Phase 1b trial separates what GIP and GLP-1 each contribute in humans

Researchers have published a small human trial that tries to answer a question that has hung over tirzepatide (Mounjaro/Zepbound) since it launched: what does the GIP half of the molecule actually do? In the Phase 1b study, published June 2, 2026 in Diabetes, Obesity and Metabolism, adults with type 2 diabetes taking metformin received a GIP-receptor-only drug called macupatide, the GLP-1 drug dulaglutide, or both together for 12 weeks ^[1].

The answer, based on glucose clamp testing: the two hormones do different jobs. Macupatide alone mainly improved insulin sensitivity — how well the body responds to insulin — while dulaglutide alone mainly increased insulin secretion. Combining them raised both ^[1].

What the trial measured

This was a two-part, randomized, double-blind study at a single site in Germany (NCT05407961), with most authors employed by Eli Lilly and Company ^[1]. Part A tested ascending doses of macupatide for safety; Part B used a 20 mg macupatide dose to compare the three treatment arms ^[1].

Instead of relying on HbA1c or weight, the researchers used hyperinsulinemic euglycemic and hyperglycemic clamps — intensive lab procedures that measure insulin sensitivity (the "M value"), insulin secretion rate (ISR), and a combined measure called the clamp disposition index (cDI), which reflects how well beta cells perform relative to the body's insulin resistance ^[1].

At baseline, average values across the study were a cDI of 0.3, an M value of 5.8 mg/kg/min, and an ISR of 116.8 pmol/min/m² ^[1].

After 12 weeks, the increases broke down like this ^[1]:

• Insulin sensitivity (M value): up 1.3 (33.1%) with macupatide, 0.2 (10.4%) with dulaglutide, 2.0 (38.3%) with the combination.
• Insulin secretion (ISR): up 37.6 (32.2%) with macupatide, 191.1 (163.6%) with dulaglutide, 225.3 (192.9%) with the combination.
• Clamp disposition index: up 0.2 (63.2%), 0.5 (181.0%) and 1.0 (321.5%), respectively.

The authors report that macupatide alone significantly increased the M value, and increased ISR to a smaller but still significant degree. The combination was described as "numerically greater" than either single drug on all three measures — wording that signals the comparison was not framed as a formal statistical win ^[1].

The most common side effects were injection site reactions and diarrhea. Discontinuations because of treatment-emergent adverse events were uncommon, and there were no deaths ^[1].

Why it matters for patients

Tirzepatide has outperformed semaglutide on weight and glucose control in head-to-head trials, and this study is part of the effort to explain why ^[1]. The authors note that a previous clamp study by the same lead investigator found tirzepatide improved both insulin sensitivity and secretion more than semaglutide, but that study could not tell which receptor deserved credit ^[1]. GIP is known to affect fat tissue — including blood flow and the breakdown of fat — which could plausibly improve insulin sensitivity separately from its effect on the pancreas ^[1].

These results suggest the GIP component is not just a second insulin-releasing signal. That has practical implications for future drug design: it supports the idea that pairing GIP with GLP-1 does something neither can do alone, and it raises the question of whether GIP-only drugs have a role of their own.

Important caveats: this was a Phase 1b trial, meaning it was small and designed to study mechanism rather than to prove clinical benefit. It ran 12 weeks in people with type 2 diabetes on metformin, so it says nothing directly about longer-term blood sugar outcomes, weight loss in people without diabetes, or how macupatide would compare with approved drugs on those measures. The published abstract text available here does not report the number of participants, the dulaglutide dose used, or changes in HbA1c and body weight, so those details are not yet clear from this material ^[1].

Eligibility was also narrow: adults aged 18 to 70 with stable body weight, weight up to 150 kg, BMI of 23 to 45, and HbA1c between 7.0% and 10.0% on diet, exercise and stable metformin (or 6.0% to 9.5% under other conditions described in the full paper) ^[1].

What happens next

The study appears in Diabetes, Obesity and Metabolism, Volume 28, Issue 8, pages 7196–7206, first published June 2, 2026 ^[1]. The sources do not describe any later-stage trial plans, regulatory filings or timelines for macupatide, so the drug's future development path is not yet known from this publication. Macupatide is not an approved medicine.