Year: 2013
Type: letter
Abstract: I read with interest the report by Kumar et al published recently in this journal 1. The work used voxel-based relaxometry (VBR) methods and presented an extensive summary of the regional transverse relaxation time (T2) group average values and their correlation with age in more than 30 bilateral regions obtained from the brains of 60 healthy adults (32 women; ages 30–64 years). This work extends another recent report by the authors using VBR applied on 31 typically developing children and young adults (11 females, 7–24 years). Some aspects of these two works are important as they pertain to the experimental design of future clinical studies. For example, the authors did not compare left and right hemispheric values and information about handedness was not provided. Relaxation time has been reported to be lateralized 3, 4 and the data tabulated may also indicate some testable lateralization effects 1, 2. In addition, some interesting observations may be gleaned upon comparing and pooling the tabulated data presented on middle age and older adults 1 with the data collected from healthy children and young adults 2. For example, the genu or anterior corpus callosum (gCC) has a significantly smaller T2 relaxation time in children 2 compared to the posterior or splenium (sCC). This important callosal spatial heterogeneity trend, which is reported in previous reports on healthy children 5, 6, is apparently or paradoxically lost in healthy older adults 1. The finding of T2(sCC) ∼ T2(gCC) in healthy adults is in conflict with several reports 7-12, and this may be related to the averaging of data from adult subjects with a wide age range or due to the use of VBR, which requires spatial normalization. A close look at the data tabulated on healthy adults 1 shows that the T2 relaxation time vs. age demonstrates more scatter and may have a complex structure than a simple linear dependence on age. Therefore, the best-fit linear model adopted may be inadequate to capture the data scatter in the least-squares sense. The need for more healthy men and women is clear to facilitate curvilinear models. Pooling the data collected from all subjects would have provided some additional clues. To illustrate this effect we reconstructed and plotted the best fit data from the group mean and linear correlations tabulated by Kumar et al 1, 2 in Fig. 1. Figure 2 provides the best fit curve on several relevant left-hemispheric brain structures that include the hippocampus (HC), amygdale (AM), caudate (CN), putamen (PT), globus pallidus (GP), anterior cingulate, thalamus, gCC, and sCC. Figure 2 demonstrates clearly based on the ages 30–64 years 1 and 7–24 years 2 that T2 relaxation time is indeed curvilinear in 20+ adults, and hence simple linear models may be inadequate when adult subjects have a wide age range. Note the universal trend in healthy controls: T2 (HC) > T2(AM) > T2(CN) > T2(PT) > T2(GP) at all ages. The curvilinear model provides the expected smooth transition and plateau in middle ages. The curvilinear dependence of relaxation time on age has been previously documented across the healthy human lifespan 12-19. The use of simple group averages and comparisons in patient studies is clearly inadequate, as the relaxation time vs. age scatter is not random. It is important to include a sufficient number of healthy controls to test for gender and side, age dependencies along with their interactions with pathology 19, 20 and to study the yet unresolved interplay of relaxation rate with biophysical contributors such as myelin 10, 20, iron content 11, 13, 15, 21, and blood flow 22. Khader M. Hasan, PhD Department of Diagnostic and Interventional ImagingUniversity of Texas Health Science Center at HoustonHouston, Texas, USA E-mail: [email protected]
DOI: 10.1002/jmri.24134
Author: Khader M. Hasan
Language: English
Subfield: Radiology, Nuclear Medicine and Imaging
Field: Medicine
Domain: Health Sciences
Open Access status: bronze
OpenAlex ID: W1983103584