacies for individual patients, and have not been studied in an analysis like this one. Anyone who tells you this research proves a benefit of a compounded product is stretching a study past what it covers.
What is reasonable to take from it
Without overreading it: this is one more observational signal that weight change and chronic condition risk move together, which fits the direction of earlier work. It is a reason for researchers to design stronger studies, and a reason to talk with a licensed provider about your own history and goals rather than chase a number.
Questions about how provider-guided care works are answered in our how it works page and the FAQ.
This post is general information, not medical advice. It does not describe a benefit of any NoBsRx program and is not a claim that any medication will prevent or treat any condition. Compounded medications are not FDA-approved. Treatment is never guaranteed, and individual results vary. Eligibility and every treatment decision are made by independent licensed providers based on an individual clinical assessment, and availability varies by state. Do not start, stop, or change any treatment based on an article.