ABSTRACT
Aims
Tirzepatide is a once‐weekly glucose‐dependent insulinotropic polypeptide and glucagon‐like peptide‐1 receptor agonist (GLP‐1 RA) approved in the US for the treatment of Type 2 diabetes (T2D), weight loss and obstructive sleep apnoea. This study aimed to assess the utilisation and effectiveness of tirzepatide among individuals in a large US commercially insured population.
Materials and Methods
This retrospective, observational study used administrative claims and electronic health records from the Healthcare Integrated Research Database (HIRD). Adults with ≥ 1 prescription claim for tirzepatide (for obesity/overweight, without T2D), between November 2023–June 2024 and continuous medical/pharmacy enrollment for ≥ 12 months pre‐index and ≥ 6 months post‐index were included. Descriptive analyses included baseline demographic and clinical characteristics, treatment patterns and tirzepatide effectiveness (change from baseline in weight, BMI, cardiometabolic risk factors evaluated among persistent individuals) and were stratified by individuals who were GLP‐1 RA naïve and individuals who switched to tirzepatide from a GLP‐1 RA obesity medication.
Results
Of the 22 512 individuals who initiated tirzepatide (mean age: 46 years, 73% female), 55% (n = 12 292) were adherent (PDC ≥ 80%) and 68% (n = 15 208) were persistent during the 6‐month follow‐up period. Of the overall population, 91% (n = 20 554) had ≥ 1 obesity‐related complication at baseline, with dyslipidaemia, hypertension and anxiety being the most frequent complications. Clinically meaningful mean weight reduction (10.5%; n = 808) and numerical reductions in most cardiometabolic risk factors were observed among persistent individuals with available pre‐ and post‐index measurements, including 12.0% body weight reduction in the GLP‐naïve subgroup.
Conclusions
Over half of US adults with obesity or overweight who initiated tirzepatide were adherent (PDC ≥ 80%) and had a favourable persistence profile during a 6‐month follow‐up period. Persistent users achieved clinically meaningful weight reduction and numerical reductions in most cardiometabolic risk factors.
Keywords: obesity, real‐world, tirzepatide, weight reduction
1. Introduction
Obesity is a chronic, relapsing and progressive disease that contributes significantly to morbidity, mortality and healthcare burden [1, 2, 3]. In the United States (US), obesity affects over 40% of adults as of 2023, a figure projected to rise to nearly 50% by 2030 [4, 5, 6]. Obesity is a major risk factor for various complications, including Type 2 diabetes (T2D), cardiovascular and renal diseases, chronic respiratory conditions, mental health disorders such as depression and anxiety, and several types of cancer, all of which contribute to a decline in quality of life [7, 8, 9, 10]. Beyond its direct health consequences, obesity also imposes a substantial economic burden on healthcare systems, with annual medical costs exceeding $170 billion in the US alone [11].
Current treatment guidelines for obesity recommend multicomponent lifestyle interventions as the first‐line approach [12]. When these measures are inadequate for individuals to achieve a clinically meaningful weight reduction (≥ 5%), pharmacotherapy is recommended as an adjunctive treatment, particularly for individuals with a body mass index (BMI) of ≥ 30 or ≥ 27 kg/m2 with at least one obesity‐related complication (ORC) [13, 14, 15, 16]. In recent years, obesity medications (OM) such as glucagon‐like peptide‐1 receptor agonists (GLP‐1 RAs) have been shown to be effective in managing obesity and its complications [17]. More recently, tirzepatide, a glucose‐dependent insulinotropic polypeptide (GIP) and GLP‐1 RA, initially approved in the US in May 2022 for T2D, has entered real‐world clinical use for obesity or overweight following approval in November 2023 based on pivotal findings from the SURMOUNT clinical trial programme [15, 18]. Tirzepatide consistently demonstrated a significant mean reduction in body weight of more than 20% after 72 weeks of therapy in people with obesity or overweight (PwO) [19, 20, 21, 22]. Additionally, tirzepatide treatment has resulted in significant improvements in BMI, waist circumference and cardiometabolic risk factors, such as glycated haemoglobin (HbA1c), blood pressure and lipid profiles [19, 20, 21, 22]. In December 2024, tirzepatide was approved in the US for the treatment of obstructive sleep apnoea (OSA) in adults with obesity and it is currently under investigation for its cardiovascular effects in adults with obesity without T2D [23, 24].
There is now a growing interest in understanding the use of tirzepatide for treating and managing obesity in routine, real‐world clinical settings. Previous claims‐based real‐world analyses conducted in the US have demonstrated moderate to favourable adherence and persistence profiles among individuals initiating tirzepatide for obesity or overweight management [25, 26, 27, 28]. However, these analyses were conducted prior to the approval of tirzepatide for this indication and were therefore limited to off‐label use in this population. This study aimed to evaluate the early utilisation (6‐month adherence, persistence and dosing patterns) and effectiveness of tirzepatide in commercially insured individuals with overweight or obesity without T2D, in a US real‐world setting by analysing a large, retrospective claims database.
2. Methods
2.1. Data Source and Study Design
This retrospective, observational cohort study was conducted using the Healthcare Integrated Research Database (HIRD), Carelon Research's administrative claims database covering over 80 million individuals from a large commercially insured population, across all regions of the United States. The HIRD also includes a fully integrated electronic health record (EHR) and outpatient laboratory data for a subset of the HIRD population. EHR data were obtained from multiple large health systems and clinics, individual provider network EHR systems, and health information exchanges. In US commercial plans with integrated medical and pharmacy benefits, coverage is managed through formulary tiering, prior authorisation, step therapy requirements and cost sharing. The database includes fully adjudicated medical and pharmacy claims, including service/fill dates, diagnosis, procedure and drug codes, quantities dispersed, days' supply, place of service and plan‐ and patient‐paid amounts, along with enrollment data. The claims‐based HIRD population is representative of the US Census population in terms of sex, age and geographic region of residence; representativeness for race/ethnicity is more limited [29]. Additional information about the HIRD, including data contents, quality and provenance, has been previously published [30].
Diagnoses and procedures for both outpatient and inpatient visits were identified by the International Classification of Diseases, Tenth Revision, Clinical Modification (ICD‐10‐CM) codes, Current Procedural Terminology and Healthcare Common Procedure Coding System codes. Outpatient pharmacy claims were captured using National Drug Codes (Table S2). EHR and laboratory data were used to evaluate information on weight‐related and cardiometabolic risk factors. All study data were accessed in compliance with applicable federal and state laws and regulations and were fully compliant with the Health Insurance Portability and Accountability Act (HIPAA) of 1996.
2.2. Time Periods
The study period spanned from November 8, 2022, to December 31, 2024 (Figure S1). Individuals initiating tirzepatide (specifically, Zepbound) were identified between November 8, 2023, and June 30, 2024 (identification period). The date of the first observed tirzepatide claim during the identification period was designated as the index date and the 12‐month pre‐index period was defined as the baseline period (inclusive of the index date). The follow‐up period was defined from the index date to 6 months post‐index. For change from baseline comparisons, pre‐index measures, including weight, BMI and cardiometabolic risk factors, data were collected from 90 days prior to the index date through 3 days after, using the value closest to the index date when multiple values were available. Follow‐up/post‐index measures were collected around the end of the 6‐month period, specifically between index date +150 days and index date +210 days, using the value closest to and prior to index date +180 days when multiple values were available.
2.3. Study Population
The study included adults aged ≥ 18 years with ≥ 1 prescription claim for tirzepatide during the identification period. Individuals were required to have continuous medical and pharmacy enrollment for ≥ 12 months pre‐index and ≥ 6 months post‐index. Individuals were excluded if they had any medical claim with a T2D diagnosis or HbA1c ≥ 6.5% or a pharmacy claim for antidiabetic medication (except metformin) during the 12‐month pre‐index period. Metformin use was permitted as it is commonly prescribed for various conditions, including polycystic ovarian syndrome, prediabetes, type 1 diabetes, gestational diabetes and metabolic dysfunction‐associated steatotic liver disease. Additionally, use of non‐GLP‐1 RA OMs was permitted. Individuals were also excluded if they had any conditions associated with unintentional weight change during the study period (Figure S2).
