Genetic risk of dementia modifies the impact of obesity on limbic white matter and spatial navigation behavior in cognitively healthy adults

Abstract

A family history (FH) of dementia, APOE-e4 genotype, and obesity are major risk factors for developing Alzheimer's disease but their combined effects on the brain and cognition remain elusive. We tested the hypothesis that these risk factors affect apparent white matter (WM) myelin and cognition including spatial navigation and processing speed in 166 asymptomatic individuals (38-71 years). Microstructure in temporal [fornix, parahippocampal cingulum, uncinate fasciculus], motor and whole-brain WM was assessed with myelin-sensitive indices from quantitative magnetization transfer [macromolecular proton fraction (MPF)] and axon density from diffusion imaging. Individuals with the highest genetic risk compared to those with FH+ alone showed obesity-related reductions in MPF and axon density in the right parahippocampal cingulum. No effects were present for those without FH.Furthermore, FH modulated obesity-related effects on spatial navigation behaviour. In summary, an individual's genetic dementia risk influenced the impact of obesity on WM myelin and cognition. quantitative magnetization transfer, NODDI, white matter, myelin, parahippocampal cingulum 7 ResultsInsert Table 2 and Figures 2 and 3 here 2.1 Multivariate covariance analysis (MANCOVA) of white matter microstructural indices MANCOVA tested for the effects of APOE genotype (e4+, e4-), family history (FH+, FH-) and central obesity (WHR+, WHR-) on all microstructural indices (MPF, kf, R1, ISOSF, ICSF, ODI) in all white matter pathways (left PHC, right PHC, left UF, right UF, left CST, right CST and fornix) and WBWM regions whilst controlling for age, sex, and years of education. Omnibus effects: Main effects were observed for age [F(48,60) = 2.64, p < 0.001, hp 2 = 0.68] and sex [F(48,60) = 2.61, p < 0.001, hp 2 = 0.67]. Interaction effects were present between APOE and WHR [F(48,60) = 1.6, p = 0.04, hp 2 = 0.56] and between APOE, FH and WHR [F(48,60) = 1.7, p = 0.02, hp 2 = 0.58].Post-hoc effects: Ageing was associated with reductions in MPF, kf, and R1 in WBWM, fornix and right UF (Table 2). MPF and kf reductions were also present in the left UF and bilateral CST respectively. In addition, age-related increases in ISOSF were observed in WBWM and fornix and increases in ODI in the fornix (Table 2, Figure 2A).Women relative to men showed lower kf, ISOSF and ODI in WBWM and the fornix and larger fornix MPF (Table 2, Figure 2B). Two-way interaction effects between APOE and WHR were present for right PHC ( Figure 3A) and bilateral CST but additional post-hoc group comparisons to assess the directions of these effects were non-significant. Three-way interaction effects between FH, APOE and WHR were observed for MPF in WBWM, bilateral CST, bilateral UF and right PHC ( Figure 3B). For right PHC there was also an effect on ICSF. These three-way interaction effects were followed up by testing for APOE and WHR interactions in FH+ and FH-individuals separately ( Figure 3C). FH-individuals did not show any interaction effects B from the Alzheimer's Society and the BR...