Question 1

What did the trials record as results?

Accepted Answer

The fullest single picture comes from a 72-week randomised trial in adults who had both obesity and type 2 diabetes. Baseline mean bodyweight was 100.7 kg, body-mass index 36.1 and HbA1c 8.02%, which is the starting position every percentage below is measured from. Least-squares mean change in bodyweight at week 72 was -12.8% on 10 mg and -14.7% on 15 mg, against -3.2% with placebo. A least-squares mean is an average adjusted for the way participants were distributed between the comparison groups. That is a result in the only form a randomised investigation produces. It is a group average recorded at a scheduled visit, with a placebo comparison beside it and a confidence interval around it. No participant in that trial published an individual figure, and none was asked to.

Question 2

Why do the published results differ from trial to trial?

Accepted Answer

Because the populations differed. At the same assignment of 15 mg once a week, three separate trials produced three different averages, and the distance between them is wider than most summaries acknowledge. In adults who had obesity and no diabetes, the mean percentage change in weight at week 72 was -20.9%. In a 52-week trial in Chinese adults with obesity, the equivalent figure was -17.5%. In adults who also carried type 2 diabetes, it was -14.7% over 72 weeks. Three figures, one assigned amount. What separates them is the population recruited and the duration of the investigation, rather than anything about the injection itself. That is the first reason a percentage encountered online may describe nobody a particular reader resembles.

Question 3

What does a trial average say about one person's result?

Accepted Answer

Less than it appears to, and the published reports are precise about what they are claiming. The 2022 obesity trial gives the mean percentage change in weight at week 72 as -15.0%, with a 95% confidence interval of -15.9 to -14.2 on the smallest assignment. That interval is routinely misread. It describes the precision of the average itself, not the range of outcomes individual participants experienced. The interval narrows as a trial recruits additional participants, rather than as those participants become more similar to one another. So a reader cannot locate themselves inside one of these figures. An average describes what a population did collectively. It does not indicate where within that population an individual would land, and no published analysis covered here attempts a prediction for one person.

Question 4

Did everyone in the trials lose weight?

Accepted Answer

The trials counted that question separately, which is itself the answer to whether an average is enough. A 2023 trial set coprimary endpoints of the percent change in bodyweight from baseline and bodyweight reduction of 5% or higher. A 2025 trial went further and set thresholds at 10%, 15%, 20% and 25%, alongside a change in waist circumference. Those are five separate lines drawn through one population. Thresholds like that exist because the participants inside an average did not all move together. A trial reports how many crossed each line precisely because the mean conceals the distribution, and a headline percentage carries the average while leaving those counts behind.

Question 5

Is there a result at one month?

Accepted Answer

No. No randomised trial covered here reports a weight figure earlier than week 40, and nothing published describes the opening month independently. The 2022 obesity trial ran 72 weeks including a 20-week dose-escalation period, so its participants spent those opening months climbing toward the amount they had been assigned. The earliest measurement of any description comes from routine care rather than from a randomised trial. A 2024 cohort study measured percentage change in weight at 3, 6 and 12 months in matched adults drawn from health records. Read what that figure actually is. It was taken from the records of people who were never randomised, at an interval no trial among these sources was designed to report. It describes whoever was still receiving a prescription on that date, and a one-month number is absent from this literature rather than buried inside it.

Question 6

Who was in these trials at the start?

Accepted Answer

Substantially heavier than most people picturing a before-and-after photograph. In the 2022 obesity trial, at baseline the mean body weight was 104.8 kg, the mean body-mass index was 38.0, and 94.5% of participants had an index of 30 or higher. The maintenance trial recruited a comparable population. Its participants had a mean age of 48 years, 71% were female, and their mean weight was 107.3 kg. Those characteristics are the population the percentages belong to. A percentage is a proportion of a starting weight, so an identical percentage represents a different quantity of kilograms for a different body. Twenty per cent of 105 kg is not equivalent to twenty per cent of 75 kg, and almost every participant in that investigation began above an index of 30 rather than below it.

Question 7

What did the trials supply besides the injection?

