Oral semaglutide dropout rates in Japan

A retrospective database study from Jikei University Hospital in Japan examined why some patients stop taking oral semaglutide (marketed as Rybelsus) shortly after starting it. The study, published in Diabetology International by Mizuki Ishiguro and Rimei Nishimura, tracked 358 people with type 2 diabetes who were newly prescribed oral semaglutide between 2022 and 2023^[1]. A correction to the article was published in May 2026^[2].

This article covers the original study's findings, what the correction changed, and what the results mean for people taking or considering oral semaglutide.

What the study found

Of the 358 participants, 65 (about 18%) discontinued oral semaglutide within 180 days, primarily because of gastrointestinal side effects. The remaining 293 continued treatment past that mark^[1].

The most notable finding was that two factors predicted discontinuation specifically at the 3 mg starting dose: higher baseline HbA1c levels and concurrent metformin use. Patients taking high-dose metformin (1,000 mg per day or more) were significantly more likely to stop oral semaglutide due to gastrointestinal symptoms at this early stage^[1]. The association was statistically significant (p = 0.005) in the 3 mg group.

At the higher maintenance doses of 7 mg and 14 mg, no meaningful predictive factors for discontinuation emerged^[1]. That pattern suggests the problem is concentrated at treatment initiation, when the GI system has not yet adapted to the drug, and when additive GI burden from metformin may tip the balance.

What the correction changed

The May 2026 correction^[2] fixed a formatting error in the original paper. A paragraph describing Figure 1 (which displayed metformin dosing categories, semaglutide dose strata, and treatment outcomes) had been placed in the Conclusion section instead of the figure legend. No data, results, or conclusions were altered. The correction moved that descriptive text to its proper location.

Corrections like this are routine in academic publishing. The study's findings remain as originally reported.

Why GI side effects matter for oral semaglutide

Gastrointestinal symptoms are the most common reason people stop taking GLP-1 receptor agonists. In the PIONEER clinical trial programme that led to Rybelsus approval, nausea occurred in about 15% of participants, diarrhoea in 10%, and vomiting in 7%. Overall, about 4% of trial participants discontinued due to GI adverse effects^[3].

Most of these events were mild to moderate and clustered during the dose-escalation period^[3]. That aligns with what the Ishiguro and Nishimura study found: the 3 mg starting dose is where the trouble concentrates, not the higher maintenance doses.

The metformin question

Metformin itself causes GI side effects, particularly at higher doses. Nausea, diarrhoea, and abdominal discomfort are well-documented metformin effects. When a patient already on high-dose metformin adds oral semaglutide, both drugs affect the gut at the same time. The Japanese data suggests this overlap may make early discontinuation more likely^[1].

Pharmacokinetic studies show no clinically relevant interaction between semaglutide and metformin in terms of drug levels^[3]. The issue appears to be additive GI burden rather than a drug-drug interaction. The practical implication for clinicians is that patients on high-dose metformin may need closer monitoring during the oral semaglutide initiation phase, or may benefit from a slower dose-escalation schedule.

Context and limitations

This was a single-centre, retrospective study with 358 participants. It did not randomise patients or control for all possible confounders. The population was entirely Japanese, and body composition and metabolic responses to GLP-1 agonists can vary between populations. Results from a single Japanese university hospital should not be assumed to apply directly to other populations or clinical settings.

The 18% discontinuation rate within 180 days falls within the range seen in other real-world studies of oral semaglutide, though direct comparisons across studies are complicated by different follow-up periods and definitions of discontinuation. A separate Japanese analysis from the same research group examined dosing effects on glycaemic control and weight in the same hospital database^[4].

What this means if you are taking oral semaglutide

If you are on metformin and your clinician is considering adding oral semaglutide, this study does not say the combination is unsafe. It says the first weeks may be harder on the gut than for someone not on metformin, and the 3 mg starting phase is when that matters most. Talking to your prescriber about pacing the introduction is reasonable.

This article is for educational purposes. It does not constitute medical advice. If you are experiencing side effects from any medication, consult your healthcare provider.