The study population was further categorised into two nested cohorts: (1) overall cohort: individuals meeting the inclusion and exclusion criteria outlined above and who were OM‐eligible (BMI ≥ 30 or ≥ 27 kg/m2 with ≥ 1 ORC during the 12‐month pre‐index period (Table S2), ascertained using EHR or claims using ICD‐10‐CM diagnosis codes; if multiple BMI values were available, the one closest to the index date was chosen); (2) persistence cohort: a subset of the overall cohort consisting of individuals who were persistent on tirzepatide for ≥ 6 months (no gap between prescription fills of ≥ 45 days). Individuals in both cohorts were stratified using a claims‐based presence/absence definition (any baseline claim) into two subgroups (1) GLP‐1 RA‐naïve (no baseline GLP‐1 RA OM therapy claims; only tirzepatide allowed at index) and (2) GLP‐1 RA‐switchers (≥ 1 baseline GLP‐1 RA OM therapy claim and tirzepatide at index).
2.4. Study Outcomes
Baseline demographics and clinical characteristics were assessed during the 12‐month pre‐index period for both cohorts. Post‐index tirzepatide utilisation (during the 6‐month follow‐up period) was evaluated in the overall cohort and included adherence, persistence/discontinuation and post‐discontinuation treatment. Adherence was assessed using the proportion of days covered (PDC; amount of medication filled in days (i.e., sum of days supplied across all pharmacy claims) divided by the length of the post‐index period) and individuals were classified as adherent if PDC was ≥ 80%. Persistence was defined as continuous tirzepatide use from initiation until end of 6 months follow‐up, allowing a fixed gap of up to 45 days between refills. Stockpiling was permitted in the adherence and persistence analyses (i.e., if an individual had overlapping days of supply of a medication, it was assumed that they used all days of supply of the first prescription fill prior to starting the next). Discontinuation was defined as failure to refill tirzepatide within 45 days after depletion of the previous fill. Discontinuation date was defined as the last day of supply of the most recent fill prior to the 45‐day gap. Switching was defined as at least one prescription fill for a non‐index OM (regardless of dose) during the period from the discontinuation date through the end of the post‐index period, provided there was no claim for tirzepatide in the same time period. Restart was defined as a tirzepatide prescription fill claim following the gap period among individuals who were identified as discontinuers. Effectiveness of tirzepatide was evaluated during the 6‐month follow‐up period among individuals in the persistence cohort who had clinical measurements during the restricted time window in the pre‐index period and 6 months post‐index. Changes from baseline in weight and BMI (absolute and percent) and the proportion of individuals achieving weight reduction targets of ≥ 5%, ≥ 10%, ≥ 15% and ≥ 20% were reported. Additionally, changes (absolute and percent) in cardiometabolic risk factors (HbA1c, systolic and diastolic blood pressure [SBP, DBP], low‐ and high‐density lipoprotein cholesterol [LDL‐C, HDL‐C], total cholesterol and triglycerides) were also reported in this cohort.
2.5. Statistical Analyses
All demographics, baseline clinical characteristics and study outcomes were reported using descriptive statistics. Categorical variables were presented as frequencies and percentages, and continuous variables as means and standard deviations (SDs). Kaplan Meier curves were plotted to examine the time‐to‐discontinuation of tirzepatide. A sensitivity analysis was conducted by extending the permissible refill gap to 60 days, compared with 45 days in the primary analysis, to assess how estimates vary under more lenient versus stricter metrics and to better reflect real‐world refill delays. Additionally, Sankey diagrams were created to illustrate changes in tirzepatide doses with each prescription fill among persistent individuals, enabling visualisation of dose‐escalation patterns, common treatment trajectories and heterogeneity in dosing over time. No hypothesis testing was conducted. All analyses were conducted using a combination of the Instant Health Data platform (Panalgo, Boston, MA), R, Version 4.3.2 (R Foundation for Statistical Computing, Vienna, Austria) and Statistical Analysis Software (SAS) Enterprise Guide, Version 9.3 (SAS Institute Inc., Cary, NC, USA). No imputation of missing values was performed.
3. Results
3.1. Demographics and Clinical Characteristics
A total of 22 512 individuals who filled a prescription for tirzepatide during the baseline period were included in the overall cohort (Figure S2). Mean (SD) age at index was 46.3 (10.5) years; the majority were female (73%) and non‐Hispanic/Latino White (79%). Mean (SD) BMI among individuals who had ≥ 1 valid baseline BMI measurement during the pre‐index period (n = 4839) was 37.6 (6.8) kg/m2. A total of 91% of individuals had at least one ORC and 74% had two or more ORCs. At baseline, initiators had been prescribed a mean (SD) of 9.1 (6.7) unique medications, including all prescriptions regardless of condition or duration of use, and 34% were previously on an OM. Overall, 15 625 (69%) individuals were GLP‐1 RA‐naïve at baseline, while 6887 (31%) had prior GLP‐1 RA exposure and switched to tirzepatide at baseline. Baseline demographic and clinical characteristics of individuals included in the overall cohort, including both subgroups, are presented in Table S1.
A total of 15 208 individuals in the overall cohort (68%) were persistent users of tirzepatide during the 6‐month follow‐up period (persistence cohort, Figure S2). Within this cohort, 10 455 (69%) individuals were GLP‐1 RA‐naïve at baseline and 4753 (31%) had prior GLP‐1 RA exposure and switched to tirzepatide at baseline (Table 1 ). The majority of individuals in the persistence cohort were female (73%) and non‐Hispanic/Latino White (79%), with a mean (SD) age of 46.4 (10.4) years. During the pre‐index period, 21% of individuals (n = 3230) had ≥ 1 valid BMI measurement, with a mean (SD) BMI of 37.7 (6.8) kg/m2. Of these, 91% (n = 2951) were classified as having Class 1, 2 or 3 obesity (Table 1). Overall, the most prevalent ORCs were dyslipidemia (45%), hypertension (43%) and anxiety (32%). At baseline, 91% (n = 13 892) of individuals had at least one ORC and 74% had two or more ORCs. Baseline prescription fills for OMs were reported in 34% (n = 5231) of individuals. Only 4% of individuals had a history of bariatric surgery (Table 1). Baseline demographic and clinical characteristics for GLP‐1 RA‐naïve and GLP‐1 RA‐switchers subgroups are presented in Table 1.
TABLE 1.
Baseline demographics and clinical characteristics in the persistence cohort.
