Skip to main content
Journal of Managed Care & Specialty Pharmacy logoLink to Journal of Managed Care & Specialty Pharmacy
. 2026 Aug;32(8):903–910. doi: 10.18553/jmcp.2026.32.8.903

Discontinuation of glucagon-like peptide-1 receptor agonists among US adults: A national survey

Michael J DiStefano 1,2,✉, Juliet Zon 3, Erik G Olson 4, Joseph F Levy 3, Gerard F Anderson 3, G Caleb Alexander 5,6
PMCID: PMC13403254  PMID: 42504813

Abstract

BACKGROUND:

Although many people discontinue glucagon-like peptide-1 receptor agonists (GLP-1 RAs), the reasons for discontinuation are unclear.

OBJECTIVE:

To identify self-reported reasons for GLP-1 RA discontinuation among US adults.

METHODS:

Cross-sectional survey of US adults fielded in June 2025 using Ipsos Omnibus and incorporating survey weights calibrated to census benchmarks. We included individuals who initiated a GLP-1 RA between January 2022 and June 2025 and subsequently discontinued therapy. Outcomes included duration of use, self-reported reason(s) for discontinuation, medication source, out-of-pocket payment, and insurance coverage. Descriptive statistics and chi-square tests for statistical significance were used.

RESULTS:

Of 7,035 individuals screened, 440 respondents met eligibility criteria and completed the survey. Of these 440 participants, 51.1% were female, 42.3% were aged between 18 and 34 years, and 23.8% reported having type 2 diabetes. More than half of these discontinuers (54.6%) reported discontinuing a GLP-1 RA within 6 months of initiation, and 79.7% reported discontinuation within 12 months of use. Approximately 1 in 7 (14.4%) reported sourcing GLP-1 RAs from a friend or family member, and more than half reported using a copayment coupon (35.4%) or free sample (22.0%). Many respondents (31.3%) reported first filling a prescription without insurance. Approximately 1 in 5 reported sourcing their medicine from a compounding pharmacy (9.8%) or Internet-based (9.1%) source. The most frequently cited reasons for discontinuation were cost (36.1%), side effects (33.3%), no insurance coverage (28.2%), and achieving a weight-loss goal (26.1%). Compared with those without diabetes, respondents with diabetes were significantly more likely to discontinue because of side effects (50.1% vs 28.1%, P = 0.0008). Respondents who reported sourcing their GLP-1 RA from a compounding pharmacy or Internet-based source were significantly more likely to discontinue because they perceived the medication to be ineffective compared with those who obtained their medication from a retail or mail-order pharmacy (17.5% vs 6.6%, P = 0.009). Neither diabetes status nor GLP-1 RA source was significantly associated with early discontinuation (<6 months).

CONCLUSIONS:

In this national survey, most US adults who stopped GLP-1 RAs did so within 6 months because of cost, adverse effects, lack of coverage, or having achieved a weight loss goal. Differences in coverage, costs, and sourcing are modifiable targets that may decrease high rates of GLP-1 RA discontinuation.

Plain language summary

Based on findings from a national, cross-sectional survey, many US adults who stop glucagon-like peptide-1 receptor agonist (GLP-1 RA) drugs do so within 6 months. The most common reasons are high cost, side effects, lack of insurance coverage, and achieving a weight loss goal. Better insurance coverage and lower costs may facilitate access to these products and help more people stay on treatment.

Implications for managed care pharmacy

This study suggests that early GLP-1 RA discontinuation is often driven by modifiable access barriers, especially out-of-pocket costs and insurance coverage. For managed care pharmacy, these findings support benefit designs that reduce cost sharing, improve coverage continuity, and limit reliance on nontraditional sources such as compounding pharmacies and friends and family members. Plans should consider nonpharmacologic weight maintenance programs for enrollees who have discontinued GLP-1 RAs after meeting their weight loss goals. Plan decisions should also be informed by the recent trajectory toward lower prices for GLP-1 RA medications in response to government intervention and increasing competition.


