Corpus callosum size, reaction time speed and variability in mild cognitive disorders and in a normative sample

Abstract

Intra-individual variability in reaction time increases with age and with neurological disorders, but the neural correlates of this increased variability remain uncertain. We hypothesized that both faster mean reaction time (RT) and less intra-individual RT variability would be associated with larger corpus callosum (CC) size in older adults, and that these associations would be stronger in adults with mild cognitive disorders. A normative sample (n = 432) and a sample with mild cognitive disorders (n = 57) were compared on CC area, RT mean and RT variability adjusting for age, sex, education, APOE genotype, smoking, alcohol consumption, grip strength, visual acuity, handedness and lung function. Samples did not differ in CC area or intra-cranial volume. In the normative sample, simple RT (SRT) and choice RT (CRT) were negatively associated with CC area but there were minimal associations between CC area and intra-individual RT variability. In the mild cognitive disorders sample, SRT, CRT and intra-individual variability on the SRT task were associated with CC area. Increased RT variability explained up to 12.7 percent of the variance in CC area in the sample with mild cognitive disorders, but less than 1 percent of the variance in CC area in the normative sample. There were no associations with APOE genotype. We conclude that intra-individual variability is associated with CC area in mild cognitive disorders, but not in normal aging. We propose that biological limits on reserve capacity must occur in mild cognitive disorders that result in stronger brain–behavior relationships being observed.

Introduction

The corpus callosum (CC), a thick band of white matter fibers, is the major interhemispheric commissure connecting the left and right neocortices. It interconnects cortical regions in one lobe with homotopic regions in the opposite lobe, and its size is related to the number of fibers available for bilateral transmission of information between cortical regions. CC size does not increase proportionately with brain size, and larger brains have relatively smaller mid-sagittal CC (Luders, Steinmetz, & Jancke, 2002). The mid-sagittal slice is commonly used as a measure of CC area, which is further segmented into different sub-regions, the three main regions being the anterior (including the rostrum, genu and anterior midbody), the midbody, and the posterior (isthmus and splenium: Aboitiz, Ide, & Olivarres, 2003). The anterior region (genu) connects the anterior cerebral hemispheres and the prefrontal cortical regions, which are crucial for motor functions (Mathias et al., 2004). Control of hand and finger co-ordination occurs in the contralateral hemisphere, and damage to the anterior CC has been shown to compromise control of bimanual hand and finger movements (Eliassen, Baynes, & Gazzaniga, 2000). The CC midbody connects to the motor, somatosemsory and auditory cortices (Aboitiz et al., 2003). The posterior structure connects the temporal, parietal and occipital lobes. The splenium is involved in interhemispheric communications along with auditory, visual information and language processes (De Guise & Lassonde, 2001).

Recent theories of cognitive aging based on results of functional neuroimaging studies suggest that interhemispheric transfer and activation differs in old compared to young subjects and may play a compensatory role (Cabeza, Anderson, Locantore, & McIntosh, 2002). It has been hypothesized that increased bilateral activation during cognitive testing among old participants is due to compensation for age-related deficits. If this is the case, then the efficiency of interhemispheric transfer may become especially important for performance on psychomotor tasks in old age. It could be hypothesized that larger corpus callosa would provide additional potential by compensating for the psychomotor slowing observed in late life, and contributing to brain reserve (Desmond, Moroney, Sano, & Stern, 2002).

Reaction time (RT) tasks have been used to investigate both interhemispheric transfer, and age related changes in cognition. RT on simple perceptual and decision tasks is a basic indicator of information processing speed (Snodgrass, Luce, & Galanter, 1967) and variability in RTs reflects processing fluctuations when cognitive processes are modeled as diffusion processes (Ratcliff & Smith, 2004). More recent findings and theories suggest that RT fluctuations or intra-individual variability could reflect non-robust information processing, which may be attributable to deficient neuronal functions (Anstey, 1999, Christensen et al., 1999; Hultsch, MacDonald, & Dixon, 2002; Li & Wu, 1999; Li et al., 2004; Li, Lindenberger, & Sikstrom, 2001). Patient groups with neurological conditions show increased variability in performance on reaction time tasks, lending support to the view that intra-individual variability is associated with pathological aging processes (Hultsch, MacDonald, Hunter, Levy-Bencheton, & Strauss, 2000).