| Variables | Persistence cohort, n = 15 208 | GLP‐1 RA‐naïve, n = 10 455 | GLP‐1 RA‐switchers, a n = 4753 |
|---|---|---|---|
| Demographic characteristics at the index date | |||
| Age at index, years, mean (SD) | 46.4 (10.4) | 46.2 (10.6) | 46.8 (9.9) |
| Age groups, n (%) | |||
| 18–24 years | 420 (2.8) | 319 (3.1) | 101 (2.1) |
| 25–44 years | 5873 (38.6) | 4096 (39.2) | 1777 (37.4) |
| 45–64 years | 8623 (56.7) | 5838 (55.8) | 2785 (58.6) |
| ≥ 65 years | 292 (1.9) | 202 (1.9) | 90 (1.9) |
| Sex, n (%) | |||
| Male | 4082 (26.8) | 3091 (29.6) | 991 (20.8) |
| Female | 11 125 (73.2) | 7363 (70.4) | 3762 (79.2) |
| Race/ethnicity, n (%) | |||
| Non‐Hispanic/Latino White | 12 022 (79.1) | 8146 (77.9) | 3876 (81.5) |
| Non‐Hispanic/Latino Black or African American | 1717 (11.3) | 1200 (11.5) | 517 (10.9) |
| Non‐Hispanic/Latino Asian | 327 (2.2) | 241 (2.3) | 86 (1.8) |
| Non‐Hispanic/Latino Other races b | 89 (0.6) | 69 (0.7) | 19 (0.4) |
| Hispanic/Latino of any race | 1053 (6.9) | 798 (7.6) | 255 (5.4) |
| Geographic region, n (%) | |||
| Northeast | 2034 (13.4) | 1488 (14.2) | 546 (11.5) |
| Midwest | 3395 (22.3) | 2458 (23.5) | 937 (19.7) |
| South | 8085 (53.2) | 5203 (49.8) | 2882 (60.6) |
| West | 1694 (11.1) | 1306 (12.5) | 388 (8.2) |
| Health plan type, n (%) | |||
| Health maintenance organisation | 1672 (11.0) | 1053 (10.1) | 619 (13.0) |
| Preferred provider organisation | 8810 (57.9) | 6029 (57.7) | 2781 (58.5) |
| Consumer‐driven health plan | 4726 (31.1) | 3373 (32.3) | 1353 (28.5) |
| Clinical characteristics during the 12‐month pre‐index period | |||
| Individuals with ≥ 1 valid BMI value during the pre‐index period, n (%) | 3230 (21.2) | 2141 (20.5) | 1089 (22.9) |
| BMI, kg/m2, mean (SD) c | 37.7 (6.8) | 38.0 (6.9) | 37.1 (6.4) |
| BMI categories, kg/m2, n (%) | |||
| Overweight: 25–29.9 kg/m2 | 279 (8.6) | 173 (8.1) | 106 (9.7) |
| Class 1 Obesity: 30–34.9 kg/m2 | 1042 (32.3) | 664 (31.0) | 378 (34.7) |
| Class 2 Obesity: 35–39.9 kg/m2 | 895 (27.7) | 609 (28.4) | 286 (26.3) |
| Class 3 Obesity: ≥ 40 kg/m2 | 1014 (31.4) | 695 (32.5) | 319 (29.3) |
| Individuals with ≥ 1 valid weight value during the pre‐index period, n (%) | 3564 (23.4) | 2337 (22.4) | 1227 (25.8) |
| Weight, kg, mean (SD) c | 107.0 (23.5) | 108.5 (24.1) | 104.3 (21.9) |
| Obesity‐related comorbidities, n (%) | n = 15 208 | n = 10 455 | n = 4753 |
| Dyslipidaemia | 6781 (44.6) | 4612 (44.1) | 2169 (45.6) |
| Hypertension | 6483 (42.6) | 4408 (42.2) | 2075 (43.7) |
| Anxiety | 4924 (32.4) | 3217 (30.8) | 1707 (35.9) |
| Prediabetes | 4641 (30.5) | 3241 (31.0) | 1400 (29.5) |
| Back pain | 4108 (27.0) | 2707 (25.9) | 1401 (29.5) |
| Depression | 3158 (20.8) | 2115 (20.2) | 1043 (21.9) |
| Obstructive sleep apnoea | 3004 (19.8) | 2082 (19.9) | 922 (19.4) |
| Gastroesophageal reflux disease | 2982 (19.6) | 1897 (18.1) | 1085 (22.8) |
| Asthma | 1569 (10.3) | 1068 (10.2) | 501 (10.5) |
| Osteoarthritis‐knee and/or hip | 1319 (8.7) | 901 (8.6) | 418 (8.8) |
| Metabolic dysfunction‐associated steatotic liver disease/Metabolic dysfunction‐associated steatohepatitis | 935 (6.1) | 636 (6.1) | 299 (6.3) |
| Polycystic ovarian syndrome | 680 (4.5) | 446 (4.3) | 234 (4.9) |
| Coronary artery disease | 468 (3.1) | 337 (3.2) | 131 (2.8) |
| Number of obesity‐related comorbidities, n (%) | |||
| 0 | 1316 (8.7) | 955 (9.1) | 361 (7.6) |
| 1 | 2571 (16.9) | 1817 (17.4) | 754 (15.9) |
| 2 | 3133 (20.6) | 2157 (20.6) | 976 (20.5) |
| 3 | 3044 (20.0) | 2123 (20.3) | 921 (19.4) |
| ≥ 4 | 5144 (33.8) | 3403 (32.5) | 1741 (36.6) |
| Obesity medication, n (%) d | 5231 (34.4) | 478 (4.6) | 4753 (100.0) |
| GLP‐1 RA | 4753 (31.3) | 0 (0.0) | 4753 (100.0) |
| Non GLP‐1 RA | 726 (4.8) | 478 (4.6) | 248 (5.2) |
| Anti‐hypertensives, n (%) | 6796 (44.7) | 4586 (43.9) | 2210 (46.5) |
| Anti‐anxiety medications, n (%) | 6792 (44.7) | 4414 (42.2) | 2378 (50.0) |
| Anti‐depressants, n (%) | 6019 (39.6) | 3905 (37.4) | 2114 (44.5) |
| Anti‐hyperlipidemics, n (%) | 3165 (20.8) | 2148 (20.5) | 1017 (21.4) |
| Metformin n (%) | 1122 (7.4) | 776 (7.4) | 346 (7.3) |
| Lifestyle modifications, n (%) | 1098 (7.2) | 715 (6.8) | 383 (8.1) |
| History of bariatric surgery, n (%) | 645 (4.2) | 386 (3.7) | 259 (5.4) |
| Overall medication burden, mean (SD) e | 9.2 (6.6) | 8.3 (6.3) | 11.0 (7.1) |
Abbreviations: BMI, body mass index; GLP‐1 RA, glucagon‐like peptide‐1 receptor agonist; N, total number of individuals; n, number of individuals in the specified category; SD, standard deviation.
GLP‐1 RA‐switchers subgroup includes individuals who switched to tirzepatide from a GLP‐1 RA OM.
Other includes American Indian, Alaskan Native, Native Hawaiian or other Pacific Islander, and other races, including unknown or undisclosed.
BMI and weight values were collected from 90 days prior to the index date through 3 days after, using the value closest to the index date when multiple values were available.
GLP‐1 RAs include semaglutide and liraglutide. Non‐GLP‐1 RAs include phentermine‐topiramate, naltrexone‐bupropion, orlistat, phentermine, benzphetamine hydrochloride, diethylpropion hydrochloride and phendimetrazine tartrate.
Overall medication burden refers to the mean total number of unique drug products/medications (assessed at the 8‐digit GPI level, ‘Drug Base Name’) [31].
3.2. Post‐Index Utilisation of Tirzepatide
Treatment patterns were assessed among individuals in the overall cohort during the 6‐month follow‐up period (Table 2). Overall, 55% (n = 12 292) of individuals were adherent to tirzepatide (PDC ≥ 80%), with a mean (SD) PDC of 0.7 (0.3) and a median (IQR) of 0.9 (0.6–1.0). Additionally, 68% (n = 15 208) remained persistent on tirzepatide for ≥ 6 months using a 45‐day gap. Among those who discontinued tirzepatide (n = 7304), 29% (n = 2087) restarted tirzepatide, 11% (n = 833) switched to a GLP‐1 RA (semaglutide) and fewer than 2% switched to a non GLP‐1 RA OM. The mean (SD) time to discontinuation was 59.6 (32.0) days, and the time between discontinuation and restart of tirzepatide was 71.5 (23.9) days. At the end of the follow‐up period, 81% (n = 18 128) of individuals remained on an incretin‐based OM, either through continued persistence on tirzepatide or, after discontinuation, restarting tirzepatide or switching to a GLP‐1 RA OM (semaglutide). Using a 60‐day gap, persistence increased to 73% (n = 16 334). Detailed treatment outcomes following discontinuation for this sensitivity analysis are presented in Table 2. Kaplan Meier curves for time‐to‐discontinuation (45‐day gap) are presented in Figure S3.
TABLE 2.
