Abstract

Slowed information processing speed (IPS) is frequently reported in those with multiple sclerosis (MS), and at least 20% are compromised on some aspect of executive functioning also. However, any relationship between these two processes has not been examined. The Sternberg Memory Scanning Test, Processing Speed Index (WAIS-III), Delis Kaplan Executive Function System (D.KEFS), and Working Memory Index (WMS-III) were administered to 90 participants with MS. Their performance on the PSI was significantly below the normative scores but no deficits in memory scanning speed were evident. The initial response speed of the Sternberg and the PSI were more closely related to D.KEFS performance, particularly in timed tasks with a high cognitive demand (switching tasks). In contrast, memory scanning speed was related to working memory. This study reinforces the link between IPS and working memory in MS, and supports the suggestion that IPS is not a unitary construct.

Introduction

Multiple sclerosis (MS) is a demyelinating disease that affects the central nervous system, and in New Zealand approximately 4000 people (approximately 1% of the population) suffer from this debilitating disease. The deficits arising from this disease include sensory and motor symptoms and a significant proportion (65%) of individuals with MS are affected by some degree of cognitive impairment (Rao, Leo, Bernardin, & Unverzagt, 1991).

One of the most consistently reported impairments in MS is slowed information processing speed (IPS). Indeed, slowed IPS was included as part of the ‘known’ rather than as part of what is ‘believed’ about MS (Fischer et al., 1994), and many authors have demonstrated a significant difference in processing speed between those with MS and healthy controls (e.g., Archibald & Fisk, 2000; Demaree, DeLuca, Guadino, & Diamond, 1999; Diamond, Johnson, Kaufman & Graves, 2008; Kail, 1998; Lengenfelder et al., 2006; Rao, St. Aubin-Faubert & Leo, 1989). Deficits and differences in IPS have been documented in several MS subtypes. For example, both chronic progressive and relapsing-remitting groups showed significantly slower information processing (scanning) speed using the Sternberg Memory Scanning Test (Sternberg, 1967) compared with healthy controls, but only the chronic progressive groups showed deficits in sensory/motor reaction time (Archibald & Fisk, 2000). Other studies suggest that those with secondary progressive MS show IPS deficits using the Paced Auditory Serial Addition Test (PASAT) compared to those with relapsing-remitting MS and healthy controls (DeLuca, Chelune, Tulsky, Lengenfelder, & Chiaravalloti, 2004; Snyder, Cappelleri, Archibald & Fisk, 2001), while Denney, Lynch, Parmenter, and Horne (2004) demonstrated that the general speed of processing was slower for a sample of those with relapsing-remitting MS than for a sample of those with primary progressive MS and both groups were significantly slower compared with healthy controls. A few studies have suggested that IPS deficits were not as apparent in their MS samples (Litvan, Grafman, Vendrell, & Martinez, 1988; Janculjak et al., 1999; Santiago, Guardia, Casado, Carmona, & Arbizu, 2007). Of these, Litvan and coworkers (1988) only found deficits on the two fastest levels of the PASAT, but not on the Sternberg memory scanning speed, whereas Janculjak and coworkers (1999) used a non-standard recording method (the examiner made the response rather than the participant), which may have biased their data. The third study (Santiago et al., 2007) reported no deficits in IPS in their MS sample (assessed using the Stroop test), but deficits in performance on the PASAT were apparent, which was used as a measure of working memory. But taken together, there appears to be a general consensus that impairments in IPS are common in those with MS.

Some researchers have suggested that slowed IPS was one characteristic of the cognitive impairment thought to typify those with MS (Archibald et al., 2004; Mahler & Benson, 1990; Randolph et al., 2005; Rao, 1986; White, Nyenhuis, & Sax, 1992). However, others suggested that rather than being a separate deficit, slowed IPS underlies all other cognitive impairment in this population (De Sonneville et al., 2002; Kail, 1998; Kujala, Portin, Revonuo, & Ruutianen, 1994; Mohr & Cox, 2001). If this were the case then a relationship between IPS and other cognitive functions would be expected. Indeed, it has been suggested that impaired reasoning and remembering abilities seen in those with MS could well be explained by a general slowing of IPS (Arnett, Higginson, & Randolph, 2001; Demaree et al., 1999; Gontkovsky & Beatty, 2006; Kail, 1998). Further, Kujala and coworkers (1994/1995) found that a cognitively ‘impaired MS’ group were impaired on aspects of IPS, whereas a cognitively ‘unimpaired MS’ group showed only minor IPS deficits. Several studies have also examined (and demonstrated) an association between working memory and IPS in those with MS (Archibald & Fisk, 2000; DeLuca et al., 2004; Demaree et al., 1999; D'Esposito et al., 1996). In particular, IPS deficits in MS are more pronounced in tasks with high working memory demands (Legnenfelder et al., 2006).

One other issue which is apparent in the IPS literature is the variety of assessment tools that are used. In MS one of the most common measures of IPS is performance on the PASAT, but others have used simple reaction time, Sternberg Memory Scanning Test, Stroop, or the Trail Making Test (Archibald & Fisk, 2000; D'Esposito et al., 1996; Santiago et al., 2007). These tasks differ in their complexity, cognitive demands, and also in their reliance upon a motor response, which together suggests that all these tasks cannot measure the same cognitive construct (Chiaravalloti, Christodoulou, Demaree, & DeLuca, 2003). Indeed, in a factor analytical study, Chiaravalloti and coworkers (2003) demonstrated that different measures of IPS load on different factors indicating that IPS is not a unitary construct. This is clearly an issue that needs to be considered when interpreting results from studies using different IPS measures.

Previous research indicates that at least 15–20% of those with MS have impaired executive functions (Drew, 2005; Drew, Tippett, Starkey, & Isler, 2008; Fischer, 2001; Fischer et al., 1994; Rao et al., 1991). The type of executive dysfunction observed in those with MS varies with researchers reporting deficits across a range of functions including verbal fluency (Benedict et al., 2006), color word naming (Denney et al., 2004), reasoning (Benedict et al., 2006), temporal ordering, semantic encoding and planning (Arnett et al., 1997, 2001), and metamemory (Beatty & Monson, 1991; Scarrabelotti & Caroll, 1999).

Few studies have been conducted which assess the link between IPS and executive function in the MS population. Of those that have the range of executive functions that have been assessed are limited. However, IPS is reported to correlate with verbal fluency (Barker-Collo, 2006; Diamond et al., 2008). Similar to the findings with IPS and working memory, it has been suggested that IPS is predictive of performance in executive function tasks that require effortful processing (Diamond et al., 2008) and that IPS slowing is more evident with complex timed tasks, for example attentional set-shifting (De Sonneville et al., 2002). Thus, there is preliminary evidence for a relationship between IPS and some executive functions, indicting that more studies assessing a wider range of executive functions is warranted. Indeed, other researchers have suggested that this is an area worthy of further investigation (DeLuca et al., 2004).

The overall aim of the larger project was to assess general ability, memory, executive functions, IPS, and attention in a community-based sample of participants with MS and to examine the relationships between these functions (Drew, 2005; Drew et al., 2008). The current study focused on two of these areas. The first aim was to examine the level of IPS in our participants with MS and secondly to examine the relationships between IPS, executive function, and working memory.

Materials and Methods

Participants

Ninety-five individuals with MS living in the community around the greater Waikato and Bay of Plenty regions of New Zealand participated in this study. Of the 98 interested participants, three were excluded because of an aneurysm, epilepsy, and malingering, respectively. This sample made up approximately 2.5% of the MS population in New Zealand. Of the remaining 95 participants, 90 completed all of the psychological assessments and the following analyses are based on these participants.

