A Study of Tirzepatide at Different Injection Sites in Participants With Different Body Sizes: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study of Tirzepatide at Different Injection Sites in Participants With Different Body Sizes (NCT04050670), a randomized phase 1 study. The primary outcome is Pharmacokinetics (PK): Area Under the Concentration Versus Time Curve of Tirzepatide From Time Zero to Infinity (AUC[0-∞]).

2. Clinical Question

3. Trial Design

The registry classifies NCT04050670 as a randomized, open-label crossover Phase 1 study. It enrolled 54 participants and is marked completed. The listed study dates are 2019-09-13 for start and 2020-04-03 for primary completion. The brief summary describes healthy males and females with different body sizes and says that each participant’s study lasted about 20 weeks, including screening.

The intervention was tirzepatide, with 5mg administered by subcutaneous injection at the upper arm, thigh, or abdomen. The crossover classification indicates that the design was intended to compare injection sites within a crossover framework. The registry does not report the treatment sequence, timing between administrations, or other details of the crossover periods. It therefore does not support a more specific description of how each participant moved between sites.

Masking is listed as none. The study purpose included comparing pharmacokinetics and evaluating tolerability and side effects. The primary outcome was AUC[0-∞], assessed using samples collected predose and at the listed post-dose times through 480 hours. The primary analysis population was participants who were randomized, received at least one dose, and had evaluable pharmacokinetic samples.

The record lists 54 participants for the abdomen and upper-arm outcome groups and 53 for the thigh group. It does not explain the difference in these outcome counts. Nor does it report whether the body-size groups were prespecified strata or provide site comparisons within any such groups. The study is marked completed, but a detailed protocol, treatment sequence, or fuller account of conduct is not reported in the registry record.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Before assignment: There was a washout period of at least 35 days between Tirzepatide injections.

Period 1

Period 2

Period 3

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: All randomized participants who received at least one dose of study drug and had evaluable PK samples.

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

Population: All randomized participants.

6. Primary Endpoint

The registry lists the primary outcome as Pharmacokinetics (PK): Area Under the Concentration Versus Time Curve of Tirzepatide From Time Zero to Infinity (AUC[0-∞]), measured over Predose, 8, 12, 24, 36, 48, 72, 96, 120, 144, 168, 336, 480 hours post dose.

The primary endpoint was the area under the concentration-versus-time curve for tirzepatide from time zero to infinity, AUC[0-∞], reported in h*ng/mL. This pharmacokinetic measure represents drug exposure across the measured concentration-time profile, with the endpoint extending from dosing to an estimated infinite time horizon. The registry lists predose sampling and post-dose sampling at 8, 12, 24, 36, 48, 72, 96, 120, 144, 168, 336, and 480 hours.

General explanation: An area-under-the-curve measure combines information about both concentration and time. It is used in pharmacokinetic comparisons because a difference in exposure may not be captured by a single concentration reading. AUC[0-∞] includes the observed profile and an estimate of the remaining area beyond the final measurement; the registry does not describe how that final portion was estimated.

The endpoint is summarized using geometric means, with geometric coefficients of variation as the dispersion measure. The reported site comparisons use ratios of geometric least squares means from a linear mixed effects model. These summaries describe exposure, not clinical benefit or harm. The registry’s primary endpoint does not by itself answer whether injection sites differ in tolerability, side effects, or other outcomes.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

General explanation: The statistical methods help distinguish descriptive summaries from model-based comparisons. A geometric mean summarizes positive, typically right-skewed measurements on a multiplicative scale. It is obtained by working with logarithms and then returning to the original scale. This can be useful for pharmacokinetic exposure, where relative rather than absolute differences are often of interest. The registry reports geometric means for each site and geometric coefficients of variation as measures of dispersion.

What does the crossover classification imply?

General explanation: In a crossover design, participants receive more than one treatment condition over time, and each participant can contribute information under multiple conditions. Comparisons can therefore account for participant-level differences that would otherwise add variability between groups. Period and sequence can also matter: measurements may vary over time, and the order in which conditions are received can affect interpretation. The registry classifies this study as crossover but does not report the sequence, period structure, or details of how these features were handled. No particular implementation should be assumed from the classification alone.

