Real-world effectiveness of tirzepatide versus semaglutide for weight loss in overweight or obese patients in an ambulatory care setting

Abstract

Obesity is a significant health burden as it is associated with comorbidities and an increased risk of mortality.¹ Weight loss and chronic weight management can improve comorbidity risk, quality of life and life expectancy,² but patients often fail to achieve weight targets with lifestyle change alone.³ Thus, clinical guidelines recommend adding pharmacotherapy with proven weight loss properties to lifestyle modification for obesity,^3-5 including GLP-1 receptor agonists (GLP-1RA) such as semaglutide,^{6, 7} and the dual GLP-1/GIP receptor agonist agent, tirzepatide.⁸ Randomised controlled trials have shown that semaglutide and tirzepatide have substantial potential for weight reduction, and results favour tirzepatide, but real-world experience may not yield the same results. This study contributes real-world insight into weight outcomes over 6 months of follow-up with tirzepatide versus semaglutide in obese and overweight patients.

This retrospective cohort study, based on deidentified electronic health record (EHR) data, included obese or overweight adults (BMI ≥27 kg/m²) with an incident prescription (index date) for semaglutide or tirzepatide between 13 May 2022 and 1 October 2023. Included patients were treated as outpatients at an academic medical centre. The study excluded pregnant patients, those with a history of bariatric surgery or solid organ transplant, or with chemotherapy for cancer treatment. It also excluded patients missing baseline and/or follow-up weight or if baseline and follow-up weights were <90 days apart. The health system's institutional review board (IRB) reviewed and approved the observational study protocol.

The primary study outcome was mean body weight change and percent weight change from baseline to the last recorded weight during the 6-month follow-up period. Secondary outcomes included the proportion of patients who achieved weight loss of more than 5%, 10% and 15%. Subgroup analyses of percent weight change were conducted by diabetes diagnosis.

Descriptive statistics were used to compare baseline characteristics between treatment groups. Independent t-tests and chi-square tests were used to assess treatment group differences in the change and percentage of weight change, as well as the proportion of patients with weight loss of >5%, >10% and >15% during the 6-month follow-up period. Univariate and bivariate linear regression analyses were used to determine the adjusted association between percent weight change from baseline and treatment, with the bivariate analyses controlling for diabetes status. Multivariate linear regression analyses controlling for other potential confounders were also conducted overall and by diabetes status. Statistical analyses were conducted using Stata v. 17 (StataCorp, LLC. College Station, TX).

A total of 945 adults were included (semaglutide n = 836, tirzepatide n = 109) (Figure S1). Semaglutide patients were older and more likely to have diabetes than those prescribed tirzepatide. Baseline weight was significantly higher in the tirzepatide group than in the semaglutide group (126.5 vs. 115.2 kg, p = 0.001) with BMI (44.1 vs. 40.1 kg/m², p < 0.001). (Baseline characteristics are provided in Supplemental Table S1).

Over the 6-month follow-up period, patients prescribed tirzepatide experienced significantly greater weight loss compared with semaglutide patients (mean loss 6.6 kg vs. 3.1 kg, p < 0.001, mean percent loss 5.3% vs. 2.7%, p < 0.001) (Table 1). In an unadjusted linear regression analysis, tirzepatide was significantly associated with an additional 2.6% (95% CI −4.0%, −1.3%, p < 0.001) weight loss compared with semaglutide. In bivariate analyses, diabetes was a strong modifying factor on percent weight loss. Thus, multivariable linear regression run by diabetes status found that in patients without diabetes that tirzepatide was associated with an additional 3.6% weight loss (95% CI −5.5%, −1.8%, p < 0.001) relative to semaglutide. In patients with diabetes, tirzepatide was not associated with weight loss (0.4% additional weight loss; 95% CI −2.7%, 1.8%, p = 0.70) (Table S2) Overall, significantly greater proportions of patients prescribed tirzepatide lost >5% (48.6% vs. 26.0%), >10% (22.9% vs. 9.3%) and >15% (10.1% vs. 3.1%) of body weight compared with the semaglutide group, respectively (p < 0.001 for all comparisons) (Table 2).

In the subgroup of patients without diabetes (n = 416), patients prescribed tirzepatide experienced a larger loss in body weight compared with those prescribed semaglutide (7.0% vs. 3.4%, p < 0.001). In the population with diabetes (n = 529), the percentage of weight loss was not statistically different between the tirzepatide and semaglutide groups (−3.3% vs. −2.2%, p = 0.30).

In this study of patients who were overweight or obese, tirzepatide was associated with greater weight loss over 6 months of follow-up than semaglutide overall and in patients without diabetes. However, weight loss did not differ between the two agents in patients with diabetes.

The results of this real-world study are generally consistent with clinical trials. This includes SURMOUNT-1, which established the weight loss efficacy of tirzepatide over 72 weeks in obese patients without diabetes,⁹ and preliminary data from SURMOUNT-5, a head-to-head trial comparing tirzepatide to semaglutide over 78 weeks in overweight/obesity patients without diabetes.¹⁰ SURPASS-2, a study comparing tirzepatide to semaglutide in patients with uncontrolled type 2 diabetes, found greater weight loss with tirzepatide over 40 weeks of follow-up than with semaglutide.⁸ However, the current real-world study was of a shorter follow-up time than SURPASS-2, which may not have been sufficient to observe a significant weight loss difference between treatments in the diabetes cohort.

The current study is also consistent with a recent retrospective cohort study by Rodriquez et al., which observed greater weight loss with tirzepatide than with semaglutide in overweight/obese patients.¹¹ The mean percentage change in body weight between tirzepatide and semaglutide in the Rodriquez study was similar to the findings of the current study. The Rodriquez study also suggested that those without diabetes experienced a larger reduction in weight loss than those with diabetes treated with either tirzepatide or semaglutide.

Key limitations of the current study include the risk of unmeasured confounding due to data not captured in EHRs such as lifestyle and medication adherence and patient out-of-pocket costs. Weight values were not consistently captured, leading to attrition, and we did not analyse outcomes by medication dose or titration schedules. Finally, our study was of short duration and fairly limited sample size; thus, it may have been underpowered to detect a true difference in weight loss between semaglutide and tirzepatide at 6 months in patients with diabetes. A future study with longer follow-up and a larger sample size is warranted to compare the weight effects between diabetes and non-diabetes cohorts prescribed semaglutide and tirzepatide, accounting for dosage changes/titrations and medication adherence.

Patients prescribed tirzepatide in a real-world setting achieved more significant weight loss compared with those prescribed semaglutide overall and in patients without diabetes. Future work is warranted to further examine the difference in weight reduction between these agents over a longer follow-up period and by diabetes status.

The authors declare no conflicts of interest.