Abstract
BACKGROUND:
Despite the clinical efficacy of glucagon-like peptide-1 receptor agonists (GLP-1RAs) for weight loss, real-world persistence remains substantially lower than seen in clinical trials. Since 2021, GLP-1RA shortages and changes in product availability have complicated access and persistence.
OBJECTIVE:
To describe by calendar year of initiation the 1-year treatment persistence and adherence trends among commercially insured members without diabetes newly initiating high-potency, weight loss–indicated GLP-1RA products, semaglutide (Wegovy) and tirzepatide (Zepbound).
METHODS:
This cohort study used Prime Therapeutics’ integrated medical and pharmacy claims data from an average monthly membership of 17.9 million commercially insured individuals to identify members without diabetes who newly initiated high-potency, weight loss–indicated GLP-1RA treatment between January 1, 2021, and June 30, 2024 (index date period), defined in year increments, 2021, 2022, 2023, and the first half of 2024 (1H 2024). Members were included in the study sample if they were continuously enrolled with no more than 15 days total enrollment gap during the 365-day preperiod and the 365-day postperiod after their index GLP-1RA claim with no prior GLP-1RA drug claim 365 days before index. Members were excluded if they had a preperiod diabetes diagnosis medical claim or diabetes drug claim or if they were aged less than 19 years at index. The primary outcome of GLP-1RA persistence was measured as no more than a 60-day gap between the end of a GLP-1RA claim’s days supply and the subsequent claim fill date in the 365-day period following index. Adherence was defined as proportion of days covered of at least 80%. Across all measures GLP-1RA product switching was allowed.
RESULTS:
Among the 62,650 members without diabetes newly initiating a high-potency, weight loss–indicated GLP-1RA between 2021 and June 30, 2024, 33,607 (53.6%) met the full study criteria. The mean age was 45.7 years, and 75.5% were female. Across the index years, 1-year persistence increased from 33.2% in 2021 to 60.9% in 1H 2024. Semaglutide 1-year persistence rates from 2021 to 1H 2024 were 33.2%, 34.1%, 39.8%, and 58.6%, respectively. For tirzepatide, 1-year persistence rates in 2023 and 1H 2024 (the only years the product was available) were 64.0% and 64.8%, respectively.
CONCLUSIONS:
This real-world analysis of high-potency, weight loss–indicated GLP-1RA products among individuals without diabetes found that 1-year treatment persistence has nearly doubled from 33.2% in 2021 to 60.9% in 1H 2024. GLP-1RA product shortage resolution may have improved persistence rates. Other potential influencing factors include improved GLP-1RA dose escalation and side effect management, as well as weight loss lifestyle management programs. Additional research is needed to understand treatment discontinuation reasons and long-term GLP-1RA cost-effectiveness.
Plain language summary
This study assessed if people without diabetes starting a weight loss glucagon-like peptide 1 receptor agonist (GLP-1RA), either Wegovy or Zepbound, remained on therapy during their first year and how trends in these rates changed in recent years. Among 33,607 people starting GLP-1RA treatments, 1 in 3 remained on therapy in 2021, which nearly doubled to 6 in 10 in the first half of 2024. This increase likely reflects resolved GLP-1RA supply shortages and better care management.
Implications for managed care pharmacy
Among commercially insured individuals without diabetes, 1-year persistence to high-potency, weight-loss–indicated GLP-1RAs nearly doubled from 33% in 2021 to 61% in the first half of 2024. With the GLP-1RA shortages resolved, these findings will help in assessing GLP-1RA products’ cost-effectiveness and forecast future GLP-1RA use and cost trends. Continued real-world analysis of GLP-1RA persistence will be essential to understanding and addressing nonpersistence.
Obesity persists at epidemic proportions in the United States, representing a significant public health and economic challenge.1–4 This chronic, progressive disease impacts approximately 40% of US adults3 with an estimated economic and societal impact of $1.39 trillion annually.4 Since 2021, the therapeutic landscape for obesity has evolved with the introduction of new antiobesity medications (AOMs), including glucagon-like peptide 1 receptor agonists (GLP-1RAs) and dual glucose-dependent insulinotropic polypeptide/GLP-1 receptor co-agonists. This drug class has been used for 2 decades in the management of type 2 diabetes. In addition, the once-weekly, high-potency formulations—semaglutide (Wegovy) and tirzepatide (Zepbound)—are highly efficacious pharmacologic options for weight loss and weight management. Semaglutide, approved by the US Food and Drug Administration (FDA) in June 2021,5 demonstrated average weight reductions of 15.0% in clinical trials,6 whereas tirzepatide, approved by the FDA in November 2023,7 demonstrated average weight reductions between 15.0% and 20.9%.8
Driven by surging demand for GLP-1RA therapies,9,10 formulations of semaglutide and tirzepatide faced supply shortages beginning in 2022.11 These shortages persisted through early 2025, creating widespread access barriers and disrupting prescribed treatment regimens.12 In response, many patients and providers turned to compounded alternatives or experienced prolonged treatment gaps.13 In parallel with rising demand and supply chain interruptions, real-world studies have reported poor persistence and adherence to weight loss GLP-1RA therapies. Persistence, defined as the continuation of GLP-1RA treatment over time, is estimated in prior studies at 32% to 50% over 1 year.14–17 Adherence, defined as the extent to which patients take medication as prescribed, has been shown to average 51% to 54% using the proportion of days covered (PDC) measure,14,15 with some evidence that patients reinitiate therapy after temporary gaps in treatment.17 Other research has identified high rates of GLP-1RA discontinuation within the first few months of therapy initiation.15,18,19 These GLP-1RA use patterns raised concerns about the real-world short- and long-term clinical value.
