Wisconsin Card Sorting Task

5.2.1 Wisconsin Card Sorting Task

The WCST (Grant and Berg, 1948) is a test of set-shifting (cognitive flexibility) and is one of most widely used neuropsychological tests in the history of clinical psychology, used to diagnose dysfunction of the frontal lobes (Retzlaff, Butler and Vanderploeg, 1992). The test is a card-matching task in which a person must match a presented stimulus card to one of four reference cards on the basis of an unknown rule (criterion). For example, the cards may be matched on the basis of the shape, colour or number of the objects on the card. The job of the participant is to use a method of trial and error to work out what the correct criterion is (shape, colour or number). The participant choses a reference card to match the stimulus card with and an experimenter says whether or not the participant is correct. The matching criterion will stay the same for a certain period of selections, after which the experimenter will change the rule. The participant then has to learn the new rule through trial and error. As a result the fundamental process required in the WCST is rule learning. Performance on the WCST is measured in a number of ways, including: preservative errors

(responding on the basis of a previously ‘correct’ criteria which no longer applied despite having been informed that the selection was incorrect), set loss errors (errors occurring after three consecutively correct responses), any other errors made, as well as the number of dimensions found, the number of dimensional changes achieved (Heaton et al., 1993).

The WCST was originally developed and utilised as a measure of abstract reasoning ability and was also found to be a measure of executive functions (Luria, 1973). The WCST is thought to measure executive functions such as: abstract reasoning, problem solving, planning, organised search, response maintenance, using feedback to shift cognitive sets and change behavioural responses, directing behaviour towards achieving a goal and controlling impulsive responding (Chelune and Baer, 1986; Lezak, 2004). It was not until later that the WCST began being used as a neuropsychological test to diagnose patients with frontal lobe dysfunction (Milner, 1963), but is now considered to be a key measure in the diagnosis of frontal lobe dysfunction (Braff et al., 1991).

Patients with diverse damage to the frontal lobes have been found to perform poorly on the WCST in comparison to healthy controls (for example Axelrod et al., 1996; Janowsky et al., 1989). Barcelo and Knight (2002) used the WCST on stroke sufferers who had damage to the frontal lobes and normal controls. The results revealed that those with damage to their frontal lobes performed significantly worse than the normal group. These findings are supported by research that tested individuals with frontal lobe lesions caused by penetrative brain injuries (Grafman et al., 1990). Lesion studies have also compared WCST performance between people with frontal lesions and non-frontal lesions, with frontal lesions found to result in significantly poorer performance than non-frontal lesions (Bornstein, 1986; Eslinger and Grattan, 1993; Heaton et al., 1993). However a number of studies failed to find significant differences between those with frontal and non-frontal lesions (Anderson et al., 1991; Axelrod et al. 1996; Grafman et al., 1990), indicating that impaired WCST performance may be sensitive to but not specific to frontal lobe damage.

Neuroimaging research has also provided evidence that the frontal lobes are activated during the WCST, with activity in the DLPFC and ventromedial prefrontal cortex (VMPFC) found in a number of studies. For example, Haines et al. (1994), Kawasaki et al. (1993) and Weinberger (1986) reported activation of

the DLPFC. Meanwhile Mentzel et al. (1998), Tien et al. (1998) and Volz et al.

(1997) reported activation of the DLPFC and VMPFC. There is no clear evidence of laterality from neuroimaging research, with some studies finding greater activity in the left hemisphere (e.g. Haines, 1994) whereas other studies have reported greater activity in the right hemisphere (e.g. Volz, 1997). As with the finding from lesion studies, neuroimaging data indicates sensitivity for areas of the frontal lobes to the WCST but not specificity. For example, Mentzel et al. (1998) also found activity in the thalamus and basal ganglia; Nagahama et al. (1996) found activity in the striate cortex and inferior parietal cortex, while Tien et al. (1998) also reported increased activity in the inferior parietal cortex.

These neuroimaging studies suggest that the DLPFC and VMPFC are recruited during the WCST. This is supported by lesion studies that have investigated the effects of localised DLPFC lesions. As mentioned in Chapter 4, the executive functions involved in the WCST, such as task switching, are associated with the DLPFC and there is evidence that the DLPFC is a prefrontal area heavily involved in the WCST. Milner (1963) proposed that the DLPFC is important for performance on the WCST. She had patients who had undergone cortical excisions complete the WCST. The study revealed that frontal excisions had a much greater affect on task performance than posterior excisions. The results of the study also revealed that removal of the inferior and orbital sections of the frontal lobe had no effect on WCST performance, whereas all removals that encroached upon Brodmann Area 9, which includes a part of the DLPFC, resulted in poor performance on the WCST. Other neuropsychological studies on patients with DLPFC excisions provide further support to this theory (Rezai et al., 1993;

Weinberger, Berman and Zec, 1986). This position is further supported by lesion studies that have reported that patients with specific damage to the DLPFC have inferior performance on the WCST in comparison to patients with lesions in other prefrontal areas such as the orbito-frontal cortex (Ettlinger, Teuber and Milner 1975; Grafman et al., 1994).

Lie et al. (2006) carried out an fMRI study on the WCST in order to understand the neural substrates utilised during the task. They used a modified design in which participants completed three WCST tasks: a standard WCST, a WCST in which the participants were informed of the sorting criteria every four trials and a baseline WCST in which the stimulus card was an exact match for one

of the reference cards. The ultimate difference in these tasks regards cognitive switching, with uninstructed, instructed and no cognitive switching required for the different tasks. The behavioural results revealed that performance was best on the baseline task, followed by the informed WCST and then the uniformed WCST.

Performance was measured in terms of cognitive costs (a comparative measure of reaction times for the different conditions), errors (preservative, set-loss, other errors) as well as the number of dimensions and dimensional changes achieved.

The fMRI results revealed a complex neural network underlying performance on the WCST. This network included frontoparietal regions and the striatum. The right VLPFC was found to be related to simple WM operations, while right DLPFC was related to more complex WM operations. The rostral anterior cingulate cortex and temporoparietal areas were responsible for attention and error detection. The caudal anterior cingulate cortex and the right DLPFC were associated with increased attentional control with regards to the increase of demands of working memory and cognitive control. Non-frontal neural activity was related to set-shifting (cerebellum) and working memory representations (superior parietal cortex, retrosplenium).

One important takeaway from this study is that while the WCST has been found to be sensitive to frontal lobe damage and elicits activity in frontal areas, it is not specific to frontal areas. However there is extensive evidence that performance in the WCST is reliant upon the frontal lobes, in particular the DLPFC.