Archives of Physical Medicine and Rehabilitation journal homepage: www.archives-pmr.org Archives of Physical Medicine and Rehabilitation 2014;-:-------
ORIGINAL ARTICLE
Improved Physical Fitness Correlates With Improved Cognition in Multiple Sclerosis Meghan Beier, PhD,a Charles H. Bombardier, PhD,a Narineh Hartoonian, PhD,a Robert W. Motl, PhD,b George H. Kraft, MD, MSa From the aDepartment of Rehabilitation Medicine, University of Washington, Seattle, WA; and bDepartment of Kinesiology and Community Health, University of Illinois, Urbana-Champaign, IL.
Abstract Objective: To determine whether there is an association between improvements in objective measures of physical fitness and performance on cognitive tests in people with multiple sclerosis (MS). Design: Post hoc correlational analysis in which people demonstrating physical improvement were compared with those not demonstrating physical improvement. Setting: Individuals with MS residing in the community. Participants: Adults with clinically confirmed MS (NZ88) who participated in a controlled trial of a telephone-based health promotion intervention, chose to work on exercise, and completed the pre- and postintervention assessments. Interventions: Participants were measured for strength (isokinetic dynamometer), aerobic fitness (bicycle ergometer), and cognition (Paced Auditory Serial Addition Test [PASAT], Trail Making Test [TMT]) at baseline and 12 weeks later. Change in fitness was calculated by subtracting each participant’s baseline score from the outcome score, and then transforming the difference to a z score. Individuals with a z score 1 on any fitness measure were placed in the physically improved group (nZ25). All others were in the physically not improved group (nZ57). Main Outcome Measures: TMT, PASAT. Results: After controlling for covariates (age, sex, ethnicity, education, disease activity, MS type), there was a significant group-by-time interaction, suggesting that cognitive functioning changed over time based on level of fitness. Participants in the physically improved group demonstrated improved performance on measures of executive functioning after 12 weeks of exercise. Conclusions: The results of this study lend support to the hypothesis that change in fitness is associated with improved executive functioning in people with MS. Archives of Physical Medicine and Rehabilitation 2014;-:------ª 2014 by the American Congress of Rehabilitation Medicine
Cognitive impairment is one of the most common and disabling features of multiple sclerosis (MS). Prevalence studies1-3 have reported rates of cognitive impairment ranging from 22% to 60%, depending in large part on whether the data were acquired from community or hospitalized samples and also on the assessment tool used. The cognitive deficits seen in MS appear most often in 3 domains: (1) information processing speed and working memory, Presented to the European Committee for Treatment and Research in Multiple Sclerosis October 4 2013, Copenhagen, Denmark. Supported by the Multiple Sclerosis Research and Training Grant funded by the Department of Education, National Institute on Disability and Rehabilitation Research (grant nos. H133B031129, H133B080025) and the National Multiple Sclerosis Society (grant no. MB 0008). No commercial party having a direct financial interest in the results of the research supporting this article has conferred or will confer a benefit on the authors or on any organization with which the authors are associated.
(2) new learning and episodic memory, and (3) executive functioning. MS-related cognitive dysfunction has a major impact on quality of life,4,5 activities of daily living,6,7 vocational functioning,8,9 and rehabilitation outcomes.10 Medical and rehabilitation approaches to improving cognition are disappointing. Studies of donepezil,11 psychostimulants,12 aminopyridines,13,14 ginkgo biloba,15 and memantine16 have not been consistently effective for improving cognition. Research on cognitive rehabilitation in MS suggests that learned strategies can improve cognitive performance; however, this area of investigation is in its infancy and largely focused on memory.17-19 Based on the older adult literature, exercise training represents a promising approach to improve cognitive functioning in MS.20
0003-9993/14/$36 - see front matter ª 2014 by the American Congress of Rehabilitation Medicine http://dx.doi.org/10.1016/j.apmr.2014.02.017
2
M. Beier et al
Systematic reviews and meta-analyses demonstrate that both aerobic and strength training significantly improve cognitive functioning in older adults and people with mild cognitive impairment.21-25 The most profound benefits of exercise training occurred in the domain of executive functioning,22,26,27 and the improvements seemingly depend on changes in physical fitness (defined as health-related elements such as aerobic and muscular endurance, muscle strength, body composition, and flexibility28). Much less is known about exercise training and cognition in MS. Correlational studies have reported a link between physical fitness and cognition in people with MS. For example, both higher cardiorespiratory endurance and greater muscle strength have been associated with better performance on tests of speed of information processing and working memory.29,30 Furthermore, cardiorespiratory endurance is positively associated with functional magnetic resonance imaging indices of both white and gray matter integrity.31 Nevertheless, 2 existing trials32,33 have failed to demonstrate benefits of exercise training on cognition in MS. Those trials might have been ineffective because there was poor compliance and no documentation of change in fitness levels. This underscores the importance of future research that examines the possibility that improvements in physical fitness can contribute to improved cognition in people with MS. We analyzed data from a previously published randomized controlled trial34 examining the effects of telephone counseling on health promotion in people with MS. In that trial, a subgroup of 88 people chose to work on exercise and the subjects were randomly assigned to an intervention group or waitlist control condition. Unfortunately, there were no between-group differences on objective measures of leg muscle strength, aerobic endurance, or cognition. However, the availability of pre- and postintervention data on these measures provided an opportunity to examine the relationship between objective improvement in muscle strength or aerobic endurance and improvement in cognitive functioning within the study group as a whole. To do this we created a “physically improved” group (improvement in muscle strength, aerobic endurance, or both) and a “physically not improved” group. We compared changes in speed of information processing, psychomotor speed, and executive functioning between these 2 groups. Our primary hypothesis was that after controlling for demographic and disease-related variables, the physically improved group would demonstrate significantly greater improvement on measures of speed of information processing, psychomotor speed, and executive functioning compared with the physically not improved group.
