Microscopic tissue properties investigated by double diffusion encoding (DDE) experiments (or formerly called double wave vector experiments, DWV) in that two diffusion weighting periods are applied successively in a single acquisition have regained interest in the past years [1-6]. With a short mixing time $$$\tau_m$$$ between the two diffusion weightings, the experiment yields cell or compartment sizes [1, 5], on the other hand for a long $$$\tau_m $$$ the diffusion anisotropy present on a microscopic scale can be investigated [1], even in region-of-interests (ROIs) that appeared isotropic in a DTI experiment [6] demonstrating the additional information that DDE can provide compared to the widely-used diffusion tensor imaging (DTI). Recent studies showed that a measure of the microscopic diffusion anisotropy, the MA index, can be determined in the living human brain [7,8], and normal values and their variation in groups of young (< 33 y) as well as old (> 60 y), healthy volunteers have been reported [9,10]. The between group-comparison of MA and FA values yielded some promising age-related changes [11]. In this study, the formerly determined whole-brain MA and FA values in brain white matter were further analysed in order to investigate the differential cross-correlation with age within the aged group.Experiments were performed on a 3T whole-body MR system (TIM Trio, Siemens Healthcare) with a 32-channel head coil and have been presented recently [9,10,11]. 18 elderly, healthy volunteers (60 -79 y, 67.69 ± 4.65 y) were measured after their informed consent was obtained. DDE measurements were performed with spin-echo echo-planar imaging using an isotropic resolution of 3.0 mm (TE/TR = 150 ms/6.5 s) in 35 slices (gap 0.5 mm). The b value for each diffusion-weighting period was 500 s mm^-2, other timing parameter were a diffusion time $$$\Delta$$$ of 25 ms, a mixing time $$$\tau_m$$$ of 45 ms, and a gradient pulse duration $$$\delta$$$ of 22 ms. 96 combinations of 18 directions of the diffusion weighting were used to determine the MA [8]. The total scan acquisition time resulted in 11 min 10 s with six images without diffusion weighting acquired in-between. Four DDE measurements, one standard DTI experiment (60 directions, TE/TR = 100 ms/4.8 s, isotropic resolution 3.0 mm), and a T1-weighted anatomical measurement (MPRAGE, voxel size 1x1x1mm³) were performed for each volunteer. Motion-correction, coregistration and DARTEL [12] normalization were performed with SPM08 as data post-processing. The resulting MA and FA whole brain measures where analysed with age as a covariate.In Fig. 1, voxels in that the MA has a significant negative correlation with age (p < 0.001, uncorr.) in the old group (volunteers > 60 years) in all 35 slices, respectively, can be observed. For illustrative purposes, the statistical parameter maps (t-values) of the age-correlation were overlaid on the mean FA map, respectively. No WM clusters showing an increase with age where detected. The MA correlations reveal both, some larger coherent patterns in some WM ROIs such as the capsulae and the centrum semi-ovale and some differential changes in rather wide-spreaded and small ROIs. In Fig. 2, the same analysis is depicted for the FA (p < 0.001, uncorr. in the old group). It seems that the FA measure is relatively uncorrelated: very few and rather scattered voxels with a significant (negative) FA correlation can be spotted. Considering the recently determined global MA group differences, MA values were found smaller in the aged group in many typical WM ROIs which could be addressed to be consistent with axonal loss [11]. Similar spatial patterns of the corresponding significant changes can be viewed in Fig. 1 though they are much less distinct. It is very promising that in general, WM structures revealing a differential correlation with age also revealed an inter-group difference in the recent analysis. MA might therefore be a promising candidate as a sensitive marker of global tissue anisotropy changes. The FA on the other hand being extremely dependant on the current fiber orientation might in this case be less appropriate.DDE experiments may be able to provide information about the tissue microstructure that is complementary to that of DTI and, thus, could help to unravel the microstructural effects that are reflected by changes of the diffusion properties. With fields of implementation such as aging as well as pathologies, it seems that even for small sample sizes, the detection of highly significant structural differences is feasible.