Immigration timing modifies the association between medication non-adherence and cognitive function in older adults

Abstract

Background

The aging U.S. immigrant population exhibits higher rates of dementia than U.S.-born residents, yet traditional diagnostic methods often fail to detect decline due to cultural and linguistic barriers. Unintentional medication non-adherence (MNA) may serve as an early, noninvasive marker of cognitive decline, but little is known about how immigration timing and other modifiers shape this relationship.

Methods

We analyzed data from 10,854 participants aged ≥65 years in the Health and Retirement Study (2012–2016). Cognitive function was assessed using standardized latent cognitive scores. Immigration status was classified as U.S.-born or immigrant, with immigrants further stratified by age at arrival (child/adolescent, early, middle, and late adulthood). Unintentional and intentional MNA were investigated separately, and a moderated moderation regression on cognitive function tested interactions between unintentional MNA, immigration status/timing, and education adjusted for demographic, socioeconomic, and health covariates.

Results

Unintentional MNA, being an immigrant (compared to non-immigrants), and older age at immigration were associated with lower cognitive scores, while higher education was protective. Significant two-way interactions indicated that immigration status (β = −0.38 ± 0.14, p = 0.0006), timing (β = −0.13 ± 0.03, p < 0.0001), and education (β = −0.09 ± 0.03, p = 0.003) moderated the MNA-cognition association, though three-way interaction was not significant. Subgroup analysis showed that unintentional MNA was generally linked to poorer cognition, but the moderating roles of immigration timing and education varied across ethnicity-immigration subgroups. The potential for residual confounding and other study limitations should be considered when interpreting these results.

Conclusion

In conclusion, unintentional MNA is a sensitive marker of cognitive decline, with risk magnified among immigrants who arrived in middle-late adulthood. Integrating adherence and immigration history into clinical assessments may improve early identification and guide targeted interventions for diverse older adults.