Methods Participants Subjects were 88 community-dwelling individuals with clinically confirmed MS who wanted help with exercise as part of the previously published health promotion trial.34 Participants were recruited from the University of Washington MS Center, MS
List of abbreviations: MS PASAT rpm TMT-A TMT-B TMT-BA
multiple sclerosis Paced Auditory Serial Addition Test revolutions per minute Trail Making TestePart A Trail Making TestePart B Trail Making TestePart B minus Part A
newsletters, newspaper advertisements and articles, support groups in the Puget Sound region, and from flyers mailed to local neurology offices. Participants were also recruited from a large survey study34 of persons with MS. The inclusion criteria for participation were (1) age 18 years; (2) diagnosis of MS confirmed by a neurologist or physiatrist specializing in MS; (3) ability to walk 300ft (90m) without assistance, equating to an Expanded Disability Status Scale score of 5.535; and (4) interest in receiving help to exercise more. All MS types (relapsing-remitting, secondary progressive, primary progressive, progressive-relapsing) were included. Participants were excluded if they (1) had a medical condition that contraindicated exercise, or (2) screened positive for depression on the Primary Care Evaluation of Mental Disorders.36 The Primary Care Evaluation of Mental Disorders is a 26-item screener of “yes” or “no” questions for a variety of symptoms. The 2 depression questions ask about depressed mood and anhedonia. Six participants were excluded from the final analysis because of missing data. Two individuals were missing all fitness measures, and 4 were missing cognitive data. This resulted in 82 participants who were included in the final analysis.
Procedures The University of Washington Institutional Review Board approved the study protocol. Before data collection, all participants provided written informed consent. Participants meeting inclusion criteria during a telephone interview were invited to the study center for a baseline assessment. Trained research assistants, supervised by study investigators, conducted baseline assessments in dedicated laboratory space at the medical center. Afterward, participants chose a health promotion activity (exercise in the present analysis) for which they wanted help. Trained research assistants conducted in-person outcome assessments 12 weeks postrandomization. Baseline and outcome assessments were conducted in the same laboratory space with the temperature controlled between 19 C and 22 C.