Eighteen (20%) were men and 72 (80%) were women. They were aged between 17 and 78 years [mean = 52.5 years, standard deviation (SD) = 11.6], and the mean number of years of education was 12.8 (SD = 3.3). The average number of years since the participants detected the first symptoms of MS was 24.2 years (SD = 13.0). In contrast, the average number of years since MS was first diagnosed was only 11.6 years (SD = 10.5). The classification of disease course (self-reported) showed a split that was consistent with the MS literature. Forty-six (51.1%) had the relapsing-remitting form of MS, 26 (28.9%) had acute or secondary progressive, with only 15 (16.7%) indicating they were chronic progressive and very few, three (3.3%) indicating the benign form. At the time of testing, all relapsing-remitting participants self-reported that they were in remission. The mean Expanded Disability Status Scale (EDSS, Kurtzke, 1983) score for this sample was 4.8 (SD = 2.0) which is in the ‘not severe’ range. The mean composite score for the Chicago Multiscale Depression Inventory (CMDI; Nyenhuis et al., 1998) was 55.4 (SD = 13.1), suggesting that the cognitive performance of the sample should not have been significantly influenced by depressive illness.

Seventy-four friends or family of the participants acted as a control group for the Sternberg Memory Scanning Test. Thirty (40.5%) were men and 44 (59.5%) were women. Their ages ranged from 15 to 78 years (mean = 51.7, SD = 14.7), and the mean number of years of education was 13.0 (SD = 2.7). There was no significant difference between the MS group and control group in relation to age or years of education (p > .05), however, a chi-squared (χ2) analysis comparing the gender ratios in each group was significant (χ2 = 8.27, p < .01) with the MS group containing a significantly greater number of women.

Measures

The results of the larger project are reported elsewhere (Drew, 2005; Drew et al., 2008). This study focused on the data for IPS, assessed using the Sternberg Memory Scanning Test (Sternberg, 1967) and the Processing Speed Index (PSI) of the Wechsler Adult Intelligence Scale III (WAIS III; Wechsler, 1997a), executive functions which were assessed using the Delis Kaplan Executive Function System (D.KEFS, Delis, Kaplan, & Kramer, 2001) and working memory function which was determined from scores on the Working Memory Index (WMI) of the WMS III (Wechsler, 1997b). Further details of these measures are provided subsequently. In addition, ratings of levels of physical disability were obtained using the Expanded Disability Status Scale (EDSS) (Kurtzke, 1983), and the CMDI (Nyenhuis et al., 1998) was administered to determine levels of depression (Drew, 2005).

Various tests have been used earlier as a measure of IPS in MS, including the PASAT (e.g., D'Esposito et al., 1996), the Sternberg Memory Scanning Test (e.g., Archibald & Fisk, 2000), and in some studies, for example the Stroop Test (e.g., Santiago et al., 2007). Several reports suggest that participants find the PASAT rather stressful, and it has been reported to increase negative mood. Furthermore, in some studies a high proportion of patients refuse to take the test or stop half way through (Lezak, Howieson, & Loring, 2004; Strauss, Sherman, & Spreen, 2006). Thus, in order to obtain data from a maximum number of participants, and to avoid jeopardizing the participants' willingness to complete the neuropsychological assessments for the rest of the study, we chose to use the Sternberg Memory Scanning Test as the primary measure of IPS (Sternberg, 1967).

The Sternberg Test requires participants to memorize a set of 1, 2, or 4 digits and then they have to indicate whether numbers which are subsequently displayed on the computer screen were one of the previously ‘remembered’ numbers. Thus, this test requires less manipulation of numbers compared with the PASAT, and there is little evidence that participants find this task stressful. The Sternberg Test was administered on a Compaq Presario laptop computer and the test parameters (e.g., stimulus display, negative, and positive responses) were based on those used in Sternberg's original publication (1967). For each trial the participant was required to memorize a set of 1, 2, or 4 digits, which were presented together in the center of the computer screen (as used by Rao et al., 1989). There was a maximum time of 2 min for this ‘memorization’ stage, however the next stage of the trial could be started when the participant indicated they were ready. Previous studies suggest that the length of this memorization phase does not affect the scanning speed results obtained from the test (Sternberg, 1975) and we wanted to minimize any memory confounds, and allow sufficient time for the stimuli to be adequately encoded. The ‘test’ stimuli consisted of a single digit (from 1 to 9) which was presented in the center of the screen and the participant was asked to press the left mouse key if the number on the screen was one of those previously remembered, or press the right mouse key if not. There were a total of 12 trials, four for each set size. Each trial comprised the presentation of the target number(s) followed by the 15 test digits, presented individually on the screen in a random order. The ‘to be remembered’ numbers were different for each trial (to avoid possible ‘practice’ effects) and there were four positive responses required on each trial. The test was preceded by two practice trials each of which contained a memory set of three digits. After each key press response, visual feedback of ‘correct’ or ‘incorrect’ appeared on the screen for 0.5 s followed by a ‘ + ’ sign which remained in the center of the screen for 1 s prior to the presentation of the next digit. The time taken to respond to each of the stimuli was recorded.

The Sternberg Memory Scanning Test provides several measures of interest. Previous studies show that the time taken to respond to the stimuli increases in a positive linear fashion as the size of the remembered set increases (Sternberg, 1967, 1969, 1975). The slope produced when reaction time is plotted against memory set size represents the time taken to compare the test digit with those in memory (memory scanning speed), which is independent of motor function or the initial perception of the stimulus. Thus, a steeper slope is indicative of slower memory scanning speed (Fig. 1). The point, at which the slope crosses the y-axis, gives a measure of the time taken to encode the test stimulus, make a yes/no decision, and produce a response. Thus, this measure is susceptible to deficits in perception or slowed motor responding and is similar to a simple reaction time (this is termed ‘initial response speed’). In addition to these main measures, accuracy (number correct out of 60 responses) in each of the 1, 2, and 4 digits conditions was also recorded.

Fig. 1.

Sternberg Memory Scanning Test: the effect of set size on mean reaction time for participants with multiple sclerosis and controls (positive and negative trials combined). The slope of the line provides a measure of memory scanning speed, whereas the intercept provides the initial responses speed.

Fig. 1.

Sternberg Memory Scanning Test: the effect of set size on mean reaction time for participants with multiple sclerosis and controls (positive and negative trials combined). The slope of the line provides a measure of memory scanning speed, whereas the intercept provides the initial responses speed.

In addition to the Sternberg Test, scores from the WAIS-III (Wechsler, 1997a) PSI (comprising Digit Symbol Coding and Symbol Search) are also reported to provide an additional measure of IPS for the MS participants. The complete WAIS III results are reported elsewhere (Drew, 2005; Drew, Tippett, Starkey, & Isler, 2008). PSI provides a measure of mental and motor speed associated with solving of non-verbal problems. It comprises two subtests, Digit Symbol Coding and Symbol Search. In Digit Symbol Coding, the participant is given a series of numbers each of which is paired with a different symbol. The participant is provided with a grid of numbers, onto which they have to copy the appropriate symbols as quickly as they can. In Symbol Search, the participant has to scan two groups of symbols and then indicate if either of the target stimuli are in the other group. Digit Symbol Coding primarily measures psychomotor speed, whereas Symbol Search provides a measure of mental speed (Kaufman & Lichtenberger, 1999). PSI has a mean of 100 and an SD of 15, whereas the subtest scores have a mean of 10 and SD of 3.

Executive functioning was assessed using the D.KEFS (Delis et al., 2001). This is a relatively new assessment battery that assesses a wide range of executive functions. It consists of nine tests, which are adaptations of tests currently used for assessing executive functions. These tests are Trail Making, Verbal Fluency, Design Fluency, Color-Word Interference, Card Sorting Test, Twenty Questions, Word Context, Tower Test, and Proverbs. For this assessment battery, the standardized, age-adjusted scaled scores that are provided for all tests, are based on normative data from a large sample (n = 1750), which is representative of the U.S.A. population. For each test the mean standardized score is 10 and SD is 3. A list of validity studies that have demonstrated the sensitivity of the D.KEFS to executive function deficits in a variety of clinical populations has recently been published by Delis, Kramer, Kaplan, and Holdnack (2004). Although reliability coefficients for the D.KEFS tests were generally <.80, this is comparable with other neuropsychological tests, and it is probable that for these assessments, test complexity underlies performance variability. Twenty primary scores that are generated by the D.KEFS were used as the measures executive functioning. As the object of the current analysis was to determine related levels of performance, detailed categorization of the nature of any executive dysfunction was not attempted.

