Compared to these monogenic forms, it is reasonable to assume that sporadic VaD might have a more complex mode of inheritance, in which multiple genetic variants with small effects predispose to disease (similar to other causes of dementia). Additionally, many of the risk factors for VaD, such as hypertension, dyslipidemia, and smoking habits, are partly genetically determined and this may further complicate the picture [
22]. Most of our current genetic understanding of VaD is based on candidate gene studies, while genome-wide associations studies (GWAS) have so far only contributed modestly.
Candidate gene studies
The candidate genes for association studies may encompass both genes implicated in stroke and Alzheimer disease and genes that affect critical biological processes to VaD. Given the multitude of candidate gene studies and the common failure of replicating the findings, we focus only on candidates that show consistent effect in two or more published studies.
From a pathophysiological view, the genes involved in lipid metabolism, especially the apolipoprotein E (
APOE), have been the focus of genetic research over the last two decades. Several meta-analyses, including a recent one which consists of 44 studies (2481 cases, 7490 controls), found a significant association between ε4 allele carriers and increased risk of VaD, irrespective of ethnicity [
23‐
25]. However, there was no difference in risk for VaD between ε2 and ε3 allele carriers nor in the
APOE promoter T-427C [
2]. Two polymorphisms (Q192R and L55M) of
PON1 were associated with a higher risk of VaD in Indian and French populations [
26‐
28], but these findings for Q192R were not confirmed in two other studies on European populations [
29,
30]. Several genes related to inflammation have also been related to VaD, including two polymorphisms in the
TNF-α in Europeans [
31] and Asians [
32]. In addition, an association between VaD and
TGF-β1 was described in two Asian reports [
33,
34]. Finally, the variant C677T of methylenetetrahydrofolate-reductase (MTHFR), which influences the level of homocysteine, was associated with VaD in Asians in three meta-analyses [
35‐
37], and more recently also showing stronger evidence in Europeans [
25].
However, candidate gene studies have not identified robust associations for various features of sporadic CSVD such as lacunar infarction, intracerebral hemorrhage, or white matter hyperintensities. Methodological issues, particularly related to inaccurate or heterogeneous phenotyping and insufficient sample sizes, have been invoked as possible reasons for this [
6].
GWAS
To identify genetic risk factors for VaD, two GWAS have been reported, firstly, a retrospective study in Koreans (84 VaD patients, 200 controls) that did not detect any genetic association [
38], and a prospective study in the Netherlands (67 patients, 5700 controls) that identified an association with the variant rs12007229 near the androgen receptor gene on the X chromosome, which was replicated in a German population (221 cases, 213 controls) [
39]. However, large-scale hypothesis-free studies are lacking, thereby hampering gene discovery in VaD.
A more recent line of research has been to consider genes that have been identified through GWAS of stroke and AD. With respect to stroke, the Cohorts of the Heart and Aging Research in Genetic Epidemiology (CHARGE) consortium [
40,
41], International Stroke Genetics Consortium (ISGC) [
42,
43], NINDS Stroke Genetics Network (SiGN) [
43], and Wellcome Trust Case Control Consortium 2 (WTCCC2) [
42] have conducted separate studies totaling tens of thousands of stroke patients and hundreds of thousands of controls. A variant near
FOXF2 was found to increase risk of all-stroke, and was also associated with white matter hyperintensities. Furthermore, multiple robust associations have been reported in an apparent subtype-specific manner –
TSPAN2 was associated with large artery atherosclerosis-related stroke,
PITX2 and
ZFHX3 with cardioembolic stroke,
HDAC9 with large artery atherosclerosis stroke, and
ALDH2 with small artery stroke. However, these genes have yet to be investigated for their relation with VaD. Currently, even larger GWAS on stroke are underway. Linking all these novel genes with VaD would crucially advance our knowledge of its genetic architecture. Similarly, the International Genomics of Alzheimer’s Project has identified various genes for AD through GWAS [
44], but these have yet to be associated with VaD. Interestingly though, not all these novel genes fit into known AD pathways involving amyloid or tau processing [
45]. Indeed, several genes have been implicated in cardiometabolic pathways, most notably
APOE,
CLU, and
ABCA7, which are related to lipid metabolism [
45].
In contrast to the candidate gene studies, GWAS have recently been carried out successfully to identify novel genes underlying CSVD. For white matter hyperintensities, for instance, five loci have been identified by studying over 21,000 individuals from the CHARGE consortium [
46]. Interestingly, genes at these loci are implicated in both AD and in (hemorrhagic) stroke [
46]. Earlier findings have been replicated in independent population-based [
47,
48] and clinical [
49] samples. The CHARGE consortium also investigated whether genetic influences could be detected for the progression of white matter hyperintensities over time [
50]. While a very small genetic contribution was identified, larger sample sizes and improved methods for assessing change could yield more results [
50]. GWAS of MRI-defined brain infarcts have been less robust, with an identified association in the
MACROD2 gene showing inconsistent effects [
51]. Nevertheless, whether the genetic associations with CSVD can be effectively reflective of those with VaD remains uncertain.
Instead of restricting genetic overlap to the top genes, an exciting alternative approach is to test for genetic overlap between clinical endpoints on a genome-wide scale. Novel methods, including linkage disequilibrium score regression [
52], now make such investigations possible. Currently, a study is ongoing for stroke and AD and, if the results are positive, this will provide important evidence of a causal link between these two clinical distinct entities, which overlap clinically in the realm of VaD.