As a direct result of GLP-1RA weight loss treatment demand, total GLP-1RA spending has become a major cost driver for plan sponsors.20–22 With 2025 annual gross GLP-1RA costs exceeding $12 000 per patient, plan sponsors face mounting pressure to balance clinical benefit with budgetary constraints. The International Foundation of Employee Benefit Plans reported that on average 10.5% of employers’ entire drug expense went toward weight loss GLP-1RAs in 2025, an increase from 8.9% in 2024 and 6.9% in 2023.22 Given the extensive utilization of these products, GLP-1RA spending remains a critical concern. To further elaborate, prior estimates in an employer-based population suggest that a 1% increase in the population using weight loss–indicated GLP-1RAs adds approximately $10 in per member per month expense, equating to a near 5% increase in the total drug budget.23 Short-term costs are compounded by long-term growth in health insurance premiums, which rose by 7% in 2024, reaching an average of $25 572 annually for family coverage.24 With substantial evidence of weight gain following GLP-1RA discontinuation, the cost associated with GLP-1RA treatment may persist for many years as the optimal duration of therapy is at least 6 months and may be lifelong.25–27
Escalating costs combined with current real-world evidence showing poor persistence and adherence to GLP-1RAs for weight loss raise cost-effectiveness and affordability concerns.28,29 Given these concerns and the importance of understanding real-world utilization patterns, this study aims to evaluate 1-year persistence and adherence trends to high-potency, weight loss–indicated GLP-1RA therapy among commercially insured adults without diabetes, stratified by year of treatment initiation from January 1, 2021, through June 30, 2024. By including the most recent first half of 2024 (1H 2024) cohort initiators, this analysis provides timely insight into past and present use patterns to inform real-world utilization of these products.
Methods
Integrated medical and pharmacy claims plus enrollment data from January 1, 2020, to September 14, 2025, across 19 commercial health plans covering all regions of the United States were obtained for the cohort study. The Strengthening of the Reporting of Observational Studies in Epidemiology (STROBE) cohort study checklist was used to guide this study.30 Data obtained for this study included medical claims (date of service, diagnoses received, and procedures performed), pharmacy claims (fill dates, day supply, and National Drug Code numbers), and eligibility information (member demographics and enrollment history).
This retrospective study (Figure 1) identified members newly initiating a high-potency, weight loss–indicated GLP-1RA medication between January 1, 2021, and June 30, 2024 (study index period) for the following GLP-1RA products: semaglutide (Wegovy) and tirzepatide (Zepbound) (Supplementary Table 1, available in online article). During the study index period, the source population contained an average of 17.9 million eligible members monthly.
FIGURE 1.
Study Design
GLP-1RA = glucagon-like peptide 1 receptor agonist.
Study inclusion criteria required members to be continuously enrolled with no more than 15 days of total enrollment gap during the 365-day preperiod before the index date and 365-day period after their index GLP-1RA claim. A new user design was employed requiring no prior GLP-1RA drug claim in the 365 days before index (ie, pre-index period). We identified 4 index periods representing the years 2021, 2022, 2023, and 1H 2024. Members were assigned to an index period using their first high-potency, weight loss–indicated GLP-1RA medication claim date (index date). Members were excluded if they had a preperiod medical claim indicating a diabetes mellitus (DM) diagnosis (type 1 DM, type 2 DM, gestational diabetes, diabetes caused by an underlying condition, chemical-induced diabetes, and other specified diabetes; Supplementary Table 2), had a pharmacy claim for an antidiabetic medication during the 365-day preperiod, or were aged less than 19 years at index. Given that obesity is significantly undercoded in claims data, we chose to proxy obesity diagnoses by excluding patients with diabetes.31 Members with diagnoses for HIV/AIDS, hemophilia, sickle cell disease, malignant cancer, or end-stage renal disease as identified by diagnosis codes in medical claims during the 365 days before the study index date were also excluded given the severity of these comorbidities, which may influence persistence and adherence.
The primary outcome of the study was 1-year persistence. All members were observed until the earliest occurrence of either GLP-1RA discontinuation or the end of the 365-day observation window. Persistence was defined as the absence of a treatment gap exceeding 60 days following the runout date of the previous claim’s days supply within the first 365 days after treatment initiation.14 Treatment gaps were analyzed after adjusting GLP-1RA claims for overlapping days supply. Following prior published research, members with a gap in therapy of 60 days or more were considered to have discontinued therapy and were therefore nonpersistent. The last day of supply before the discontinuation event gap was defined as the member’s discontinuation date.
