Dementia and its comorbidities : genetic and epigenetic influences

Sammanfattning: Dementia is a multifactorial disorder of late life, characterized by memory deficits, personality changes, and impaired reasoning abilities. There is considerable co-morbidity between dementia, cardiovascular disease (CVD), and late-life depression, but the nature of the associations remains elusive. We therefore seek to investigate how genetic and epigenetic factors act, independently and in concert, to contribute to dementia as well as to its association with CVD and depression. The first two studies focused on what role specific genes play in the association between dementia, depression, and CVD. In study I, we investigated how apolipoprotein E (APOE) genotype affects the association between depression and dementia, and whether the timing of depression onset is of importance. Utilizing a nested case-control design with 804 dementia cases and 1,600 matched controls, we found that depression within ten years of dementia onset was associated with disease regardless of APOE genotype, while depression more distal to dementia was a risk factor only in carriers of the ε4 risk allele. Study II focused on the shared genetic architecture between dementia and CVD, and entailed two parts. In the first part we used data from 13,231 Swedish twins, and found that genetically predisposed CVD was a stronger risk factor for dementia compared to CVD with a lower genetic risk. In the second part of the study we utilized summary statistics from previously published genome-wide association studies to investigate the genetic overlap between Alzheimer´s disease (AD), the most common form of dementia, and coronary artery disease. We found no evidence of genetic overlap between the disorders, but that both diseases have a significant number of genes in common with lipid levels. The last two studies focused on epigenetic factors and investigated how gene specific methylation is associated with dementia. Study III focused on the APOE gene, and how methylation levels in leukocytes relate to the risk of dementia, AD, and CVD. Using data from 447 Swedish twins, we demonstrated that hypermethylation in the promoter region of the gene was associated with dementia and AD, but not with CVD. Results were similar within discordant twin pairs, and did not differ as a function of APOE genotype. In study IV, we focused on five other AD related genes that are differentially methylated in post-mortem brain samples from AD patients compared to controls. The aim was to investigate whether these differences could also be detected in blood samples collected pre-mortem. There was a significant difference in methylation of SORL1 in leukocytes from dementia patients and of BIN1 in leukocytes from AD patients. Findings were stronger in discordant twin pairs, indicating that the association cannot be attributed to genetic factors. In conclusion, the studies included in this thesis highlight the complexity of late-life comorbidities, and the importance of taking both genetic factors and the timing of disease into account when studying these associations. Furthermore, methylation of genes related to AD is of importance for dementia, and has the potential to serve both as a biomarker and identify mechanisms of disease development.