Tirzepatide utilisation during the 6‐month post‐index period in the overall cohort.
| Description | Overall cohort, N = 22 512 | GLP‐1 RA‐naïve, n = 15 625 | GLP‐1 RA‐switchers, a n = 6887 |
|---|---|---|---|
| Adherence, PDC ≥ 80%, n (%) | 12 292 (54.6) | 8546 (54.7) | 3746 (54.4) |
| Adherence, PDC | |||
| Mean (SD) | 0.73 (0.28) | 0.73 (0.29) | 0.74 (0.26) |
| Median (IQR) | 0.85 (0.56–0.97) | 0.85 (0.52–0.97) | 0.84 (0.61–0.96) |
| 45‐day gap | |||
| Persistence b , n (%) | 15 208 (67.6) | 10 455 (66.9) | 4753 (69.0) |
| Discontinuation, n (%) | 7304 (32.4) | 5170 (33.1) | 2134 (31.0) |
| Days to discontinuation, mean (SD) | 59.6 (32.0) | 58.6 (31.9) | 62.0 (32.1) |
| Post‐discontinuation treatment c | |||
| Restarted tirzepatide d , n (%) | 2087 (28.6) | 1349 (26.1) | 738 (34.6) |
| Days to restart, mean (SD) | 71.5 (23.9) | 71.2 (23.4) | 72.2 (24.9) |
| Switched to a non‐index GLP‐1 RA or non‐GLP‐1 RA OM e , n (%) | 999 (13.7) | 568 (11.0) | 431 (20.2) |
| Days to non‐index GLP‐1 RA or non‐GLP‐1 RA OM, mean (SD) | 36.3 (32.3) | 36.1 (33.6) | 36.5 (30.4) |
| Switched to a non‐index GLP‐1 RA f , n (%) | 833 (11.4) | 439 (8.5) | 394 (18.5) |
| Days to non‐index GLP‐1 RA, mean (SD) | 34.2 (30.4) | 32.1 (30.0) | 36.5 (30.7) |
| Persistent b or restarted tirzepatide, n (%) | 17 295 (76.8) | 11 804 (75.5) | 5491 (79.7) |
| Persistent b or restarted tirzepatide or switched to a GLP‐1 RA OM, n (%) | 18 128 (80.5) | 12 243 (78.4) | 5885 (85.5) |
| 60‐day gap | |||
| Persistence b , n (%) | 16 334 (72.6) | 11 192 (71.6) | 5142 (74.7) |
| Discontinuation, n (%) | 6178 (27.4) | 4433 (28.4) | 1745 (25.3) |
| Days to discontinuation, mean (SD) | 57.2 (30.0) | 56.2 (29.9) | 59.6 (30.2) |
| Post‐discontinuation treatment c | |||
| Restarted tirzepatide d , n (%) | 1142 (18.5) | 744 (16.8) | 398 (22.8) |
| Days to restart, mean (SD) | 86.6 (23.0) | 85.7 (22.5) | 88.3 (23.8) |
| Switched to a non‐index GLP‐1 RA or non‐GLP‐1 RA OM e , n (%) | 997 (16.1) | 553 (12.5) | 444 (25.4) |
| Days to non‐index GLP‐1 RA or non‐GLP‐1 RA OM, mean (SD) | 36.8 (32.5) | 36.6 (33.9) | 37.2 (30.8) |
| Switched to a non‐index GLP‐1 RA | |||
| Liraglutide, n (%) | 27 (0.4) | < 5 | > 20 (> 1.1) |
| Semaglutide, n (%) | 817 (13.2) | 426 (9.6) | 391 (22.4) |
| Days to non‐index GLP‐1 RA, mean (SD) | 34.7 (30.7) | 32.3 (30.3) | 37.2 (31.1) |
| Persistent b or restarted tirzepatide, n (%) | 17 476 (77.6) | 11 936 (76.4) | 5540 (80.4) |
| Persistent b or restarted tirzepatide or switched to an OM GLP‐1 RA, n (%) | 18 315 (81.4) | 12 364 (79.1) | 5951 (86.4) |
Abbreviations: GLP‐1 RA, Glucagon‐Like Peptide‐1 Receptor Agonist; HbA1c, glycated haemoglobin; OM, obesity medication; PDC, proportion of days covered; T2D, Type 2 diabetes.
GLP‐1 RA‐switchers subgroup includes individuals who switched to tirzepatide from a GLP‐1 RA OM.
Persistence is defined as no gap in medication > 45 or 60 days from the time the days' supply runs out.
ops‐discontinuation treatment percentages were calculated among discontinuers.
Proportion of patients who have a tirzepatide (Zepbound) claim after the gap period after being identified as discontinuers.
GLP‐1 RAs include semaglutide and liraglutide. Non‐GLP‐1 RAs include phentermine‐topiramate, naltrexone‐bupropion, orlistat, phentermine, benzphetamine hydrochloride, diethylpropion hydrochloride and phendimetrazine tartrate.
When using a 45‐day gap analysis, all individuals who switched to a GLP‐1 RA switched to semaglutide.
Dose escalation patterns among persistent individuals were also assessed. A total of 95% (n = 14 491) of individuals had a dose escalation, with a mean (SD) time between all escalations of 44.8 (16.8) days. Among these, 23% (n = 3286) of individuals had one escalation, 29% (n = 4247) had two escalations and 48% (n = 6958) had three escalations. Similar trends were observed among both GLP‐1 RA‐naïve individuals and GLP‐1‐switchers. Most individuals (83%) were initiated on a 2.5 mg dose of tirzepatide. This was consistent in the GLP‐1 RA‐naïve subgroup (91%), but lower among individuals who switched to tirzepatide from a GLP‐1 RA OM (67%). At the sixth prescription refill, most individuals (75%) were receiving tirzepatide doses of ≤ 10 mg, including 2.5 mg (4%), 5 mg (24%), 7.5 mg (23%) and 10 mg (24%). Corresponding proportions were 81% among GLP‐1 RA‐naïve individuals and 63% among individuals who switched to tirzepatide from a GLP‐1 RA OM. Sankey plots displaying changes in tirzepatide doses with each prescription fill for all three groups are presented in Figure 1.
FIGURE 1.
Sankey plots of tirzepatide dose escalation among persistent users of tirzepatide. (a) Persistence cohort (n = 15 208). In this figure, each column represents a fill, starting at left with the index fill, and ending at right with the 6th fill. Each column shows the dose distribution for that fill, and the transitions from the prior and into the following fill. For example, at their second fill, 4498 tirzepatide users (i.e., 29.6% of the 15 207 users who had a second fill) had a dose of 2.5 mg, while 8699 users (57.2%) had a dose of 5 mg. Almost all of the 5 mg users escalated from 2.5 mg at their first (index fill), while the small remainder continued from 5 mg at their first fill. (b) GLP‐1 RA‐naïve subgroup (n = 10 455). (c) GLP‐1 RA‐switchers subgroup (n = 4753). GLP‐1 RA switchers subgroup includes individuals who switched to tirzepatide from a GLP‐1 RA OM. GLP‐1 RA, glucagon‐like peptide‐1 receptor agonist.
