According to our results, women in Chile aged 60 in 2016 could expect to live 0.8 years longer than women who were aged 60 in 2003. Men aged 60 in 2016 could expect to live 1 year longer than their 2003 counterparts. Chilean women at age 60 expected to live more years without CI, compared to men across all waves of the survey. For both men and women, there was an increase in the years expected to be lived free of CI and in the proportion of years free of CI between 2003 and 2016–17.
The increase in LE between 2003 and 2016 in Chile was lower than in Brazil [
10] and the United States [
23]. Mathers et al. [
24] have shown that mortality rates in old age in middle and high income countries, including Chile, have been consistently decreasing. Others have suggested that increases in LE at 50 years of age in Latin America are slowing, attributable to the impact of smoking [
25]. As observed in other studies [
9,
10,
26] we found that women at 60 years expected to live more years free of CI, compared to men. The greater LE free of CI among women, compared to men could be related to a later CI onset among older women [
27]. The number of years expected to be lived free of CI was similar to what has been reported by other studies in Latin America [
9,
10]. Studies from England [
26] and the United States [
23] report fewer years of LE free of CI at age 60 compared to our study. One explanation is that these studies included people aged 65 and over, at least 5 years older than the population in our study. Two studies from the United States [
23,
28] found a higher prevalence of CI in older adults than we did. Our sensitivity analyses suggest that this is attributable to differences in the way that CI was defined. We used the scores of the MMSE alone to determine CI. With this measure, prevalence of CI increased and LE free of CI was lower for men and women. However, sex differences and time trends were similar to those observed in our primary analysis. The combination of MMSE and PFAQ scores which we used are a more accurate screening tool of CI for the Chilean older population than MMSE alone [
19]. We found an increase in LE free of CI and a decrease in LE with CI between 2003 and 2016–17. It has been previously suggested that education level could have an impact on LE free of CI [
10,
29]. Garcia et al. [
30] found that more educated older adults, from different ethnic background, living in the United States were more likely to be free of CI. Nitrini et al. [
31] observed that the prevalence of CI among illiterate older people in Latin America was twice as high compared to literate people. In line with this, we observed an improvement in educational attainment among Chilean older men and women during this period and the risk of CI was significantly higher among people with less years of education. The effect of education on LE free of CI among the Chilean older population warrants further examination. Some limitations of this study must be pointed out. In Chile, there are few longitudinal health surveys and the ones that exist are limited to specific regions of the country [
32‐
34]. Only cross-sectional data were available to answer our research question. Our approach did not allow us to consider transitions between states of health or duration of each state of health. Instead, we based our calculations on prevalence at a specific point in time. Nevertheless, the Sullivan method is the most widely used method to estimate health expectancies, since population studies with longitudinal data are less common [
35]. Second, institutionalized people are not included in the Chilean NSH, which might mean we have overestimated LE free of CI. In 2002, the estimation was 1.6% in the city of Santiago [
36]. There are no updated estimations of the percentage of older people living in institutions. The first National Policy on Aging in Chile focused on extending family care, in order to delay entry into institutional settings [
37]. The current National Plan of Dementia [
38] emphasises family and community support. Hence, it is likely that the proportion of older people with CI who are institutionalised is low. Finally, the screening test for cognitive impairment was an adapted version of the MMSE in combination with the PFAQ. As discussed above, this makes international comparisons difficult. However, the cut-off point of the adapted version of the MMSE for the Chilean population has been validated [
18], and the combination of the MMSE and the PFAQ results have been established as a screening tool for CI with good sensitivity and specificity [
19]. Hence, our results provide an appropriate estimate of the years to be lived with CI among Chilean older people, to inform public health decisions. Our report also provides estimations of trends in LE with or without CI, which is an important strength of our study.
Unlike European countries and the United States, Latin American countries have recently started to estimate health expectancies among the older population. Our goal as a region should be to generate a similar set of knowledge to inform health policy decisions. This process calls for a joint effort, including researchers and public health authorities. It is necessary to identify existing data sources and to design population studies to gather information to estimate health expectancies on a regular basis, in order to complement the LE indicator with information about the health status of the population. It is important to determine and standardise specific measures of health status, to enable proper time-trend analyses and international comparisons.