New Model Explains Tirzepatide's Edge

A new modeling study traces tirzepatide's edge over semaglutide on fasting glucose to one specific mechanism: greater suppression of glucose production by the liver, not a bigger boost to insulin-making beta cells and not extra weight loss^[1]. The paper, published ahead of print in The Journal of Clinical Endocrinology and Metabolism, built a mathematical model from data collected in a 28-week trial where people with type 2 diabetes took tirzepatide 15 mg or semaglutide 1 mg, then used that model to test which biological lever actually explains the gap between the two drugs^[1].

What the model actually found

The liver constantly releases stored glucose into the blood, a process called hepatic glucose production. It ramps up between meals and overnight to keep blood sugar from dropping too low, but in type 2 diabetes it runs too high, which is a major reason fasting glucose stays elevated. The researchers built a mechanistic model, a set of equations describing how fasting glucose, fasting insulin, insulin sensitivity, and beta-cell function move together over the course of treatment, then used it to isolate how much of each drug's effect on fasting glucose came from suppressing hepatic glucose production independent of insulin's own effect^[1]. Tirzepatide suppressed it more. That single difference accounted for most of the gap in fasting glucose control between the two drugs^[1].

Why this isn't just about weight loss

It's tempting to assume tirzepatide wins because it produces more weight loss, and weight loss improves glucose control on its own. The model tested that directly by simulating what would happen if weight loss were artificially increased. The result: extra simulated weight loss sped up how quickly fasting glucose improved early in treatment, but it did not lower the long-term fasting glucose level any further^[1]. Weight loss changed the timeline, not the destination. The destination was set by how much hepatic glucose production came down.

Why beta-cell function isn't the story either

The model also tracked beta-cell function using HOMA-B, a standard index that estimates how well the insulin-producing cells in the pancreas are working from fasting glucose and insulin levels. Beta-cell function initially rose on both drugs, then declined as insulin sensitivity improved elsewhere in the body, which is the reverse of the usual pattern seen as type 2 diabetes progresses untreated^[1]. That pattern held for both drugs. It didn't distinguish tirzepatide from semaglutide, which is part of why the researchers concluded the liver, not the pancreas, was doing the differentiating work.

The people who didn't respond as well

Both drug groups included people who responded poorly on fasting glucose. The model found that low responders shared one trait: their hepatic glucose production wasn't suppressed as much as everyone else's^[1]. Running simulations, the researchers found that pushing hepatic glucose suppression further improved fasting glucose in these low responders without triggering the kind of sustained insulin overproduction that can wear out beta cells over time^[1]. That's a specific, testable target for the next generation of incretin-based drugs, or for how existing ones might eventually be dosed.

How this fits with the head-to-head trial data

This modeling paper doesn't stand alone. A separate, larger head-to-head trial, SURPASS-2, randomized 1,879 people with type 2 diabetes to tirzepatide or semaglutide 1 mg for 40 weeks and found tirzepatide produced a bigger drop in HbA1c, a measure of average blood glucose over roughly three months^[4]. The prescribing label for Mounjaro, the brand name Eli Lilly sells tirzepatide under, reports that a 15 mg dose cut fasting glucagon by 28% and after-meal glucagon by 43% compared with no change on placebo^[2]. Glucagon is the hormone that tells the liver to release glucose, so a bigger cut to glucagon is consistent with the bigger cut to hepatic glucose production the new model found. Semaglutide's label describes the same glucagon-suppressing mechanism at a smaller scale^[3], which lines up with the modeled difference between the two drugs.

What we don't yet know

This is one model built on one trial's worth of data, and it was posted ahead of print, meaning it hasn't yet appeared in a final print issue. Mechanistic models are built to fit the data they're given, so the specific numbers behind the low-responder simulations need to hold up when tested against new, independent patient data before anyone changes clinical practice on the strength of them. The model also doesn't test a new drug or a new dose. It re-reads existing trial data through a different lens to suggest where future research should look. Whether hepatic glucose suppression can be targeted more directly, and whether doing so is safe over years rather than months, are open questions the paper itself does not answer.

This article is for educational and journalistic purposes and does not constitute medical advice. If you are managing type 2 diabetes, decisions about which medication to use belong with your prescribing clinician, who can weigh this kind of mechanistic evidence against your own history.