3.3. Post‐Index Effectiveness of Tirzepatide
The effectiveness of tirzepatide was evaluated based on changes in weight‐related and cardiometabolic risk factors among individuals in the persistence cohort who had measurements available at both baseline and 6 months post‐index; these analyses were therefore limited to a small proportion (< 5%) of the persistence cohort. Using a restricted time window around index to collect data, mean (SD) baseline BMI (n = 797) was 38.5 (6.9) kg/m2. At 6 months post‐index, a mean (SD) reduction of 4.0 (2.9) kg/m2 was observed, corresponding to a 10.4% decrease in BMI (Figure 2a). Mean (SD) weight at baseline (n = 808) was 110.3 (25.1) kg (243.2 [55.3] lbs). At 6 months, mean (SD) reduction in weight was 11.6 (8.3) kg (25.6 [18.3] lbs), reflecting a 10.5% decrease (Figure 2a). Between baseline and 6 months post‐index, 79%, 52%, 24% and 8% of individuals achieved weight reductions of ≥ 5%, ≥ 10%, ≥ 15% and ≥ 20%, respectively (Figure 2b). Among individuals in whom effectiveness was evaluated (n = 808), 82% (n = 660) had a sixth prescription refill. Of these, 80% (n = 531) were receiving tirzepatide doses of ≤ 10 mg, distributed as follows: 2.5 mg (5%, n = 31), 5 mg (24%, n = 158), 7.5 mg (29%, n = 189) and 10 mg (23%, n = 153). The remaining individuals received doses of either 12.5 mg (11%, n = 70) or 15 mg (9%, n = 59). Individuals in the GLP‐1 RA‐naïve subgroup (baseline BMI: 38.9 (7.2) kg/m2; baseline weight: 111.5 (26.0) kg (245.8 [57.4] lbs)) experienced an 11.9% reduction in BMI and 12.0% reduction in weight (Figure 2a). Individuals who switched to tirzepatide from a GLP‐1 RA OM (baseline BMI: 37.8 (6.2) kg/m2; baseline weight: 107.7 (22.8) kg (237.5 [50.3] lbs)), experienced a 7.1% reduction in BMI and 7.3% reduction in weight since initiation of tirzepatide (Figure 2a). GLP‐1 RA‐naïve individuals achieved weight reduction thresholds during the 6‐month follow‐up period: ≥ 5% (88%), ≥ 10% (62%), ≥ 15% (29%) and ≥ 20% (11%) (Figure 2b). Individuals who switched to tirzepatide from a GLP‐1 RA OM, achieved weight reduction thresholds of: ≥ 5% (60%), ≥ 10% (31%), ≥ 15% (11%) and ≥ 20% (4%) at 6‐months (Figure 2b).
FIGURE 2.
Weight‐related outcomes at 6 months among persistent users of tirzepatide (a) Changes in BMI and weight (b) Categorical change in weight. Pre‐index measures were collected from 90 days prior to the index date through 3 days after, using the value closest to the index date when multiple values were available. Post‐index measures were collected around the end of the 6‐month period, specifically between index date +150 days and index date +210 days, using the value closest to and prior to index date +180 days when multiple values were available. BMI and weight changes were evaluated among individuals in the persistence cohort who had measurements available at both baseline and 6 months post‐index. BMI, body mass index; GLP‐1 RA, glucagon‐like peptide‐1 receptor agonist; SD, standard deviation.
Among individuals with pre‐ and post‐index measurements, baseline cardiometabolic risk factor values were generally within the normal to mildly elevated range. Mean (SD) HbA1c was 5.7% (0.4). Blood pressure readings showed a mean (SD) SBP value of 128.3 (15.4) mmHg and a DBP of 80.5 (9.1) mmHg. Lipid profiles showed a mean (SD) total cholesterol of 193.5 (40.4) mg/dL, LDL‐C of 116.5 (33.3) mg/dL, HDL‐C of 52.1 (15.2) mg/dL and triglycerides of 142.1 (85.0) mg/dL. At 6 months post‐index, numerical reductions were observed across most outcomes. HbA1c decreased by 0.3% (0.3), while SBP and DBP dropped by 6.5 (15.6) mmHg and 2.4 (9.7) mmHg, respectively. Total cholesterol levels reduced by 12.9 (31.1) mg/dL, LDL‐C by 6.1 (27.1) mg/dL and triglycerides by 26.7 (67.7) mg/dL. However, HDL‐C declined by 2.3 (10.0) mg/dL (an undesirable change) (Table 3). Absolute and percent reductions in cardiometabolic risk factors for individuals in the GLP‐1 RA‐naïve and GLP‐1 RA switchers subgroups are provided in Table 3.
TABLE 3.
Cardiometabolic risk factors among individuals with pre‐index and post‐index data.
| Outcome | Persistence cohort, n = 15 208 | GLP‐1 RA‐naïve, n = 10 455 | GLP‐1 RA‐switchers, a n = 4753 | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Pre‐index | Post‐index | Absolute Change | % Change | Pre‐index | Post‐index | Absolute change | % Change | Pre‐index | Post‐index | Absolute change | % Change | |
| HbA1c, % | n = 384 | n = 286 | n = 98 | |||||||||
| 5.7 (0.4) | 5.4 (0.3) | −0.3 (0.3) | −5.1 (4.6) | 5.7 (0.4) | 5.4 (0.3) | −0.3 (0.3) | −5.8 (4.5) | 5.5 (0.4) | 5.4 (0.3) | −0.2 (0.3) | −3.1 (4.6) | |
| n = 717 | n = 483 | n = 234 | ||||||||||
| Systolic BP, mmHg | 128.3 (15.4) | 121.8 (13.5) | −6.5 (15.6) | −4.3 (11.5) | 128.9 (14.6) | 121.1 (13.1) | −7.8 (14.6) | −5.4 (10.9) | 127.0 (17.0) | 123.2 (14.0) | −3.8 (17.3) | −2.1 (12.2) |
| Diastolic BP, mmHg | 80.5 (9.1) | 78.1 (8.8) | −2.4 (9.7) | −2.3 (12.0) | 80.9 (9.4) | 77.8 (8.8) | −3.1 (9.6) | −3.1 (11.7) | 79.9 (8.4) | 78.9 (8.9) | −1.0 (9.7) | −0.6 (12.5) |
| Total cholesterol, mg/dL | n = 479 | n = 351 | n = 128 | |||||||||
| 193.5 (40.4) | 180.6 (37.9) | −12.9 (31.1) | −5.3 (16.1) | 196.6 (39.7) | 181.2 (38.2) | −15.3 (32.0) | −6.6 (16.4) | 185.2 (41.1) | 178.9 (37.1) |
−6.3 (27.8) |
−2.0 (14.9) | |
| LDL‐C, mg/dL | n = 488 | n = 356 | n = 132 | |||||||||
| 116.5 (33.3) | 110.3 (32.7) | −6.1 (27.1) | −2.4 (26.1) | 118.7 (34.2) | 110.6 (33.8) | −8.2 (28.1) | −4.1 (27.0) | 110.3 (30.2) | 109.7 (29.7) | −0.6 (23.4) | −2.1 (22.9) | |
| n = 477 | n = 349 | n = 128 | ||||||||||
| HDL‐C, mg/dL | 52.1 (15.2) | 49.8 (13.0) | −2.3 (10.0) | −2.4 (16.2) | 52.4 (15.3) | 49.8 (13.0) | −2.5 (10.2) | −2.8 (16.2) | 51.4 (14.8) | 49.8 (12.9) | −1.6 (9.2) | −1.2 (16.3) |
| Triglycerides, mg/dL | 142.1 (85.0) | 115.4 (60.6) | −26.7 (67.7) | −10.9 (33.2) | 144.7 (87.8) | 115.0 (64.5) | −29.6 (72.2) | −13.1 (33.2) | 135.0 (76.8) | 116.3 (49.0) | −18.7 (52.9) | −4.7 (32.6) |
Note: All values presented as mean (SD). Pre‐index measures were collected from 90 days prior to the index date through 3 days after, using the value closest to the index date when multiple values were available. Post‐index measures were collected around the end of the 6‐month period, specifically between index date +150 days and index date +210 days, using the value closest to and prior to index date +180 days when multiple values were available.
Abbreviations: BMI, Body mass index; BP, blood pressure; dL, deciliter; GLP‐1 RA, glucagon‐like peptide 1 receptor agonist; HbA1c, Haemoglobin A1c; HDL, high density lipoprotein cholesterol; LDL‐C, low‐density lipoprotein cholesterol; mg, milligramme; mm Hg, millimetre of mercury; N, total number of individuals; n, number of individuals in the specified category; SD, standard deviation.