Measures Self-reported demographic and medical variables such as age, ethnicity, marital status, educational level, date of diagnosis, and MS subtype37 were collected during a telephone screen. Measure of disease activity The Timed 25-Foot Walk38 is a clinical tool used to measure the stability of MS disease. Research suggests this is a useful tool for detecting stability of disease and the presence of an exacerbation.38 The 25-Foot Walk was measured at baseline and follow-up. The change score was used as a covariate for controlling for disease activity. Measures of strength and fitness Right and left leg flexion and extension muscle strength were measured on a Cybex 6000 isokinetic dynamometer.a The Cybex was calibrated weekly using the manufacturer’s procedures. Each subject was given 5 submaximal trial repetitions before testing. To obtain peak torque output for knee flexion and extension, 10 maximal repetitions were performed by each leg at a rate of 60 per minute. Leg strength was defined as peak torque averaged over 10 repetitions for the right and left legs and for both extension and flexion. www.archives-pmr.org
Fitness and cognition in multiple sclerosis Aerobic fitness was estimated using a Schwinn Airdyne bicycle ergometer.b Subjects were asked to pedal the ergometer at 60 revolutions per minute (rpm) for as long as they could while pedal resistance was increased. Resistance began at 0W and increased incrementally by 0.5W every 2 minutes until the participant could no longer sustain 60rpm. The number of seconds the subject could sustain pedaling at 60rpm was the measure of aerobic fitness. The original trial collected minutes of self-reported physical activity. However, as the original study was done to examine behavior change based on a motivational interviewing protocol, individuals were asked to choose any fitness activity in which they wanted to engage. This could have included activities such as stretching, which may not be associated with an objective change in physical fitness. We determined that a significant change on objective measures of physical fitness would be a more effective way to categorize changes in physical fitness rather than relying on self-report. Measures of strength and fitness: physically improved versus physically not improved groups Group definitions were based on the combination of changes in leg muscle strength and aerobic fitness. Baseline scores were subtracted from the outcome score for each individual. The change scores were then converted to z scores. We defined significant improvement as a z score 1. Participants who demonstrated significant improvement in leg muscle strength or aerobic fitness were included in the physically improved group (nZ25). Participants who did not demonstrate significant improvement in leg strength or fitness were categorized as physically not improved (nZ57). Measures of cognition The Paced Auditory Serial Addition Test (PASAT) measures information processing speed, flexibility, calculation ability,39 divided attention, and working memory.40 The PASAT requires the individual to listen to a series of single-digit numbers presented at 3-second intervals during the first trial and 2-second intervals during a second trial. The subject states the sum of each pair of contiguous numbers to the examiner. The outcome variable is the number of correct responses. The Trail Making Test (TMT) is a widely used 2-part test. Part A (TMT-A) is a test of psychomotor speed, processing speed, and visual scanning. It requires the subject to connect a series of numbered circles in a paper-and-pencil format. Part B (TMT-B) is similar but requires the subject to follow an alternating sequence (1-A-2-B-3-C.) and is thought to measure working memory, setswitching, and executive functioning. The main dependent factor is time to complete each sequence.41 Subtracting the time to complete TMT-A from TMT-B (TMT-BA) minimizes the contribution of visual scanning, working memory, and psychomotor speed demands, producing a purer measure of executive functioning.41,42
Statistical analysis Statistical analyses were performed using IBM SPSS Statistics version 20.c Mahalanobis and Cook’s distance were used to screen for multivariate outliers. No univariate or multivariate outliers were detected, and assumptions regarding normality of sampling distributions were met. Data were analyzed using repeatedmeasure multivariate analysis of covariance, with physical improvement (at 2 levels, physically improved and physically not improved) as the between-subjects factor and time (at 2 levels,
www.archives-pmr.org
3 preintervention and postintervention) as the within-subjects factor. To test the hypothesis that improved physical fitness was associated with cognitive improvement, we examined the 2-way physical improvement (improved vs not improved) by time interaction. Within physical improvement, change was assessed by a 1 degree of freedom contrast. Two-tailed P values were reported for the relevant effects. Age, sex, ethnicity, education, disease activity, and MS type were chosen as covariates because previous research demonstrated they could have a relationship with cognition.43-46 Significant interaction effects were followed by custom nonorthogonal planned contrasts. Significant interaction effects are decomposed using simple effects post hoc analyses. Because SPSS is not amenable to nonstandard contrast analyses within a mixed-model design, planned comparisons were performed using PSY: A Program for Contrast Analysis.d PSY allows for the control of family-wise error rates for custom-planned contrast analysis. PSY can also supply standard or Bonferroni-adjusted critical values. Power analyses were conducted with GPower 3,47,e evaluating the a priori sample size required to detect a medium-sized effect (fZ.25)48 for the group time interaction. Assuming 2-tailed aZ.05 and autocorrelations between repeated measures of rZ.50, adequate power (.80) to detect a medium-sized group time effect (fZ.25) would be achieved with a total sample size of 34. Given the sample size of 82, observed power is more than adequate power (.99).
Results Baseline participant characteristics Of the 82 participants included in the final analysis, 80.5% (nZ66) were women and 98.8% (nZ81) were white, with 1.2% (nZ1) being African American. Most participants (59.3%, nZ48) were married or living with a partner, 70.7% (nZ58) had received a diagnosis of relapsing-remitting MS, and 70.7% (nZ58) had a bachelor’s degree or higher. Table 1 provides a comparison of baseline demographics and exercise measures for the physically improved versus physically not improved groups using independent-samples t tests and chi-square analyses. The 2 groups were equivalent at baseline on all measures.
