Drug facts

Tirzepatide benefits, as the label and the trials state them

By David Chen·Editor·Reviewed September 2026

What Zepbound’s label approves¹

Section 1 is short enough to read in full. Zepbound is indicated “in combination with a reduced-calorie diet and increased physical activity”:

“to reduce excess body weight and maintain weight reduction long term in adults with obesity or adults with overweight in the presence of at least one weight-related comorbid condition.”

“to treat moderate to severe obstructive sleep apnea (OSA) in adults with obesity.”

Two details in that wording do work. The diet and activity are part of the indication, not a footnote to it: every figure below comes from trials in which both arms got dietary counseling. And maintenance is in the indication — the approved use is keeping weight off, not only losing it. The label adds one limitation of use: taking Zepbound with another tirzepatide product, or with any GLP-1 receptor agonist, is not recommended.

What Mounjaro’s label approves²

Mounjaro is the same molecule with its own trials and its own approvals, and its section 1 now reads:

“as an adjunct to diet and exercise to improve glycemic control in adults and pediatric patients 10 years of age and older with type 2 diabetes mellitus.”

“to reduce the risk of major adverse cardiovascular (CV) events (CV death, non-fatal myocardial infarction (MI), or non-fatal stroke) in adults with type 2 diabetes mellitus who are at high risk for these events.”

Both are recent additions, and the label dates them itself. Its Recent Major Changes table lists the pediatric wording at 12/2025 and the cardiovascular indication at 08/2026. That second line is the one worth knowing about: until this year, no tirzepatide product carried a cardiovascular indication anywhere. The trial it rests on, and what it actually showed ↓

These are Mounjaro indications. Zepbound’s label carries neither of them, and Mounjaro’s carries neither of Zepbound’s. Why one molecule has two labels →

Weight³

The pivotal weight trial, SURMOUNT-1, randomized 2,539 adults with obesity, or overweight with a weight-related complication, to one of three doses or to placebo for 72 weeks. People with diabetes were excluded. At week 72 mean weight change was -15.0% on 5 mg, -19.5% on 10 mg and -20.9% on 15 mg, against -3.1% on placebo. In the 15 mg group 57% of participants lost 20% of body weight or more, against 3% on placebo.

That is the approved use, measured. What it does not tell you is what happens next: the withdrawal trial that stopped the drug is where the maintenance half of the indication comes from, and does tirzepatide work covers how often it fell short.

Blood pressure⁴

This is the most-searched of the unapproved effects, and the label does report it. Table 3 gives the change at week 72 in the two pivotal weight trials. In Study 1, which is SURMOUNT-1, blood pressure fell further on tirzepatide than on placebo at every dose:

Every one of those rows carries the same footnote in the label: “Not controlled for type I error rate.” In plain terms, the trial was built to prove the weight endpoints, and these were measured alongside without the statistical protection that would let them stand as proven findings. In the diabetes trial, Study 2, the systolic differences were 4.4 mmHg at 10 mg and 5.9 mmHg at 15 mg, with the same footnote.

A substudy measured it properly over a whole day.⁵ 600 SURMOUNT-1 participants wore 24-hour monitors, and 494 had usable readings at both baseline and week 36. Their mean 24-hour systolic pressure was 124.6 mmHg to start; entry required pressure under 140/90 and, if treated for hypertension, a stable prescription for three months. Against placebo, 24-hour systolic pressure fell 7.4 mmHg at 5 mg, 10.6 mmHg at 10 mg and 8.0 mmHg at 15 mg, day and night alike. Diastolic pressure fell 2.0 mmHg at 5 mg and 2.9 mmHg at 10 mg, but not significantly at 15 mg. The authors calculated that about 70% of the systolic change tracked weight loss.

The same substudy is where the other half of the story is.⁶ Heart rate rose against placebo by 2.1, 2.3 and 5.4 beats per minute at week 36, and the label reports a mean increase of 1 to 3 beats per minute across the pooled weight trials against no increase on placebo. The label also counts low blood pressure as an adverse reaction in 1.6% of participants against 0.1% on placebo — 2.2% among those already taking blood pressure medicines, against 1.2% among those who were not, and it notes hypotension occurring alongside dehydration from stomach upset. What the label says about other medicines →

No tirzepatide product is approved to treat high blood pressure. A measured fall in a weight trial is not a treatment claim, and it is not a reason to change an antihypertensive prescription.

Cholesterol, triglycerides and blood sugar⁴

The same label table reports lipids as a percentage change relative to placebo at week 72 in Study 1, alongside waist circumference. All of the lipid rows carry the “not controlled for type I error rate” footnote.

Blood sugar divides by who was in the trial. Study 1 excluded people with diabetes, and their glycated hemoglobin started at 5.6% and fell 0.3 to 0.4 points further than placebo — a small move from a normal starting point, again uncontrolled for type I error. Study 2 enrolled people with type 2 diabetes, starting at 8.0%, and there HbA1c fell 1.6 percentage points further than placebo at both 10 mg and 15 mg. That figure is one of the few in the table the label marks as controlled for type I error and statistically significant for superiority.