The secondary outcome was 1-year adherence measured using the PDC method. Unlike the standard PDC methodology used by the Centers for Medicare & Medicaid Services, which requires a minimum of 2 claim fills, our approach included members with 1 or more fills.32 This methodological difference allowed adherence to be assessed beginning with the member’s index GLP-1RA claim. All prescription fill dates within the 365 days following each member’s study index date were included in the PDC calculation. Members with a PDC greater than or equal to 80% were classified as adherent; those with a PDC less than 80% were considered nonadherent. An 80% threshold was chosen for adherence as it is the most common cut point used in the literature and represents an acceptable level of adherence in Healthcare Effectiveness Data and Information Set and Star Ratings specifications for 3 common chronic classes of medication.32
Given the evolving GLP-1RA weight loss treatment landscape during the study observation period, we allowed for GLP-1RA product switching in our persistence and adherence measures to capture real-world use patterns. GLP-1RA product switching was defined as a change from 1 GLP-1RA product to another between consecutive claims for a given member, regardless of brand name or indication. For example, if a member initiated therapy with Zepbound (index drug) and then switched to Mounjaro (another tirzepatide product), this was counted as a switch, even though both products contain tirzepatide. All subsequent changes from 1 GLP-1RA product to another between later consecutive claims were counted as additional switches. The total number of product switches per member was categorized into 3 groups: 0, 1, or at least 2 switches.
Demographic and clinical characteristics of members were compared across index GLP-1RA products by year of initiation as well as for the overall study cohort. The characteristics assessed included sex, age, and health status. Age and sex were determined at index using plan enrollment data. Health status was evaluated using the Charlson comorbidity index (CCI), calculated from medical claims during the pre-index period.33 Differences in demographic and CCI scores across index years were assessed using analysis of variance for continuous variables and chi-square tests for categorical variables.
GLP-1RA therapy persistence was evaluated using 2 approaches. First, median time-to-discontinuation and corresponding 95% CIs were evaluated using the Kaplan-Meier method, with overall differences across initiation years assessed using the log rank test.34 Second, persistence status at the end of 1 year was summarized as a binary outcome. Overall group differences in 1-year persistence rates across index year periods were assessed using a chi-square test. Pairwise comparisons of nonpersistence odds between index years were conducted using a generalized linear model (GLM) with a binomial distribution and logit link function. Pairwise contrasts of estimated log odds computed between index years were derived from the model’s covariance matrix, and P values were adjusted for multiple comparisons using the Bonferroni method to maintain a family-wise error rate of α equals 0.05 (adjusted significance threshold: α = 0.0083 for 6 comparisons). Odds ratios (ORs) and 95% CIs were reported. The same GLM-based approach was used to compare adherence and switching across index year periods, with each GLM estimating the log odds of being adherent or having a GLP-1RA product switch.
All statistical analyses were conducted using Python version 3.11.7 with the following packages: lifelines v0.3 and statsmodels v0.14.3. This study was conducted to address the objectives stated for health insurance business purposes and, therefore, is institutional review board exempt. All authors with data access are employees of Prime Therapeutics, a pharmacy benefits manager. Prime Therapeutics adjudicates and pays pharmacy claims, storing all data according to the Health Insurance Portability and Accountability Act regulations, from which a limited dataset was created to answer the study objectives.
Results
During the index study period, a total of 33 607 members without diabetes who newly initiated a high-potency, weight loss–indicated GLP-1RA product (semaglutide [Wegovy] or tirzepatide [Zepbound]) between January 1, 2021, and June 30, 2024, met all predefined inclusion and exclusion criteria (Figure 2). As shown in Table 1, 75.5% of the cohort were female. The proportion of women varied across index years, ranging from 73.2% in 1H 2024 to 79.9% in 2022. Age and CCI scores remained stable across index years. Mean age at index was 45.7 (SD = 10.6) with a mean CCI score of 0.2 (SD = 0.5). More than three-fourths (79.9%) of individuals had a CCI score of 0.
FIGURE 2.
Cohort Attrition
Rare or disabling disease includes HIV, AIDS, hemophilia A, hemophilia B, sickle cell disease, malignant cancer, and end-stage renal disease.
GLP-1RA = glucagon-like peptide 1 receptor agonist.
TABLE 1.