Obesity and type 2 diabetes mellitus remain major public health challenges in the United States. Glucagon-like peptide-1 receptor agonists (GLP-1 RAs), including GLP-1 and gastric inhibitory polypeptide dual agonists like tirzepatide, have emerged as highly effective treatments for both conditions, generating widespread clinical and commercial enthusiasm because of their positive impact on weight loss, glycemic control, cardiovascular health, and renal outcomes.1–4

Despite this enthusiasm, previous studies focused on GLP-1 RA use have reported 1-year discontinuation rates ranging from 36% to 65%, with rates at the higher end among individuals without type 2 diabetes.5–8 Other studies have identified 1-year persistence rates ranging from 40% to 61% among those using these drugs for weight loss.9,10 Although GLP-1 RA supply shortages that began in 2022 have contributed to access barriers,11 recent data suggest that treatment stopping remains common.12

To date, studies describing GLP-1 RA discontinuation or persistence have used electronic health records or administrative claims, providing little insight into the many potential reasons why so many individuals are stopping treatment. In addition, claims-based analyses and even electronic health records often fail to capture the experiences of individuals who access GLP-1 RA drugs without insurance, such as those turning to cash-only options or compounded GLP-1 RA products in response to access and affordability challenges,13 despite recent estimates that compounded products may account for as much as one-fifth of all GLP-1 RA use.14,15

To address these gaps, we conducted a national survey of US adults who had initiated and subsequently discontinued GLP-1 RA therapy within the past 3 years. We focused on the reasons for discontinuation and medication access experiences (eg, insurance coverage, out-of-pocket cost, and medication sourcing), with a goal of better informing pricing decisions, coverage policy, and clinical care.

Methods

STUDY DESIGN AND DATA SOURCE

We conducted a cross-sectional, Internet-based survey in June 2025 using the Ipsos Omnibus approach. In an omnibus survey, a single interview instrument combines questions from multiple independent client organizations; each client submits its own question module, and all clients share the common sociodemographic data collected during the interview. Survey responses were collected using a nonprobability sample drawn from Ipsos’ online panel, partner online panel sources, and river sampling.16 All respondents provided informed consent electronically prior to survey initiation. The study was approved by the Johns Hopkins Medicine Institutional Review Board.

SURVEY INSTRUMENT

The instrument, developed by clinicians, epidemiologists, and health policy experts in consultation with Ipsos, was refined through cognitive testing (Supplementary Exhibit 1 (263.2KB, pdf) , available in online article). It included items assessing insurance coverage for GLP-1 RAs, medication source, duration of GLP-1 RA use, out-of-pocket costs, and discontinuation reasons, using multiple-choice, ordinal, and free-text formats. These survey data were linked with basic sociodemographic information regarding each participant that is maintained by Ipsos as part of their survey panel.

STUDY POPULATION AND FIELDING

We defined eligible participants as individuals aged 18 years or older residing in the United States who reported starting a GLP-1 RA between January 2022 and June 2025 and who subsequently discontinued therapy. After explaining our goal and listing examples of GLP-1 RA medicines, we screened individuals using the following questions: (1) “Within the past three years, did a prescriber start you on a GLP-1 medicine?” and (2) “Are you currently taking a GLP-1 medicine?” Participants answering “Yes” to (1) and “No” to (2) qualified. The survey was fielded June 5 to 16, 2025. Data collection occurred over multiple daily recruitment waves, meaning respondents were not all interviewed simultaneously but were drawn from the panel in successive batches until the target sample was reached.

WEIGHTING AND REPRESENTATIVENESS

Ipsos calibrated respondent characteristics to be representative of the US population using raking-ratio adjustments.17 The source of these population targets is US Census 2021 American Community Survey data.18 The sample drawn for this study reflects fixed sample targets on age and gender. Post hoc weights were derived based on age, gender, region, and household income. We report results that incorporate these weights to approximate national estimates for US adults who discontinued a GLP-1 RA between January 2022 and June 2025.

STATISTICAL ANALYSIS

To assess the risk of nonresponse bias, we compared the demographics of eligible respondents with those who were eligible but abandoned the survey before completion using chi-square tests and a threshold for statistical significance of P < 0.05. All analyses were performed using Stata 18.0 (StataCorp, College Station, TX).

Before summarizing the data, we visually inspected each variable and confirmed there were no missing values. We used descriptive statistics to characterize the primary findings. Additionally, we compared reported reasons for discontinuation and the likelihood of early discontinuation (<6 months) by diabetes status and source of first fill (ie, retail/mail-order pharmacy vs compounding pharmacy/Internet-based source) using chi-square tests and a threshold for statistical significance of P < 0.05.

