Elsevier

Cortex

Volume 47, Issue 4, April 2011, Pages 441-450
Cortex

Research report
Cerebellar brain volume accounts for variance in cognitive performance in older adults

https://doi.org/10.1016/j.cortex.2010.01.001Get rights and content

Abstract

Introduction

Frontal lobe atrophy is implicated in patterns of age-related cognitive decline. However, other brain areas, including the cerebellum, support the work of the frontal lobes and are also sensitive to the effects of ageing. A relationship between cerebellar brain volume and cognitive function in older adults is reported, but no study has separated variance associated with cerebellar gray matter volume and cerebellar white matter volume; and no study has examined whether or not brain volume in the cerebellum is related to cognitive function in older adults after statistical control for frontal lobe volume of gray and white matter.

Method

We used voxel based morphometry (VBM) and structural equation modelling (SEM) to analyse relations between general cognitive ability (G) and volume of GM and WM in frontal areas and cerebellum in a sample of 228 older adults (121 males and 107 females).

Results

Results indicate that GM volume in the cerebellum predicts G, even when total intracranial volume (TICV) and GM gray and WM volumes in frontal lobes are statistically controlled. However, results differ for males and females, with males showing a stronger relationship between brain volume in the cerebellum and G.

Conclusions

Results are discussed in light of neurological models of cognitive ageing and the significance of the cerebellum in models of cognitive functioning.

Introduction

There is mounting evidence that cognitive functions are supported by the neocerebellum containing the phylogenetically more recent cerebellar regions (Desmond and Fiez, 1998, Cabeza and Nyberg, 2000, Timmann and Daum, 2007, Ito, 2008). Functional imaging research provides evidence for cerebellar activity during a broad range of cognitive activities including: rule-based word-generation (Petersen et al., 1989), mental imagery (Ryding et al., 1993, Parsons et al., 1995), cognitive flexibility (Kim et al., 1994), sensory discrimination (Gao et al., 1996), motor learning (Jenkins et al., 1994, Rauch et al., 1995), verbal memory (Grasby et al., 1994), and working memory (Klingberg et al., 1995, Hayter et al., 2007).

Neuroanatomical evidence suggests that the pre-frontal cortex and lateral cerebellum are the last structures to develop during maturation and the first to undergo involution in later life (Arenberg, 1987, Haug, 1991, Allman et al., 1993, Grady et al., 1999, Raz et al., 2001). Potentially, the cerebellum works in conjunction with the frontal and pre-frontal cortex in the modulation of behaviour (Schmahmann, 1996) but the mechanism of such possible interaction is less clear (Timmann and Daum, 2007, Ito, 2008). The frontal lobes are thought to support executive control processes, that is, a sub-set of cognitive skills requiring effortful attention and the ability to manage one's thoughts, memories and actions in accordance with task-relevant goals (Stuss and Benson, 1984, Chudasama and Robbins, 2006, Mohr et al., 2006, Stuss and Alexander, 2007). Executive function declines throughout middle to late adulthood and frontal lobe atrophy may play a central role in these patterns of age-related cognitive decline (West, 1996). The cerebellum may support the resource and capacity base of the frontal lobes by modulating the speed, variability, and automaticity of information processing operations (Leiner et al., 1989, Leiner et al., 1991, Ito, 1993, Schmahmann, 1998, Schmahmann, 2004, Hogan, 2004, Ramnani, 2006), which in turn may facilitate executive control and general cognitive ability (G) (Salthouse, 1996; Anderson et al., 1998).

In order to understand how the cerebellum might work in conjunction with the frontal lobes in the modulation of behaviour, it is potentially useful to examine the effect of ageing on both the frontal lobes and the cerebellum and to model these effects conjointly when examining brain-behaviour relations. While this strategy of analysis can extend to other brain areas that play a role in patterns of ageing cognition (e.g., the temporal lobes), the current study focuses on G and the hypothesised role of cerebellar-frontal lobe associations with G (Hogan, 2004).

