Correlations Between Cholesterol-Related Genetic Risk Scores and Lower Brain-Imaging Measurements of Regional Glucose Metabolism

Eric M. Reiman,Kewei Chen,Richard Caselli,Daniel Bandy,Wendy Lee,Gene Alexander,Dietrich Stephan,Andreas Papassotiropoulos
DOI: https://doi.org/10.1016/j.jalz.2005.06.281
2005-01-01
Abstract:In a positron emission tomography (PET) study of cognitively normal, late-middle-aged persons, we demonstrated a correlation between apolipoprotein E (APOE) ϵ4 gene dose, an established contributor to the risk of late-onset Alzheimer's disease (AD), and lower measurements of the cerebral metabolic rate for glucose (CMRgl) in AD-affected brain regions. We proposed the use of PET as a quantitative, pre-symptomatic endophenotype to evaluate the individual and aggregate effects of putative genetic and non-genetic modifiers of AD risk. In a genetic study of European AD cases and controls, Papassotiropoulos recently identified a cluster of single nucleotide polymorphisms (SNPs) in 7 cholesterol-related genes (APOE, SOAT1, APOE 5’-untranslated region, OLR1, CYP46A1, LPL, LIPA, and APOA4) associated with the risk of AD. He proposed the calculation of a person's cholesterol-related genetic AD risk scores from the risk-weighted summation of his or her implicated SNPs. To test the hypothesis that the proposed cholesterol-related AD genetic risk scores in cognitively normal, late-middle persons are correlated with lower CMRgl measurements in AD-affected brain regions. A brain mapping algorithm (SPM99) was used to compute significant correlations between cholesterol-related genetic risk scores and lower PET CMRgl measurements in 118 previously studied, cognitively normal, late-middle-aged persons. They included 25 APOE ϵ4 homozygotes, 35 ϵ4 heterozygotes, and 58 ϵ4 non-carriers and were 57±5 years of age. As predicted, significant correlations between cholesterol-related genetic risk scores and lower CMRgl were observed in each of the AD-affected brain regions; several of the correlations observed in the overall group were again observed in the aggregate APOE ϵ4 carrier group and in the ϵ4 homozygotes, heterozygotes, and non-carriers; significant correlations between the risk scores and higher CMRgl were not observed in any location. Furthermore, many of the observed correlations were again noted after excluding effects of APOE ϵ4 gene dose from the genetic risk scores. This study provides further support for the role of a cluster of cholesterol-related genes in the risk of AD, and it provides further support for the role of brain imaging in the pre-symptomatic assessment of putative modifiers of AD risk.
What problem does this paper attempt to address?