The precision of Ipsos online polls is measured using a credibility interval. In this case, the survey has a credibility interval of plus or minus 5.8 percentage points for all respondents. Ipsos calculates a design effect for each study based on the variation of the weights, following the formula of Kish.19 For this study, the sample of 440 qualified respondents has an unadjusted credibility interval of plus or minus 4.7 percentage points. After applying a typical design effect of 1.5, the adjusted credibility interval was plus or minus 5.8 percentage points.

We followed CROSS reporting guidelines for survey studies (Supplementary Table 1 (263.2KB, pdf) ).20

Results

CHARACTERISTICS OF RESPONDENTS REPORTING GLP-1 RA DISCONTINUATION

Of 7,035 individuals screened, 670 respondents met eligibility criteria and 436 completed the full survey. Eligible respondents who completed the full survey were younger, had higher household income, were more likely to have a college degree, and were more likely to be employed compared with eligible respondents who did not complete the survey (Supplementary Table 2 (263.2KB, pdf) ).

Among the 440 weighted respondents who discontinued a GLP-1 RA, 42.3% were aged 18 to 34 years and 51.1% were female (Table 1). Most reported annual household incomes under $100,000 (68.3%), approximately half (51.1%) reported a college degree, and nearly two-thirds were employed full-time (61.5%). The sample was composed of 56.4% identifying as non-Hispanic White, 17.0% as Hispanic, and 16.4% as non-Hispanic Black. Approximately three-quarters (76.2%) of respondents reported they did not have diabetes.

TABLE 1.

Characteristics of Respondents Reporting Glucagon-Like Peptide-1 Receptor Agonist Discontinuation (N = 440)

Characteristics %
Age, years
 18-34 42.3
 35-54 35.0
 55 or older 22.7
Sex
 Female 51.1
 Male 47.3
 Other/not reported 1.5
Household income
 Under $50k 24.5
 $50k to <$100k 43.8
 $100k+ 31.7
Education
 No college degree 48.9
 College degree 51.1
Employment status
 Full-time 61.5
 Part-time 10.4
 Not employed 15.9
 Retired 12.1
Race and ethnicity
 White (non-Hispanic) 56.4
 Black (non-Hispanic) 16.4
 Hispanic 17.0
 Asian 7.3
 Other race (non-Hispanic) 3.0
Diabetes
 Yes 23.8
 No 76.2

Because of rounding, percentages may not add to 100. Reported results incorporate survey weights.

SOURCING, OUT-OF-POCKET COST, AND COVERAGE OF GLP-1 RAs

Among adults who discontinued a GLP-1 RA, 14.4% reported using a medication they obtained from a nonprescriber source, such as a friend or family member (Table 2). More than one-third (35.4%) reported using a drug or copay coupon, and 22.0% reported receiving a free sample. Most respondents filled their first prescription at a retail pharmacy (60.7%), followed by mail order (eg, Express Scripts, CVS Caremark, OptumRx, Amazon) (12.4%), compounding pharmacies (9.8%), and Internet-based sources (eg, Hims/Hers, Ro, Zealthy, Levity, Mochi Health) (9.1%). A small proportion obtained their medication internationally (1.2%) or from another source (6.9%), with free-text responses indicating sources such as medical spas, weight loss clinics, or direct provision by a clinician.

TABLE 2.

Sourcing, Out-of-Pocket Payment, and Insurance Coverage for GLP-1 RA Use Among Respondents Reporting Discontinuation (N = 440)

Question %
Did you use a GLP-1 from another source than a licensed prescriber (eg, friend, family member)?
 Yes 14.4
 No 85.6
Have you ever used any of the following to help get your GLP-1 medicine?
 A drug or copay coupon 35.4
 A free sample 22.0
When you filled your first prescription for a GLP-1 medicine, how much did you have to pay out of pocket? (n = 282 a )
 $0 15.0
 $1-$24 14.6
 $25-$49 19.8
 $50-$99 17.5
 $100-$149 12.9
 $150 or more 20.3
When you filled your first prescription for a GLP-1 medicine, was at least some of the cost covered by insurance?
 Yes 57.7
 No 31.3
 Don’t know/not sure 11.0
When you filled your first prescription for a GLP-1 medicine, where did you fill it?
 A retail pharmacy (eg, CVS, Walgreens, Walmart, Rite Aid) 60.7
 A mail-order pharmacy (eg, Express Scripts, CVS Caremark, OptumRx, Amazon) 12.4
 A compounding pharmacy 9.8
 An Internet-based pharmacy (eg, Hims/Hers, Ro, Zealthy, Levity, Mochi Health) 9.1
 In another country (eg, Canada, Mexico) 1.2
 Otherb 6.9