GLP‐1 RA‐switchers subgroup includes individuals who switched to tirzepatide from a GLP‐1 RA OM.
4. Discussion
This real‐world, observational study utilised data from the HIRD claims and EHR database to examine the 6‐month utilisation and effectiveness of tirzepatide in individuals with overweight or obesity without T2D. To our knowledge, this is among the first studies to examine real‐world treatment patterns following the FDA approval of tirzepatide for obesity or overweight management in November 2023. As such, our findings reflect prescribing behaviour under approved labelling, evolving guideline support and expanded insurance coverage. At baseline, the mean BMI of individuals with continuous health plan enrollment initiating tirzepatide for obesity or overweight was 37.6 kg/m2, consistent with class 2 obesity. Notably, nearly three‐fourths of these individuals had two or more ORCs, highlighting the presence of multimorbidity. Additionally, polypharmacy was observed, and over one‐third of individuals had prior exposure to OMs, most commonly GLP‐1 RAs. Collectively, these findings are suggestive of a population with advanced obesity and considerable treatment experience, either with OMs or therapies for obesity‐related comorbidities, prior to initiating tirzepatide. Six months post‐initiation, more than two‐thirds of individuals remained persistent on tirzepatide. Clinically meaningful weight reduction and numerical reductions in most cardiometabolic risk factors were also observed among persistent individuals, with reductions observed among GLP‐1 RA‐naïve individuals and individuals who switched to tirzepatide from a GLP‐1 RA OM.
In this study, tirzepatide persistence at 6 months when stockpiling was permitted was 68% using a 45‐day gap and 73% using a 60‐day gap, which is higher than reported persistence rates of other GLP‐1 RAs. Previously reported real‐world GLP‐1 RA persistence rates among PwO without T2D were 46% (60‐day gap) at 6 months (January–December 2021) and 35% (60‐day gap) at 12 months (January 2018–December 2023) [32, 33]. However, these findings reflect different study periods, follow‐up durations and were potentially influenced by drug shortages, limiting their comparability to the current analysis. Database studies using similar eligibility criteria conducted prior to tirzepatide's approval for weight loss reported 6‐month post‐index persistence rates ranging between 53% and 74% when stockpiling was permitted and between 61% and 72% when not permitted (45‐day gap) [25, 26, 27, 28]. Of those who discontinued tirzepatide in the current study, 29% re‐initiated tirzepatide post‐discontinuation, resulting in 77% of individuals either remaining persistent on tirzepatide or reinitiating it post‐discontinuation. Reasons for discontinuation are not available in the database we used; however, potential drivers of discontinuation or switching may include supply shortages, adverse effects and cost or insurance‐related barriers [34, 35]. Recent studies have identified high costs and insurance‐related issues as the most common reasons for treatment discontinuation with tirzepatide among individuals with obesity [33, 36].
Individuals who were persistent on tirzepatide, with pre‐ and post‐index clinical data available, experienced a mean weight reduction of 10.5% at 6 months post‐index, despite 80.5% of individuals receiving tirzepatide doses of ≤ 10 mg at the sixth refill. While the magnitude of weight reduction was lower than that reported in the SURMOUNT‐1 trial following 24 weeks of tirzepatide treatment (dose‐dependent; approximately 10% for tirzepatide 5 mg, 14% for tirzepatide 10 mg and 15% for tirzepatide 15 mg), this is noteworthy as individuals in this study could have had gaps in treatment of up to 45 days and were predominantly on lower doses, which may have impacted real‐world treatment effectiveness [19]. Moreover, earlier real‐world studies assessing tirzepatide's effectiveness among PwO without T2D prior to its approval for obesity treatment have demonstrated comparable results [26, 27]. Overall, 79% of persistent individuals, 88% of GLP‐1 RA‐naïve individuals and 60% of GLP‐1 RA switchers achieved a weight reduction of ≥ 5%, a threshold associated with meaningful improvements in obesity‐related health outcomes [13]. Most cardiometabolic risk factors, including blood pressure and lipid profiles, also showed favourable trends, both among GLP‐1 RA‐naïve individuals and among those who switched to tirzepatide from a GLP‐1 RA OM. Sustained reductions in SBP, DBP and lipid levels that were maintained through 72 weeks of tirzepatide treatment were observed in the SURMOUNT‐1 trial as well. Moreover, in a SURMOUNT‐1 post hoc analysis, improvements in cardiometabolic risk factors were positively associated with greater weight reduction [37]. Although HDL‐C levels decreased after 6 months of tirzepatide treatment in this study, this pattern is documented in literature as a common early response to weight loss, especially among women, with levels often increasing after the initial phase [38, 39]. Extended follow‐up is required to fully characterise the long‐term impact of tirzepatide on cardiometabolic risk factors.
As per dosing recommendations, tirzepatide was initiated at 2.5 mg in over 90% of individuals in the GLP‐1 RA‐naïve subgroup [15]. Although dose escalation to 15 mg is possible over a 6‐month period, by the sixth prescription fill, 76% of individuals in this subgroup were prescribed intermediate doses (5, 7.5 or 10 mg), and only 19% reached higher doses (12.5 or 15 mg). Among individuals who switched to tirzepatide from a GLP‐1 RA OM, only 67% initiated treatment at a dose of 2.5 mg, and by the sixth fill, 37% had escalated to 12.5 or 15 mg. These differences may reflect greater confidence among prescribers in escalating doses for GLP‐1 RA‐exposed individuals. Conversely, slower escalation in naïve individuals may be aligned with label recommendations regarding maintenance dosing and management of side effects, which can prompt delayed or paused up‐titration, while some patients may remain at intermediate doses due to adequate clinical response. However, the specific reasons for dose adjustments cannot be directly observed in claims data.
This study has a few limitations. First, since it relies exclusively on data from commercially insured individuals, the findings are representative of this population segment and may not be representative of the broader US population, particularly those who are uninsured or covered by other insurance types. Individuals were also required to have continuous insurance enrollment, which inherently excludes individuals who discontinue coverage. Moreover, the limited sample size, especially the individuals who had available pre‐ and post‐index data, raises uncertainty about the generalisability of the effectiveness findings to both the broader population and to the individuals who were persistent. It is also possible that persistent users were more likely to continue therapy due to perceived or experienced benefits, which could have introduced selection bias. Second, the results are influenced by the specific enrollment periods and the availability of weight and BMI data. As with any analysis based on administrative claims databases, interpretation should account for common issues such as coding errors and incomplete data. Additionally, supply shortages of tirzepatide during the study period may have impacted treatment persistence and dose escalation. Lastly, tirzepatide samples (2.5 mg doses) provided by healthcare providers to their patients are not captured in claims databases, which may have led to underestimation of the number of individuals who initiated treatment at this dose. Dose‐dependent weight reduction was also not assessed in this study.
Despite these limitations, our findings demonstrate that tirzepatide is being used in the real‐world management of obesity, yielding clinically meaningful weight reduction outcomes. These findings offer valuable insights for healthcare providers, payers and policymakers as the adoption of tirzepatide for weight management continues to evolve in routine clinical practice. The treatment patterns observed in this study can help inform patient counselling by setting expectations for treatment duration, identifying when discontinuation is most likely, and guiding the timing of clinical support. Moreover, the study findings could inform payer policies, including prior authorisation durations, renewal timing and refill grace periods that better align with real‐world use, thereby minimising avoidable interruptions in therapy. Further research is warranted to evaluate long‐term outcomes and to examine treatment patterns across diverse subpopulations.
5. Conclusion
Results from this real‐world study of commercially insured US adults with obesity or overweight indicate that over half of individuals initiating tirzepatide had severe obesity accompanied by two or more obesity‐related complications. During the 6‐month follow‐up period, findings indicated moderate levels of patient adherence and a favourable persistence rate of nearly 70% among tirzepatide initiators. Among those who remained persistent for ≥ 6 months, clinically meaningful improvements in weight and BMI and numerical reductions in HbA1c and other cardiometabolic risk factors were observed. These results support the effectiveness of tirzepatide for the treatment of obesity in real‐world settings, demonstrating benefits in both GLP‐1 RA‐naïve individuals and individuals who switched to tirzepatide from a GLP‐1 RA OM.