Although working memory processes are utilized when performing the D.KEFS tests, working memory is not measured directly (or separately) by this battery. Therefore the age-adjusted, scaled WMI score from the Wechsler Memory Scale, Third Edition (WMS-III; Wechsler, 1997b), was used as a measure of these processes. The WMI score (mean = 100, SD = 15) is a composite of two subtests from the WMS-III; Letter Number Sequencing and Spatial Span (mean = 10, SD = 3). In Letter Number Sequencing, participants listen to a series of alternating letters and numbers. They have to repeat these back to the examiner with the numbers in ascending order and the letters in alphabetical order. For the spatial span task, the examiner points to a series of wooden block on a board, the participant has to point to them in the same order. In later trials, the participant has to point to the block in reverse order (Wechsler, 1997b).

Procedure

Ethical approval was obtained from the Psychology Research and Ethics Committee at the University of Waikato. Individuals with a diagnosis of MS who belonged to the local MS society were sent an information letter and contacted the researcher directly if they were interested in taking part in the research. The same information sheet was sent to MS field officers who covered the outlying areas, and they forwarded the names of those willing to participate. All participants signed a consent form prior to testing.

The assessments for the larger project were carried out over two, 3-h sessions. General ability (WAIS-III), memory (WMS-III), and premorbid intelligence (Wechsler Test of Adult Reading [WTAR]) were assessed in the first session. The IPS assessment (Sternberg Memory Scanning Test) was administered at the beginning of the second session followed by the D.KEFS battery. The WAIS-III, the WMS-III, and the D.KEFS were administered and scored according to the standardized instructions in their respective manuals.

The time between the first and second sessions for the MS participants varied but the average interval was 12 days (SD = 13.8). The control group completed all required assessments in the first session.

Data Analysis

Data were analyzed using SPSS (version 14) and various data analysis approaches were utilized. The first part of the analysis focused on the Sternberg Memory Scanning Test, with both correct and incorrect responses included in the analysis. Initially, pre-analyses data screening was conducted to determine if the ‘yes’ and ‘no’ responses could be pooled and to determine if men and women performed similarly on this task. Then a multivariate analysis of covariance (MANCOVA; with age and number of errors as covariates) was conducted to examine the effects of MS on the measures from the Sternberg. For the PSI and its subtests, one-sample t-tests were carried out to compare the MS participants' performance to normative data (as control participants were not administered a WAIS III). Subsequently, correlations were conducted to examine the relationship between the measures from the Sternberg Test and the PSI and its subtests. Further analysis examined the link between level of disability and processing speed. Finally, correlations and a series of multiple regressions were conducted to examine the relationship between the measures of processing speed, executive function, and working memory.

Results

Pre-analysis Data Screening for the Sternberg Memory Scanning Test

In previous studies, the ‘yes’ and ‘no’ response times increased in a similar linear fashion as the digit number increased (i.e. the memory scanning speed did not differ between ‘yes’ and ‘no’ responses) and thus data from both type of responses had been pooled (e.g., Sternberg, 1975). This was also the case here (‘Yes’ memory scanning speed = 69.62 ms/digit; ‘No’ memory scanning speed = 69.86 ms/digit), suggesting that the data for both responses could be pooled. But first, we needed to determine if the MS and control groups differed in their response time to the ‘yes’ and ‘no’ stimuli. Thus, a 2 (Group: MS vs. Control) × 2 (Response type: yes/no) × 3 (number of digits: 1, 2, or 4) mixed analysis of variance (ANOVA) was conducted on response time. The overall ANOVA indicated significant main effects for number of digits, F(2,161) = 131.42, p < .001; response type, F(1,162) = 15.96, p < .001; and group, F(1,162) = 60.27, p < .001, but no significant interaction effects (all ps > .05). Closer examination of the data confirmed that, as described, response time significantly increased as the number of digits in the set increased (ps < .05 in all cases), that the ‘yes’ responses were significantly faster than ‘no’ responses and that MS participants' response times were significantly slower than those of the controls. Thus, as the time taken to respond to ‘yes’ and ‘no’ stimuli were the same for the MS and control groups, this indicated that data from these responses could be pooled for subsequent analysis. This is in keeping with other studies using the Sternberg Test with MS populations (e.g., Archibald & Fisk, 2000; Rao et al., 1989).

As there was a greater proportion of women in the MS group compared with the control group, the performance of men and women on the two measures of the Sternberg Test were also compared and no significant differences were found (p > .05); thus gender was not included as a covariate in subsequent analyses.

The Effects of MS on Performance in the Sternberg Memory Scanning Test

As earlier studies have indicated that reaction time is correlated with age (e.g., De Sonneville et al. 2002), this was included as a covariate in the analysis described subsequently. In addition, as both correct and incorrect responses were retained in our data set, accuracy of responding was also included as a covariate. Descriptive data for the MS and control groups can be found in Table 1.

Table 1

Means (and standard deviations) of multiple sclerosis participants and controls for demographic and Sternberg test variables

Group MS (n = 90) Control (n = 74) 
Demographics   
 Gender** M = 18; F = 72 M = 30; F = 44 
 Age 52.5 (11.6) 51.7 (14.7) 
 Years of education 12.8 (3.3) 13.0 (2.7) 
Sternberg measures   
 Memory scanning speed (ms/digit) 73.8 (67.0) 65.0 (39.6) 
 Initial response speed (ms)** 749.7 (253.7) 569.3 (122.5) 
 1 digit no correct (from 60) 58.7 (1.70) 59.2 (1.49) 
 2 digit no correct (from 60) 59.0 (1.41) 59.4 (1.01) 
 4 digit no correct (from 60) 58.3 (1.74) 58.7 (1.58) 
Group MS (n = 90) Control (n = 74) 
Demographics   
 Gender** M = 18; F = 72 M = 30; F = 44 
 Age 52.5 (11.6) 51.7 (14.7) 
 Years of education 12.8 (3.3) 13.0 (2.7) 
Sternberg measures   
 Memory scanning speed (ms/digit) 73.8 (67.0) 65.0 (39.6) 
 Initial response speed (ms)** 749.7 (253.7) 569.3 (122.5) 
 1 digit no correct (from 60) 58.7 (1.70) 59.2 (1.49) 
 2 digit no correct (from 60) 59.0 (1.41) 59.4 (1.01) 
 4 digit no correct (from 60) 58.3 (1.74) 58.7 (1.58) 

Notes: **p < .01 between the MS and control group on this measure. M = Male; F = Female.

To examine the effect of MS on the performance in the Sternberg Test, a between groups (MS/control) MANCOVA with covariates of age and accuracy was conducted on the two measures of the Sternberg Test; memory scanning speed (slope) and initial response speed (intercept). The overall MANCOVA revealed a significant difference between the MS and control groups, F(2, 157) = 13.51, p < .001. Subsequent analyses showed that the MS group had a significantly slower initial response speed compared with the control group, F(1, 158) = 25.81, p < .001, but there was no significant difference in the memory scanning speed, F(1,158) = .30, p > .05. Fig. 1 shows how memory scanning speed increased as the memory set size increased. It can be clearly seen that the initial response speed (intercept) is slower for the MS group compared with the controls but the slope of the lines, which is the measure of memory scanning speed, are very similar.