Because of rounding, percentages may not add to 100. Reported results incorporate survey weights.

a

Limited to respondents who reported filling a 1-month supply.

b

Other sources provided in free-text responses included medical spas, weight loss clinics, or direct provision by a clinician.

GLP-1 RA = glucagon-like protein-1 receptor agonist.

Out-of-pocket costs for a 1-month supply varied widely at initiation; 29.6% paid less than $25, including 15.0% who paid nothing, whereas 20.3% paid $150 or more. Just under 60% reported that their first fill was at least partially covered by insurance (57.7%), whereas nearly one-third (31.3%) reported no insurance coverage for their first fill of a GLP-1 RA, and 11% were unsure.

PATTERNS AND CAUSES OF GLP-1 RA DISCONTINUATION

Among adults who discontinued a GLP-1 receptor agonist, 54.6% had used the medication for less than 6 months, including 20.2% who stopped within the first month (Table 3). Only 6.6% reported using the medication for 2 years or more. The most commonly reported reasons for discontinuation were high cost (36.1%), side effects (33.3%), lack of insurance coverage (28.2%), and achieving a weight loss goal (26.1%). Less commonly reported reasons included perceiving the medication to be ineffective (8.3%), coupons running out (6.4%), or the pharmacy being unable to refill a prescription (6.1%). Other reasons for discontinuation provided in free-text responses included pregnancy or family planning, burdensome clinical monitoring/follow-up requirements, and unrelated medical reasons (eg, cancer diagnosis).

TABLE 3.

Patterns and Reasons for GLP-1 RA Discontinuation Among US Adults (N = 440)

Question %
How long did you use your GLP-1 before discontinuing?
 Less than 1 month 20.2
 1-5 months 34.4
 6-11 months 25.1
 12-23 months 13.8
 24 or more months 6.6
Why did you stop your GLP-1 medicine? a
 High cost 36.1
 Side effects 33.3
 No insurance coverage for medication 28.2
 Weight loss goal achieved 26.1
 Medication ineffective 8.3
 Coupons ran out 6.4
 Pharmacy could not refill 6.1
 Otherb 7.4

Because of rounding, percentages may not add to 100. Reported results incorporate survey weights.

a

Response options were not mutually exclusive; respondents could select multiple responses.

b

Other reasons for discontinuation provided in free-text responses included pregnancy or family planning, burdensome clinical monitoring/follow-up requirements, and unrelated medical reasons (eg, cancer diagnosis).

GLP-1 RA = glucagon-like protein-1 receptor agonist.

Respondents with diabetes were significantly more likely to report discontinuing GLP-1 RA medication because of side effects compared with those without diabetes (50.1% vs 28.1%, P = 0.0008) (Table 4). Respondents without diabetes were significantly more likely to discontinue because they met their weight loss goal compared with those with diabetes (30.2% vs 12.9%, P = 0.0039). Other reasons for discontinuation did not differ significantly by diabetes status.

TABLE 4.

Reasons for GLP-1 RA Discontinuation Among US Adults by Diabetes Status and Source of First Fill

Question Diabetes status, n (%) GLP-1 RA source, n (%)
Participant reported having diabetes (n = 105) Participant reported not having diabetes (n = 335) Retail/mail-order pharmacy (n = 321a) Compounding pharmacy/Internet-based source (n = 83a)
Why did you stop your GLP-1?b
 High cost 39 (37.7) 119 (35.6) 109 (33.8) 34 (41.4)
 Side effects 52 (50.1)c 94 (28.1)c 103 (32.0) 34 (41.0)
 No insurance coverage for medication 32 (30.5) 92 (27.5) 90 (27.9) 19 (23.4)
 Weight loss goal achieved 14 (12.9)d 101 (30.2)d 93 (29.0) 20 (23.6)
 Medication ineffective 10 (9.9) 26 (7.8) 21 (6.6)d 15 (17.5)d
 Coupons ran out 5 (4.9) 23 (6.9) 22 (6.8) 5 (6.2)
 Pharmacy could not refill 10 (9.9) 17 (5.0) 16 (5.1) 5 (6.0)
 Othere 7 (7.0) 25 (7.6) 18 (5.5) 2 (2.3)

