Tirzepatide beats dulaglutide on A1c and weight in two Japanese phase 3 trials

Two phase 3 trials in Japan found that weekly tirzepatide lowered A1c and body weight in adults with type 2 diabetes and that adding it to a single oral diabetes medication was generally tolerated over a year. In the first trial, tirzepatide produced larger reductions in both A1c and weight than dulaglutide. Both studies were funded by Eli Lilly and Company ^[1][2].

SURPASS J-mono: tirzepatide compared with dulaglutide

J-mono enrolled 636 adults at 46 medical research centers and hospitals in Japan, with enrollment between May 7, 2019, and March 31, 2021. Participants were 20 or older and had type 2 diabetes. They were assigned by chance in equal groups to once-weekly tirzepatide 5 mg (159 people), 10 mg (158) or 15 mg (160), or to dulaglutide 0.75 mg (159); 615 people, or 97%, finished the study. The average age was 56.6 years, and 481 participants (76%) were male ^[1].

At week 52, A1c fell by 2.4 points with tirzepatide 5 mg, 2.6 points with 10 mg and 2.8 points with 15 mg, compared with 1.3 points with dulaglutide. The estimated differences versus dulaglutide were 1.1 points for 5 mg, 1.3 points for 10 mg and 1.5 points for 15 mg, all favoring tirzepatide (p<0.0001 for each) ^[1]. Body weight fell by 5.8 kg (7.8%), 8.5 kg (11.0%) and 10.7 kg (13.9%) in the three tirzepatide groups, compared with 0.5 kg (0.7%) for dulaglutide ^[1].

Nausea was the most commonly reported side effect and was more frequent at higher doses: 12% of the 5 mg group, 20% of the 10 mg group and 20% of the 15 mg group, against 8% of the dulaglutide group. Constipation was reported by 15%, 18% and 14% of the tirzepatide groups versus 11% of the dulaglutide group, and nasopharyngitis by 18%, 16% and 14% versus 16% ^[1].

SURPASS J-combo: adding tirzepatide to one oral drug

J-combo enrolled 443 adults at 34 centers in Japan who had A1c between 7.0% and just under 11.0% while taking a single oral diabetes medicine — a sulfonylurea, biguanide, alpha-glucosidase inhibitor, thiazolidinedione, glinide or SGLT2 inhibitor — for at least three months. Participants had a BMI of 23 or higher. They were assigned by chance to tirzepatide 5, 10 or 15 mg weekly, and 398 (90%) completed the study. The trial was open-label, meaning participants and researchers knew which dose was given, and its main goal was safety rather than a comparison with another drug ^[2].

In total, 343 of 443 participants (77%) had at least one treatment-emergent adverse event. These were more frequent in the 15 mg group, at 125 of 148 people (84%), than in the 5 mg group (109 of 148, 74%) and the 10 mg group (109 of 147, 74%). The most common were reported as mild or moderate: nasopharyngitis (75 people, 17%), nausea (74, 17%), constipation (54, 12%), diarrhea (51, 12%) and decreased appetite (44, 10%). No adjudication-confirmed deaths were reported ^[2].

At week 52, body weight fell by 3.8 kg (5.1%) with 5 mg, 7.5 kg (10.1%) with 10 mg and 10.2 kg (13.2%) with 15 mg. Average A1c dropped from 8.5% to 6.0% in the 5 mg group and from 8.6% to 5.6% in both the 10 mg and 15 mg groups ^[2].

Why it matters for patients

Both trials studied only Japanese adults, and the J-combo researchers wrote that potential ethnic differences in the biology of type 2 diabetes are a reason new treatments are tested in Japanese patients ^[2]. Whether the same results would appear in a US patient population has not been reported here. Tirzepatide is the active ingredient in Mounjaro and Zepbound.

J-mono compared tirzepatide with dulaglutide 0.75 mg, a GLP-1 drug, not with semaglutide or with another tirzepatide dose. The weight changes in both trials occurred in people with type 2 diabetes, a group that typically loses less weight on these drugs than people without diabetes. Gastrointestinal complaints such as nausea, constipation and diarrhea were the most common side effects in both studies, and reported side effects generally rose with dose ^[1][2]. Neither report describes US pricing, insurance coverage or availability, and neither offers guidance on choosing between these medicines.

What happens next

Both trials have finished and their 52-week results have been published, so no further results are expected from them. One item remains open: a correction notice for the J-mono paper appeared in the December 2025 issue of the journal that published it, posted online October 15, 2025. That notice carries no abstract, and the specific change has not been detailed in it ^[1].