There are significant inconsistencies across cortical thickness and volume studies regarding the localization and extent of age effects. This is despite the revelation of use of effects on large areas of the mind by cross-sectional magnetic resonance imaging (MRI) system. These dissimilarities impede research on effects of aging on brain, and reduce the probable worth of MR in studies on age-related brain alterations.
The authors used six independent samples comprising of 883 participants to examine consistency of age effects on the thickness of cortical. Surface-based method of segmentation helped in computation of cortical thickness across the surface of the brain. Consistent age effects were revealed in various cortices across samples (Fjell, Westlye, Amlien, Espeseth, Reinvang, Raz,… & Walhovd, 2009).
Introduction
Many studies in MRI agree that as one ages, there is a reduction of brain volumes as well as enlargement of the ventricular system. However, there is growing proof that brain aging is significantly heterogeneous rather than homogeneous across brain regions. Regrettably, this heterogeneity is hard to analyse because of discrepancies between results.
Fjell et al. (2009) in their study endeavoured to tackle this predicament by testing the consistency of age effects on cortical thickness across six samples drawn from four research centers, with 883 participants. Fjell et al. (2009) argues that recent semiautomated and automated dissection methods have allowed studies of age effects continuously across the cortical mantle without manually defining ROIs. The study objective was to examine the consistency of age effects on regional cortical thickness across samples.
Method
The study involved 883 participants with an age range of 75 years. Scanning of all study subjects was by use of 1.5T magnets from two different manufacturers. The researchers used four different models of scanners. A single scanner generated data for both sample 4 and 5. For the Siemens scanners, T1-weighted sequences were acquired while GE used pulse sequences known as 3D spoiled gradient recalled (SPGR).
Slice thickness were between 1.25 mm (samples 4 and 5) and 1.5 mm (sample 1), with acquisition matrices of 256 3 192 (samples 1, 3, and 6) or 256 3 256 (samples 2, 4, and 5). The need to increase the strength of the signal-to-noise ratio (SNR) resulted in development of several scans within equal scanning time and then averaging for all samples.
General linear models (GLMs) were the statistical tools used to analyse the correlation between age and cortical thickness at each vertex across the entire cortical mantle in each independent sample. The foundation of the computation for the level of overlap between results from the different samples was the number of samples in which each of the P value thresholds was attained for each surface vertex. This information was color coded and projected onto a template brain. Next, all samples were included simultaneously in one GLM.
Results
The authors observed widespread age differences in cortical thickness across samples. However, the magnitude of effects varied among samples and brain regions. Results of using FDR < 0.05 as threshold revealed that there was a relationship between age and thinning of the cortex across nearly the whole brain surface.
However, bilateral thickening in the medial frontal cortex including anterior cingulate gyrus emerged in samples 4–6. Sample 3 had smaller age effects than in the other samples, but even in that sample, thinning was more noticeable in the prefrontal regions. Using a P value scale from 10–3 to 10–9 permitted assessment of regionally differential effects. The most strongly affected areas by age across all samples were the frontal cortices.
While some thinning was evident in four of the samples, age generally had more moderate effects on the medial-temporal cortices. Thickening in the medial frontal cortex emerged in the left hemisphere of samples 4 and 5 after using higher P value threshold. An area was a region of consistent age effects when effects emerged in five or all six of the samples.
Conversely, an area was an area of preservation when effects emerged in none or one of the samples. When FDR < 0.05 was used, large areas showed consistent age effects across studies, especially frontal cortices, where effects were seen in superior, middle, and inferior frontal cortices in all 6 samples.
MRI parameters
Note: FOV, field of view; FA, flip angle; TR, repetition time; TE, echo time; TI, inversion time
Source: Fjell et al. (2009)
Limitations
An important limitation of MRI studies of aging is the achievement of highest CNR possible. Lower CNR decreases the accuracy of the thickness estimation, and may cause variation across the cortical surface. Besides, the thin gap between hippocampus and the neighbouring cortical GM may make it difficult to locate the GM/WM surface around the insular and entorhinal cortical regions thus intensifying the inconsistency of the thickness estimates.
Data smoothing is a problem despite its role in reducing noise and thus improving reliability of the thickness estimates. As smoothing level increases, thickness measurement variability becomes smaller. Manual methods cannot reliably measure thickness because for proper measurements to be obtained, both the localization and the orientation of the white and pial surfaces must be known.
Discussion
The study investigated age effects on the cerebral cortex across multiple large samples enabling assessment of consistency. A considerable component of interstudy variability reduced because of using similar pre-processing procedures for all brains. One may conclude there is a relationship between progression in age and widespread thinning of the cerebral cortex.
Additionally, there was discrepancy of the size of age differences in thickness across cortical regions. Age effects were strongest in the prefrontal cortex, especially in superior, lateral, and medial regions. The other area that recorded heavy effects is the superior temporal gyri at the lower bank of the Sylvian fissure. Although there was generally good agreement among the different samples, some discrepancies cropped up.
For instance, in sample 3, the effects of age were weaker than in other samples. Across all or five of the six samples, large areas had consistent affects. The sensitivity to age effects may increase by using data from different studies and scanners, even in regions where consistency across samples is not impressive (Fjell et al., 2009).
Reference
Fjell, M.A, Westlye, T.L, Amlien, I, Espeseth, T., Reinvang, I., Raz, N,.… &Walhovd, B.K. (2009). High Consistency of Regional Cortical Thinning in Aging across Multiple Samples. Cerebral Cortex September, 19 (9), 2001—2012. doi:10.1093/cercor/bhn232