Demographics, Clinical Characteristics, and Outcomes by Index Year
| All membersa (N = 33,607) | 2021 initiators (n = 889) | 2022 initiators (n = 2,480) | 2023 initiators (n = 14,665) | 1H 2024 initiators (n = 15,573) | P valueb | |
|---|---|---|---|---|---|---|
| Characteristic c | ||||||
| Female, n (%) | 25,370 (75.5) | 704 (79.2) | 1,981 (79.9) | 11,283 (76.9) | 11,402 (73.2) | <0.001 |
| Male, n (%) | 8,237 (24.5) | 185 (20.8) | 499 (20.1) | 3,382 (23.1) | 4,171 (26.8) | |
| Age, mean (SD), years | 45.7 (10.6) | 45.6 (10.2) | 46.0 (9.9) | 45.5 (10.4) | 45.8 (10.9) | 0.029 |
| Charlson comorbidity index, mean (SD) | 0.2 (0.5) | 0.2 (0.5) | 0.2 (0.5) | 0.2 (0.5) | 0.2 (0.6) | 0.141 |
| 0, n (%) | 26,836 (79.9) | 727 (81.8) | 1,997 (80.5) | 11,732 (80.0) | 12,380 (79.5) | 0.041 |
| 1, n (%) | 5,683 (16.9) | 132 (14.8) | 410 (16.5) | 2,465 (16.8) | 2,676 (17.2) | |
| 2+, n (%) | 862 (2.6) | 27 (3.0) | 56 (2.3) | 390 (2.7) | 389 (2.5) | |
| Study outcome d | ||||||
| Persistent without 60-day gap, n (%) | 16,515 (49.1) | 295 (33.2) | 845 (34.1) | 5,898 (40.2) | 9,477 (60.9) | <0.001 |
| Adherence (PDC), mean (SD) | 63.9 (33.2) | 52.6 (32.9) | 51.6 (33.9) | 57.3 (33.3) | 72.8 (28.8) | <0.001 |
| Adherent (PDC ≥80%), n (%) | 14,968 (39.3) | 271 (30.2) | 769 (30.7) | 5,304 (35.9) | 8,624 (55.5) | <0.001 |
| Number of switches, n (%) d , e | ||||||
| 0 | 28,771 (85.6) | 750 (84.4) | 2,021 (81.5) | 12,619 (86.0) | 13,381 (85.9) | <0.001 |
| 1 | 3,730 (11.1) | 113 (12.7) | 292 (11.8) | 1,448 (9.9) | 1,877 (12.1) | |
| 2+ | 1,106 (3.3) | 26 (2.9) | 167 (6.7) | 598 (4.1) | 315 (2.0) | |
Commercially insured adults without diabetes newly initiating a high-potency, weight loss–indicated glucagon-like peptide 1 receptor agonist between January 1, 2021, and June 30, 2024.
P values for continuous variables are derived from the F test resulting from the generalized linear model. P values for categorical variables are derived from the chi-square test.
Clinical characteristics were measured in a 365-day period before glucagon-like peptide-1 receptor agonist initiation (preperiod). Age and sex were ascertained on the member’s study index date.
Measured at 365 days after glucagon-like peptide-1 receptor agonist initiation.
Count of member-level product switches.
1H 2024 = first half of 2024, PDC = proportion of days covered.
Persistence increased from 33.2% in 2021 to 34.1% in 2022, 40.2% in 2023, and 60.9% in 1H 2024 (Table 1). Semaglutide (Wegovy) 1-year persistence rates from 2021 to 1H 2024 were 33.2%, 34.1%, 39.8%, and 58.6% (Table 2) across the respective years, whereas tirzepatide (Zepbound; first available in November 2023) rates were 64.0% and 64.8% for 2023 and 1H 2024, respectively. Median time to GLP-1RA discontinuation (Figure 3) differed significantly across index years (P < 0.001) and ranged from a low of 144 days (95% CI = 135-156) among 2022 initiators to not reached (95% CI = NR-NR) among 1H 2024 initiators, indicating that fewer than 50% of members had discontinued therapy by the end of the observation period. Pairwise comparisons (Supplementary Table 3 (248.2KB, pdf) ) across index years found the odds of being persistent at the end of 1 year was more than 3 times higher among 1H 2024 initiators compared with 2021 (OR = 3.13, 95% CI = 2.71-6.31) and 2022 initiators (OR = 3.01, 95% CI = 2.75-3.29) and more than 2 times higher compared with 2023 initiators (OR = 2.31, 95% CI = 2.21-2.42).
TABLE 2.
Percentage of Members Persistent at 1 Year by GLP-1RA Index Product
| Index GLP-1RA product, n (%) | |||
|---|---|---|---|
| Overall, n (%) | Semaglutide (Wegovy) | Tirzepatide (Zepbound) | |
| 2021 initiators | 889 (33.2) | 889 (33.2) | NA |
| 2022 initiators | 2,480 (34.1) | 2,480 (34.1) | NA |
| 2023 initiators | 14,665 (40.2) | 14,429 (39.8) | 236 (64.0) |
| 1H 2024 initiators | 15,573 (60.9) | 9,861 (58.6) | 5,712 (64.8) |
1H 2024 = first half of 2024, GLP-1RA = glucagon-like peptide 1 receptor agonist; NA = not applicable as tirzepatide (Zepbound) was not approved by the US Food and Drug Administration until November 2023.
FIGURE 3.
High-Potency, Weight Loss–Indicated Glucagon-Like Peptide-1 Receptor Agonists: Kaplan-Meier 1-Year Therapy Persistence (N = 33,607)
The Kaplan-Meier curve represents 33,607 commercially insured adults without diabetes initiating a high-potency, weight loss–indicated glucagon-like peptide-1 receptor agonist product defined as semaglutide (Wegovy) or tirzepatide (Zepbound) between January 1, 2021, through June 30, 2024. All persistency measurements were conducted at the glucagon-like peptide-1 receptor agonist product level. Members were considered persistent if they did not have a 60-day gap in therapy and were censored at the end of the 365-day period.
1H 2024 = first half of 2024; NR = not reached.