There are few studies localizing the brain structures associated with intra-individual variability in performance within reaction time tasks. One recent functional neuroimaging study of intra-individual variability suggested that the pre-frontal cortex is activated during task-switching and that activation patterns are strongly influenced by trial to trial fluctuations (Braver, Reynolds, & Donaldson, 2003). It has also been proposed that attentional blocks or lapses (Bunce, Warr, & Cochrane, 1993; MacDonald, Nyberg, & Backman, 2006) may account for increased variability on RT tasks observed in later life. Banich (1998) argues that the CC plays an important role in attentional control through interhemispheric interaction by allowing for cognitive tasks to be carried out across the two hemispheres (Banich, 1998).

Complex RT tasks, being more dependent on interhemispheric transfer, would be expected to show stronger relationships with corpus callosum size or integrity of corpus callosum fibers. However, even simple RT tasks may involve interhemispheric transfer in order to inhibit responses, and there is evidence that unimanual movements are associated with CC area, even though this is a weaker association than that between bimanual movements and CC measures (Stancak, Cohen, Seidler, Duong, & Kim, 2003).

Recently, fractional anisotropy (FA), an indicator of white matter integrity, and mean diffusivity (MD; another measure of brain tissue change), obtained through diffusion tensor imaging, has been related to interhemispheric information processing speed (Schulte, Pfefferbaum, & Sullivan, 2004). Specifically, low FA is associated with larger differences in crosshemispheric versus non-crosshemispheric visuomotor processing speed, indicative of less efficient interhemispheric information processing (Schulte et al., 2004).

Accumulating evidence demonstrates that the CC size is significantly reduced in Alzheimer's disease (AD) as compared to healthy aging (Gootjes et al., 2006; Lyoo, Satlin, Lee, & Renshaw, 1997; Teipel et al., 2002; Thomann, Wustenberg, Pantel, Essig, & Schroder, 2006; Wang et al., 2005) and a few studies have shown significant differences in CC atrophy in patients with mild dementia, mild cognitive impairment (MCI) and cognitive complaints relative to healthy control subjects (Hensel et al., 2002, Teipel et al., 2002, Wang et al., 2005). FA has been shown to be lower and MD higher in the splenium of the CC for patients with AD as compared to healthy control subjects, indicating decreased structural integrity of the CC (Duan et al., 2006). It is possible that very early cognitive decline is associated with reduced integrity of callosal fibers, prior to an association being evident at the gross anatomical level.

It is also possible that decreased efficiency of interhemispheric transfer in AD, from either structural or functional changes to the corpus callosum or changes to neurotransmitter systems, is mediated by genetic factors, such as APOE genotype. Recently Persson et al. (2006) demonstrated a decrease in FA in the CC posterior region (isthmus and splenium) for APOE*E4 allele carriers compared to non-carriers (Persson et al., 2006). Another study conducted by Bartzokis et al. (2006) involving healthy participants 55–75 years of age found that age-related myelin breakdown in the genu region of the CC, but not the splenium, was associated with APOE genotype (Bartzokis et al., 2006). Individuals who were APOE*E4 carriers had an increased rate of myelin breakdown in contrast to carriers of the protective APOE*E2 allele. Based on the findings from these studies APOE genotype was included as a possible marker for CC size in the current study.

To date, most studies of behavioral or functional correlates of CC area have involved clinical samples or small experimental studies. To our knowledge, there has been no large study of the corpus callosum using a randomly selected community sample. The aim of this study was to evaluate the relationship between CC area and RT performance within a normative sample, and a sample of individuals meeting criteria for a number of mild cognitive disorders. Given that the integrity of the CC is related to interhemispheric information transfer and information processing speed in general, we expected that individual differences in CC area may be associated with information processing speed and its variability, and the association would be stronger in samples that are at risk of cognitive impairment or dementia. We expected that the associations would be strongest in the anterior regions of the CC as these connect the frontal cortices (Aboitiz et al., 2003) that are thought to govern consistency of responses to reaction time tasks. Health behaviors that affect both brain aging and cognitive performance (e.g., alcohol consumption and smoking) and biomarkers that indicate both brain and cognitive aging (APOE genotype, vision, pulmonary function and grip strength) were considered as potentially confounding variables. In addition, sex and handedness were investigated as covariates because they have been previously associated with differences in corpus callosum size and reaction time.