Sporadic cerebral small vessel disease and cognitive abilities
Hamilton, Olivia Kathryn Leigh
Cerebral small vessel disease (SVD) is a leading cause of vascular cognitive impairment, contributing to multiple neurological disorders ranging from stroke, to mild cognitive impairment and dementia. However, despite a huge number of studies on the subject, we have a limited understanding of how SVD affects cognitive ability. This thesis aims to address this knowledge gap, by examining domain-specific cognitive abilities in a range of clinical and non-clinical presentations of SVD. In the introductory chapters of this thesis I will discuss what is meant by the term cerebral small vessel disease (SVD), describing key radiological features of SVD and its varied clinical and non-clinical presentations. However, before considering the current consensus on how SVD impacts different domains of cognitive ability, I will first consider what happens to these abilities in the context healthy cognitive ageing. Finally, I will consider the current consensus on the pattern of cognitive changes that occur in SVD and will examine the vast and often conflicting evidence that underpins this. To gain a comprehensive overview of the published literature examining cognitive abilities in SVD, Chapter 4 presents a systematic review and meta-analysis of 69 studies presenting cognitive data for at least one cohort with SVD (n=3679) and one comparison control group without SVD (n=3229). Results indicated that relative to controls, cohorts with SVD performed more poorly on cognitive tests in all of the cognitive domains examined. Meta-regression analyses suggested that fewer years of education in the SVD vs. control groups accounted for a proportion of the differences in their test scores in some cognitive domains. Further meta-regression analyses suggested that cohorts with SVD-related cognitive impairment or dementia performed more poorly on tests in certain cognitive domains than cohorts with stroke or non-clinical presentations of SVD. Overall, however, SVD cohorts performed more poorly than controls on cognitive tests in all domains, regardless of their SVD presentation. Chapters 5 and 6 focus more closely on the key radiological markers of SVD and their associations with cognitive test scores using data from the Lothian Birth Cohort 1936 (LBC1936): a cohort of relatively healthy, community-dwelling, older individuals. To increase the fidelity with which SVD is typically measured, I combined computational volumes and visually-rated MRI markers of SVD to construct a variable representing the total MRI-visible burden of SVD. The study in Chapter 5 presents the results of cross-sectional associations between this latent SVD variable and latent variables of processing speed, verbal memory and visuospatial ability, within a structural equation modelling framework (SEM; n=540; mean age 72.6±0.7 years). Age, sex, vascular risk, depression status, and age-11 IQ were included as covariates. The latent SVD variable was negatively associated with all cognitive factors, in line with the results of the systematic review and meta-analysis. However, after accounting for the shared variance between the different cognitive domains (a construct described as general cognitive ability, which previous studies have not accounted for), only the association between the latent SVD variable and processing speed remained significant. This suggests that SVD’s association with slowed processing speed is not driven by, but is independent of its association with poorer general cognitive ability. In Chapter 6 this work is developed further by exploring associations between the latent SVD variable and decline in the same latent cognitive factors over a period of 9 years, from the age of around 73 to 82, again in the LBC1936. This was carried out using latent growth curve modelling within a SEM framework. Age, sex, vascular risk, and age-11 IQ were included as covariates. Results indicated that the latent SVD variable was associated with greater decline in general cognitive ability and processing speed. However, after accounting for the covariance between tests of processing speed and general cognitive ability, only the association between greater SVD burden and decline in general cognitive ability remained significant. Whereas the results of Chapter 5 suggested that SVD burden at age 73 may have specific and independent effects on processing speed measured at the same age, the results of our longitudinal analyses suggest that SVD burden at age 73 associates with declining processing speed due to SVD’s overarching association with general cognitive decline. In the final chapter of this thesis, I summarise the findings of these three studies, discuss their limitations, and make recommendations for future research.