Relationship of physical improvement to cognition After controlling for covariates, there was a significant group (physically improved vs physically not improved) time (baseline to 12wk) interaction (F4,71Z2.60, PZ.04, h2pZ.13). This effect indicated that cognitive functioning at baseline and 12 weeks differed based on the level of fitness. To clarify this interaction, contrasts compared time 2 (12wk) with baseline across the 2 groups for each cognitive test. There was a nonsignificant change for TMT-B (F1,74Z2.78, PZ.1, dZ.036). Significant results were seen for TMT-BA (F1,74Z4.54, PZ.04, dZ.58) and TMT-A (F1,74Z3.78, PZ.05, dZ.05). Custom nonorthogonal planned contrasts using PSY were conducted with TMT-A and TMT-BA, as well as TMT-B, because the interaction was approaching significance. Although there was a significant interaction between fitness groups for TMT-A, PSY-planned comparisons revealed no significant difference in the baseline scores or the 12-week scores between the 2 fitness groups. This significant interaction involving TMT-A is more likely attributable to regression to the mean rather than a change in cognition associated with change in fitness.
4
M. Beier et al Table 1
Baseline characteristics of each fitness group
Characteristics Sex Female Male Education Less than a 4-y degree 4-y degree Graduate degree MS type RRMS PPMS or SPMS Benign or unknown Ethnicity White Nonwhite Marital status* Married/living with partner Other Age (y) Timed 25-Foot Walk (s) Baseline Change score Baseline exercise measures Bicycle ergometer time (s) Isokinetic dynamometer leg extension peak torque (ft/lb) Isokinetic dynamometer leg flexion peak torque (ft/lb)
Physically Improved (nZ25)
Physically Not Improved (nZ57)
P .26
88 (22) 12 (3)
77.2 (44) 22.8 (13)
26 (15) 44 (25) 30 (17)
36 (9) 36 (9) 28 (7)
72 (18) 8 (2) 20 (5)
70.2 (40) 17.5 (10) 12.3 (7)
100 (25) 0
98.2 (56) 1.8 (1)
.66
.41
.51
.56 64 (16) 36 (9) 44.610.87
57.1 (32) 42.9 (24) 47.79.87
.21
4.621.33 .523.46
5.032.22 .03.93
.4 .26
522.17154.69 58.2935.31 29.3824.65
524.87153.79 67.1831.5 35.6818.84
.94 .28 .22
NOTE. Values are % (n), mean SD, or as otherwise indicated. Abbreviations: PPMS, primary progressive MS; RRMS, relapsing-remitting MS; SPMS, secondary progressive MS. * Missing value for one participant.
There was no significant difference in baseline scores for TMTB or TMT-BA in the physically improved group versus the physically not improved group. However, 12-week comparisons demonstrated a significant difference between the 2 groups: TMTB (PZ.05) and TMT-BA (PZ.02). As seen with the interaction graphs (fig 1), the physically improved group demonstrated better performance (decreased time to completion) at follow-up (12wk) on both TMT-B and TMT-BA, suggesting that greater physical fitness is associated with improved executive functioning on these measures. Table 2 provides baseline and follow-up scores on each cognitive outcome measure for each group.
Discussion We observed that among people with MS who wanted to exercise more, improvement on objective measures of physical fitness (muscle strength and/or aerobic endurance), was associated with improvement on measures of executive functioning (TMT-BA and TMT-B), but not on a traditional measure of speed of information processing (PASAT). This extends previous evidence supporting a connection between physical fitness and cognition in MS, which came from cross-sectional studies.20,29-31 Our results add to the literature through the use of a longitudinal design and indicate that change in physical fitness is associated with change in cognitive function. These findings parallel conclusions from a meta-analysis22 of exercise training effects on cognition in older adults suggesting that executive functioning is more sensitive to the effects of changes in physical fitness than other cognitive domains. Our findings support the notion that research should investigate
exercise as a viable treatment option for improving cognition in MS, through pilot studies and clinical trials.
Study limitations A number of study limitations should be acknowledged. First, the physically improved and physically not improved groups could have other unknown characteristics or experiences that biased them toward the differential improvement in cognitive functioning that was observed. This limitation is inherent to any correlational study. While we emphasize that improved physical fitness may be the factor driving improved executive functioning, other variables that improve with exercise, such as mood,49 may also mediate improved cognition.50 Second, our measures of strength, physical fitness, and cognition were limited. Our measure of strength was limited to leg flexion and extension tests rather than a more comprehensive strength assessment. The fitness measure consisted of a graded bicycle ergometer protocol but lacked the assessment of expired oxygen essential to standard cardiorespiratory fitness testing. Nevertheless, such an approach has been validated in MS.51 The neuropsychological evaluations in the present study were limited to only a small number of tests and to only a few cognitive domains. In addition, the TMTs are not often used in people with MS because impaired motor functioning may confound the interpretation of test performance that is partially dependent on upper extremity speed and coordination.52,53 However, the use of TMTBA partially overcomes this limitation. Future studies should include a more thorough neuropsychological assessment that takes www.archives-pmr.org
Fitness and cognition in multiple sclerosis
5
Fig 1 Time-by-group interaction for cognitive measures. *P
ORIGINAL ARTICLE
Improved Physical Fitness Correlates With Improved Cognition in Multiple Sclerosis Meghan Beier, PhD,a Charles H. Bombardier, PhD,a Narineh Hartoonian, PhD,a Robert W. Motl, PhD,b George H. Kraft, MD, MSa From the aDepartment of Rehabilitation Medicine, University of Washington, Seattle, WA; and bDepartment of Kinesiology and Community Health, University of Illinois, Urbana-Champaign, IL.