As shown in Table 1, mean 1-year PDC was lowest among 2022 initiators at 51.6% (SD = 33.9), and highest among 1H 2024 initiators at 72.8% (SD = 28.8). Overall, 39.3% of members were adherent during the follow-up period, with adherence rates ranging from 30.2% to 55.5% across index years. Odds of adherence rates were significantly higher among 1H 2024 initiators compared with all other index years (Supplementary Table 4 (248.2KB, pdf) ). The odds of being adherent at 1 year were nearly 3 times higher for 1H 2024 initiators compared with 2021 (OR = 2.83, 95% CI = 2.45-3.28) and 2022 initiators (OR = 2.76, 95% CI = 2.52-3.02) and more than 2 times higher compared with 2023 initiators (OR = 2.19, 95% CI = 2.09-2.29).
Overall, 14.4% of members had at least 1 GLP-1RA switch during the follow-up period. Members having 1 or more GLP-1RA switches varied by index year and ranged from 14.0% among 2023 initiators to 18.5% among 2022 initiators. Pairwise comparisons (Supplementary Table 5 (248.2KB, pdf) ) of having at least 1 GLP-1RA product switch showed that members initiating therapy in 2023 (OR = 0.71; 95% CI = 0.64-0.80) or in 1H 2024 (OR = 0.72; 95% CI = 0.65-0.81) were significantly less likely to have at least 1 GLP-1RA product switch during the 1-year follow-up compared with those initiating in 2022. No statistically significant differences in the likelihood of having a GLP-1RA product switch were observed between other index years.
Discussion
This real-world analysis of commercially insured members without diabetes using high-potency, weight loss–indicated GLP-1RA drug therapy revealed significant improvements in persistence and adherence over time, particularly among 1H 2024 initiators. One-year persistence nearly doubled, increasing from 33% among 2021 initiators to 61% among those who began therapy in 1H 2024. This finding suggests a significant reduction in the number of patients discontinuing treatment among patients initiating treatment in more recent years. Furthermore, adherence during the first year of therapy also improved over time, with the highest rates observed among the most recent initiators (55.5% in 1H 2024) relative to rates in earlier years (eg, 30.2% in 2021). This suggests that in addition to greater persistence, more patients can regularly take medication as prescribed throughout the year in more recent years as well.
Prior real-world studies have largely focused on off-label use of weight loss GLP-1RAs treatment and found high rates of discontinuation and poor persistence.14–19 Our findings support and extend existing real-world data among similar patient cohorts initiating therapy before 2024. To our knowledge, this is the first study to assess persistence, adherence, and switching among individuals initiating GLP-1RA weight loss products as recently as 1H 2024. By including the 1H 2024 index period, our study provides insight into persistence and adherence with the contemporary high-potency, weight loss–indicated GLP-1RA product mix. Although 1-year persistence and adherence to high-potency, weight loss–indicated GLP-1RA therapy has improved since 2021, both remain below the levels seen in randomized controlled trials of more than 80%.6,8
Additionally, our findings suggest that users of high-potency, weight loss–indicated GLP-1RA medications are generally younger, healthier adults, and predominately female. Although the proportion of women declined modestly over time—from 79.2% in 2021 to 73.2% in 1H 2024—women represented three-quarters of the overall cohort. The mean age of initiators was stable at 46 years, and the comorbidity burden was low, with nearly 80% of individuals having a CCI score of 0. Consistent age and comorbidity profiles over time may reflect consistent prescribing patterns, and the gradual shift in distribution by sex may indicate increased treatment among male patients. Also, the exclusion of members with diabetes reduces the observed comorbidity burden across all study periods.
The observed improvements in persistence and adherence across index years and lower switching rates in more recent years coincide with a period of evolving GLP-1RA supply chain issues.11 Semaglutide and tirzepatide first appeared on the FDA shortage list in 2022. Manufacturers implemented corrective measures throughout 2023 and 2024, with the FDA officially declaring the tirzepatide shortage resolved in October 2024 and the semaglutide shortage resolved in February 2025. As a result, supply constraints may contribute to greater GLP-1RA use missingness in our data during earlier index years, resulting in misclassification of nonpersistence and nonadherence. Members initiating therapy in 2021 through early 2023 may have experienced more frequent treatment interruptions or sought GLP-1RA products outside of their pharmacy benefit because of limited product availability through traditional channels. In contrast, those initiating in late 2023 and early 2024 likely benefited from improved supply chain stability, resulting in higher observed persistence and adherence owing to the removal of supply barriers and the subsequent more complete GLP-1RA record seen in pharmacy claims. Of note, the weight loss–indicated semaglutide (Wegovy) product came to market in August 2021 and tirzepatide (Zepbound) product came to market in November 2023. It is also worth noting that in 2023, tirzepatide persistence far exceeded semaglutide persistence (39.8% vs 64.0%, respectively). However, moving to 1H 2024, the persistence gap diminished with 58.6% of semaglutide patients persisting on treatment and 64.8% of tirzepatide patients persisting on treatment. This timing coincides with an end to shortages, which lasted an additional 5 months longer among semaglutide users.11