The reported comparison method is a linear mixed effects model. General explanation: A mixed effects model can represent both fixed effects, such as the injection-site contrast of interest, and random effects that account for correlated observations or participant-level variation. This is a natural class of methods for many crossover analyses because observations from the same participant are not independent. The registry names the method but does not provide the model formula, covariance structure, terms included, or diagnostic checks. Those details are not reported in the registry record.

How should the ratios be read?

The model output is described as a ratio of geometric least squares means. Least squares means are model-based estimates, adjusted for the terms included in a model. A ratio expresses the relative scale of two such estimates rather than their arithmetic difference. The posted ratios are 0.953 for the abdomen-thigh comparison and 0.990 for the abdomen-upper-arm comparison. The record does not explicitly state the ratio numerator in the result descriptions. The exact direction should therefore be taken from a fuller model specification, which is not reported in the registry record.

General explanation: A confidence interval communicates statistical uncertainty around an estimate. Its width reflects the precision under the model and assumptions; it is not a guarantee that the true value falls inside the interval. The registry reports 90% confidence intervals for the ratios: 0.935 to 0.970 for the abdomen-thigh comparison, and 0.972 to 1.01 for the abdomen-upper-arm comparison. These intervals should be interpreted on the ratio scale. No p-values are listed for these analyses.

What does the non-inferiority label establish?

The registry labels both reported comparisons as non-inferiority analyses. General explanation: A non-inferiority analysis evaluates whether an estimate is compatible with a prespecified limit on how much worse one condition could be than another. The margin and the direction of the comparison are essential to that judgment. A confidence interval must be assessed against the stated margin according to the specified decision rule. The registry does not report the non-inferiority margin or decision rule, so the label and interval alone do not establish whether a non-inferiority criterion was met.

The endpoint is AUC[0-∞], with a geometric coefficient of variation reported for each site summary. General explanation: A coefficient of variation describes variability relative to the scale of a measurement. Geometric coefficients of variation are suited to multiplicative summaries and help describe dispersion around a geometric mean. They are not confidence intervals for the mean and do not show the range of participant values. The registry does not provide individual-level measurements or raw concentration profiles, so the summaries cannot be used to reconstruct those distributions.

Why can variability affect interpretation?

General explanation: When measurements vary considerably between participants or within participants across periods, estimates are less precise unless the design and analysis effectively account for that variation. In some settings, over-dispersion means that observed variation exceeds what a simpler model assumes. Failing to represent relevant variability can make uncertainty estimates unreliable. The posted result gives geometric coefficients of variation, but it does not report a specific assessment of over-dispersion or model fit. It would be inappropriate to assert that over-dispersion occurred in this trial.

The sample-size assumptions describe a planning illustration based on the observed effect, two-sided significance level 0.05, equal allocation between two arms, and a normal approximation. The effect to detect is described as a stated fraction of the observed effect. General explanation: Such calculations illustrate how a hypothetical future study might be planned under chosen assumptions. They are not the same as the probability that the completed study detected an effect, and they should not be described as observed power. The record does not give a sample-size justification for the original trial, and no planning result should be treated as evidence about the trial’s actual design rationale.

Finally, the registry states that values were missing or not estimable for a two-arm comparison for another endpoint. That note does not change the reported primary AUC comparisons, but it signals that not every listed analysis could be summarized. Details of missingness, estimability, and any sensitivity analyses are not reported in the registry record.

10. Analyses Reported by the Sponsor

The posted injection-site comparisons are reported as linear mixed effects model analyses, with ratios of geometric least squares means and 90% confidence intervals. These are the registry’s model-based results. The record does not include a separate set of unadjusted recalculated comparison estimates from the summary statistics, so no such estimate can be presented alongside the reported model output.

In general, a simple recalculation from group summaries would not necessarily reproduce a model-based crossover analysis. Such a recalculation may not account for within-participant correlation, period effects, or other terms in the fitted model. The registry does not give enough model detail to reproduce the reported analysis independently. The reported ratios are 0.953 for the abdomen-thigh comparison and 0.990 for the abdomen-upper-arm comparison, with 90% confidence intervals of 0.935 to 0.970 and 0.972 to 1.01, respectively. The registry provides no p-values.