Abstract Objective: To determine whether there is an association between improvements in objective measures of physical fitness and performance on cognitive tests in people with multiple sclerosis (MS). Design: Post hoc correlational analysis in which people demonstrating physical improvement were compared with those not demonstrating physical improvement. Setting: Individuals with MS residing in the community. Participants: Adults with clinically confirmed MS (NZ88) who participated in a controlled trial of a telephone-based health promotion intervention, chose to work on exercise, and completed the pre- and postintervention assessments. Interventions: Participants were measured for strength (isokinetic dynamometer), aerobic fitness (bicycle ergometer), and cognition (Paced Auditory Serial Addition Test [PASAT], Trail Making Test [TMT]) at baseline and 12 weeks later. Change in fitness was calculated by subtracting each participant’s baseline score from the outcome score, and then transforming the difference to a z score. Individuals with a z score 1 on any fitness measure were placed in the physically improved group (nZ25). All others were in the physically not improved group (nZ57). Main Outcome Measures: TMT, PASAT. Results: After controlling for covariates (age, sex, ethnicity, education, disease activity, MS type), there was a significant group-by-time interaction, suggesting that cognitive functioning changed over time based on level of fitness. Participants in the physically improved group demonstrated improved performance on measures of executive functioning after 12 weeks of exercise. Conclusions: The results of this study lend support to the hypothesis that change in fitness is associated with improved executive functioning in people with MS. Archives of Physical Medicine and Rehabilitation 2014;-:------ª 2014 by the American Congress of Rehabilitation Medicine
Cognitive impairment is one of the most common and disabling features of multiple sclerosis (MS). Prevalence studies1-3 have reported rates of cognitive impairment ranging from 22% to 60%, depending in large part on whether the data were acquired from community or hospitalized samples and also on the assessment tool used. The cognitive deficits seen in MS appear most often in 3 domains: (1) information processing speed and working memory, Presented to the European Committee for Treatment and Research in Multiple Sclerosis October 4 2013, Copenhagen, Denmark. Supported by the Multiple Sclerosis Research and Training Grant funded by the Department of Education, National Institute on Disability and Rehabilitation Research (grant nos. H133B031129, H133B080025) and the National Multiple Sclerosis Society (grant no. MB 0008). No commercial party having a direct financial interest in the results of the research supporting this article has conferred or will confer a benefit on the authors or on any organization with which the authors are associated.
(2) new learning and episodic memory, and (3) executive functioning. MS-related cognitive dysfunction has a major impact on quality of life,4,5 activities of daily living,6,7 vocational functioning,8,9 and rehabilitation outcomes.10 Medical and rehabilitation approaches to improving cognition are disappointing. Studies of donepezil,11 psychostimulants,12 aminopyridines,13,14 ginkgo biloba,15 and memantine16 have not been consistently effective for improving cognition. Research on cognitive rehabilitation in MS suggests that learned strategies can improve cognitive performance; however, this area of investigation is in its infancy and largely focused on memory.17-19 Based on the older adult literature, exercise training represents a promising approach to improve cognitive functioning in MS.20
0003-9993/14/$36 - see front matter ª 2014 by the American Congress of Rehabilitation Medicine http://dx.doi.org/10.1016/j.apmr.2014.02.017
2
M. Beier et al
Systematic reviews and meta-analyses demonstrate that both aerobic and strength training significantly improve cognitive functioning in older adults and people with mild cognitive impairment.21-25 The most profound benefits of exercise training occurred in the domain of executive functioning,22,26,27 and the improvements seemingly depend on changes in physical fitness (defined as health-related elements such as aerobic and muscular endurance, muscle strength, body composition, and flexibility28). Much less is known about exercise training and cognition in MS. Correlational studies have reported a link between physical fitness and cognition in people with MS. For example, both higher cardiorespiratory endurance and greater muscle strength have been associated with better performance on tests of speed of information processing and working memory.29,30 Furthermore, cardiorespiratory endurance is positively associated with functional magnetic resonance imaging indices of both white and gray matter integrity.31 Nevertheless, 2 existing trials32,33 have failed to demonstrate benefits of exercise training on cognition in MS. Those trials might have been ineffective because there was poor compliance and no documentation of change in fitness levels. This underscores the importance of future research that examines the possibility that improvements in physical fitness can contribute to improved cognition in people with MS. We analyzed data from a previously published randomized controlled trial34 examining the effects of telephone counseling on health promotion in people with MS. In that trial, a subgroup of 88 people chose to work on exercise and the subjects were randomly assigned to an intervention group or waitlist control condition. Unfortunately, there were no between-group differences on objective measures of leg muscle strength, aerobic endurance, or cognition. However, the availability of pre- and postintervention data on these measures provided an opportunity to examine the relationship between objective improvement in muscle strength or aerobic endurance and improvement in cognitive functioning within the study group as a whole. To do this we created a “physically improved” group (improvement in muscle strength, aerobic endurance, or both) and a “physically not improved” group. We compared changes in speed of information processing, psychomotor speed, and executive functioning between these 2 groups. Our primary hypothesis was that after controlling for demographic and disease-related variables, the physically improved group would demonstrate significantly greater improvement on measures of speed of information processing, psychomotor speed, and executive functioning compared with the physically not improved group.