The Effects of MS on Performance on the PSI (from WAIS III)

Control data were only collected for the Sternberg Memory Scanning Test, therefore, to examine performance levels of the MS participants on the WAIS-III measures of processing speed, one-sample t-tests were conducted using the normative data as a comparison (normative mean = 10, SD = 3). For these measures, the MS sample obtained significantly lower scores compared with the normative data: Digit Symbol Coding (M = 8.3, SD = 3.0), t(84) = 5.14, p < .001; Symbol Search (M = 8.6, SD = 3.4), t(90) = 3.9, p < .001; PSI (M = 92.6. SD = 15.6), t(84) = 25.0, p < .001. Closer inspection of these data revealed that approximately one-third of participants obtained scores more than 1SD below the normative mean, of these, around 13% obtained scores more than 2 SDs below the norms and only one participant obtained scores more than 3 SDs below the normative mean.

Relations between the Processing Speed Measures

Pearson's correlations revealed that the scores on the two Sternberg measures, initial response time (intercept), and memory scanning speed (slope) was low and not significant (r = −0.05, p > .05), suggesting that these two scores assess different processes. In contrast, the scores for Digit Symbol Coding and Symbol Search correlated quite highly (r = .77, p < .001), indicating that there may be some similarity in the functions assessed by these two tests.

Pearson's correlations were also conducted between the Sternberg measures and the WAIS-III PSI measures and are summarized in Table 2. Memory scanning speed was significantly (but weakly) correlated with Digit Symbol Coding and Symbol Search but not with the overall PSI. In contrast, response speed correlated more highly with all PSI measures. Fisher's r-to-z transformations indicated that there was a significant difference in the correlations between the two Sternberg measures and Symbol Search (p < .05), whereas the difference between the correlations for the overall PSI approached significance (p = .07). This is likely to be reflective of the motor component in both the response speed of the Sternberg test and the subtests of the WAIS-III PSI.

Table 2

Correlations between the Sternberg test measures and the WAIS III processing speed measures in multiple sclerosis participants

 Sternberg test measures
 
WAIS-III processing speed measures Memory scanning speed Response speed 
Digit symbol coding −0.220* −0.435** 
Symbol search −0.217* −0.425**,a 
Processing speed index −0.204 −0.454** 
 Sternberg test measures
 
WAIS-III processing speed measures Memory scanning speed Response speed 
Digit symbol coding −0.220* −0.435** 
Symbol search −0.217* −0.425**,a 
Processing speed index −0.204 −0.454** 

*p < .05 for correlations; **p < .01 for correlations.

aIndicates instances where the memory scanning speed and the response speed correlations are significantly different (p < .05). See text for further details.

In order to determine if the level of disability influenced the processing speed assessments, correlations were also conducted between the EDSS scores and the processing speed measures. EDSS scores correlated significantly with the PSI and its subtests (Digit Symbol Coding, r = −0.39, p < .001; Symbol Search r = −0.33, p < .01; PSI, r = −0.32, p < .01) and the initial response speed of the Sternberg (r = .43, p < .001). However, the correlation between memory scanning speed and EDSS was small and not statistically significant (r = .06, p > .05). This is in keeping with the suggestion that only the intercept measure of the Sternberg relates to motor response speed and that the memory scanning measure is relatively free from this confound.

Executive Function and Working Memory in Participants with MS

As these data have been reported elsewhere they will only be summarized here (Drew, 2005; Drew et al., 2008). On the D.KEFS, the mean scores for the MS group were close to the standardized mean of the test. The group mean ranged from as low as 8.5 for Trails (SD = 3.9) to 10.5 for Design Fluency (SD = 3.2), compared with the standardized mean of 10 (SD = 3). Around one-third of participants showed no apparent deficits on any of the D.KEFS measures, obtaining scores within 1 SD of the mean. However, around 16% of the participants obtained scores more than 1 SD below the mean on more than six D.KEFS measures, indicating they had widespread executive function difficulties.

Similarly, for the WMI, the group mean was 97 (SD = 15.1), indicating that overall the group showed little evidence of working memory difficulties compared with the standardized mean of 100 (SD = 15). However, around one-third of participants obtained WMI scores that were significantly lower than that predicted from pre-morbid intelligence measures (WTAR). This suggests that for some of the participants there had been a significant decline in working memory ability.

Relations between Executive Functions, Working Memory and the Processing Speed Measures

In order to examine the link between the two processing speed measures, executive functioning, and working memory, Pearson's correlations between the Sternberg Memory Scanning Test measures, the 20 primary scores of the D.KEFS tests and the working memory measures from the WMS-III (WMI) and its two subtests (Letter Number Sequencing and Spatial Span) were conducted. In addition, correlations were also conducted between these scores and the PSI measures from the WAIS III (overall PSI, Digit Symbol Coding, and Symbol Search). The results of these analyses are presented in Table 3.

Table 3

Correlations between the primary scores from the Delis Kaplan Executive Function System tests, the working memory measures from the WMS-III and the information processing speed (IPS) measures from the Sternberg and WAIS III Processing Speed Index (PSI) for the multiple sclerosis participants

Test Sternberg IPS measures
 
WAIS III IPS measures
 
 Memory scanning speed Response speed Processing speed index Digit symbol coding Symbol search 
1. Trail making      
 Switching −0.288** −0.573**, 0.620**,†† 0.584**, 0.665**,†† 
2. Verbal fluency      
 Letters −0.289** −0.409** 0.428** 0.385** 0.511** 
 Category −0.304** −0.481** 0.464** 0.469** 0.514** 
 Switching −0.231* −0.397** 0.522**,† 0.430** 0.608**,†† 
3. Design fluency      
 Switching −0.207 −0.426** 0.490**,† 0.415** 0.581**,†† 
4. C-W interference      
 Inhibition −0.360** −0.459** 0.565** 0.488** 0.648**,†† 
 Switching −0.322** −0.375** 0.642**,†† 0.624**,†† 0.616**, 
5. Card sorting test      
 Correct sorts −0.196 −0.320** 0.332** 0.315** 0.429** 
 Description −0.183 −0.328** 0.335** 0.346** 0.410** 
 Recognition −0.105 −0.325** 0.456**,†† 0.403**, 0.505**,†† 
6. 20 Questions      
 Abstraction −0.149 −0.248* 0.220* 0.118 0.697**,†† 
 Total questions −0.082 −0.463**,†† 0.433**,† 0.429**,†† 0.492**,†† 
 Total −0.088 0.454**,†† 0.301** 0.303** 0.423**,†† 
7. Word context      
 Total −0.168 0.375** 0.228* 0.199 0.354** 
8. Tower test      
 Total −0.006 0.420**,†† 0.291* 0.263* 0.396**,†† 
9. Proverbs      
 Total −0.361** −0.263* 0.390** 0.358** 0.467** 
 Common −0.263* −0.204 0.309** 0.249* 0.430** 
 Uncommon −0.347** −0.282** 0.350** 0.346** 0.405** 
 Accuracy −0.313** −0.173 0.268* 0.220* 0.368** 
 Abstraction −0.278** −0.357** 0.387** 0.371** 0.473** 
WMS-III      
 Letter numbering sequencing −0.409** −0.245* 0.363** 0.359** 0.415** 
 Spatial span −0.183 −0.342** 0.352** 0.242* 0.506**,†† 
 Working memory index −0.357** −0.314** 0.402** 0.339** 0.502** 
Test Sternberg IPS measures
 