No discrepancy between a recalculation and the reported analysis is recorded. The sponsor’s reported analysis is therefore the relevant posted inferential result; the available record does not permit a detailed numerical reconciliation against an unadjusted calculation.

11. Multiplicity, Missing Data and Analysis Populations

The registry reports two primary-endpoint comparisons: abdomen with thigh, and abdomen with upper arm. Both are labelled non-inferiority analyses. No multiplicity adjustment or testing hierarchy is reported in the registry record. Without that information, it is not possible to determine whether the comparisons were treated as a family, whether one was designated as the main contrast, or how the chance of false-positive conclusions across analyses was addressed.

The record does not report a non-inferiority margin or the decision rule for the labelled analyses. Those details are necessary to evaluate the formal non-inferiority claim. The reported 90% confidence intervals describe uncertainty around the ratios, but the intervals cannot be compared with an unreported margin. No p-values are listed.

The analysis population for the primary outcome is described as all randomized participants who received at least one dose of study drug and had evaluable pharmacokinetic samples. The outcome summaries list 54 participants for abdomen, 54 for upper arm, and 53 for thigh. The record does not explain the difference in counts, identify reasons for missing or unevaluable samples, or report how missing data were handled. It also does not describe sensitivity analyses assessing the impact of missingness.

The study description refers to participants with different body sizes, but subgroup analyses by body size are not reported. The record also does not report interim analyses or a detailed statistical analysis plan. These omissions limit assessment of prespecification, multiplicity control, and the robustness of the reported comparisons. They do not invalidate the posted estimates, but they constrain how far conclusions can extend beyond the specific reported model results and analysis population.

12. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 5% reporting threshold.

Most frequent non-serious adverse event terms

The registry lists zero participants with serious adverse events in each injection-site group: abdomen, upper arm, and thigh. The listed groups at risk are 54 for abdomen, 54 for upper arm, and 53 for thigh. These figures describe the serious adverse event counts recorded in the registry for the listed groups; they do not establish that serious events are impossible or that the treatment is free of risk.

The record also lists participants affected by other adverse events: 37 in the abdomen group, 31 in the upper-arm group, and 23 in the thigh group. No serious adverse event terms are listed. The record does not provide event descriptions, severity, timing, relationship to study drug, or participant-level overlap across groups. It is therefore not possible to determine from these counts how many events occurred, whether events were repeated, or whether the same participant contributed to more than one group’s count.

The study was designed to collect tolerability and side-effect information, but these brief counts do not provide a full safety assessment. They should be interpreted as descriptive registry reporting, alongside the study’s limited size and the absence of additional event detail in the record.

13. What the Result Does, and Does Not, Mean

The posted results describe pharmacokinetic exposure after tirzepatide was injected at different sites in a randomized crossover Phase 1 study. The primary endpoint was AUC[0-∞]. The registry reports geometric mean values for abdomen, upper arm, and thigh, and model-based ratios with 90% confidence intervals for the abdomen-thigh and abdomen-upper-arm comparisons. Those estimates provide a quantitative description of the reported comparisons and their statistical precision.

The results do not by themselves establish a clinical benefit, a clinical harm, or a difference in how participants feel or function. AUC is a pharmacokinetic measure, not a direct clinical outcome. The study description also mentions tolerability and side effects, but the posted adverse-event counts do not provide enough detail to characterize the nature or seriousness of non-serious events beyond the recorded counts.

The analyses are labelled non-inferiority, but the registry does not report the margin or decision rule. It is therefore not possible to determine from the posted results whether a prespecified non-inferiority criterion was satisfied. The ratio convention is also not explicitly described in the result text, so the direction of each ratio should not be inferred without the model specification.

The results do not establish that injection-site comparisons are the same across body sizes. While the study description refers to participants with different body sizes, the registry does not report body-size subgroup estimates. Nor do the posted results establish that the same pattern would apply to people outside the described healthy study population or to other dosing circumstances. The analysis population required evaluable pharmacokinetic samples, and reasons for any missing or unevaluable data are not reported.