Beyond supply chain dynamics, affordability remains a key concern because of 2025 gross annual costs exceeding $12 000 per patient when obtained through health insurance. At their current price point, weight loss GLP-1RAs pose challenges to cost-effectiveness and sustained affordability, particularly given the growing scale of use.28,29 Employers and payers face mounting pressure to balance clinical benefit with budgetary constraints, prompting increased scrutiny of coverage policies, formulary management, and the development of value-based purchasing agreements. Fiscal pressures are mounting at the federal level as well. If Medicare adds AOMs to the benefit, the Congressional Budget Office estimates 12.5 million Medicare beneficiaries would newly qualify for AOM therapy, resulting in net new federal spending of $35 billion over the next 10 years.35
Lastly, refined clinical management and GLP-1RA dose escalation are factors likely contributing to improvements in medication-taking behavior over the study period.36 Expert and World Health Organization recommendations support both health care providers and patients in minimizing the frequency and intensity of GLP-1RA–related side effects and in achieving the target maintenance dose—all aimed at preventing therapy discontinuation and maximizing weight loss benefit.27,37 Nutritional and lifestyle interventions were a component of all FDA approval GLP-1RA clinical trials, and a growing body of evidence supports the need for nutritional and lifestyle interventions to mitigate side effects, prevent nutrient deficiencies, and preserve lean body mass.38,39 These strategies are essential not only for maximizing treatment effectiveness but also for sustaining long-term adherence and preventing weight regain.40
LIMITATIONS
Data were sourced from administrative health care claims; therefore, misclassification bias may have occurred because of using medical and pharmacy claims to exclude individuals without diabetes. The data used to measure adherence and persistence also assume that the medication dispensed represents actual consumption, which cannot be measured directly in claims data. This study allowed for GLP-1RA product switching in the measurement of persistence and adherence to capture dynamic, real-world use patterns across different GLP-1RA products. Although this approach improves accuracy in capturing time on therapy, it may obscure product-specific discontinuation patterns—although the overall switch rate was below 15%. Additionally, members obtaining GLP-1RA medications outside of their pharmacy benefit (eg, compounded product, cash purchases at pharmacies, or direct-to-consumer purchase options) may have reduced observed persistence and adherence.
Although this study provides valuable insights into real-world GLP-1RA utilization patterns, benefit design and coverage decisions were outside the scope of the analysis. These factors may influence initiation, persistence, and adherence behaviors, particularly for weight loss–indicated GLP-1RAs. Among the study cohort, more than 80% had a GLP-1RA prior authorization on record during follow-up, with 53% having a prior authorization with a duration of at least 365 days at time of index. These findings suggest broad on-benefit coverage among our study cohort.
This analysis did not account for potential differences in GLP-1RA dosing. Specifically, the study did not assess if members achieved maximum therapy dose, describe maximum tolerated doses, or assess microdosing,41 which all may influence treatment persistence, adherence, and switching. Additionally, lack of weight loss or achieving weight loss goals before 1 year of therapy may have resulted in some of the therapy discontinuation seen in this study. Neither body mass index nor weight were available in the claims data to determine lack of response, and weight loss care management enrollment data were not available. This study examined commercially insured members, and therefore, results are not generalizable to Medicare or Medicaid populations. The impact of cost sharing; other diagnoses; social determinants of health such as income, education, and health literacy; and other member characteristics are outside the scope of this analysis but may be worthy of future consideration.
Conclusions
Our analysis of high-potency, weight loss–indicated GLP-1RA products found 1-year persistence nearly doubling from 33.2% in 2021 to 60.9% in 2024. This improvement likely reflects the resolution of GLP-1RA product shortage supply chain issues, which were addressed throughout 2024 and into 2025. Other contributing factors may include better GLP-1RA dose escalation protocols, improved side effect management, and adoption of care and lifestyle management programs. Despite the observed improvements in persistence and adherence, further research is needed to understand the drivers of treatment discontinuation and to evaluate the long-term cost-effectiveness of weight loss–indicated GLP-1RA therapies. Additional research should also examine the impact of nonpersistence and nonadherence on health outcomes, health resource utilization, and health spending. Our study provides valuable real-world evidence to inform policy decisions, support actuarial planning, and guide future value-based assessments in the evolving landscape of obesity pharmacotherapy.
Disclosures
Drs Marshall, Gleason, Friedlander, and Urick are employed by the study sponsor.
