CAPITULO 2 MARCO DE REFEREN CIA
2.2. Marco de Referencia Teórico
number of regions of interest (ROI) is unlikely to clarify complex relationships between axonal injury and cognitive impairment after TBI, given the diffuse pathology. As the connecting white matter pathways of structures involved in cognitive processes sustain damage in TBI, reduced
efficiency of information transfer within functional brain networks may follow. It is also likely that structural abnormalities in white matter tracts have differential effects on cognitive functions depending on which specific tracts are damaged. Verbal learning and memory, executive function and information processing speed are all believed to depend on distributed network function. An ROI approach would thus critically limit the analysis of the structural causes of acquired impairments in these cognitive domains. These issues are compounded by our limited knowledge of how white matter tract structure relates to cognitive function in the normal brain, making it important to assess white matter abnormalities after TBI with as comprehensive spatial coverage as possible. However, a downside of a whole-brain approach is that statistical power to detect significant relationships can be weak, particularly if sample sizes are small and a large number of variables studied, but this naturally depends on the size of the effect studied and the variance in the measures of interest. This study is the first to use TBSS, a voxelwise whole-brain approach, to investigate how the structure of specific white matter tracts may relate to particular cognitive impairments after TBI.
Voxelwise analysis of white matter properties typically relies on scalar measures derived, for example, from a tensor model fit to diffusion-weighted MRI data (Jbabdi et al., 2010). Several voxel-based methods can be used, of which VBM has previously been used in TBI (Salmond, Menon, Chatfield, Williams et al., 2006). The problems associated with VBM- type analysis have been discussed in previous chapters, and Chapter 4 found tract-based spatial statistics (TBSS) to be sensitive to white matter abnormalities after TBI. As detailed in Chapter 2, TBSS uses nonlinear registration of each individual’s data that is optimised for the registration of diffusion data, and these data are subsequently projected onto an alignment invariant ‘skeleton’ representing the group’s average white matter tract structure. This has some clear advantages over the VBM method. Briefly, TBSS reduces alignment and registration errors between the individual and standard brain spaces and across participants. This reduces partial white matter volume effects, making the results more interpretable (Jones, Symms, Cercignani, & Howard, 2005; Smith et al., 2006). Finally, as opposed to restricting the analysis to a limited number of small ROIs, TBSS allows complex patterns of white matter disruption to be identified across the brain’s white matter tracts, and relationships of tract-specific abnormalities with specific cognitive functions to be studied.
5.1.6 Aims of the present study. Here, TBSS is used for the first time to investigate relationships between anatomically distributed white matter damage after TBI and impairments of verbal learning and memory, executive function, and information processing speed. Specifically, this study explores whether structural properties of particular white matter tracts (FA, MD, axial diffusivity and radial diffusivity) are correlated with cognitive functions within the three domains commonly impaired following TBI.
5.1.7 Hypotheses. It is hypothesised that greater white matter disruption following TBI will be associated with greater cognitive impairment, and that particular patterns of white matter disruption will be associated with distinct types of impairment. The following hypotheses are tested:
1) Structural abnormalities of hippocampal and prefrontal connections (more than of other tracts) will be correlated with worse associative learning and memory performance. 2) Abnormalities of medial temporal lobe white matter tracts and tracts interconnecting the
frontal and more posterior temporal and parietal cortices (more than of other tracts) will be associated with worse logical memory performance.
3) Abnormalities of frontal white matter tracts and tracts connecting the frontal to posterior medial/parietal regions (more than of other tracts) will be associated worse executive function.
4) Abnormalities of fronto-parietal, motor, and interhemispheric pathways (more than the integrity of other tracts) will be associated with slower information processing speed.
5.2 Methods and Materials
5.2.1 Design. A cross-sectional correlational study was carried out to investigate the relationships between DTI indices of white matter structure and measures of cognitive function in TBI patients and healthy controls.
5.2.2 Participants. The 28 post-acute/chronic TBI patients and 26 age-matched healthy controls who participated in the present study were the same as those in the study reported in
Chapter 4. The reader is therefore referred to the previous chapter for clinical details on each patient (p. 136) and summary demographic information about each group (p. 139). Chapter 2 described the procedures for participant recruitment, as well as the general and group-specific inclusion and exclusion criteria. All participants gave a written informed consent according to the Declaration of Helsinki (World Medical Association, 2008). The present study was approved by the Hammersmith, Queen Charlotte's and Chelsea Research Ethics Committee and the Departmental Ethics Committee of Goldsmiths Psychology Department.
5.2.3 Neuropsychological assessment. All participants completed a battery of neuropsychological tests sensitive to cognitive impairments associated typically with TBI. This has been fully described in Chapter 2. The specific tests and measures used in the current study are listed below.
5.2.3.1 Measures of general intellectual ability:
• Premorbid IQ: Wechsler Test of Adult Reading (WTAR; Wechsler, 2001), age- scaled score
• Current verbal and nonverbal reasoning ability: Similarities and Matrix Reasoning (WASI; Wechsler, 1999), age-adjusted T-scores
5.2.3.2 Measures of theoretical interest:
• Verbal (associative) learning and memory: People Test (Doors and People battery; Baddeley et al., 1994), immediate recall index
• Verbal learning and memory (logical/structured material): Logical Memory I (WMS- III; Wechsler, 1997), first recall total score
• Set-shifting ability: Trail Making Test (TMT; Reitan, 1958; Reitan & Wolfson, 1985), completion time for TMT-B minus completion time for TMT-A = ‘alternating switch- cost index’
• Cognitive flexibility/susceptibility to interference: Color-Word Interference test (D- KEFS; Delis et al., 2001), inhibition/switching completion time minus a baseline of
the average completion time for the color naming and word reading conditions = ‘Color-Word interference index’
• Word generation fluency: D-KEFS Verbal Fluency/letter fluency, total correct score for letters F, A and S
• Information processing speed: A computerised choice-reaction task (CRT; see Chapter 2 for details), median reaction time on correct trials
Subcomponents of the TMT and D-KEFS Color-Word Interference test tapped different aspects of processing speed (visual search: completion time for TMT-A; complex information: completion time for TMT-B; and speed of naming and reading: completion times for Color-Word Interference baseline conditions, color naming and word reading), but these were not measures of interest in the present analyses.
5.2.4 Structural imaging. The conventional T1- and T2*-weighted MRI sequences