Methods Participants Subjects were 88 community-dwelling individuals with clinically confirmed MS who wanted help with exercise as part of the previously published health promotion trial.34 Participants were recruited from the University of Washington MS Center, MS
List of abbreviations: MS PASAT rpm TMT-A TMT-B TMT-BA
multiple sclerosis Paced Auditory Serial Addition Test revolutions per minute Trail Making TestePart A Trail Making TestePart B Trail Making TestePart B minus Part A
newsletters, newspaper advertisements and articles, support groups in the Puget Sound region, and from flyers mailed to local neurology offices. Participants were also recruited from a large survey study34 of persons with MS. The inclusion criteria for participation were (1) age 18 years; (2) diagnosis of MS confirmed by a neurologist or physiatrist specializing in MS; (3) ability to walk 300ft (90m) without assistance, equating to an Expanded Disability Status Scale score of 5.535; and (4) interest in receiving help to exercise more. All MS types (relapsing-remitting, secondary progressive, primary progressive, progressive-relapsing) were included. Participants were excluded if they (1) had a medical condition that contraindicated exercise, or (2) screened positive for depression on the Primary Care Evaluation of Mental Disorders.36 The Primary Care Evaluation of Mental Disorders is a 26-item screener of “yes” or “no” questions for a variety of symptoms. The 2 depression questions ask about depressed mood and anhedonia. Six participants were excluded from the final analysis because of missing data. Two individuals were missing all fitness measures, and 4 were missing cognitive data. This resulted in 82 participants who were included in the final analysis.
Procedures The University of Washington Institutional Review Board approved the study protocol. Before data collection, all participants provided written informed consent. Participants meeting inclusion criteria during a telephone interview were invited to the study center for a baseline assessment. Trained research assistants, supervised by study investigators, conducted baseline assessments in dedicated laboratory space at the medical center. Afterward, participants chose a health promotion activity (exercise in the present analysis) for which they wanted help. Trained research assistants conducted in-person outcome assessments 12 weeks postrandomization. Baseline and outcome assessments were conducted in the same laboratory space with the temperature controlled between 19 C and 22 C.
Measures Self-reported demographic and medical variables such as age, ethnicity, marital status, educational level, date of diagnosis, and MS subtype37 were collected during a telephone screen. Measure of disease activity The Timed 25-Foot Walk38 is a clinical tool used to measure the stability of MS disease. Research suggests this is a useful tool for detecting stability of disease and the presence of an exacerbation.38 The 25-Foot Walk was measured at baseline and follow-up. The change score was used as a covariate for controlling for disease activity. Measures of strength and fitness Right and left leg flexion and extension muscle strength were measured on a Cybex 6000 isokinetic dynamometer.a The Cybex was calibrated weekly using the manufacturer’s procedures. Each subject was given 5 submaximal trial repetitions before testing. To obtain peak torque output for knee flexion and extension, 10 maximal repetitions were performed by each leg at a rate of 60 per minute. Leg strength was defined as peak torque averaged over 10 repetitions for the right and left legs and for both extension and flexion. www.archives-pmr.org
Fitness and cognition in multiple sclerosis Aerobic fitness was estimated using a Schwinn Airdyne bicycle ergometer.b Subjects were asked to pedal the ergometer at 60 revolutions per minute (rpm) for as long as they could while pedal resistance was increased. Resistance began at 0W and increased incrementally by 0.5W every 2 minutes until the participant could no longer sustain 60rpm. The number of seconds the subject could sustain pedaling at 60rpm was the measure of aerobic fitness. The original trial collected minutes of self-reported physical activity. However, as the original study was done to examine behavior change based on a motivational interviewing protocol, individuals were asked to choose any fitness activity in which they wanted to engage. This could have included activities such as stretching, which may not be associated with an objective change in physical fitness. We determined that a significant change on objective measures of physical fitness would be