Reported results incorporate survey weights.

a

Excluding respondents who reported obtaining their GLP-1 medication in another country or from another source (“other”).

b

Response options were not mutually exclusive; respondents could select multiple responses.

c

P<0.001.

d

P<0.01.

e

Other reasons for discontinuation provided in free-text responses included pregnancy or family planning, overly burdensome clinical monitoring/follow-up, and unrelated medical reasons (eg, cancer diagnosis).

Respondents who reported sourcing their GLP-1 RA from a compounding pharmacy or Internet-based source were significantly more likely to report discontinuing because the medication was perceived to be ineffective compared with those who obtained their medication from a retail or mail-order pharmacy (17.5% vs 6.6%, P = 0.009) (Table 4). Other reasons for discontinuation did not differ significantly by medication source.

Neither diabetes status nor GLP-1 RA source was significantly associated with early discontinuation (<6 months) (Supplementary Table 3 (263.2KB, pdf) ).

Discussion

Although many people discontinue GLP-1 RAs, the reasons for this discontinuation are unclear. We conducted a cross-sectional, Internet-based survey of US adults who initiated and later stopped a GLP-1 RA. Most individuals who stopped GLP-1 RAs did so within 6 months of initiation because of cost, adverse effects, lack of coverage, or having achieved a weight loss goal. These findings help explain the basis for previously reported high discontinuation rates of GLP-1 RAs, and they suggest a variety of modifiable barriers to improve GLP-1 RA use.

Medication sourcing was heterogeneous, with substantial minorities of respondents reporting using nontraditional sources such as free samples, compounding or Internet-based sources, and personal contacts, potentially indicating efforts to remain adherent in the face of medication affordability or availability challenges. The varied sources individuals reported using to obtain GLP-1 RAs also reflect fragmentation within the pharmaceutical supply chain and the economic incentives that motivate parties such as compounding and Internet-based sources (many of which sell compounded formulations) to dispense this highly sought-after medicine at presumably a lower cost compared with the branded medications. Although the compounded market developed in response to GLP-1 RA drug shortages, ongoing litigation and recent US Food and Drug Administration warning letters sent to telehealth companies confirm this market remains robust despite the initial shortages having been resolved.21 Patients and clinicians should be aware of the potentially wide variety of product quality in the compounded GLP-1 RA market. For example, research describing the market for compounded GLP-1 RAs identified GLP-1 RA products compounded with B vitamins, additives with an unknown effect on medication safety or effectiveness.13 Concerns have also been raised regarding the source manufacturers of semaglutide bulk powder for use in compounding.22 Notably, survey respondents who reported receiving GLP-1 RA medications from a compounding pharmacy or Internet-based source were more likely to discontinue because of the perceived ineffectiveness of the drug, potentially reflecting these product quality issues. Individuals receiving GLP-1 RA medications from friends or family members may also face increased safety risks if they are using these medications without clinical guidance and ongoing monitoring. Given these findings, clinicians who prescribe these medications should discuss the potential risks of sharing doses with their patients.