References
- 1.Mitchell NS, Catenacci VA, Wyatt HR, Hill JO. Obesity: Overview of an epidemic. Psychiatr Clin North Am. 2011;34(4):717-32. doi: 10.1016/j.psc.2011.08.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Temple NJ. The origins of the obesity epidemic in the USA-Lessons for today. Nutrients. 2022;14(20):4253. doi: 10.3390/nu14204253 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.U.S. Centers for Disease Control and Prevention . About obesity. Published January 23, 2024. Accessed August 7, 2025. https://www.cdc.gov/obesity/php/about/index.html
- 4.Lopez C, Bendix J, Sagynbekov K. Weighing down America: 2020 update. Milliken Institute. Published 2020. Accessed August 8, 2025. https://milkeninstitute.org/sites/default/files/reports-pdf/Weighing%20Down%20America%20v12.3.20_0.pdf
- 5.Center for Drug Evaluation and Research. Approval package for: Wegovy injection (semaglutide). US Food and Drug Administration. Published June 2021. Accessed August 7, 2025. https://www.accessdata.fda.gov/drugsatfda_docs/nda/2021/215256Orig1s000Approv.pdf
- 6.Wilding JPH, Batterham RL, Calanna S, et al. ; STEP 1 Study Group. STEP 1 Study Group. Once-weekly semaglutide in adults with overweight or obesity. N Engl J Med. 2021;384(11):989-1002. doi: 10.1056/NEJMoa2032183 [DOI] [PubMed] [Google Scholar]
- 7.Center for Drug Evaluation and Research. Approval package for: Zepbound injection (tirzepatide). US Food and Drug Administration. Published November 2023. Accessed August 7, 2025. https://www.accessdata.fda.gov/drugsatfda_docs/nda/2024/217806Orig1s000Approv.pdf
- 8.Jastreboff AM, Aronne LJ, Ahmad NN, et al. ; SURMOUNT-1 Investigators. Tirzepatide once weekly for the treatment of obesity. N Engl J Med. 2022;387(3):205-16. doi: 10.1056/NEJMoa2206038 [DOI] [PubMed] [Google Scholar]
- 9.Watanabe JH, Kwon J, Nan B, et al. Trends in glucagon-like peptide 1 receptor agonist use, 2014 to 2022. J Am Pharm Assoc (2003). 2024;64(1):133-8. doi: 10.1016/j.japh.2023.10.002. [DOI] [PubMed] [Google Scholar]
- 10.Yeo YH, Rezaie A, Hsieh TY, et al. ; Gastrointestinal Motility and Metabolic Pharmacoepidemiology Group. Shifting trends in the indication of glucagon-like peptide-1 receptor agonist prescriptions: A nationwide analysis. Ann Intern Med. 2024;177(9):1289-91. doi: 10.7326/M24-0019 [DOI] [PubMed] [Google Scholar]
- 11.US Food and Drug Administration. FDA clarifies policies for compounders as national GLP-1 supply begins to stabilize. Published April 28, 2025. Accessed August 7, 2025. https://www.fda.gov/drugs/drug-safety-and-availability/fda-clarifies-policies-compounders-national-glp-1-supply-begins-stabilize
- 12.Whitley HP, Trujillo JM, Neumiller JJ. Special report: Potential strategies for addressing GLP-1 and dual GLP-1/GIP receptor agonist shortages. Clin Diabetes. 2023;41(3):467-73. doi: 10.2337/cd23-0023 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Mattingly TJ II, Conti RM. Marketing and safety concerns for compounded GLP-1 receptor agonists. JAMA Health Forum. 2025;6(1):e245015. doi: 10.1001/jamahealthforum.2024.5015 [DOI] [PubMed] [Google Scholar]
- 14.Gleason PP, Urick BY, Marshall LZ, Friedlander N, Qiu Y, Leslie RS. Real-world persistence and adherence to glucagon-like peptide-1 receptor agonists among obese commercially insured adults without diabetes. J Manag Care Spec Pharm. 2024;30(8):860-7. doi: 10.18553/jmcp.2024.23332 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Do D, Lee T, Peasah SK, Good CB, Inneh A, Patel U. GLP-1 receptor agonist discontinuation among patients with obesity and/or type 2 diabetes. JAMA Netw Open. 2024;7(5):e2413172. doi: 10.1001/jamanetworkopen.2024.13172 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Polonsky WH, Arora R, Fernandes J, et al. ; Demographic and clinical characteristics associated with real-world persistence on semaglutide for weight management in the USA. Diabetes. 2025;74(Supplement_1):786-P. doi: 10.2337/db25-786-P [DOI] [Google Scholar]
- 17.Rodriguez PJ, Zhang V, Gratzl S, et al. Discontinuation and reinitiation of dual-labeled GLP-1 receptor agonists among US adults with overweight or obesity. JAMA Netw Open. 2025;8(1):e2457349. doi: 10.1001/jamanetworkopen.2024.57349 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Blue Health Intelligence. Real-world trends in GLP-1 treatment persistence and prescribing for weight management. Blue Cross Blue Shield Association. Accessed August 7, 2025. https://www.bcbs.com/media/pdf/BHI_Issue_Brief_GLP1_Trends.pdf
- 19.Xu Y, Carrero JJ, Chang AR, et al. Titration and discontinuation of semaglutide for weight management in commercially insured US adults. Obesity (Silver Spring). 2025;33(7):1243-8. doi: 10.1002/oby.24315 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.United Healthcare and Health Action Council. The obesity epidemic: Growing health impacts and cost implications. Accessed August 11, 2025. https://healthactioncouncil.org/media/i5qj3qah/2025-hac-white-paper.pdf
- 21.Williams E, Rudowitz R, Bell C. Medicaid coverage of and spending on GLP-1s. Kaiser Family Foundation. Published November 4, 2024. Accessed August 4, 2025. https://www.kff.org/medicaid/issue-brief/medicaid-coverage-of-and-spending-on-glp-1s/