a more effective way to categorize changes in physical fitness rather than relying on self-report. Measures of strength and fitness: physically improved versus physically not improved groups Group definitions were based on the combination of changes in leg muscle strength and aerobic fitness. Baseline scores were subtracted from the outcome score for each individual. The change scores were then converted to z scores. We defined significant improvement as a z score 1. Participants who demonstrated significant improvement in leg muscle strength or aerobic fitness were included in the physically improved group (nZ25). Participants who did not demonstrate significant improvement in leg strength or fitness were categorized as physically not improved (nZ57). Measures of cognition The Paced Auditory Serial Addition Test (PASAT) measures information processing speed, flexibility, calculation ability,39 divided attention, and working memory.40 The PASAT requires the individual to listen to a series of single-digit numbers presented at 3-second intervals during the first trial and 2-second intervals during a second trial. The subject states the sum of each pair of contiguous numbers to the examiner. The outcome variable is the number of correct responses. The Trail Making Test (TMT) is a widely used 2-part test. Part A (TMT-A) is a test of psychomotor speed, processing speed, and visual scanning. It requires the subject to connect a series of numbered circles in a paper-and-pencil format. Part B (TMT-B) is similar but requires the subject to follow an alternating sequence (1-A-2-B-3-C.) and is thought to measure working memory, setswitching, and executive functioning. The main dependent factor is time to complete each sequence.41 Subtracting the time to complete TMT-A from TMT-B (TMT-BA) minimizes the contribution of visual scanning, working memory, and psychomotor speed demands, producing a purer measure of executive functioning.41,42
Statistical analysis Statistical analyses were performed using IBM SPSS Statistics version 20.c Mahalanobis and Cook’s distance were used to screen for multivariate outliers. No univariate or multivariate outliers were detected, and assumptions regarding normality of sampling distributions were met. Data were analyzed using repeatedmeasure multivariate analysis of covariance, with physical improvement (at 2 levels, physically improved and physically not improved) as the between-subjects factor and time (at 2 levels,
www.archives-pmr.org
3 preintervention and postintervention) as the within-subjects factor. To test the hypothesis that improved physical fitness was associated with cognitive improvement, we examined the 2-way physical improvement (improved vs not improved) by time interaction. Within physical improvement, change was assessed by a 1 degree of freedom contrast. Two-tailed P values were reported for the relevant effects. Age, sex, ethnicity, education, disease activity, and MS type were chosen as covariates because previous research demonstrated they could have a relationship with cognition.43-46 Significant interaction effects were followed by custom nonorthogonal planned contrasts. Significant interaction effects are decomposed using simple effects post hoc analyses. Because SPSS is not amenable to nonstandard contrast analyses within a mixed-model design, planned comparisons were performed using PSY: A Program for Contrast Analysis.d PSY allows for the control of family-wise error rates for custom-planned contrast analysis. PSY can also supply standard or Bonferroni-adjusted critical values. Power analyses were conducted with GPower 3,47,e evaluating the a priori sample size required to detect a medium-sized effect (fZ.25)48 for the group time interaction. Assuming 2-tailed aZ.05 and autocorrelations between repeated measures of rZ.50, adequate power (.80) to detect a medium-sized group time effect (fZ.25) would be achieved with a total sample size of 34. Given the sample size of 82, observed power is more than adequate power (.99).
Results Baseline participant characteristics Of the 82 participants included in the final analysis, 80.5% (nZ66) were women and 98.8% (nZ81) were white, with 1.2% (nZ1) being African American. Most participants (59.3%, nZ48) were married or living with a partner, 70.7% (nZ58) had received a diagnosis of relapsing-remitting MS, and 70.7% (nZ58) had a bachelor’s degree or higher. Table 1 provides a comparison of baseline demographics and exercise measures for the physically improved versus physically not improved groups using independent-samples t tests and chi-square analyses. The 2 groups were equivalent at baseline on all measures.
