A SELECTIVE AGEING EFFECT ON THE FRONTAL LOBE CONNECTIONS

Abstract

The study of the frontal connections is of particular interest for the neuroscience of aging. With aging the neuronal loss and small vessel alteration lead to progressive white matter damage associated with cognitive decline in the elderly (Pantoni, 2010). Cognitive decline affects predominantly executive functions, and brain changes seem to distribute unevenly, concerning predominantly the frontal region (Bishop et al., 2010). Previous studies using tract specific measurements revealed a slow decrease with aging in fractional anisotropy (FA) for frontal callosal tracts (Lebel et al., 2010) and for the average value of long tract connecting the frontal lobe (Jones et al., 2006) affecting more prominently the frontal portion of these tracts (Davis et al., 2009). Using voxel-based statistics, FA has been found to correlate negatively with aging, especially within the frontal lobes (Cabeza and Dennis, 2012). Whether this progressive decline affects the whole frontal white matter or specific tracts is unknown. Diffusion weighted datasets from 47 healthy volunteers aged 22-70 (M:F 24:23) were acquired on a 3T Siemens Verio TIM system equipped with a 32-channel head coil with the following parameters: voxel size 2x2x2 mm, matrix 128x128, slices 60, NEX 1, TE 90 ms, b-value 1500 s/mm2, 60 diffusion-weighted directions and 6 non-diffusion-weighted volumes, using a spin-echo EPI sequence. Cardiac Gating was applied with effective TR of 24 R-R intervals. Diffusion datasets were corrected for motion and eddy current distortions (Smith et al., 2004) and then processed with a Spherical Deconvolution algorithm based on a damped version of the Richardson-Lucy algorithm (Dell'acqua et al., 2010, Dell'Acqua et al., 2013). Tractography was performed following the method described in (Catani et al., 2012). We dissected fifty-five frontal tracts including U-shaped fibers. For each dissection FA (Basser and Pierpaoli, 1996) and hindrance modulated oriented anisotropy (HMOA) (Dell'Acqua et al., 2013) were extracted as an indirect measure of the tract integrity and correlated with the age of the participants regressing out the level of education. P values are presented after false discovery rate (FDR) correction for multiple comparisons (* p < 0.05 ; ** p < 0.01 ; *** p < 0.001). Aging was significantly associated with a decrease of FA (r = - 0.501*) and OA (r = - 0.508**) in the frontal projections of the corpus callosum. Aging was also associated with a decrease of OA in the right frontal lobe including the SLF I (r = - 0.401**) and SLF III (r = - 0.576***) branches of the superior longitudinal fasciculus, inferior fronto-occipital fasciculus (r = - 0.331*), fronto-thalamic projections (r = - 0.515***). In the left hemisphere, OA measure also decreased with aging for the frontal inferior longitudinal fasciculus (r = - 0.542***) and the frontal orbito-polar tract (r = - 0.542***). Results are summarized in Figure 1. We confirmed preliminary evidences reporting reduced integrity in the frontal portion of the corpus callosum associated with aging (Lebel et al., 2010). This commissural decline may explain the increased reaction times associated with aging reported in tasks requiring interhemispheric transfer (Reuter-Lorenz and Stanczak, 2000). Our results also suggest for the first time that aging alters significantly other tracts in the right hemisphere which brings up interesting pathophysiological hypotheses for ageing decline in visuospatial and verbal working memory, memory encoding and retrieval, reward-based associative learning that can be tested in the elderly (Cabeza and Dennis, 2012). Correlation between age of the participants and tract specific measurements, a) frontal corpus callosum b) first branch of the right superior longitudinal fasciculus; c) third branch of the right superior longitudinal fasciculus; d) right inferior fronto-occipital fasciculus; e) right fronto-thalamic projections; f) left frontal orbito-polar tract; g) left frontal inferior longitudinal fasciculus. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001 false discovery rate corrected for multiple comparison.