A central aim in this project has been to evaluate the importance of vascular health for cognitive functioning in old age (i.e., from age 70 to death) using data from Swedish population-based representative longitudinal studies. The project has mainly consisted of reports from secondary data analyses of already collected data (although new data has also been collected and added to the analyses during the project).

Some of the main project findings include a robust non-convergence of between-person and within-person associations across serum total cholesterol and variety of measurements of cognitive functioning in old age, implying that many (if not all) previous reports of the associations between serum total cholesterol and cognition in old age are more or less biased—to what degree needs further evaluation (Thorvaldsson, Skoog, & Johansson, under review). These findings are in line with our previous reports concerning the associations between diastolic and systolic blood pressure and cognition in old age (see e.g., Thorvaldsson et al., 2012) and underline a methodological nuance that must be addressed in analyses with focus on the associations between vascular health and cognitive functioning in old age if the principal aim is to obtained unbiased parameter estimates of the strength and direction of the associations. As part of this project, we have outlined and demonstrated how this can be conducted and achieved using already existing longitudinal cognitive aging studies.

On a more theoretical notes, we found a consistent non-linear between-person association across serum total cholesterol and cognitive functioning, such that older individuals with relatively high but in particular low serum total cholesterol levels showed, on average, a poorer cognitive functioning in old age. Our findings therefore pinpoint toward an interval of serum total cholesterol levels that is associated with optimal cognitive function in old age. Our findings also revealed that decline in serum total cholesterol in the general population of older individuals is extensive, i.e., our data implies that more that 90% of the general Swedish population that reach the age of 70 are expected to show decline prior to death. Further, our findings revealed that the rate of this decline, i.e., in serum total cholesterol, is systematically associated with decline in cognitive functioning, implying a potential causal mechanism.

Yet, we found that the relative importance of vascular health for cognitive functioning in old age is somewhat reduced in cohort born in 1930, then in comparison with an earlier born cohorts in 1901/02 (Karlsson, Johansson, Skoog, Skoog, Rydén & Thorvaldsson, 2018). Although birth cohort differences in these effects sizes was found to be relatively small, they may imply an effect of substantial significant for future healthcare planning (then given the premises that the findings can be replicated in other studies).

Our findings have further revealed that those individuals with higher IQ (used as a proxy of cognitive/brain reserve) at age 70 can better compensate age-related neuropathology (e.g., of vascular origin) later in life in comparison to the less cognitively able by delaying the onset of accelerated decline prior to death (i.e., terminal decline) (Thorvaldsson, Skoog, & Johansson, 2017).

The project has stimulated to several new lines of research inquires with focus on the role of vascular health as a significant determinant of cognitive functioning, and dementia risk, in old age. One such line of inquires has led to research questions concerned with unwanted cognitively related side effects as associated with statin usage, and other cholesterol lowing drugs, that are common among older individuals in western societies. Other lines of inquires relates to the timing of the potential causal effect of vascular related impairment and how, according to our data, this seems to be non-invariant across the birth cohorts. Yet another line of research questions relate to the potential mediational role of vascular health for the association between brain reserve and onset of acceleration in cognitive decline prior to dementia diagnosis or death. On the methodological side, the findings derived from the project clearly emphasis the need for more data collections with a specific focus on within-person (i.e., intra-individual) variability, both short- and long-term, in the form of intensive measurements designs. Our findings underline that between-person associations are generally a poor, and potentially misleading, proxy of the within-person associations—which is of substantial theoretical and practical interest.   

All research reports from this project have been published in peer review professional journals (two papers are still under review) and introduced and presented at both national and international professional conferences. The project has been conducted in collaborations with several members from the AGECAP - Center of Aging and Health (see https://agecap.gu.se/english) research network at the University of Gothenburg and the international research network IALSA – Integrative Analysis of Longitudinal Studies of Aging and Dementia (see https://www.maelstrom-research.org/mica/network/ialsa#/).