Explore this article's sources with AI
Follow PeptideMethods on Google
Oral Semaglutide and CVR R
A small retrospective study linked lower deep breathing CVR R with early oral semaglutide discontinuation. It cannot predict an individual's tolerance.
Why we wrote this. A new observational study could be mistaken for a screening rule, so this explainer separates its association from a usable clinical prediction.
In this article (5 sections)
A small retrospective study has found that lower variation in heartbeats during deep breathing was associated with stopping oral semaglutide early because of nausea or vomiting in adults with type 2 diabetes. It is an interesting signal, not a test that can tell an individual whether they will tolerate the medicine. The study analysed 61 people at Japanese centres, including 21 who stopped oral semaglutide within six months for nausea or vomiting[1]. Its result needs external validation before a deep breathing measurement could be treated as a clinical decision tool.
The measure is called CVR R during deep breathing, or CVR R DB. It describes variation between successive heartbeats while a person breathes deeply, and the authors used it as a marker of parasympathetic autonomic function. The parasympathetic nervous system helps regulate involuntary functions, including parts of gastrointestinal motility. That proposed link is plausible, but the researchers did not measure gastric emptying or semaglutide concentrations, so it remains an explanation rather than a demonstrated mechanism[1].
What the study found
Researchers screened 203 consecutive patients who began oral semaglutide between February 2021 and December 2024. After exclusions, 40 people remained on treatment for at least six months and 21 had discontinued because of nausea or vomiting within that period. The discontinuation group had a lower median CVR R DB than the maintenance group: 1.93% versus 3.26%[1].
In the authors' age and sex adjusted analysis, lower CVR R DB was associated with discontinuation. Their model had an area under the curve of 0.832, a statistic that describes how well it separated the two groups in this dataset. An internally derived threshold also divided six month persistence in this same cohort. These are signals from the people used to build the model, not proof that the threshold will work in another clinic or population[1].
The study did not find a statistically significant difference between the discrimination of the deep breathing model and the resting CVR R model. It also did not show that changing treatment on the basis of CVR R DB improves persistence, nausea, vomiting, blood glucose, or any other patient outcome. Those are different research questions. Our semaglutide research overview has context on the broader evidence base.
Why an oral formulation matters
This paper concerns oral semaglutide only. The authors note that the tablet relies on conditions in the stomach for absorption and suggest that autonomic dysfunction could affect gastric motility, exposure, or tolerability. But the study had no injectable GLP 1 receptor agonist comparison and did not measure either gastric emptying or drug levels. It therefore cannot establish that the association is unique to the tablet, or that altered absorption caused the nausea and vomiting[1].
That formulation boundary matters because results from one semaglutide product should not automatically be moved to another. The current US prescribing information describes Rybelsus as oral semaglutide and lists gastrointestinal adverse reactions among important safety information. It also says the product is not recommended in patients with severe gastroparesis[2]. That label is a product document, not evidence that CVR R DB testing identifies who should receive it.
For readers comparing formats, the important distinction is not that one route is universally easier to tolerate. It is that this study did not test an injectable comparator. The semaglutide safety section and the regulation overview explain why product, indication, and jurisdiction should stay attached to any claim.
What the result can and cannot mean
An association is not a diagnosis and is not a cause. People in the lower CVR R DB group may have differed in ways this study could not fully capture. The analysis adjusted for age, sex, and neuropathy stage, but retrospective studies remain vulnerable to selection bias and unmeasured differences between groups. Only 61 people were analysed, with 21 discontinuation events. That is a useful starting point for a hypothesis, not a basis for self screening.
The authors list further limitations. CVR R assessments could have been obtained as much as six months before semaglutide was started. The machine learning work was exploratory. Patients were excluded when relevant assessments or a clearly defined six month outcome were unavailable. The study population consisted only of people with type 2 diabetes, so its relevance to people using semaglutide for obesity without diabetes is uncertain[1].
There is also a practical clinical boundary. A heart rhythm variability result needs context, including the reason it was measured and the rest of a person's diabetes care. It should not be used to start, stop, or alter a prescribed medicine without the clinician managing that care. The FDA approved label remains the source for the product's authorised use and safety information, while a single observational study can only add an early research signal[2].
What we do not yet know
We do not know whether CVR R DB predicts early oral semaglutide discontinuation outside these centres, in other ethnic and clinical populations, or under prospective follow up. We do not know whether it performs better than ordinary clinical history, adverse effect monitoring, or other autonomic tests. Most importantly, no study has shown that testing CVR R DB before treatment improves a patient's experience or treatment outcome.
The proposed biology is also unfinished. The paper connects lower deep breathing heart rate variation with parasympathetic function and gastric motility, but did not directly measure gastric emptying, drug exposure, or a causal pathway to nausea. A future prospective study would need to test those steps rather than assume them. That is why the paper is best read as a candidate predictor study, not a new rule for oral semaglutide care.
What readers can take from this
The headline is narrow: in one retrospective cohort of adults with type 2 diabetes, lower CVR R during deep breathing tracked with early oral semaglutide discontinuation for nausea or vomiting. It did not prove cause, establish a universal cutoff, or compare oral and injectable treatment. It does give researchers a concrete question to test in a larger prospective study.
If nausea or vomiting is affecting a prescribed medicine, contact the clinician or pharmacist responsible for it rather than relying on a screening statistic. For background, see our semaglutide overview and our guide to reading peptide research.
Medical disclaimer: This article is for educational and journalistic purposes only and does not constitute medical advice. Peptides discussed may be classified as prescription medicines or research chemicals depending on your jurisdiction. Always consult a qualified healthcare professional before using any peptide product. PeptideMethods.com does not sell, distribute, or facilitate the sale of any peptide product.
Frequently asked
What did the oral semaglutide CVR R study find?
In 61 analysed adults with type 2 diabetes, people who stopped oral semaglutide within six months because of nausea or vomiting had lower median CVR R during deep breathing than people who continued. This was a retrospective association, not proof that low CVR R caused intolerance.
Can a deep breathing CVR R test predict who will tolerate oral semaglutide?
Not yet. The threshold and prediction model were derived within one small retrospective cohort. They need validation in new populations, and no trial has shown that testing CVR R before treatment improves symptoms, persistence, or other patient outcomes.
Does this finding apply to injectable semaglutide?
The study did not include an injectable GLP 1 receptor agonist comparison. Its result concerns oral semaglutide in adults with type 2 diabetes, so it should not be assumed to apply to injectable products or to people using semaglutide for obesity without diabetes.
Should someone change treatment because of this study?
No. The study does not create a treatment rule or establish a self screening test. Anyone with nausea, vomiting, or concerns about a prescribed medicine should discuss them with the clinician or pharmacist responsible for their care.
Sources
- [1]Ogiso K, Deguchi T, Kawade S, et al. Parasympathetic dysfunction assessed by deep breathing CVR R is associated with early discontinuation of oral semaglutide in type 2 diabetes: a multicenter retrospective study. Journal of Diabetes Investigation. 2026. PMID 42921716.Tier 1 · primary↩
- [2]RYBELSUS oral semaglutide tablets: FDA approved prescribing information via DailyMed, revised October 2025.Tier 1 · primary↩
No revisions yet. First published .