Substantive clues about the importance of connectivity and interaction between several cortical and sub-cortical sites come from studies of sub-cortical dementia (Albert et al., 1974, Filley et al., 1989, Rao et al., 1989, Grafman et al., 1990, Junque et al., 1990, Gunning-Dixon and Raz, 2000), but little is known about the relationship between cognition and cerebellar volumes in healthy older adults. In a notable exception, MacLullich et al. (2004) observed significant correlations between size of the neocerebellar areas of the vermis and several cognitive tests in a sample of healthy elderly men. However, they did not assess the relative contribution of brain volume in frontal regions in accounting for cognitive performance, nor did they separate out the effects of gray matter (GM) volume and WM volume in their analysis – an important distinction in light of research demonstrating a relationship between WM atrophy and cognitive decline (Gunning-Dixon and Raz, 2000, Murray et al., 2005, Medina et al., 2006). Furthermore, MacLullich et al. (2004) examined the relationship between brain area (not volume) in the cerebellum and cognition in males only and, in light of recent evidence pointing not only to a reduction in the size of the cerebellum over the adult lifespan but also to gender differences in the size of the cerebellum (Raz et al., 2001) and gender differences in the rate of cell loss in the cerebellum with age (Xu et al., 2000), we argue that it is important to model the relationship between frontal lobe volume, brain volume in the cerebellum and cognitive performance in both males and females separately. Specifically, research suggests that the cerebellum shrinks proportionally more with age in men than women (Xu et al., 2000). Therefore, in samples of healthy older males and females matched on age, education, and childhood intelligence, we might expect that any observed relationship between cerebellar volume and G would be stronger for males than females, that is, after statistically controlling for gender differences in total intracranial volume (TICV).

In the present study, we used voxel based morphometry (VBM) and structural equation modelling (SEM) to analyse relations between G and volume of GM and WM in both frontal lobes and the cerebellum in a large sample of males and females (N = 228) who are participating in the Aberdeen Longitudinal Study of Cognitive Ageing (Murray et al., 2005). Males and females were matched on age, education, and childhood intelligence. We used VBM to first establish whether or not there is any relationship between brain volume and G. We then used Multi-Group SEM to examine the direct and indirect effects on G of both GM and WM brain volume in the cerebellum and the frontal lobes while controlling for TICV. We used several indicators of G and cerebellar volume in our SEM analysis, and as part of our analysis we tested for measurement invariance for G, across both males and females. We predicted gender invariance in G and thus specified an a priori multi-group measurement model with a single latent variable and equal factor loadings for males and females. Also, given the conclusion by Raz et al. (2001) that age-related reduction in the cerebellum is uniform, we defined and tested two a priori measurement models that used a single latent variable to represent cerebellar gray matter (Cg) volume and cerebellar white matter (Cw) volume, again with equal factor loadings for males and females (see SEM Results section, Models 1–3). In Model 4 we combined the three measurement models into a fuller path analysis which specified a priori relations between the core independent and dependent variables of interest. Consistent with the findings of MacLullich et al. (2004), we hypothesised a direct effect on G of brain volume in the cerebellum, and we expected this effect to be stronger in males when compared with females. As part of our effort to ensure rigorous testing of the relationship between cerebellar brain volume and G, our VBM analysis examined the relationship between G and brain volume across the whole brain (i.e., without reference to any a priori hypothesis and by using statistical controls that correct for multiple comparisons in the analysis of brain-behaviour relations). Similarly, in our SEM analysis of the effect on G of brain volume in the cerebellum we controlled for both brain volume in the frontal lobes and TICV. Finally, in light of the role of the frontal lobes in age-related reduction in G (West, 1996, Hedden and Gabrieli, 2004), we remained open to the possibility that any effect of the volume of cerebellum on G would vanish in the context of statistical controls for frontal lobe volume.

Section snippets

Participants

Between 2000 and 2001, we recruited, from the local community, volunteers who were all born in 1936. All had participated in the Scottish Mental Survey of 1947 when about 95% of children born in 1936 and at school in Scotland on 1st June 1947 were tested. 986 of the 1823 Aberdeen city children, who sat the test in 1947, were traced to an address in North East Scotland. The 986 who were traced were linked to local general medical practices. General practitioners were selected at random to help

Results

Descriptive statistics and gender difference t-test statistics are reported in Table 1.

Discussion

The findings of the current study extend those of MacLullich et al. (2004), who observed a relationship between the neocerebellar vermian areas and cognitive performance. The results of our multi-group SEM analysis revealed a significant relationship between cerebellum GM volume and G in men but not women. The VBM analysis also revealed a relationship between GM and G in men but not women. VBM analysis also identified several GM regions associated with G in the full sample that are in part

References (69)

  • M.J. Hogan

    The cerebellum in thought and action: A fronto-cerebellar aging hypothesis

    New Ideas in Psychology

    (2004)
  • M.J. Hogan et al.