WAIS III IPS measures
 
 Memory scanning speed Response speed Processing speed index Digit symbol coding Symbol search 
1. Trail making      
 Switching −0.288** −0.573**, 0.620**,†† 0.584**, 0.665**,†† 
2. Verbal fluency      
 Letters −0.289** −0.409** 0.428** 0.385** 0.511** 
 Category −0.304** −0.481** 0.464** 0.469** 0.514** 
 Switching −0.231* −0.397** 0.522**,† 0.430** 0.608**,†† 
3. Design fluency      
 Switching −0.207 −0.426** 0.490**,† 0.415** 0.581**,†† 
4. C-W interference      
 Inhibition −0.360** −0.459** 0.565** 0.488** 0.648**,†† 
 Switching −0.322** −0.375** 0.642**,†† 0.624**,†† 0.616**, 
5. Card sorting test      
 Correct sorts −0.196 −0.320** 0.332** 0.315** 0.429** 
 Description −0.183 −0.328** 0.335** 0.346** 0.410** 
 Recognition −0.105 −0.325** 0.456**,†† 0.403**, 0.505**,†† 
6. 20 Questions      
 Abstraction −0.149 −0.248* 0.220* 0.118 0.697**,†† 
 Total questions −0.082 −0.463**,†† 0.433**,† 0.429**,†† 0.492**,†† 
 Total −0.088 0.454**,†† 0.301** 0.303** 0.423**,†† 
7. Word context      
 Total −0.168 0.375** 0.228* 0.199 0.354** 
8. Tower test      
 Total −0.006 0.420**,†† 0.291* 0.263* 0.396**,†† 
9. Proverbs      
 Total −0.361** −0.263* 0.390** 0.358** 0.467** 
 Common −0.263* −0.204 0.309** 0.249* 0.430** 
 Uncommon −0.347** −0.282** 0.350** 0.346** 0.405** 
 Accuracy −0.313** −0.173 0.268* 0.220* 0.368** 
 Abstraction −0.278** −0.357** 0.387** 0.371** 0.473** 
WMS-III      
 Letter numbering sequencing −0.409** −0.245* 0.363** 0.359** 0.415** 
 Spatial span −0.183 −0.342** 0.352** 0.242* 0.506**,†† 
 Working memory index −0.357** −0.314** 0.402** 0.339** 0.502** 

Notes:Bold indicates the measure which was the primary predictor in the multiple regression. See text for further details.

*p < .05 for correlation with the PSI measures; **p < .01 for correlation with the PSI measures.

p < .05, significantly different from memory scanning speed correlation; ††p < .01, significantly different from memory scanning speed correlation.

Overall, it can be seen from this table that those D.KEFS tests which were timed (Trail Making, Fluency, and Color-Word Interference) correlated with both measures of the Sternberg test. The measures for the Proverbs and the Letter Numbering Sequencing and the overall WMI measures also showed significant correlations with memory scanning speed. In contrast, all measures except for the Common Proverbs and the Proverbs Accuracy were significantly correlated with initial response speed. Fisher's r-to-z transformation was used to test for significant differences between the correlation coefficients from the two Sternberg Test measures. This revealed that response speed correlated significantly more highly with Trails (p < .05), 20 total questions (p < .01), 20 Questions total score (p < .01), and the Tower Test total score (p < .01) compared with memory scanning speed. There were no statistically significant differences between the correlations of the two measures from the Sternberg Test and the other D.KEFS measures.

The PSI and both its subtests showed significant correlations with the D.KEFS and WMI measures. In most cases these measures correlated more highly with the executive function test scores than those obtained from the Sternberg. Particularly high correlations (>0.6) were observed between the PSI and D.KEFS tests which incorporated inhibition or switching (Trail Making, Color-Word Interference). Fisher's r-to-z transformation indicated no significant differences between the correlation coefficients for the three processing speed scores from the WAIS III. However, compared with the Sternberg Memory Scanning Speed measure, the PSI correlated significantly more highly with several of the D.KEFS measures including all of the switching scores (Trail Making, Verbal Fluency, Design Fluency, Color-Word Interference), the Recognition score in the Card Sorting Test, 20 Questions Test, and the Tower Test.

To determine the extent to which the IPS measures (Memory Scanning Speed, Response Speed and PSI) accounted for unique variance in each of the D.KEFS scores, stepwise multiple regressions were conducted. Digit Symbol Coding and Symbol Search were not included in the regressions as they were both highly correlated with the overall PSI. As expected, the variable with the highest correlation with the D.KEFS measure was the primary predictor. These are indicated in bold in Table 3.

PSI was the primary predictor for most of the D.KEFS and WMI scores. In fact, PSI made a significant contribution to the variance of all measures except for 20 Questions Total, Word Context Total, and Tower Test Total scores. For the scores where PSI was the significant primary predictor, memory scanning speed accounted for significant additional variance in Category Fluency (ΔR2 = .04), Color-Word Interference Switching (ΔR2 = .04), Total (ΔR2 = .07), and Common Proverbs (ΔR2 = .05), and the WMI (ΔR2 = .07). The response speed measure from the Sternberg made a unique and significant contribution to the variance in Trails Switching (ΔR2 = .07), Category Fluency (ΔR2 = .06), and Total Questions on the 20 Questions Test (ΔR2 = .05).

Memory scanning speed was the primary predictor for Proverbs Accuracy and Letter Number Sequencing and the PSI score made a significant and unique contribution to the variance for each of these (ΔR2 Proverbs Accuracy = .04, ΔR2 Letter Number Sequencing = .08).

The Sternberg Response Speed measure was the primary significant predictor for the 20 Questions Total Score, Word Context Total Score, and the Tower Test Total Score. Neither memory scanning speed nor the PSI made a significant contribution to explaining the variance in these scores.

Discussion

The Sternberg Memory Scanning Test revealed no significant difference in memory scanning speed between the participants with MS and the control group. However, MS participants had a significantly slower initial response speed compared with controls.

Our findings differ from several previous studies which list reduced IPS as one factor in the cognitive profile of those with MS (e.g., Archibald et al., 2004; Arnett, 2000; Beatty, 1996; Fischer et al., 1994; Randolph et al., 2005; Rao, 1995, 1996), as well as others who have suggested that reduced IPS seems to be a key factor which possibly underlies all other cognitive impairment in those with MS (e.g., Brassington & Marsh, 1998; Demaree et al., 1999; Mohr & Cox, 2001). Although different assessment tools and definitions make direct comparisons of results difficult, three previous studies which have used versions of the Sternberg Memory Scanning Test have found evidence of a slowed IPS in their MS samples (Archibald et al., 2004; Archibald & Fisk, 2000; Rao et al.,1989), whereas two have not (Janculjak et al., 1999; Litvan et al., 1988).

The scanning speed demonstrated by our MS participants was, on average 74 ms/digit, which is relatively fast compared with values reported by others [Archibald & Fisk, 2000 (86.4 ms/digit); Archibald et al., 2004 (67.5 ms/digit); and Rao et al., 1989 (100 ms/digit)]. It is possible that as our large sample was community-based rather then accessed via a specialist clinic they have less severe symptoms than those who participated in other studies. Indeed, a limitation of using a community-based sample was that diagnosis of MS subtype was by self-report, which was not verified by a clinician. However, the possibility of our sample showing less severe MS symptoms is in keeping with Kujala and coworkers (1994) who found that a cognitively ‘impaired MS’ group were impaired on aspects of IPS, whereas a cognitively ‘unimpaired MS’ group showed only minor IPS deficits. In addition to this, the scanning speed of our control group (64 ms/digit) was relatively slow compared with others [Archibald & Fisk, 2000 (47.9 ms/digit); Rao et al., 1989 (68 ms/digit); Sternberg, 1967 (36 ms/digit)]. The control group were relatives/caregivers of the MS participants and there is a possibility that this burden may have led to significant levels of depression/fatigue in this group, however as we did not assess depression in this group, we are unable to determine if this was a contributing factor.

One point of difference between the current study and two of the earlier studies (e.g., Archibald & Fisk, 2000) was the number of digits in the memory set. While this study used a maximum of four digits, other studies have used up to six. Although these previous studies found no significant differences between their MS sample and a control group on measures of memory, the sample sizes were small and performance levels of those with MS were shown to be consistently lower on the memory measures. Thus, it is possible that the additional number of digits that were to be committed to memory may well have slowed the response (or processing) time in previous studies. Given the high level of accuracy in the performance of the MS participants in the Sternberg Scanning Test it seems unlikely that deficits in short-term memory can explain the findings.

