Cognitive reserve hypothesis in frontotemporal dementia: A FDG-PET study

Abstract

Background and objective: Reserve is defined as the ability to maintain cognitive functions relatively well at a given level of pathology. Early life experiences such as education are associated with lower dementia risk in general. However, whether more years of education guards against the impact of brain alterations also in frontotemporal dementia (FTD) has not been shown in a large patient collective. Therefore, we assessed whether education is associated with relatively high cognitive performance despite the presence of [¹⁸F]-fluorodeoxyglucose positron-emission-tomography (FDG-PET) hypometabolism in FTD.

Methods: Sixty-six FTD subjects (age 67 ± 8 years) and twenty-four cognitively healthy controls (HC) were evaluated. Brain regions with FTD-related glucose hypometabolism in the contrast against HC and brain regions that correlate with the cognitive function were defined by a voxel-based analysis and individual FDG-PET values were extracted from all frontotemporal brain areas. Linear regression analysis served to test if education is associated with residualized cognitive performance and regional FDG-PET hypometabolism after controlling for global cognition.

Results: Compared to healthy controls, patients with FTD showed glucose hypometabolism in bilateral frontal and temporal brain areas whereas cognition was only associated with deteriorated glucose metabolism in the left temporal lobe. The education level was significantly correlated with the residualized cognitive performance (residuals from regression analysis between hypometabolism and cognitive function as a quantitative index of reserve) and also negatively correlated with left temporal FDG-PET hypometabolism after controlling for cognition.

Conclusions: In patients with FTD, the education level predicts the existing left temporal FDG-PET hypometabolism at the same cognition level, supporting the cognitive reserve hypothesis in FTD.

Keywords: Cognitive reserve; FDG-PET; Frontotemporal dementia; Hypometabolism.

Fig. 1

Surface projections of the right (on the left) and left hemisphere (on the right) in the FTD cohort (n = 66) A Regions with significant hypometabolism (p < 0.05, FWE-corrected, k > 50 voxel) in FDG-PET against healthy controls (n = 24) for the whole cohort and separately for bvFTD and PPA B Correlation cluster of FDG-PET with MMSE (p < 0.05, FWE-corrected, k > 50 voxel) for the whole FTD cohort. FDG, fluordesoxyglucose; FTD, frontotemporal dementia; FWE, family-wise error rate; PPA, primary progressive aphasia; bvFTD, behavioural-variant FTD; MMSE, mini-mental-state-examination; PET, positron-emission-tomography.

Fig. 2

A Correlation plots showing the residualized FDG-PET/ MMSE scores as a function of years of education for the whole FTD cohort and separately for the bvFTD and PPA subgroup. B Regression plots showing FDG-PET in the left temporal cluster as a function of years of education after control for age, sex and MMSE for the whole FTD cohort and separately for the bvFTD and PPA subgroup.. *p-value < 0.05. FDG, fluordesoxyglucose; MMSE, mini-mental-state-examination; PET, positron-emission-tomography; FTD, frontotemporal dementia; PPA, primary progressive aphasia; bvFTD, behavioural-variant FTD.

Fig. 3

Surface projections of the right (on the left) and left hemisphere (on the right) in the FTD cohort with significant hypometabolism (p < 0.05, FWE-corrected, k > 50 voxel) in FDG-PET against healthy controls (n = 24) separately for subjects with A lower education levels (n = 37) and B higher education levels (n = 29). FDG, fluordesoxyglucose; FTD, frontotemporal dementia; FWE, family-wise error rate; PET, positron-emission-tomography.