Clinical and neuroimaging risk factors for cognitive decline in community-dwelling older adults living in rural Ecuador. A population-based prospective cohort study.

OBJECTIVE: There is limited information on factors influencing cognitive decline in rural settings from low- and middle-income countries. Using the Atahualpa Project cohort, we aimed to assess the burden of cognitive decline in older adults living in a rural Ecuadorian village. METHODS: The study included Atahualpa residents aged greater than or equal to 60 years who had a follow-up Montreal Cognitive Assessment (MoCA) repeated at least 1 year after baseline. MoCA decline was assessed by multivariable longitudinal linear models, adjusted for demographics, days between MoCA tests, cardiovascular risk factors, and neuroimaging signatures of structural brain damage. RESULTS: We included 252 individuals who contributed 923.7 person-years of follow-up (mean: 3.7 ± 0.7 years). The mean baseline MoCA was 19.5 ± 4.5 points, and the follow-up MoCA was 18.1 ± 4.9 points (P = 0.001). Overall, 154 individuals (61%) had lower MoCA scores at follow-up. The best fitted longitudinal linear model showed a decline of follow-up MoCA from baseline (ß: 0.14; 95% CI, 0.0-0.21; P < 0.001). High glucose levels, global cortical atrophy, and white matter hyperintensities were independently and significantly associated with greater MoCA decline. CONCLUSION: This study provides evidence of cognitive decline in older adults living in a rural setting. Main targets for prevention should include glucose control and the control of factors that are deleterious for the development of cortical atrophy and white matter hyperintensities.