In contrast, other researchers have also found no deficit in the memory scanning measure of the Sternberg test, but observed deficits in the initial response time (Janculjak et al., 1999; Litvan et al., 1988). The methodology of this latter study was somewhat different from other studies, as the researcher rather than the participant recorded the response, but this should have primarily affected the initial response time and not the time increment evident over increasing set sizes. In addition to the Sternberg Test, Litvan et al. (1988) also administered the Paced Auditory Serial Addition Task (PASAT) and found deficits in the two fastest levels, which is indicative of slowed IPS. This indicates that the Sternberg and the PASAT measure different constructs and reflects the suggestion of Chiaravalloti and coworkers (2003) that IPS is not a unitary construct.

In keeping with this, our data revealed that although the MS participants showed no evidence of memory scanning deficits, performance on the PSI and its subtests was significantly below that of the test norms. Thus, deficits in IPS were apparent with one measure but not the other. One possible explanation is that the Sternberg memory scanning measure reflects an underlying function quite distinct from that assessed by either the initial response speed of the Sternberg Test or the PSI. This is supported by larger correlations between the Sternberg initial response speed measure and the PSI compared with those between the memory scanning speed and the PSI. Furthermore, the level of disability was found to correlate with initial response speed and the PSI but not memory scanning speed. Initial response time is a measure similar to simple reaction time and includes motor speed, stimulus recognition time, and other initiating responses such as perceptual analysis and response selection, that remain constant regardless of the number of digits to be processed. The closer relation between the PSI and initial response speed probably reflects the motor component of the PSI subtests. In contrast, the smaller correlations between the Sternberg memory scanning measure and the PSI may be because the stimuli for both subtests of the PSI are constantly displayed and thus limited memory scanning is required. Thus, the difference in demands of the processing speed measures may explain the lack of consistency in our findings.

These findings lend further support to the suggestion by Chiaravalloti and coworkers (2003) that IPS is not a unitary construct and highlights the need for clear definitions of the functions being assessed by IPS tests. As previously mentioned, measures of IPS have ranged from simple reaction time, to the tasks which make up the PSI of the WAIS-III battery (e.g., DeLuca et al., 2004), through to the PASAT, or its variants, which require the manipulation of information, and therefore make greater demands on other cognitive resources than the less complex measures (Demaree et al., 1999, Diamond, DeLuca, Kim, & Kelley, 1997; Kalmar, Bryant, Tulsky & Deluce, 2004; Lengenfelder et al., 2006). Thus, the task requirements in the different tests differ substantially.

Not only do measures of IPS differ across studies, the name given to the construct being measured also varies, with the same tasks described as measuring either complex attention, cognitive speed, or processing speed (Chiaravalloti et al., 2003; Demaree et al., 1999, Diamond et al., 1997, Lengenfelder et al., 2006). In order to clarify this issue, Chiaravalloti and coworkers (2003) carried out factor analysis on scores from various processing speed tasks including simple reaction time, choice reaction time (CRT), and the PASAT. They found that both the simple and CRT tasks loaded on the first factor which they labeled simple speed/reaction time, the PASAT loaded on the second factor which was termed complex information processing and working memory tasks loaded on the third and final factor. On closer examination of the tasks involved in the Sternberg and the CRT, there are clear similarities—both require a key press in relation to a decision regarding stimuli presented on the screen, however the task for the Sternberg increases in difficulty, unlike the CRT. It is this increasing difficulty that allows the initial response speed to be dissociated from the memory scanning time. However, even in the 4 digit condition, there is minimal demand on the participant to manipulate information—their task is simply to compare the current stimulus with that held in short-term memory. In contrast, the PASAT requires the participant to undertake a substantial amount of information manipulation and it has been suggested that this task draws heavily on the ‘central executive’ component of working memory proposed by Baddeley and Hitch (1994) (Chiaravalloti et al., 2003). Thus, the cognitive requirements of the Sternberg Memory Scanning Test lie somewhere between that of the CRT and the PASAT, as it is neither a test of simple or CRT, nor of complex IPS. Indeed, successful performance on the Sternberg Test will probably rely primarily on one of the working memory slave systems (visuospatial sketchpad), and as the number of digits in the set increases, the central executive may also be engaged. However, the demands on the central executive during the Sternberg Test are likely to be much lower than during the PASAT. Together this suggests that the Sternberg may best be thought of as a test that assesses one of the cognitive constructs (memory scanning speed) that contributes to the larger group of abilities known as ‘IPS’ (Chiaravalloti et al., 2003).

Interestingly, an extension of the CRT test has been developed, the Computerized Test of Information Processing (CTIP), which progressively increases the amount of processing required by the participant (Reicker, Tombaugh, Walker, & Freedman, 2007). At its easiest level, the test assesses simple reaction time, the second stage assesses CRT, whereas the third stage involves conceptual processing, as the participant has to decide if a stimulus belongs to a particular group or not. This test shows remarkable similarities to the Sternberg Memory Scanning Test and has recently been used to demonstrate processing speed deficits in MS patients compared with controls. To provide additional clarification of the relationship between tests purporting to assess processing speed, further factor analytical studies based on performance on the Sternberg Test, the CTIP, and the PASAT would be warranted.

As the Sternberg memory scanning measure and the PSI appeared to be measuring quite distinct abilities, their relationship with executive functions and working memory differed. Generally, the PSI and the response speed measure from the Sternberg showed the highest correlations with the executive function measures. This is probably partly owing to the degree of overlap in the functions assessed by these two measures and may reflect the elementary response-initiating factors that are common to many of the tasks. There appeared to be some relationship between processing speed, task complexity, and time constraints as the PSI explained the greatest amount of variance in timed measures with high cognitive demands (Trails and CW interference switching). Interestingly, both of these tasks have been used to assess IPS in previous studies (DeLuca, Johnson, Beldowicz, & Natelson, 1995; Santiago et al., 2007). The relationship between IPS and timed complex tasks is in keeping with previous studies (De Sonneville et al., 2002; Diamond et al., 2008).

Thus, these results indicate that memory scanning speed does not appear to underlie successful performance on D.KEFS executive function tests. However, as our MS sample showed no significant deficits in memory scanning speed, it is possible that stronger relationship between this measure and executive function may occur in those whose scanning performance shows a greater degree of impairment.

In contrast, memory scanning speed was linked to working memory, in particular performance on the Letter Number Sequencing task. This is in keeping with Cowan (1999), who described memory scanning, as measured by the Sternberg paradigm as a search of currently activated memory and it adds support to the preceding discussion about working memory. The relationship between working memory and IPS has been well-described, for example, Mecklinger and coworkers (2003) revealed that those with a higher working memory capacity exhibit both a faster CRT and a lesser interference cost than those with a lower working memory capacity. Other studies indicate that although there is a relationship between working memory and processing speed in the MS population (e.g., Demaree et al., 1999; DeLuca et al., 2004; Lengenfelder et al., 2006; Litvan et al., 1988), it has also been shown that these processes are separate. Thus when participants are given enough time to complete a working memory task, performance generally improves to a level equivalent to that of the control group (Demaree et al., 1999; Lengenfelder, 2006).

Overall, the results of this study support the findings of Cowan (1999) and others who suggest that there is some link between memory scanning rate (and therefore an underlying construct of IPS) and working memory. Although there has been suggestion previously that slowed IPS underlies all cognitive impairment found in those with MS (e.g., Kail, 1998), this study lends greater support to the idea that if a task does not have a time constraint, there seems to be very little effect of memory scanning speed. The exception to this is the relationship between working memory and memory scanning speed, which may vary depending on the nature and complexity of the tasks.

