Determinants of Cognitive Trajectories in Lower Educated Older Adults: Evidence from Brazil.

Abstract

AbstractBackgroundPatterns of cognitive change and modifiable factors for declining versus stable cognitive trajectories have rarely been described in a lower‐educated older population. Our goal was to identify heterogeneous long‐term trajectories of cognitive functioning and explore possible factors associated with cognitive decline beyond education, among older Brazilian adults.MethodWe used data on n = 1,042 adults aged ≥ 60 from the Health, Welfare and Aging (SABE) without cognitive impairment at baseline over four waves (2000‐2015) in São Paulo, Brazil. We applied group‐based trajectory modelling to identify cognitive trajectories. Cognition was measured using the abbreviated version of the Mini‐Mental State Examination (MMSE).ResultWe identified three cognitive groups: stable (n = 754), mild declining (n = 183), and strongly declining (n = 105). The probability of being in the strongly declining group membership was 10.7%. According to the MMSE classification, the mild declining group reached the average threshold for cognitive impairment on the last follow‐up, while the strongly declining group was on the second follow‐up. Furthermore, respondents in the stable trajectory were more likely to report more than primary education, to be white and married, to earn more than four times the minimum wage/month, to live in urban areas during childhood, and to practice vigorous physical activity at least three times a week. On the other hand, those in the strongly declining trajectory were more likely to identify as black or mixed, to self‐report subjective emptiness, and to have lower BMI, diabetes, and stroke. Moreover, both mild and strongly declining participants presented higher average age and lower MMSE scores at baseline compared to the stable trajectory. The multinomial logistic regression analysis revealed that respondents who self‐reported being mixed, had higher age, had primary education, had a stroke, and smoked at baseline were more likely to be in the strongly declining group.ConclusionOur findings suggest that besides education, interventions to reduce cognitive decline might include health and social policies addressing inequalities and modifiable risk factor burdens to improve later‐life cognitive functioning of at‐risk individuals. Sample attrition suggests estimates of the magnitude of the risk factors for cognitive decline to be conservative.