Accepted Answer

A programme, and supervision. In the 2024 trial in Chinese adults, participants were randomly assigned to 10 mg or 15 mg tirzepatide or placebo, plus a lifestyle intervention, for 52 weeks. The maintenance trial describes its own object as assessing the effect of tirzepatide, with diet and physical activity, on the maintenance of weight reduction. Attrition belongs in the picture too. In that maintenance trial, 783 participants entered an open-label lead-in period and 670 were randomised at week 36. The 113 who never reached randomisation appear in no percentage reported afterwards. So a trial average describes an injection given inside a programme, to people who kept turning up, with the material supplied and the appointments booked. Those conditions are part of the result rather than a footnote to it.

Question 8

Why do personal accounts and trial results diverge?

Accepted Answer

Because an individual account has no comparison group. The 2022 obesity trial was a phase 3 double-blind, randomized, controlled trial in 2539 adults, which means neither the participants nor the investigators knew who was receiving the injection. Its placebo column is the portion worth reading twice. Participants assigned to placebo recorded a mean weight change of -3.1% at week 72, with a confidence interval of -4.3 to -1.9. Those participants were in the same programme, attending the same appointments and injecting on the same schedule, without the compound rather than with it. That 3.1% is the measured magnitude of everything else. Somebody comparing two photographs cannot separate the injection from the diet, the season, the scales or the expectation of a change, because an individual account carries no placebo comparison to subtract. That is the difference between an account and a measurement.

Question 9

Did the results last?

Accepted Answer

The maintenance trial built that question into its endpoints, and the wording is the informative part. One of its key secondary endpoints was the proportion of participants at week 88 who maintained at least 80% of the weight loss during the lead-in period. A trial that measures how many participants kept four fifths of what they had lost is a trial that did not expect everybody to keep all of it. That expectation was written into the design rather than discovered afterwards. The longest published observation after the injections stop is a 17-week off-treatment period, which followed 176 weeks of treatment in participants who had obesity and prediabetes. What happened to weight when the injection was withdrawn in the maintenance trial is set out on the main tirzepatide page.

Question 10

What do reviews of oral tirzepatide describe?

Accepted Answer

Nothing that appears in this record. Every published trial gave tirzepatide as a weekly injection under the skin. The obesity trial gave once-weekly, subcutaneous tirzepatide, the trial in adults who also had diabetes gave the same, and the heart failure trial gave up to 15 mg subcutaneously once per week. No trial among the research behind this page has given tirzepatide by mouth and measured a result, so a tablet carries no trial figure at all rather than a smaller one. That matters for anyone comparing reviews of one against reviews of another. The percentages circulating under this name were produced by weekly injections, in supervised trials, in participants whose starting weight was recorded. None of them can be lent to a tablet that no trial among the research behind this page has tested.

Question 11

What is on record about results from compounded tirzepatide?

Accepted Answer

No trial among the research behind this page has given compounded tirzepatide to anyone and reported what happened, so there is no outcome to quote rather than a poor one. What exists in its place is an enforcement record. FDA listings include a tirzepatide and cyanocobalamin injectable at 15mg/1mg/mL in a pre-filled syringe, recalled for lack of assurance of sterility. A separate listing covers a tirzepatide and niacinamide solution recalled for lack of processing controls. Cyanocobalamin is vitamin B12 and niacinamide a form of vitamin B3. Those documents describe how a vial was manufactured rather than what it did to anybody. Every percentage on this page came from material supplied for a registered trial, with its identity and its strength established before anyone injected it. What can be verified about a seller is set out on the buying page.

Question 12

What did the trials measure that a photograph cannot?

Accepted Answer

Instruments, for the most part. The first diabetes trial set its primary endpoint as the mean change in glycated haemoglobin from baseline at 40 weeks, which is a laboratory measurement rather than a visible change. Glycated haemoglobin records the average blood sugar of the preceding months. The weight programme went beyond the bathroom scales as well. A substudy of 160 of the 2539 participants underwent dual-energy X-ray absorptiometry, a scan separating fat from lean tissue, at baseline and week 72. A photograph records neither measurement, and the composition of the weight is precisely what a photograph cannot establish. What the trials measured, outcome by outcome, is set out on the benefits page, and what that scan found is on the main tirzepatide page.

Tirzepatide: reported results by outcome