Our findings have implications for patients, clinicians, and payors vested in improving care for obesity and diabetes. For patients, our findings underscore the importance of planning for long-term treatment—including budgeting and understanding coverage rules—and anticipating common adverse effects. In this study, respondents with diabetes were more likely to report discontinuing because of adverse effects. This finding possibly reflects the availability of other effective type 2 diabetes therapies with different mechanisms of action (eg, sodium-glucose cotransporter 2 inhibitors); because comparably effective alternative pharmacological therapies are not available for individuals using GLP-1 RAs for weight loss, these individuals may be more willing to tolerate adverse events. For clinicians, our findings demonstrate the importance of discussing total costs and access barriers at initiation,23 scheduling early follow-up for side-effect management, setting realistic expectations for outcomes, and counseling about when and how medication discontinuation may be achieved. This latter discussion is especially important given the finding that approximately one-quarter of survey respondents reported discontinuing GLP-1 RA medication after achieving a weight loss goal, with significantly higher rates of discontinuation for this reason among respondents who did not have diabetes. GLP-1 RAs are intended as long-term medications, and discontinuation of these medications has consistently been found to result in weight regain.24 For payors, predictable coverage with lower patient cost sharing for evidence-based use, streamlined authorization to reduce on-off cycling, and payment for longitudinal behavioral supports may improve GLP-1 RA continuity. These findings are important because they highlight actionable levers across the patient-clinician-payer ecosystem to reduce preventable discontinuation, enhance safety, and sustain clinical benefit at scale.

Looking ahead, cost-related discontinuation may lessen as manufacturers lower prices in response to governmental intervention and increasing competition, thus facilitating more generous coverage terms by payors. First, the Medicare Drug Price Negotiation Program has set a maximum fair price for semaglutide to take effect in 2027,25 a change that may exert downward pressure across the medication class and encourage broader coverage with possible spillovers to commercial and Medicaid markets. Additionally, the Trump Administration has pursued separate agreements with Eli Lilly and Novo Nordisk to lower the prices of these medications.26 Branded competition from orforglipron, a newly approved GLP-1 RA molecule,27 and other weight loss drug candidates in the pipeline28 may soon exert additional pressure on manufacturers to lower the prices of existing products. Whether these factors will translate to improved adherence and persistence will depend in part on payer choices, including formulary design; standardized, transparent coverage criteria; fewer disruptive reauthorization cycles; and payment for ongoing lifestyle management services such as dietary and exercise support. Importantly, more affordable pricing and expanded payer coverage will not address discontinuation that is unrelated to cost, such as stopping medication after achieving a weight loss goal.

LIMITATIONS

Our study has limitations. First, all measures were self-reported; social desirability bias, recall error, and misclassification of medication use are possible. Second, although our use of survey weights improves representativeness, residual selection and nonresponse bias may remain, and the survey was conducted online in English, which may underrepresent some populations. Third, we only enrolled individuals who had discontinued therapy, and thus, our study was not designed to estimate discontinuation rates or to compare experiences with current users. Fourth, our limited sample size resulted in imprecision around point estimates and prevented more extensive subgroup analysis. Finally, some respondents may have inaccurately reported the source of their medication. To increase our confidence in these responses, we confirmed that respondents who reported receiving GLP-1 RAs from compounding pharmacies and Internet-based sources were significantly less likely to report having insurance coverage for GLP-1 RAs and significantly more likely to spend a greater amount on a monthly supply than respondents who used retail and mail-order pharmacies (Supplementary Table 4 (263.2KB, pdf) ). Overall, these limitations suggest caution in interpreting point estimates but do not diminish the central implication that affordability, coverage, tolerability, and sourcing policies are key levers for improving continuity of GLP-1 RA therapy.

Conclusions

In this cross-sectional, Internet-based survey of US adults who stopped GLP-1 RAs, most individuals reported having done so within 6 months of initiation because of factors including cost, adverse effects, lack of insurance coverage, or having achieved a weight loss goal. Improving affordability, coverage, and sourcing are modifiable targets that may decrease high rates of GLP-1 RA discontinuation.

Disclosures

Dr Alexander is past chair of FDA’s Peripheral and Central Nervous System Advisory Committee and a co-founding principal and equity holder in Stage Analytics. These arrangements have been reviewed and approved by Johns Hopkins University in accordance with its conflict-of-interest policies. Mr Olson is an employee of Ipsos. There are no other disclosures to report.

This research was funded by Arnold Ventures. The funder had no role in the design, conduct, or interpretation of the study or in the decision to pursue peer-reviewed publication.

Acknowledgments

The authors gratefully acknowledge Ipsos for assistance with survey design and fielding. We used an AI-assisted writing tool (ChatGPT 4o, OpenAI; accessed July 2024) for language editing to improve clarity and readability. All authors reviewed and approved the text and take full responsibility for the content.

References


Articles from Journal of Managed Care & Specialty Pharmacy are provided here courtesy of Academy of Managed Care Pharmacy

RESOURCES