Background: Stroke patients are at high risk of dementia. While cardiovascular comorbidities, such as hyperlipidaemia and atherosclerosis, have been linked to dementia in the general population, it is unclear whether the associations persist in stroke patients, who usually suffer more cardiovascular comorbidities. Statins have been recommended for secondary prevention in stroke patients, but it is unclear whether they have any cognitive effects separate from reducing risk of vascular events. Aim: To triangulate the potential roles of major blood lipids, atherosclerosis and statins in the development of dementia after stroke. Methods: First, I conducted a systematic review to identify the gaps in previous studies on the association of the three factors with post-stroke dementia. In response to the limitations of the previous studies, I conducted three cohort studies using the Clinical Practice Research Datalink (CPRD) for blood lipids, atherosclerosis and statin use, respectively. Given the limitations in the CPRD data, I further used UK Biobank to explore whether the CPRD results could be replicated using similar analytic strategies. Results: My systematic review suggested that hyperlipidaemia was not associated with post-stroke dementia, atherosclerosis was associated with increased risk of post-stroke dementia and post-stroke statin use may reduce the risk. Only one small-scale trial was available, and all other studies were observational with key methodological limitations. My CPRD study on major blood lipids showed that low-density lipoprotein (LDL) cholesterol was associated with increased risk of post-stroke dementia (adjusted hazard ratio [aHR] 1.29, 95% confidence interval [CI] 1.14-1.47, per one log-mmol/L increase), with a linear increasing trend in the risk across quintiles and different guideline targets of LDL cholesterol (all P-trend <0.001). Conversely, triglycerides were associated with decreased risk of dementia (aHR 0.79, 95% CI 0.70-0.90, per one log-mmol/L increase), with a linear decreasing trend in the risk across quintiles of triglycerides (P-trend <0.001). No significant association was found for high-density lipoprotein (HDL) cholesterol. My CPRD study on atherosclerosis suggested that stroke patients with prior atherosclerotic cardiovascular disease (ASCVD) were more likely to develop subsequent dementia (aHR 1.18, 95% CI 1.12-1.25). After full adjustment for potential baseline confounding, however, the risk of post-stroke dementia is attenuated (aHR 1.07, 95% CI 1.00-1.13), with no linear relationship between the age of ASCVD onset or duration of pre-stroke ASCVD and post-stroke dementia (all P-trend >0.05). My CPRD study on statin use found that statin initiation within the first three months after stroke was associated with lower risk of dementia (aHR 0.70, 95% CI 0.64-0.75). After accounting for non-persistence, the risk was further decreased (aHR 0.55, 95% CI 0.50-0.62). My UK Biobank study supported the CPRD findings. The observed associations tended to be stronger than from the CPRD, while mostly not statistically significant due to low incidence of post-stroke dementia in the younger volunteer population in UK Biobank.
Conclusions: Blood lipids may affect the risk of post-stroke dementia in different ways, with higher risk associated with LDL cholesterol, lower risk associated with triglycerides, and no association with HDL cholesterol. Stroke patients with prior ASCVD were more likely to develop subsequent dementia but the direct impact of pre-stroke atherosclerosis on the occurrence of post-stroke dementia may not be substantial. Statins have the potential to reduce the risk of dementia over and above future vascular events. Further trials and more real-world studies are needed to confirm the potential benefits of statins, to determine the optimal target of LDL cholesterol and to clarify the roles of triglycerides in the prevention of dementia.