Tirzepatide 15 mg cut weight 20.9% and beat semaglutide head to head

How does tirzepatide work?

Tirzepatide is a 39-amino-acid synthetic peptide that activates both the GLP-1 receptor and the GIP receptor, unlike semaglutide, which activates only the GLP-1 receptor. A C20 fatty-diacid chain attached to tirzepatide binds circulating albumin, which slows clearance and allows once-weekly dosing.

The combined effect on insulin secretion, glucagon suppression, gastric emptying and hypothalamic appetite signaling produces both the glycemic and the weight effects. The GLP-1 side of this mechanism is well characterized. The added value of GIP agonism specifically, as opposed to a stronger GLP-1 effect from a well-engineered molecule, is still argued over in the literature.

How much weight does tirzepatide produce?

Tirzepatide's 15 mg dose produced a mean 20.9% weight loss at 72 weeks versus 3.1% on placebo in SURMOUNT-1, a trial of 2,539 people with obesity without diabetes ^[2]. SURMOUNT-1 is the trial most people quote.

Tirzepatide's evidence is grade A because the findings come from large, well-designed Phase 3 randomized controlled trials, not observational data or small pilot studies. Five trials anchor it: SURMOUNT-1, SURPASS, SURMOUNT-2, a 176-week prediabetes follow-up from SURMOUNT-1, and an obesity head-to-head trial against semaglutide ^[2]⁠^[1]⁠^[3]⁠^[4]⁠^[5].

In SURMOUNT-1, 57% of participants on the 15 mg dose lost 20% or more of body weight ^[2]. That figure matters more than the headline average: it means over 40% of participants did not reach the 20% threshold, and some lost meaningfully less. The trial average is real, but it is the average of a distribution, not a promise.

Is tirzepatide better than semaglutide?

Yes, in two head-to-head trials: tirzepatide beat semaglutide on weight loss in obesity, 20.2% versus 13.7% at 72 weeks, and on HbA1c in type 2 diabetes ^[5]⁠^[1]. The obesity trial (n=751) found superiority across every weight-loss threshold tested (p<0.001) ^[5].

In type 2 diabetes, the head-to-head SURPASS trial (n=1879) found all three tirzepatide doses, 5, 10 and 15 mg, noninferior and superior to semaglutide 1 mg for HbA1c reduction at 40 weeks ^[1]. Tirzepatide patients lost up to 5.5 kg more weight than the semaglutide group ^[1].

Two separate trials pointing the same way make the advantage a replicated finding, not spin. Semaglutide's own randomized trials show what the comparator achieves against placebo.

Does tirzepatide work as well in people with type 2 diabetes?

Tirzepatide works in type 2 diabetes with smaller but still substantial effects: 12.8% and 14.7% mean weight loss on 10 mg and 15 mg at 72 weeks ^[3]. Those figures come from SURMOUNT-2, a trial of 938 people with obesity and type 2 diabetes ^[3].

The pattern holds across the program. People with diabetes tend to lose somewhat less weight on the same dose than people without it, likely reflecting different underlying metabolic dynamics.

Does tirzepatide prevent type 2 diabetes?

Yes, in people with prediabetes: over 176 weeks, tirzepatide reduced progression to type 2 diabetes by 93% relative to placebo (1.3% versus 13.3%, hazard ratio 0.07) ^[4]. The trial followed the prediabetes subgroup of SURMOUNT-1, testing durability directly ^[4].

Weight loss held at 19.7% on the 15 mg dose for more than three years ^[4]. That is one of the stronger durability signals in the obesity drug literature and worth taking seriously.

How much of tirzepatide's weight loss is muscle?

Rapid weight loss of any kind, from a drug, surgery or aggressive caloric restriction, tends to include some loss of lean mass alongside fat mass. That loss is physiologically plausible and reported as a general concern with GLP-1 and GIP agonist therapy. If a specific percentage appears online, ask which study it comes from.

Concurrent resistance training is a reasonable, low-risk strategy to blunt lean-mass loss during any rapid weight-loss intervention. None of the five trials tests that combination with tirzepatide; the trial evidence on preserving muscle during GLP-1 weight loss comes from other designs.

What happens after stopping tirzepatide?

None of the five trial reports cited below includes a placebo-controlled discontinuation phase, so they give no controlled regain figure. Regain after stopping is the single most consequential gap in these trials for anyone planning long-term use.

Tirzepatide's effects depend on continuous receptor agonism, so appetite suppression and metabolic effects would be expected to wane once the drug is cleared, and some degree of weight regain is biologically plausible. Treat any specific regain number as unverified until it traces to a trial. What happens when people come off a GLP-1 covers the withdrawal evidence across the drug class.

Is GIP agonism why tirzepatide outperforms semaglutide?

The trials do not show it: tirzepatide wins head to head against semaglutide ^[1]⁠^[5], but that comparison does not prove GIP agonism is the reason. Forums often treat GIP and GLP-1 dual agonism as an obviously superior design over GLP-1 alone.

The advantage could reflect differences in receptor affinity, pharmacokinetics or dosing instead of GIP's specific contribution. Some researchers have also proposed that GIP receptor antagonism, not agonism, could work similarly in preclinical models, an unresolved area of the science.

Tirzepatide works better in trials. Why it works better at the receptor level is still debated by the people who study it.

Do tirzepatide stacking and microdosing protocols have trial support?

No: nothing in the tirzepatide trial data addresses stacking, microdosing or symptom-timed dosing. That covers combining tirzepatide with other peptides, such as retatrutide or cagrilintide, using doses below the studied range ("microdosing"), and timing injections for symptom control beyond standard titration.

Peptide communities discuss these as plausible ideas, and plausible is not the same as tested. Every trial number came from fixed, studied doses on a fixed weekly schedule. Conflating "tirzepatide is well studied" with "my sourced, self-dosed protocol is well studied" is the exact error to avoid.