GLP-1 Medication Persistence Among Participants in a Virtual Weight Management Program
Susan M, Devaraj, PhD, MS, RD; Jeanean B. Naqvi, PhD; Hope Chang, PharmD; Sarah Linke, PhD, MPH
Abstract
Background
GLP-1 and GIP/GLP-1 receptor agonists have demonstrated efficacy in chronic weight management, promoting weight loss of 10% or more on average after about 6 months of use. However, real-world evidence has indicated variable medication persistence, with an estimated 33-70% persisting to 24 weeks. We evaluated medication persistence and its association with weight loss among adults without diabetes enrolled in a virtual weight management or hypertension program.
Methods
Eligible members had initiated GLP-1 treatment between 60 days prior to and 30 days following opting into an Enhanced GLP-1 Care Track, which provided care-team supported, GLP-1-specific behavioral lifestyle guidance. Members were considered persistent on their medication until pharmacy claims data indicated the absence of a medication refill ≥60 days following their last prescription’s supply. We evaluated medication persistence rates through 12 and 24 weeks in the program by medication type, demographic variables, and program engagement using Fisher’s exact tests and logistic regression models, and weight loss by persistence using a t-test.
Results
Participants (n=1,124) were 71.0% white, 83.7% female, and had an average baseline BMI of 38.8 kg/m2 and age of 46.4 years, with 57.0% taking tirzepatide and 36.9% taking semaglutide. Participants who persisted on medication through 24 weeks lost more weight compared to those who were persistent for <24 weeks (-12.1% vs. -7.4%, p<.001). Overall, 93.9% of participants were persistent with GLP-1 medication use through 12 weeks, and 84.1% of participants were persistent through 24 weeks. Participants taking tirzepatide were more likely to persist than those taking semaglutide through 12 weeks (96% vs. 91% p<.01) and 24 weeks (88% vs. 77%, p<.01). Additionally, participants who were older (ps<.01) or had a higher household income (ps<.05) were more likely to persist with their medication through 12 and 24 weeks than those who were younger or had a lower household income. The odds of persisting per 10 additional weekly engagement actions were 26% higher (95% CI: 1.05-1.52) in the unadjusted model and 54% higher (95% CI: 1.16-2.10) after controlling for household income, race/ethnicity, and age through 12 weeks. Through 24 weeks, engagement remained associated with persistence in the unadjusted model (OR 1.14, 95% CI 1.02-1.28). After adjustment, there was no significant difference between white and Black members in odds of persistence, but Hispanic members (OR 0.49, 95% CI 0.26-0.96) and members of another race/ethnicity (OR 0.39, 95% CI 0.20-0.80) had lower odds of persistence than white members.
Conclusions
Rates of medication persistence were higher than estimates from published real-world evidence. Program engagement may be especially meaningful in promoting persistence during medication initiation and titration, and future research should explore how to promote more equitable medication use as indicated to drive clinical outcomes.