Blood sugar control in type 2 diabetes is a Mounjaro indication, not a Zepbound one, and it has its own head-to-head evidence: SURPASS-2, against semaglutide 1 mg.

Sleep apnea and inflammation⁷

Sleep apnea is the second Zepbound indication, and the only place an inflammation marker appears anywhere in either label. Two 52-week placebo-controlled trials enrolled 469 adults with moderate to severe obstructive sleep apnea and obesity: Study 5 for people not using positive airway pressure, Study 6 for people who were. The apnea-hypopnea index fell by 25.3 and 29.3 events an hour on tirzepatide against 5.3 and 5.5 on placebo. The sleep apnea results in full →

The label then adds one sentence that the inflammation searches are really about:

“In both Studies 5 and 6, patients treated with ZEPBOUND achieved a greater reduction in systolic blood pressure and high-sensitivity C-reactive protein levels compared to placebo.”

High-sensitivity C-reactive protein, or hsCRP, is a blood marker that rises with inflammation.⁸ The published trial gives the numbers the label leaves out, and both were among the secondary endpoints it was designed to test: hsCRP fell 0.7 mg per liter further than placebo in the first study and 1.0 mg per liter in the second, from starting values around 3 mg per liter. Systolic pressure fell 7.6 and 3.7 mmHg further than placebo at week 48. Weight fell 16.1% and 17.3% further than placebo.

A later analysis of the same trials asked what the hsCRP change followed.⁹ Part of it tracked the improvement in the sleep apnea measures themselves rather than weight alone, as did insulin resistance and triglycerides, while the systolic pressure change tracked weight. The authors conclude that treating both the breathing and the obesity is likely needed to get the full cardiometabolic effect.

What none of that supports is an inflammation claim. A marker moving in a sleep apnea trial is not an approved use, no tirzepatide product is indicated for any inflammatory or autoimmune condition, and neither label mentions arthritis, inflammatory bowel disease or any similar diagnosis.

The heart, and what SURPASS-CVOT showed¹⁰

SURPASS-CVOT has published, and its design decides how to read it. It enrolled 13,299 adults with type 2 diabetes and atherosclerotic cardiovascular disease and randomized them to weekly tirzepatide, up to 15 mg, or to weekly dulaglutide 1.5 mg — another injected GLP-1 drug, itself already approved to cut cardiovascular events. There was no placebo group. The trial was built as a noninferiority trial: it asked whether tirzepatide was not worse than an active drug with a proven benefit, and it could only claim superiority if the upper edge of the confidence interval came in under 1.00.

The primary endpoint was death from cardiovascular causes, heart attack or stroke, whichever came first. In the analysis population of 6,586 and 6,579 participants, as the paper reports it:

“A primary end-point event occurred in 801 patients (12.2%) in the tirzepatide group and 862 (13.1%) in the dulaglutide group (hazard ratio, 0.92; 95.3% confidence interval, 0.83 to 1.01; P = 0.003 for noninferiority; P = 0.09 for superiority).”

So: noninferior to dulaglutide, and superiority was not established. Adverse events were similar between the groups, with more gastrointestinal ones on tirzepatide. The Mounjaro label reports the same hazard ratio from a slightly different analysis set — 12.1% against 13.0% across a median 210 weeks of follow-up, roughly four years — and states plainly that “Superiority to dulaglutide was not established.”¹¹

The label’s table also shows all-cause death at 8.5% against 10.1%, a hazard ratio of 0.84 (95% CI 0.75 to 0.94). Its own footnote says that row was not controlled for family-wise type I error rate, which is the label’s way of saying it is not a finding the approval rests on. The approval rests on the composite above, and the indication it produced is limited to adults with type 2 diabetes at high risk of cardiovascular events.

Two things therefore cannot be said. Not that tirzepatide has been shown to prevent heart attacks compared with no treatment — the comparison was against a drug that already does. And not that this applies to weight loss without diabetes: that trial is SURMOUNT-MMO, about 15,000 adults with obesity and without diabetes, randomized to tirzepatide or placebo, with a five-part endpoint of heart attack, stroke, coronary revascularization, heart failure events or death from any cause.¹² A PubMed search on 11 September 2026 found its design paper and commentary on it, and no results publication. Until that reads out, there is no cardiovascular-outcome figure for tirzepatide in people without diabetes.

What is not an approved use¹

Everything measured above that is not in one of the four indication sentences is off-label if it is the reason for the prescription. That includes lowering blood pressure, improving cholesterol or triglycerides, lowering hsCRP, and preventing cardiovascular events in anyone who does not have type 2 diabetes. Prescribing off-label is legal and routine; describing an off-label effect as an approved benefit is not accurate, so this page does not.

Two absences are worth naming for the same reason. Neither label makes any claim about testosterone or fertility, and neither mentions bone density. Where the label is silent, the honest answer is “not studied”, not an inference from how the drug works.

What none of this covers

This page reports approved uses and measured endpoints, with the comparator attached to each one. It is not advice, and whether any of it applies to you is a question for a prescriber. See the medical disclaimer, and the other side of the ledger in the reported side effects.