Life expectancy with and without cognitive impairment in seven Latin American and Caribbean countries

Abstract

Background: The rising prevalence of cognitive impairment is an increasing challenge with the ageing of our populations but little is known about the burden in low- and middle- income Latin American and Caribbean countries (LAC) that are aging more rapidly than their developed counterparts. We examined life expectancies with cognitive impairment (CILE) and free of cognitive impairment (CIFLE) in seven developing LAC countries.

Methods: Data from The Survey on Health, Well-being and Ageing in LAC (N = 10,597) was utilised and cognitive status was assessed by the Mini-Mental State Examination (MMSE). The Sullivan Method was applied to estimate CILE and CIFLE. Logistic regression was used to determine the effect of age, gender and education on cognitive outcome. Meta-regression models were fitted for all 7 countries together to investigate the relationship between CIFLE and education in men and women at age 60.

Results: The prevalence of CI increased with age in all countries except Uruguay and with a significant gender effect observed only in Mexico where men had lower odds of CI compared to women [OR = 0.464 95% CInt (0.268 - 0.806)]. Low education was associated with increased prevalence of CI in Brazil [OR = 4.848 (1.173-20.044)], Chile [OR = 3.107 (1.098-8.793), Cuba [OR = 2.295 (1.247-4.225)] and Mexico [OR = 3.838 (1.368-10.765). For males, total life expectancy (TLE) at age 60 was highest in Cuba (19.7 years) and lowest in Brazil and Uruguay (17.6 years). TLE for females at age 60 was highest for Chileans (22.8 years) and lowest for Brazilians (20.2 years). CIFLE for men was greatest in Cuba (19.0 years) and least in Brazil (16.7 years). These differences did not appear to be explained by educational level (Men: p = 0.408, women: p = 0.695).

Conclusion: Increasing age, female sex and low education were associated with higher CI in LAC reflecting patterns found in other countries.

Publication
PLoS ONE 2015; 10(3):e0121867

Related