ASSESSING THE THEORETICAL, DIAGNOSTIC, AND THERAPEUTIC
IMPLICATIONS OF INDIVIDUAL PERFORMANCE VARIABILITY IN
PERSONS WITH BRAIN INJURY
James S. McDonnell Foundation Workshop
May 15-17, 2006
Skamania Lodge
http://www.skamania.com/about.php
Performance deficits, commonly measured as increased reaction times on
neuropsychological tests, are characteristic of subjects with traumatic
brain injury (TBI) and other brain pathologies. Patients groups have difficulty
with sustaining attention, have increased distractibility, perform poorly
on divided attention tasks, and show an increase in the degree to which
their performance varies from trial to trial. The finding of increased
variability in individual performance has not been as intensively studied
as some of the other characteristics of cognitive dysfunction but may
hold important implications for our understanding of both normal brain
function and disability.
Traditionally, in research studies with non-clinical populations, variability in reaction time between individuals is considered ‘noise’ and is deliberately manipulated out of data analysis via the overall study design. Usually, inter-individual variability rather than intra-individual variability accounts for the variance observed in healthy subject populations. Presenting data as group mean plus/minus the standard deviation fairly represents the group performance when individual variability is low, but the relationship does not hold in clinical populations with both large inter- individual differences and large intra-individual differences. This has been well demonstrated in aging studies where it is the intra-individual variability that accounts for most of the group variance and is, in fact, one of the better predictors of cognitive decline and mortality. TBI patients consistently show high intra-individual variability and the prolonged reaction times associated with attention deficits. Conversely, individuals scoring high on intelligence and working memory tests also show lower intra-individual variations on various measures including reaction times, and ERPs, and changes in cerebral metabolic rates during cognitive tasks.
What is the neural process that reduces variability?
In interacting with their environment, organisms tend to optimize performance by decreasing response variability from event to event. Various neural systems have evolved in response to this need. In mammals, for example, evidence suggests the cerebellum contributes to the smoothing of motor performance. Pathology, such as demonstrated by patients with cerebellar lesions, causes a gross increase in intra-individual variability on motor tasks. Less substantiated is the role the cerebellum may play in “smoothing” cognitive performance. An interesting result obtained during fMRI experiments is that changes in BOLD signal are detected in medial cerebellum during motor tasks and in the lateral cerebellum during cognitive tasks. Admittedly, the distinction being made in these experiments between “motor” and ‘cognitive” might be artificial. However, recent neuroanatomical mappings of the cerebellar dentate nucleus have discovered that half of the dentate maps to motor regions and the other half to the prefrontal cortex. It is interesting to speculate that the brain co-opted neural processes evolved for reducing movement variability for reducing cognitive response variability. The cerebellar mechanism to reduce variability seems to involve a feed-forward or anticipatory process. In the absence of cerebellar function (lesions, agenesis or degeneration) the prefrontal cortex can assume aspects of certain tasks associated with the cerebellum, such as anticipating timing control, with a marked increase in intra-individual variability as shown in cognitive studies (with fMRI) in aging and TBI.
Individual variability may be a useful metric for the study of attention and cognitive processing efficiency and particularly important when accessing the function of special populations. The neural network sub serving attention shares common features and purposes with reducing individual performance variability. Further investigations may give new insight into clinically relevant aspects of attention processing.
The Questions to be addressed by the workshop are:
- Is individual cognitive performance variability coincident, correlated or causal to brain injury?
- How do you measure individual variability reliably?
- Is increased intra- individual variability truly detrimental in daily life performance?
- What are the common neuroanatomical, electrophysiological and behavioral features of attention and performance variability control?
- What is the clinical value of measuring and treating variability? Does this provide new treatment insights?
- Have variability measures proven useful in aging and other clinical disorders of attention? To what extent?
