Association of Cognitive Function With Cause-Specific Mortality in Middle and Older Age: Follow-up of Participants in the English Longitudinal Study of Ageing

Abstract

We examined the little-tested associations between general cognitive function in middle and older age and later risk of death from chronic diseases. In the English Longitudinal Study of Ageing (2002-2012), 11,391 study participants who were 50-100 years of age at study induction underwent a battery of cognitive tests and provided a range of collateral data. In an analytical sample of 9,204 people (4,982 women), there were 1,488 deaths during follow-up (mean duration, 9.0 years). When we combined scores from 4 cognition tests that represented 3 acknowledged key domains of cognitive functioning (memory, executive function, and processing speed), cognition was inversely associated with deaths from cancer (per each 1-standard-deviation decrease in general cognitive function score, hazard ratio = 1.21, 95% CI: 1.10, 1.33), cardiovascular disease (hazard ratio = 1.71, 95% CI: 1.55, 1.89), other causes (hazard ratio = 2.07, 95% CI: 1.79, 2.40), and respiratory illness (hazard ratio = 2.48, 95% CI: 2.12, 2.90). Controlling for a range of covariates, such as health behaviors and socioeconomic status, and left-censoring to explore reverse causality had very little impact on the strength of these relationships. These findings indicate that cognitive test scores can provide relatively simple indicators of the risk of death from an array of chronic diseases and that these associations appear to be independent of other commonly assessed risk factors.

Keywords: aging; cancer; cardiovascular disease; cognitive function; mortality; respiratory illness.

Figure 1.

Selection of study participants, English Longitudinal Study of Ageing, 2002–2012. CVD, cardiovascular disease.

Figure 2.

Hazard ratios for the association of general cognitive function scores with total mortality rates, English Longitudinal Study of Ageing (n = 9,204), 2002–2012. Higher deciles represent higher cognitive function scores. Hazard ratios are fully adjusted for the covariates in Table 2. Bars, 95% confidence intervals.

Figure 3.

Hazard ratios for the associations of general cognitive function scores with category-specific mortality rates, English Longitudinal Study of Ageing (n = 9,204), 2002–2012. A) Deaths from cardiovascular disease (P for trend < 0.001); B) deaths from cancer (P for trend = 0.141); C) deaths from respiratory illnesses (P for trend < 0.001); and D) deaths from other causes (P for trend < 0.001). Higher quintiles represent higher cognitive function scores. Hazard ratios are fully adjusted for the covariates in Table 2. Bars, 95% confidence intervals.