Relationship of physical improvement to cognition After controlling for covariates, there was a significant group (physically improved vs physically not improved) time (baseline to 12wk) interaction (F4,71Z2.60, PZ.04, h2pZ.13). This effect indicated that cognitive functioning at baseline and 12 weeks differed based on the level of fitness. To clarify this interaction, contrasts compared time 2 (12wk) with baseline across the 2 groups for each cognitive test. There was a nonsignificant change for TMT-B (F1,74Z2.78, PZ.1, dZ.036). Significant results were seen for TMT-BA (F1,74Z4.54, PZ.04, dZ.58) and TMT-A (F1,74Z3.78, PZ.05, dZ.05). Custom nonorthogonal planned contrasts using PSY were conducted with TMT-A and TMT-BA, as well as TMT-B, because the interaction was approaching significance. Although there was a significant interaction between fitness groups for TMT-A, PSY-planned comparisons revealed no significant difference in the baseline scores or the 12-week scores between the 2 fitness groups. This significant interaction involving TMT-A is more likely attributable to regression to the mean rather than a change in cognition associated with change in fitness.
4
M. Beier et al Table 1
Baseline characteristics of each fitness group
Characteristics Sex Female Male Education Less than a 4-y degree 4-y degree Graduate degree MS type RRMS PPMS or SPMS Benign or unknown Ethnicity White Nonwhite Marital status* Married/living with partner Other Age (y) Timed 25-Foot Walk (s) Baseline Change score Baseline exercise measures Bicycle ergometer time (s) Isokinetic dynamometer leg extension peak torque (ft/lb) Isokinetic dynamometer leg flexion peak torque (ft/lb)
Physically Improved (nZ25)
Physically Not Improved (nZ57)
P .26
88 (22) 12 (3)
77.2 (44) 22.8 (13)
26 (15) 44 (25) 30 (17)
36 (9) 36 (9) 28 (7)
72 (18) 8 (2) 20 (5)
70.2 (40) 17.5 (10) 12.3 (7)
100 (25) 0
98.2 (56) 1.8 (1)
.66
.41
.51
.56 64 (16) 36 (9) 44.610.87
57.1 (32) 42.9 (24) 47.79.87
.21
4.621.33 .523.46
5.032.22 .03.93
.4 .26
522.17154.69 58.2935.31 29.3824.65
524.87153.79 67.1831.5 35.6818.84
.94 .28 .22
NOTE. Values are % (n), mean SD, or as otherwise indicated. Abbreviations: PPMS, primary progressive MS; RRMS, relapsing-remitting MS; SPMS, secondary progressive MS. * Missing value for one participant.
There was no significant difference in baseline scores for TMTB or TMT-BA in the physically improved group versus the physically not improved group. However, 12-week comparisons demonstrated a significant difference between the 2 groups: TMTB (PZ.05) and TMT-BA (PZ.02). As seen with the interaction graphs (fig 1), the physically improved group demonstrated better performance (decreased time to completion) at follow-up (12wk) on both TMT-B and TMT-BA, suggesting that greater physical fitness is associated with improved executive functioning on these measures. Table 2 provides baseline and follow-up scores on each cognitive outcome measure for each group.
Discussion We observed that among people with MS who wanted to exercise more, improvement on objective measures of physical fitness (muscle strength and/or aerobic endurance), was associated with improvement on measures of executive functioning (TMT-BA and TMT-B), but not on a traditional measure of speed of information processing (PASAT). This extends previous evidence supporting a connection between physical fitness and cognition in MS, which came from cross-sectional studies.20,29-31 Our results add to the literature through the use of a longitudinal design and indicate that change in physical fitness is associated with change in cognitive function. These findings parallel conclusions from a meta-analysis22 of exercise training effects on cognition in older adults suggesting that executive functioning is more sensitive to the effects of changes in physical fitness than other cognitive domains. Our findings support the notion that research should investigate
exercise as a viable treatment option for improving cognition in MS, through pilot studies and clinical trials.
Study limitations A number of study limitations should be acknowledged. First, the physically improved and physically not improved groups could have other unknown characteristics or experiences that biased them toward the differential improvement in cognitive functioning that was observed. This limitation is inherent to any correlational study. While we emphasize that improved physical fitness may be the factor driving improved executive functioning, other variables that improve with exercise, such as mood,49 may also mediate improved cognition.50 Second, our measures of strength, physical fitness, and cognition were limited. Our measure of strength was limited to leg flexion and extension tests rather than a more comprehensive strength assessment. The fitness measure consisted of a graded bicycle ergometer protocol but lacked the assessment of expired oxygen essential to standard cardiorespiratory fitness testing. Nevertheless, such an approach has been validated in MS.51 The neuropsychological evaluations in the present study were limited to only a small number of tests and to only a few cognitive domains. In addition, the TMTs are not often used in people with MS because impaired motor functioning may confound the interpretation of test performance that is partially dependent on upper extremity speed and coordination.52,53 However, the use of TMTBA partially overcomes this limitation. Future studies should include a more thorough neuropsychological assessment that takes www.archives-pmr.org
Fitness and cognition in multiple sclerosis
5
Fig 1 Time-by-group interaction for cognitive measures. *P