- 22.Plier R. GLP-1 drugs responsible for over ten percent of annual claims. International Foundation of Employee Benefit Plans. Published May 22, 2025. Accessed August 4, 2025. https://www.ifebp.org/resources---news/news-and-regulatory-updates/press-room/press-releases/2025/05/22/glp-1-drugs-responsible-for-over-ten-percent-of-annual-claims
- 23.Prime Therapeutics. Real-world analysis of GLP-1a drugs for weight loss finds low adherence and increased cost in first year. Published July 11, 2023. Accessed October 1, 2025. https://www.primetherapeutics.com/w/real-world-analysis-of-glp-1a-drugs-for-weight-loss-finds-low-adherence-and-increased-cost-in-first-year
- 24.Kaiser Family Foundation. Annual family premiums for employer coverage rise 7% to average $25,572 in 2024, benchmark survey finds, after also rising 7% last year. Published October 9, 2024. Accessed August 4, 2025. https://www.kff.org/health-costs/press-release/annual-family-premiums-for-employer-coverage-rise-7-to-average-25572-in-2024-benchmark-survey-finds-after-also-rising-7-last-year/
- 25.Tzang CC, Wu PH, Luo CA, et al. Metabolic rebound after GLP-1 receptor agonist discontinuation: A systematic review and meta-analysis. EClinicalMedicine. 2025;90:103680. doi: 10.1016/j.eclinm.2025.103680 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Horn DB, Linetzky B, Davies MJ, et al. Cardiometabolic parameter change by weight regain on tirzepatide withdrawal in adults with obesity: A post hoc analysis of the SURMOUNT-4 trial. JAMA Intern Med. Published online November 24, 2025. doi: 10.1001/jamainternmed.2025.6112 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Celletti F, Farrar J, De Regil L. World Health Organization guideline on the use and indications of glucagon-like peptide-1 therapies for the treatment of obesity in adults. JAMA. Published online December 1, 2025. doi: 10.1001/jama.2025.24288 [DOI] [PubMed] [Google Scholar]
- 28.Hwang JH, Laiteerapong N, Huang ES, Kim DD. Lifetime health effects and cost-effectiveness of tirzepatide and semaglutide in US adults. JAMA Health Forum. 2025;6(3):e245586. doi: 10.1001/jamahealthforum.2024.5586 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Pearson SD, Whaley CM, Emond SK. Affordable access to GLP-1 obesity medications: Strategies to guide market action and policy solutions. Institute for Clinical and Economic Review. Published April 9, 2025. Accessed August 6, 2025. https://icer.org/wp-content/uploads/2025/04/Affordable-Access-to-GLP-1-Obesity-Medications-_-ICER-White-Paper-_-04.09.2025.pdf [DOI] [PMC free article] [PubMed]
- 30.von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP; STROBE Initiative. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: Guidelines for reporting observational studies. J Clin Epidemiol. 2008;61(4):344-9. doi: 10.1016/j.jclinepi.2007.11.008 [DOI] [PubMed] [Google Scholar]
- 31.Suissa K, Schneeweiss S, Lin KJ, Brill G, Kim SC, Patorno E. Validation of obesity-related diagnosis codes in claims data. Diabetes Obes Metab. 2021;23(12):2623-31. doi: 10.1111/dom.14512 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Centers for Medicare & Medicaid Services. Medicare 2025 Part C & D Star ratings technical notes. Updated October 3, 2024. Accessed October 8, 2025. https://www.cms.gov/files/document/2025-star-ratings-technical-notes.pdf
- 33.Glasheen WP, Cordier T, Gumpina R, Haugh G, Davis J, Renda A. Charlson comorbidity index: ICD-9 update and ICD-10 translation. Am Health Drug Benefits. 2019;12(4):188-97. [PMC free article] [PubMed] [Google Scholar]
- 34.Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc. 1958;53(282):457-81. doi: 10.1080/01621459.1958.10501452 [DOI] [Google Scholar]
- 35.Congressional Budget Office. How would authorizing Medicare to cover anti-obesity medications affect the federal budget? October 8, 2024. Accessed August 6, 2025. https://www.cbo.gov/publication/60441
- 36.Dobbie LJ, Parretti HM, Fallows E, et al. Ten top tips for the management of GLP-1 receptor agonists in adults within primary care. Obes Facts. 2025;18(6):535-40. doi: 10.1159/000546472 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Gorgojo-Martínez JJ, Mezquita-Raya P, Carretero-Gómez J, et al. Clinical recommendations to manage gastrointestinal adverse events in patients treated with GLP-1 receptor agonists: A multidisciplinary expert consensus. J Clin Med. 2022;12(1):145. doi: 10.3390/jcm12010145 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Rao M, Shaughnessy F, Sokol R. Prescribing GLP-1 agonists for weight loss: Wrestling with our philosophical angst. Am Fam Physician. 2024;110(4):340-1. [PubMed] [Google Scholar]
- 39.Mozaffarian D, Agarwal M, Aggarwal M, et al. Nutritional priorities to support GLP-1 therapy for obesity: A joint advisory from the American College of Lifestyle Medicine, the American Society for Nutrition, the Obesity Medicine Association, and The Obesity Society. Obesity (Silver Spring). 2025;33(8):1475-503. doi: 10.1002/oby.24336 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Khan SS, Ndumele CE, Kazi DS. Discontinuation of glucagon-like peptide-1 receptor agonists. JAMA. 2025;333(2):113-4. doi: 10.1001/jama.2024.22284 [DOI] [PubMed] [Google Scholar]
- 41.Komé AM, Chandran MM, Tungate Lopez SS, Buse JB, Klein KR. One size does not fit all: Understanding microdosing semaglutide for diabetes in multidose pens. Diabetes Care. 2025;48(3):e25-7. doi: 10.2337/dc24-2575 [DOI] [PubMed] [Google Scholar]