    Electrophysiological and information processing variability predicts memory decrements associated with normal age-related cognitive decline and alzheimer's disease (ad)

    Brain Research

    (2006)
  • G.B. Karas et al.

    Global and local gray matter loss in mild cognitive impairment and Alzheimer's disease

    NeuroImage

    (2004)
  • H.C. Leiner et al.

    The human cerebro-cerebellar system: Its computing, cognitive, and language skills

    Behavioural Brain Research

    (1991)
  • A.M.J. MacLullich et al.

    Size of the neocerebellar vermis is associated with cognition in healthy elderly men

    Brain and Cognition

    (2004)
  • D. Medina et al.

    White matter changes in mild cognitive impairment and ad: A diffusion tensor imaging study

    Neurobiology of Aging

    (2006)
  • S.L. Rauch et al.

    Neurosurgical treatment of tourette's syndrome: A critical review

    Comprehensive Psychiatry

    (1995)
  • E. Ryding et al.

    Motor imagery activates the cerebellum regionally: A SPECT rCBF study with super (99m) tc-hmpao

    Cognitive Brain Research

    (1993)
  • R.T. Staff et al.

    Generality and specificity in cognitive aging: A volumetric brain analysis

    NeuroImage

    (2006)
  • M.L. Albert et al.

    The “Subcortical dementia” of progressive supranuclear palsy

    Journal of Neurology, Neurosurgery and Psychiatry

    (1974)
  • J.M. Allman et al.

    Brain structures and life-span in primate species

    Proceedings of the National Academy of Science U S A

    (1993)
  • N.D. Anderson et al.

    The attentional demands of encoding and retrieval in younger and older adults: I. Evidence from divided attention costs

    Psychology and Aging

    (1998)
  • D. Arenberg

    A note of differences and changes in memory with age

  • P.M. Bentler

    Comparative fit indices in structural models

    Psychological Bulletin

    (1990)
  • K.A. Bollen

    Structural Equations With Latent Variables

    (1989)
  • R. Cabeza et al.

    Imaging cognition II: An empirical review of 275 pet and fMRI studies

    Journal of Cognitive Neuroscience

    (2000)
  • JB. Carroll

    Human cognitive abilities: A survey of factor-analytic studies. Cambridge

    (1993)
  • C.M. Filley et al.

    A comparison of dementia in Alzheimer's disease and multiple sclerosis

    Archives of Neurology

    (1989)
  • J.H. Gao et al.

    Cerebellum implicated in sensory acquisition and discrimination rather than motor control

    Science

    (1996)
  • C.L. Grady et al.

    The effects of age on the neural correlates of episodic encoding

    Cerebral Cortex

    (1999)
  • J. Grafman et al.

    Implicit learning in patients with alzheimer's disease

    Pharmacopsychiatry

    (1990)
  • P.M. Grasby et al.

    A graded task approach to the functional mapping of brain areas implicated in auditory & verbal memory

    Brain

    (1994)
  • F.M. Gunning-Dixon et al.

    The cognitive correlates of white matter abnormalities in normal aging: A quantitative review

    Neuropsychology

    (2000)
  • H. Haug

    Aging of the brain

  • Cited by (62)

    • Differential vulnerability of the cerebellum in healthy ageing and Alzheimer's disease

      2021, NeuroImage: Clinical
      Citation Excerpt :

      Differential subregional vulnerability to AD pathology or to normal ageing may be due to downstream influences resulting from cerebro-cerebellar synergism (Raz and Rodrigue, 2006; Sullivan and Pfefferbaum, 2006). Structural decline in the cerebellum may also have selective effects on distant neocortical regions (Guo et al., 2016b; Hogan et al., 2011; Limperopoulos et al., 2010). Recent studies on longitudinal changes in cerebellar volume in AD have come to conflicting results, with one suggesting that cerebellar volume loss only contributes to disease progression in later stages from MCI to AD (Tabatabaei-Jafari et al., 2017), whereas another stipulates that the opposite is true (Lin et al., 2020).

    View all citing articles on Scopus
    View full text