Author Contributions
Design: Theresa Hunter Gibble, Malvika Venkataraman, Grace Stockbower, Michael Grabner, Karishma Desai; Conduct/data collection: Michael Grabner, Karishma Desai; Analysis/interpretation: Theresa Hunter Gibble, Malvika Venkataraman, Donna Mojdami, Carlos Vallarino, Michael Grabner, Mythili Ananth, Karishma Desai; Writing/critical revision of manuscript: Theresa Hunter Gibble, Malvika Venkataraman, Donna Mojdami, Carlos Vallarino, Grace Stockbower, Michael Grabner, Mythili Ananth, Karishma Desai. All named authors meet the International Committee of Medical Journal Editors (ICMJE) criteria for authorship for this article, take responsibility for the integrity of the work as a whole, and have given their approval for this version to be published.
Funding
The study and all support for the manuscript were funded by Eli Lilly and Company, Indianapolis, United States.
Ethics Statement
As this observational study used de‐identified data, a formal Consent to Release Information form and Ethical Review Board approval or waiver were not required. This study was conducted in accordance with the ethical principles that have their origin in the Declaration of Helsinki and that are consistent with Good Pharmacoepidemiology Practices and applicable laws and regulations of the country or countries where the study was conducted, as appropriate. Data accessed were compliant with applicable federal and state laws and regulations, including the Health Insurance Portability and Accountability Act of 1996.
Conflicts of Interest
Theresa Hunter Gibble, Donna Mojdami, Carlos Vallarino and Mythili Ananth are employees and stockholders of Eli Lilly and Company. Malvika Venkataraman, Grace Stockbower, Michael Grabner and Karishma Desai are employees of Carelon Research. Michael Grabner and Karishma Desai are stockholders of Elevance Health.
Supporting information
Figure S1: Study design.
Figure S2: Patient attrition figure.
Figure S3: Kaplan Meier curves for time‐to‐discontinuation among tirzepatide initiators using a 45‐day allowed gap.
Table S1: Baseline demographics and clinical characteristics in the overall cohort.
Table S2: List of diagnosis codes and national drug codes.
Acknowledgements
The authors thank Michael Kelleman for providing statistical peer review support.
Gibble T. H., Venkataraman M., Mojdami D., et al., “Tirzepatide Persistence and Associated Changes in Clinical Outcomes in US Adults With Obesity or Overweight: A 6‐Month Claims Database Analysis,” Diabetes, Obesity and Metabolism 28, no. 11 (2026): 9966–9979, 10.1111/dom.71064.
Handling Editor: Richard Donnelly
Data Availability Statement
The data that support the findings of this study are available from Carelon Research, but restrictions apply to the availability of these data to external sources, and therefore they are not publicly available. Data may be made available from the corresponding author upon reasonable request and with permission from Carelon Research.
References
- 1. Bray G. A., Kim K. K., and JPH W., “Obesity: A Chronic Relapsing Progressive Disease Process. A Position Statement of the World Obesity Federation,” Obesity Reviews 18, no. 7 (2017): 715–723, 10.1111/obr.12551. [DOI] [PubMed] [Google Scholar]
- 2. Ford J. H., Lage M. J., Boye K. S., Bae J. P., Terrell K. A., and Bunck M. C., “Five‐Year Morbidity and Mortality Rates in a U.S. Population With Obesity With and Without Prediabetes,” Journal of Diabetes and Its Complications 37, no. 5 (2023): 108454, 10.1016/j.jdiacomp.2023.108454. [DOI] [PubMed] [Google Scholar]
- 3. GBD 2021 Adult BMI Collaborators , “Global, Regional, and National Prevalence of Adult Overweight and Obesity, 1990–2021, With Forecasts to 2050: A Forecasting Study for the Global Burden of Disease Study 2021,” Lancet 405, no. 10481 (2025): 813–838, 10.1016/s0140-6736(25)00355-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Emmerich S. D., Fryar C. D., Stierman B., and Ogden C. L., “Obesity and Severe Obesity Prevalence in Adults: United States, August 2021–August 2023,” NCHS Data Brief 508 (National Center for Health Statistics, 2024), 10.15620/cdc/159281. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Lingvay I., Cohen R. V., CWL R., and Sumithran P., “Obesity in Adults,” Lancet 404, no. 10456 (2024): 972–987, 10.1016/s0140-6736(24)01210-8. [DOI] [PubMed] [Google Scholar]
- 6. Ward Z. J., Bleich S. N., Cradock A. L., et al., “Projected U.S. State‐Level Prevalence of Adult Obesity and Severe Obesity,” New England Journal of Medicine 381, no. 25 (2019): 2440–2450, 10.1056/NEJMsa1909301. [DOI] [PubMed] [Google Scholar]
- 7. Mandviwala T., Khalid U., and Deswal A., “Obesity and Cardiovascular Disease: A Risk Factor or a Risk Marker?,” Current Atherosclerosis Reports 18, no. 5 (2016): 21, 10.1007/s11883-016-0575-4. [DOI] [PubMed] [Google Scholar]
- 8. Ansari S., Haboubi H., and Haboubi N., “Adult Obesity Complications: Challenges and Clinical Impact,” Therapeutic Advances in Endocrinology and Metabolism 11 (2020): 2042018820934955, 10.1177/2042018820934955. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Campbell L. A., Kombathula R., and Jackson C. D., “Obesity in Adults,” JAMA 332, no. 7 (2024): 600, 10.1001/jama.2024.5126. [DOI] [PubMed] [Google Scholar]
- 10. Stephenson J., Smith C. M., Kearns B., Haywood A., and Bissell P., “The Association Between Obesity and Quality of Life: A Retrospective Analysis of a Large‐Scale Population‐Based Cohort Study,” BMC Public Health 21, no. 1 (2021): 1990, 10.1186/s12889-021-12009-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Ward Z. J., Bleich S. N., Long M. W., and Gortmaker S. L., “Association of Body Mass Index With Health Care Expenditures in the United States by Age and Sex,” PLoS One 16, no. 3 (2021): e0247307, 10.1371/journal.pone.0247307. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Gaskin C. J., Cooper K., Stephens L. D., Peeters A., Salmon J., and Porter J., “Clinical Practice Guidelines for the Management of Overweight and Obesity Published Internationally: A Scoping Review,” Obesity Reviews 25, no. 5 (2024): e13700, 10.1111/obr.13700. [DOI] [PubMed] [Google Scholar]
- 13. Garvey W. T., Mechanick J. I., Brett E. M., et al., “American Association of Clinical Endocrinologists and American College of Endocrinology Comprehensive Clinical Practice Guidelines for Medical Care of Patients with Obesity,” Endocrine Practice 22, no. Suppl 3 (2016): 1–203, 10.4158/ep161365.GL. [DOI] [PubMed] [Google Scholar]
- 14. Elmaleh‐Sachs A., Schwartz J. L., Bramante C. T., Nicklas J. M., Gudzune K. A., and Jay M., “Obesity Management in Adults: A Review,” Journal of the American Medical Association 330, no. 20 (2023): 2000–2015, 10.1001/jama.2023.19897. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. U.S. Food and Drug Administration , “ZEPBOUND (Tirzepatide) Injection for Subcutaneous Use,” 2023, https://www.accessdata.fda.gov/drugsatfda_docs/label/2024/217806s005s006s011s015s019lbl.pdf.
- 16. U.S. Food and Drug Administration , “WEGOVY (semaglutide) Injection for Subcutaneous Use Initial U.S. Approval,” 2017, https://www.accessdata.fda.gov/drugsatfda_docs/label/2025/215256s024lbl.pdf.