In short, the registry presents model-based comparisons of tirzepatide exposure by injection site, along with descriptive adverse-event counts. The confidence intervals indicate the precision of the reported ratios under the analysis, while unreported design and analysis details limit more definitive interpretation.

14. Why This Trial Matters Statistically

This study is statistically useful because it focuses on a pharmacokinetic question in a crossover setting: whether measured exposure varies by injection site. In general, crossover designs can reduce the influence of stable differences between participants when each participant contributes observations under multiple conditions. That advantage depends on how the periods, sequences, and within-participant correlation are handled. The registry identifies the design as crossover and reports a mixed effects model, but it does not describe the implementation or model details needed to assess those choices fully.

The outcome is summarized on a geometric scale, and the model results are expressed as ratios. This aligns the comparison with relative differences in exposure rather than raw arithmetic differences. The geometric coefficients of variation also convey information about dispersion. Together, these summaries help readers distinguish the typical exposure estimates from variation around them.

The record illustrates an important interpretive point: a non-inferiority label is not enough to establish a non-inferiority conclusion. The margin, direction, and decision rule are required to interpret the confidence intervals against a prespecified criterion. Their absence from the registry record means the posted estimates can be described, but a formal decision cannot be independently assessed from the record.

Finally, the study description highlights different body sizes, while the posted primary results do not include body-size subgroup comparisons. This makes the scope of the evidence especially important: the overall site comparisons do not show whether effects differ across body sizes. The record’s strengths are its clear endpoint and reported model estimates; its omissions constrain evaluation of subgroup patterns and the inferential framework.

15. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

Several limitations are specific to the information reported for this study. The registry classifies the design as crossover but does not report the treatment sequence, period structure, timing between administrations, or the model terms used to represent the design. These details matter for understanding how the within-participant comparisons were estimated and whether possible period or sequence effects were considered.

The reported analyses are labelled non-inferiority, but the non-inferiority margin and decision rule are not reported. The ratio convention is not explicitly identified in the result text, and no p-values are listed. These omissions limit independent interpretation of direction and formal non-inferiority conclusions. The registry also does not report multiplicity handling for the two posted comparisons.

The outcome-group counts are 54 for abdomen and upper arm and 53 for thigh. The reason for the differing thigh count is not reported. The analysis population requires evaluable pharmacokinetic samples, but the record does not describe missingness, exclusions, or sensitivity analyses. It also does not provide individual-level data or concentration profiles from which the summaries could be examined further.

Although the study description refers to participants with different body sizes, results by body size are not reported. The registry does not provide detailed participant characteristics, so it is not possible to assess how well the enrolled population represents any particular body-size category. Safety reporting is also brief: serious adverse event counts and counts affected by other adverse events are listed, but event details are not included.

Finally, the original sample-size justification and detailed statistical analysis plan are not reported in the registry record. The planning assumptions described for sample-size illustrations are not evidence of the original design rationale or of observed power.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: A Study of Tirzepatide at Different Injection Sites in Participants With Different Body Sizes — NCT04050670.
• Registry record last updated 2023-05-06; status last verified 2023-04.
• Results first posted 2023-03-24.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

19. Record Summary

NCT04050670 is a completed randomized, open-label crossover Phase 1 study of tirzepatide injection sites in healthy participants with different body sizes. The registry reports enrollment of 54 and lists 5mg injections at the abdomen, upper arm, and thigh. The primary endpoint was AUC[0-∞], measured in h*ng/mL.

The posted geometric mean AUC values were 112000.0 for abdomen, 111000.0 for upper arm, and 106000.0 for thigh. Model-based comparisons reported ratios of 0.953 for abdomen-thigh and 0.990 for abdomen-upper arm, each with a 90% confidence interval. The analyses are labelled non-inferiority, but the margin and decision rule are not reported. Serious adverse event counts were zero in each listed group; other adverse event counts are also reported. Missing-data details, model specification, and body-size subgroup results are not reported in the registry record.