One final point to note is that this study has highlighted the need to give careful consideration to the tasks used to measure processing speed. In particular, it emphasizes the need to interpret the findings to reflect the true underlying nature of the cognitive constructs being assessed rather than using the general term ‘IPS’, which has been shown not to be a unitary construct (Chiaravalloti et al., 2003).

Funding

This research was supported by PhD Research Scholarships from the Waikato Branch of the New Zealand Federation of University Women and a New Zealand Ministry of Education Top Achiever Doctoral Scholarship.

Conflict of Interest

None declared.

Acknowledgements

We wish to thank the participants for willingly giving up their time for this project.

References

Archibald
C. J.
Fisk
J. D.
Information processing efficiency in patients with multiple sclerosis
Journal of Clinical and Experimental Neuropsychology
 , 
2000
, vol. 
22
 
5
(pg. 
686
-
701
)
Archibald
C. J.
Wei
X.
Scott
J. N.
Wallace
C. J.
Zhang
Y.
Metz
L. M.
Mitchell
J. R.
Posterior fossa lesion volume and slowed information processing in multiple sclerosis
Brain
 , 
2004
, vol. 
127
 (pg. 
1526
-
1534
)
Arnett
P. A.
Neuropsychology in multiple sclerosis
 , 
2000
 
Arnett
P. A.
Higginson
C. I.
Randolph
J. J.
Depression in multiple sclerosis: Relationship to planning ability
Journal of the International Neuropsychological Society
 , 
2001
, vol. 
7
 
6
(pg. 
665
-
674
)
Arnett
P. A.
Rao
S. M.
Grafman
J.
Bernardin
L.
Luchetta
T.
Binder
J. R.
Lobeck
L.
Executive functions in multiple sclerosis: An analysis of temporal ordering, semantic encoding, and planning abilities
Neuropsychology
 , 
1997
, vol. 
11
 
4
(pg. 
535
-
544
)
Baddeley
A. D.
Hitch
G. J.
Developments in the concept of working memory
Neuropsychology
 , 
1994
, vol. 
8
 (pg. 
485
-
493
)
Beatty
W. W.
Adams
R. L.
Parsons
O. A.
Culbertson
J. L.
Nixon
S. J.
Multiple sclerosis
Neuropsychology for clinical practice: Etiology, assessment and treatment of common neurological disorders
 , 
1996
Washington DC
American Psychological Corporation
(pg. 
225
-
242
)
Beatty
W. W.
Monson
N.
Metamemory in multiple sclerosis
Journal of Clinical and Experimental Neuropsychology
 , 
1991
, vol. 
13
 
2
(pg. 
309
-
327
)
Benedict
R. H. B.
Cookfair
D.
Gavett
R.
Gunther
M.
Munschauer
F.
Garg
N.
Weinstock-Guffman
B.
Validity of the minimal assessment of cognitive function in multiple sclerosis (MACFIMS)
Journal of the International Neuropsychological Society
 , 
2006
, vol. 
12
 (pg. 
549
-
558
)
Barker-Collo
S. L.
Quality of life in multiple sclerosis: Does information processing speed have an independent effect?
Archives of Clinical Neuropsychology
 , 
2006
, vol. 
21
 (pg. 
167
-
174
)
Brassington
J. C.
Marsh
N. V.
Neuropsychological aspects of multiple sclerosis
Neuropsychology Review
 , 
1998
, vol. 
8
 
2
(pg. 
43
-
77
)
Chiaravalloti
N. D.
Christodoulou
C.
Demaree
H. A.
DeLuca
J.
Differentiating simple versus complex processing speed: Influence on new learning and memory performance
Journal of Clinical and Experimental Neuropsychology
 , 
2003
, vol. 
25
 
4
(pg. 
489
-
501
)
Cowan
N.
Miyake
A.
Shah
P.
A. Embedded-Process model of working memory
Models of working memory: mechanisms of active maintenance an executive control
 , 
1999
Cambridge
Cambridge University Press
(pg. 
62
-
101
)
Delis
D. C.
Kaplan
E.
Kramer
J. H.
The Delis-Kaplan Executive Function System.
 , 
2001
San Antonio
The Psychological Corporation
Delis
D. C.
Kramer
J. H.
Kaplan
E.
Holdnack
J.
Reliability and validity of the Delis-Kaplan Executive Function System: An update
Journal of the International Neuropsychological Society
 , 
2004
, vol. 
10
 (pg. 
301
-
303
)
DeLuca
J.
Chelune
G. J.
Tulsky
D. S.
Lengenfelder
J.
Chiaravalloti
N. D.
Is speed of processing or working memory the primary information processing deficit in multiple sclerosis?
Journal of Clinical and Experimental Neuropsychology
 , 
2004
, vol. 
26
 
4
(pg. 
550
-
562
)
DeLuca
J.
Johnson
S. K.
Beldowicz
D.
Natelson
B. H.
Neuropsychological impairments in chronic fatigue syndrome, multiple sclerosis, and depression
Journal of Neurology, Neurosurgery and Psychiatry
 , 
1995
, vol. 
58
 (pg. 
38
-
43
)
Demaree
H. A.
DeLuca
J.
Gaudino
E. A.
Diamond
B.
Speed of information processing as a key deficit in multiple sclerosis: Implications for rehabilitation
Journal of Neurology, Neurosurgery and Psychiatry
 , 
1999
, vol. 
67
 (pg. 
661
-
663
)
Denney
D. R.
Lynch
S. G.
Parmenter
B. A.
Horne
N.
Cognitive impairment in relapsing and primary progressive multiple sclerosis: Mostly a matter of speed
Journal of the International Neuropsychological Society
 , 
2004
, vol. 
10
 (pg. 
948
-
956
)
De Sonneville
L. M. J.
Boringa
J. B.
Reuling
I. E. W.
Lazeron
R. H. C.
Ader
H. J.
Polman
C. H.
Information processing characteristics in subtypes of multiple sclerosis
Neuropsychologia
 , 
2002
, vol. 
40
 (pg. 
1751
-
1765
)
D'Esposito
M.
Onishi
K.
Thompson
H.
Robinson
K.
Armstrong
C.
Grossman
M.
Working memory impairments in multiple sclerosis: Evidence from a dual task paradigm
Neuropsychology
 , 
1996
, vol. 
10
 
1
(pg. 
51
-
56
)
Diamond
B. J.
DeLuca
J.
Kim
H.
Kelley
S. M.
The question of disproportionate impairments in visual and auditory information processing in multiple sclerosis
Journal of Clinical and Experimental Neuropsychology
 , 
1997
, vol. 
19
 
1
(pg. 
34
-
42
)
Diamond
B. J.
Johnson
S. K.
Kaufman
M.
Graves
L.
Relationships between information processing, depression, fatigue and cognition in multiple sclerosis
Archives of Clinical Neuropsychology
 , 
2008
, vol. 
23
 (pg. 
189
-
199
)
Drew
M. A.
The effects of multiple sclerosis: Exploring executive dysfunction and its links with other cognitive impairment
 , 
2005
New Zealand
University of Waikato
 
Unpublished doctoral dissertation
Drew
M. A.
Tippett
L. J.
Starkey
N. J.
Isler
R. B.
Executive dysfunction and cognitive impairment in a large community based sample with multiple sclerosis from New Zealand: A descriptive study
Archives of Clinical Neuropsychology
 , 
2008
, vol. 
23
 (pg. 
1
-
19
)
Fischer
J. S.
Cook
S. D.
Cognitive impairment in multiple sclerosis
Handbook of multiple sclerosis
 , 
2001
3rd ed.
New York
Marcel Dekker Inc
(pg. 
233
-
255
)
Fischer
J. S.
Foley
F. W.
Aikens
J. E.
Ericson
D. G.
Rao
S. M.
Shindell
S.
What do we really know about cognitive dysfunction, affective disorders, and stress in multiple sclerosis? A practitioner's guide
Journal of Neurological Rehabilitation
 , 
1994
, vol. 
8
 (pg. 
151
-
164
)
Gontkovsky
S. T.
Beatty
W. W.
Practical methods for the clinical assessment of information processing speed
International Journal of Neuroscience
 , 
2006
, vol. 
116
 