- 17. Ansari H. U. H., Qazi S. U., Sajid F., et al., “Efficacy and Safety of Glucagon‐Like Peptide‐1 Receptor Agonists on Body Weight and Cardiometabolic Parameters in Individuals With Obesity and Without Diabetes: A Systematic Review and Meta‐Analysis,” Endocrine Practice 30, no. 2 (2024): 160–171, 10.1016/j.eprac.2023.11.007. [DOI] [PubMed] [Google Scholar]
- 18. U.S. Food and Drug Administration , “MOUNJARO (Tirzepatide) Injection for Subcutaneous Use,” 2022, https://www.accessdata.fda.gov/drugsatfda_docs/label/2022/215866s000lbl.pdf.
- 19. Jastreboff A. M., Aronne L. J., Ahmad N. N., et al., “Tirzepatide Once Weekly for the Treatment of Obesity,” New England Journal of Medicine 387, no. 3 (2022): 205–216, 10.1056/NEJMoa2206038. [DOI] [PubMed] [Google Scholar]
- 20. Garvey W. T., Frias J. P., Jastreboff A. M., et al., “Tirzepatide Once Weekly for the Treatment of Obesity in People With Type 2 Diabetes (SURMOUNT‐2): A Double‐Blind, Randomised, Multicentre, Placebo‐Controlled, Phase 3 Trial,” Lancet 402, no. 10402 (2023): 613–626, 10.1016/s0140-6736(23)01200-x. [DOI] [PubMed] [Google Scholar]
- 21. Wadden T. A., Chao A. M., Machineni S., et al., “Tirzepatide After Intensive Lifestyle Intervention in Adults With Overweight or Obesity: The SURMOUNT‐3 Phase 3 Trial,” Nature Medicine 29, no. 11 (2023): 2909–2918, 10.1038/s41591-023-02597-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Aronne L. J., Sattar N., Horn D. B., et al., “Continued Treatment With Tirzepatide for Maintenance of Weight Reduction in Adults With Obesity: The SURMOUNT‐4 Randomized Clinical Trial,” Journal of the American Medical Association 331, no. 1 (2024): 38–48, 10.1001/jama.2023.24945. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Lam C. S. P., Rodriguez A., Aminian A., et al., “Tirzepatide for Reduction of Morbidity and Mortality in Adults With Obesity: Rationale and Design of the SURMOUNT‐MMO Trial,” Obesity 33, no. 9 (2025): 1645–1656, 10.1002/oby.24332. [DOI] [PubMed] [Google Scholar]
- 24. U.S. Food and Drug Administration , “FDA Approves First Medication for Obstructive Sleep Apnea”.
- 25. Hunter Gibble T., Chinthammit C., Ward J. M., et al., “Real‐World Use of Tirzepatide Among Individuals Without Evidence of Type 2 Diabetes: Results From the Veradigm Database,” Diabetes, Obesity and Metabolism 27, no. 6 (2025): 3185–3194, 10.1111/dom.16330. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Hankosky E. R., Chinthammit C., Meeks A., et al., “Real‐World Use and Effectiveness of Tirzepatide Among Individuals Without Type 2 Diabetes: Results From the Optum Market Clarity Database,” Diabetes, Obesity & Metabolism 27, no. 5 (2025): 2810–2821, 10.1111/dom.16290. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Hankosky E. R., Desai K., Chinthammit C., et al., “Real‐World Use and Effectiveness of Tirzepatide Among People Without Evidence of Type 2 Diabetes in the United States,” Diabetes & Metabolism 51, no. 3 (2025): 101636, 10.1016/j.diabet.2025.101636. [DOI] [PubMed] [Google Scholar]
- 28. Gibble T. H., Hankosky E. R., Meeks A., et al., “Characteristics and Treatment Patterns of People Without Type 2 Diabetes Diagnoses Who Use Tirzepatide in the United States,” Clinical Therapeutics 47, no. 8 (2025): 572–580, 10.1016/j.clinthera.2025.05.003. [DOI] [PubMed] [Google Scholar]
- 29. Stephenson J. J., Teng C. C., and Harris K., “RWD100 Assessing the Representativeness of Real‐World Claims Databases,” Value in Health 27, no. 6 (2024): S376–S377, 10.1016/j.jval.2024.03.1754. [DOI] [Google Scholar]
- 30. Barron J. J., Willey V. J., Doherty B. T., et al., “The Healthcare Integrated Research Database (HIRD) as a Real‐World Data Source for Pharmacoepidemiologic Research,” Pharmacoepidemiology and Drug Safety 34, no. 2 (2025): e70110, 10.1002/pds.70110. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. “Medi‐Span Generic Product Identifier (GPI),” 2026, https://assets.contenthub.wolterskluwer.com/api/public/content/2610220‐medi‐span‐generic‐product‐identifier‐gpi‐pdf‐f776137639?v=ea3eb174.
- 32. Gleason P. P., Urick B. Y., Marshall L. Z., Friedlander N., Qiu Y., and Leslie R. S., “Real‐World Persistence and Adherence to Glucagon‐Like Peptide‐1 Receptor Agonists Among Obese Commercially Insured Adults Without Diabetes,” Journal of Managed Care & Specialty Pharmacy 30, no. 8 (2024): 860–867, 10.18553/jmcp.2024.23332. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Rodriguez P. J., Zhang V., Gratzl S., et al., “Discontinuation and Reinitiation of Dual‐Labeled GLP‐1 Receptor Agonists Among US Adults With Overweight or Obesity,” JAMA Network Open 8, no. 1 (2025): e2457349–e2457349, 10.1001/jamanetworkopen.2024.57349. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Ruder K., “Patients Are Flocking to Compounded Weight‐Loss Drugs, but Are They Safe?,” Journal of the American Medical Association 333, no. 8 (2025): 652–655, 10.1001/jama.2024.27251. [DOI] [PubMed] [Google Scholar]
- 35. Hwang J. H., Laiteerapong N., Huang E. S., and Kim D. D., “Lifetime Health Effects and Cost‐Effectiveness of Tirzepatide and Semaglutide in US Adults,” JAMA Health Forum 6, no. 3 (2025): e245586–e245586, 10.1001/jamahealthforum.2024.5586. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Gasoyan H., Butsch W. S., Casacchia N. J., et al., “Reasons for Discontinuation of Obesity Pharmacotherapy With Semaglutide or Tirzepatide in Clinical Practice,” Obesity 12 (2025): 2296–2303, 10.1002/oby.70058. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Linetzky B., Sattar N., Verma S., et al., “Improvements in Cardiometabolic Risk Factors by Weight Reduction: A Post Hoc Analysis of Adults With Obesity Randomly Assigned to Tirzepatide,” Annals of Internal Medicine 178, no. 8 (2025): 1095–1105, 10.7326/annals-24-02623. [DOI] [PubMed] [Google Scholar]
- 38. Aicher B. O., Haser E. K., Freeman L. A., et al., “Diet‐Induced Weight Loss in Overweight or Obese Women and Changes in High‐Density Lipoprotein Levels and Function,” Obesity 20, no. 10 (2012): 2057–2062, 10.1038/oby.2012.56. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Yatsuya H., Jeffery R. W., Erickson D. J., et al., “Sex‐Specific HDL Cholesterol Changes With Weight Loss and Their Association With Anthropometric Variables: The LIFE Study,” Obesity 19, no. 2 (2011): 429–435, 10.1038/oby.2010.216. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Figure S1: Study design.
Figure S2: Patient attrition figure.
Figure S3: Kaplan Meier curves for time‐to‐discontinuation among tirzepatide initiators using a 45‐day allowed gap.
Table S1: Baseline demographics and clinical characteristics in the overall cohort.
Table S2: List of diagnosis codes and national drug codes.
Data Availability Statement
The data that support the findings of this study are available from Carelon Research, but restrictions apply to the availability of these data to external sources, and therefore they are not publicly available. Data may be made available from the corresponding author upon reasonable request and with permission from Carelon Research.