Back

Social Determinants of Cost-Related Medication Nonadherence in the All of Us Cohort

Abegaz, T.; Sezgin, E.; Bhavnani, S. K.; Donneyong, M. M.

2025-06-01 epidemiology
10.1101/2025.05.31.25328711 medRxiv
Show abstract

BackgroundThis study aimed to examine the social determinants of health (SDH) associated with Cost-related medication nonadherence (CRMNA). MethodsA cross-sectional analysis was conducted using data from the All of Us Research Program to identify SDH features associated with CRMNA. CRMNA include inability to afford prescription medication, skipped doses, reduced dosage, selection of lower-cost alternatives, use of alternative therapies, purchasing medications from another country, and delayed prescription filling. A network analysis using the Fruchterman-Reingold algorithm was performed to visualize correlations among SDH features and CRMNA. SDH variables most strongly correlated with CRMNA were analyzed using binary logistic regression to estimate their association with each CRMNA, adjusting for demographic variables. ResultsAccording to the network analysis, the SDH features most frequently associated with CRMNAs include housing problems, worries about food insecurity, and experiencing poor attention by healthcare professionals. In the adjusted binary logistic regression model, Individuals who reported concerns about food not lasting were more than twice as likely to report overall CRMNA (adjusted odds ratio [AOR] = 2.29; 95% CI: 2.19-2.40; p < 0.05). Other significant SDH predictors of CRMNA included feeling unheard by a doctor (AOR = 1.59; 95% CI: 1.50-1.69; p < 0.05), experiencing housing problems (AOR = 1.40; 95% CI: 1.36-1.46; p < 0.05), and reporting perceived discrimination (AOR = 1.36; 95% CI: 1.32-1.40; p < 0.05). ConclusionThis study highlights the multifaceted impact of SDH on CRMNA, emphasizing the critical roles of food insecurity, housing problem, and negative healthcare experiences in shaping medication adherence behaviors.

Matching journals

The top 8 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.