11
(pg. 
1317
-
1325
)
Janculjak
D.
Mubrin
Z.
Brzovic
Z.
Brinar
V.
Barac
B.
Palic
J.
Spilich
G.
Changes in short term memory processes in patients with multiple sclerosis
European Journal of Neurology
 , 
1999
, vol. 
6
 (pg. 
663
-
668
)
Kail
R.
Speed of information processing in patients with multiple sclerosis
Journal of Clinical and Experimental Neuropsychology
 , 
1998
, vol. 
20
 
1
(pg. 
98
-
106
)
Kalmar
J. H.
Bryant
D.
Tulsky
D.
DeLuca
J.
Information processing deficits in multiple sclerosis: Does choice of screening instrument make a difference?
Rehabilitation Psychology
 , 
2004
, vol. 
49
 
3
(pg. 
213
-
218
)
Kaufman
A. S.
Lichtenberger
E. O.
Essentials of WAIS III Assessment
 , 
1999
Hoboken
NJ: John Wiley & Sons
(pg. 
34
-
35
)
Kujala
P.
Portin
R.
Revonsuo
A.
Ruutiainen
J.
Automatic and controlled information processing in multiple sclerosis
Brain
 , 
1994
, vol. 
117
 (pg. 
1115
-
1126
)
Kujala
P.
Portin
R.
Revonsuo
A.
Ruutiainen
J.
Attention related performance in two cognitively different subgroups of patients with multiple sclerosis
Journal of Neurology, Neurosurgery and Psychiatry
 , 
1995
, vol. 
59
 (pg. 
77
-
82
)
Kurtzke
J. F.
Rating neurologic impairment in multiple sclerosis: An expanded disability status scale (EDSS)
Neurology
 , 
1983
, vol. 
33
 (pg. 
1444
-
1452
)
Lengenfelder
J.
Bryant
D.
Diamond
B. J.
Kalmar
J. H.
Moore
N. M.
DeLuca
J.
Processing speed interacts with working memory efficiency in multiple sclerosis
Archives of Clinical Neuropsychology
 , 
2006
, vol. 
21
 (pg. 
229
-
238
)
Lezak
M. D.
Howieson
D. B.
Loring
D. W.
Neuropsychological Assessment
 , 
2004
4th ed.
Oxford
Oxford University Press
Litvan
I.
Grafman
J.
Vendrell
P.
Martinez
J. M.
Slowed information processing in multiple sclerosis
Archives of Neurology
 , 
1988
, vol. 
45
 
3
(pg. 
281
-
285
)
Mahler
M. E.
Benson
D. F.
Rao
S. M.
Cognitive dysfunction in multiple sclerosis: A subcortical dementia?
Neurobehavioral aspects of multiple sclerosis
 , 
1990
New York
Oxford University Press
(pg. 
88
-
101
)
Mecklinger
A.
Weber
K.
Gunter
T. C.
Engle
R. W.
Dissociable brain mechanisms for inhibitory control: effects of interference content and working memory capacity
Cognitive Brain Research
 , 
2003
, vol. 
18
 
1
(pg. 
26
-
38
)
Mohr
D. C.
Cox
D.
Multiple sclerosis: Empirical literature for the clinical health psychologist
Journal of Clinical Psychology
 , 
2001
, vol. 
57
 
4
(pg. 
479
-
499
)
Nyenhuis
D. S.
Luchetta
T.
Yamamoto
C.
Terrien
A.
Bernardin
L.
Rao
S. M.
Garron
D. C.
The development standardisation and initial validation of the Chicago Multiscale Depression Inventory
Journal of Personality Assessment
 , 
1998
, vol. 
70
 
2
(pg. 
386
-
401
)
Randolph
J. J.
Wishart
H. A.
Saykin
A. J.
McDonald
B. C.
Schuschu
K. R.
MacDonald
J. W.
Mamourian
A. C.
Fadul
C. E.
Ryan
K. A.
Kasper
L. H.
FLAIR lesion volume in multiple sclerosis: Relation to processing speed and verbal memory
Journal of the International Neuropsychological Society
 , 
2005
, vol. 
11
 (pg. 
205
-
209
)
Rao
S. M.
Neuropsychology of multiple sclerosis: A critical review
Journal of Clinical and Experimental Neuropsychology
 , 
1986
, vol. 
8
 
5
(pg. 
503
-
542
)
Rao
S. M.
Neuropsychology of multiple sclerosis
Current Opinion in Neurology
 , 
1995
, vol. 
8
 (pg. 
216
-
220
)
Rao
S. M.
White matter disease and dementia
Brain and Cognition
 , 
1996
, vol. 
31
 (pg. 
250
-
268
)
Rao
S. M.
Leo
G. J.
Bernardin
L.
Unverzagt
F.
Cognitive dysfunction in multiple sclerosis. I. Frequency, patterns and prediction
Neurology
 , 
1991
, vol. 
41
 (pg. 
685
-
691
)
Rao
S. M.
St. Aubin-Faubert
P.
Leo
G. J.
Information processing speed in patients with multiple sclerosis
Journal of Clinical and Experimental Neuropsychology
 , 
1989
, vol. 
11
 
4
(pg. 
471
-
477
)
Reicker
L. I.
Tombaugh
T. N.
Walker
L.
Freedman
M. S.
Reaction time: An alternative method for assessing the effects of multiple sclerosis on information processing speed
Archives of Clinical Neuropsychology
 , 
2007
, vol. 
22
 (pg. 
655
-
664
)
Santiago
O.
Guardia
J.
Casado
V.
Carmona
O.
Arbizu
T.
Specificity of frontal dysfunctions in relapsing-remitting multiple sclerosis
Archives of Clinical Neuropsychology
 , 
2007
, vol. 
22
 (pg. 
623
-
629
)
Scarrabelotti
M.
Carroll
M.
Memory dissociation and metamemory in multiple sclerosis
Neuropsychologia
 , 
1999
, vol. 
37
 (pg. 
1335
-
1350
)
Snyder
P. J.
Cappelleri
J. C.
Archibald
C. J.
Fisk
J. D.
Improved detection of differential information processing speed deficits between two disease course types of multiple sclerosis
Neuropsychology
 , 
2001
, vol. 
15
 
4
(pg. 
617
-
625
)
Sternberg
S.
Two operations in character recognition: Some evidence from reaction-time measurements
Perception and Psychophysics
 , 
1967
, vol. 
2
 (pg. 
45
-
53
)
Sternberg
S.
Memory scanning: mental processes revealed by reaction time experiments
American Scientist
 , 
1969
, vol. 
57
 (pg. 
421
-
457
)
Sternberg
S.
Memory scanning: New findings and current controversies
The Quarterly Journal of Experimental Psychology
 , 
1975
, vol. 
27
 
1
(pg. 
1
-
32
)
Strauss
E.
Sherman
E. M. S.
Spreen
O.
A compendium of neuropsychological tests: Administration, norms, and commentary
 , 
2006
3rd ed.
New York
Oxford University Press
Wechsler
D.
Wechsler Adult Intelligence Scales
 , 
1997
, vol. 
a
 
3rd ed.
San Antonio, TX
The Psychological Corporation
Wechsler
D.
Wechsler Memory Scale
 , 
1997
, vol. 
b
 
3rd ed.
San Antonio, TX
The Psychological Corporation
White
R. F.
Nyenhuis
D. S.
Sax
D. S.
White
R. F.
Multiple sclerosis
Clinical syndromes in adult neuropsychology: The practitioners handbook
 , 
1992
Amsterdam
Elsevier Science Publishers B.V
(pg. 
177
-
212
)