Estimating the Effect of Tirzepatide on Body Weight in Specific Subgroups: A Secondary Analysis Across Trials

Obesity is a widespread and serious health condition, affecting around 650 million adults worldwide. It happens when excess body fat increases the risk of other health problems, such as diabetes (disease that occurs when blood sugar is too high), heart disease, and high blood pressure. One promising treatment for weight loss is tirzepatide, a medication that helps control blood sugar levels and appetite. While studies show that tirzepatide can help people lose weight on average, we still don’t fully understand whether its effects differ between groups—for example, between men and women, or among people of different ages or body sizes. Understanding these differences is important because medications often work differently across populations, and tailoring treatments to specific groups could improve patient outcomes.

However, researchers face challenges when studying these subgroup effects. A single clinical trial may not have enough participants in certain groups to provide clear results. Additionally, researchers often don’t have access to detailed individual data from other trials. Instead, they may only have summary results, which show overall treatment effects without breaking them down by specific subgroups. To address this, our study compares different methods for estimating how treatment effects vary across subgroups by integrating external information in different ways.

To do this, we will use data from two different clinical trials. One trial, called the internal trial, includes detailed information about individual patients in the specific groups, like by age or gender. The other trial, called the external trial, only provides general results about how the treatment worked for all patients combined, without group-specific details. One method will only use the data from the internal trial. Another will rely completely on the external trial, assuming that it is entirely accurate. A third method will take into account any uncertainty in the external trial, recognizing that it may not be perfect. Finally, we will test a method that carefully combines both sets of data, adjusting for any differences between the two.

By comparing these methods, we aim to identify the most reliable approach for estimating how treatment effects differ across patient groups. This research can help improve the way clinical trials integrate external data, leading to more personalized and effective treatments for obesity and other health conditions.