ORIGINAL ARTICLE
Exercise and Cognitive Function: A Randomized Controlled Trial Examining Acute Exercise and Free-Living Physical Activity and Sedentary Effects Paul D. Loprinzi, PhD, and Christy J. Kane, PhD Abstract Objective: To simultaneously examine the effects of acute exercise intensity and free-living physical activity and sedentary behavior on cognitive function in young, healthy adults. Patients and Methods: Using a counterbalanced, crossover, randomized controlled design, 87 young adults (mean age, 21.4 years) completed various cognitive assessments with and without an acute bout of exercise preceding the assessment. Participants were randomized into 1 of 4 groups to complete a 30-minute bout of acute exercise: control (no exercise), light intensity (40%-50% of predicted maximum heart rate [HRmax]), moderate intensity (51%-70% of predicted HRmax), or vigorous intensity (71%-85% of predicted HRmax). Subjectively and objectively determined (accelerometry) physical activity and sedentary behavior were assessed to examine the association between these free-living behaviors and cognitive function. The study duration was August 26, 2013, to September 11, 2014. Results: Concentration-related cognition (mean SD Feature Match test score) was significantly higher after a 30-minute acute bout of moderate-intensity exercise (145.126.9) compared with cognitive assessment without exercise (121.319.2; P¼.004). Furthermore, questionnaire-determined sedentary behavior was inversely associated with visual attention and task switching (Trail Making Test A score) (b¼e0.23; P¼.04). Last, estimated cardiorespiratory fitness (volume of maximum oxygen consumption) was positively associated with reasoning-related cognitive function (Odd One Out test score) (b¼0.49; P¼.05); when adding metabolic equivalent of task minutes per week to this model, the results were not significant (b¼0.47; P¼.07). Conclusion: These findings provide some support for acute moderate-intensity exercise, sedentary behavior, and cardiorespiratory fitness being associated with executive functioningerelated cognitive function in young, healthy adults. ª 2015 Mayo Foundation for Medical Education and Research
merging research is starting to find neuroprotective effects of physical activity (PA) on the brain,1 particularly in youth2 and older adults,3,4 with fewer investigations in young, healthy adults.5 There is also some evidence indicating that some aspects of cognitive function may start to decline in the early adult years (ie, 20s).6 The underlying mechanisms through which PA may improve cognition are likely a result of PA-induced changes at the systemic, molecular, and cellular levels.1,3 For example, PA may influence neural systems (eg, improved information processing and memory encoding) involved in attention, learning, and memory7; increase molecular mediators (eg, brain-derived neurotrophic factor) by which
E
n
Mayo Clin Proc. 2015;nn(n):1-11
PA affects cognition; and promote a cellular environment that enhances cognition through PA-induced neurogenesis and vascular function.3,8 In addition to the need for further studies in young, healthy adults, delineating the effects of different exercise intensity levels on cognition is important; some evidence suggests a differential effect of exercise intensity on cognition. For example, studies have reported that lower-intensity exercise may be more beneficial regarding brain protection and restoration9,10 compared with vigorousintensity exercise, which may result in much larger increases in catecholamine levels, ultimately inducing neural noise and
Mayo Clin Proc. n XXX 2015;nn(n):1-11 n http://dx.doi.org/10.1016/j.mayocp.2014.12.023 www.mayoclinicproceedings.org n ª 2015 Mayo Foundation for Medical Education and Research
From the Department of Health, Exercise Science, and Recreation Management, School of Applied Sciences, Center for Health Behavior Research, The University of Mississippi, University (P.D.L.); and Department of Respiratory Therapy, Donna and Allan Lansing School of Nursing and Health Sciences, Bellarmine University, Louisville, KY (C.J.K.).
1
MAYO CLINIC PROCEEDINGS
inhibiting exercise-induced cognitive changes.11 Furthermore, high-intensity exercise may require greater activation of the premotor and supplementary motor areas of the brain, which may result in less cognitive adaption as these areas of the brain will be activated at the expense of the prefrontal cortex. Moderateintensity exercise should, in theory, facilitate cognition because this intensity should be high enough to elicit changes in brain neurotransmitters but low enough to prevent large increases in catecholamine levels that may be induced during vigorous-intensity PA. In addition to the potential differential effects of exercise intensity on cognition, forced vs voluntary/habitual exercise may have differential effects on brain function. Forced exercise is often defined as exercise that is augmented mechanically to assist the individual in achieving or maintaining a given heart rate (HR).12 Forced exercise has been found to have neurotransmitter-sparing effects.13 However, there is also some evidence suggesting that habitual free-living PA, compared with forced exercise, has a greater ability to elicit higher concentrations of brain-derived neurotrophic factor and to induce less of a corticosterone stress response.14 In contrast, other studies15 have found that forced exercise, compared with habitual PA, may be more beneficial in increasing cerebral blood flow and cerebral glycolysis. Most of the human intervention studies examining the effect of exercise on cognition have used a forced exercise model, often consisting of acute exercise followed by cognitive testing. We are aware of no studies that have simultaneously considered acute exercise and freeliving PA and their potential relationships with cognition. To address these gaps in the literature, the primary purposes of this study were to examine the effects of exercise on cognitive function in young, healthy adults and, specifically, to address the role of exercise intensity and acute vs habitual exercise on cognition. Furthermore, we examined whether exercise is associated with cognition independent of mood and anxiety because it is unclear whether the effects of exercise on cognition are independent of changes in mood and anxiety. That is, it is unclear whether exerciseinduced effects on cognition are due to 2
Mayo Clin Proc.
n
exercise-induced changes in affect or whether separate neural systems mediate these effects. METHODS Sample and Design Participants were eligible for the study if they were 35 years or younger, were “ready” to engage in PA as determined by the Physical Activity Readiness Questionnaire, spoke English, and provided written informed consent. Participants were excluded from the study if they perceived having any difficulty completing all the tests or had a current illness. No recruited participants were excluded. The study was conducted between August 26, 2013, and September 11, 2014. Eighty-seven young (mean SD age, 21.42.1 years) (Table 1), healthy (as determined by the Physical Activity Readiness Questionnaire) adults were recruited to participate, with all the participants providing informed consent; all the study procedures were approved by Bellarmine University’s institutional review board. Participants were recruited by the student researchers using a nonprobability convenience sampling approach at Bellarmine University (ie, student researchers proposed the study to students enrolled in university courses). Participants completed 2 visits (at approximately the same time of day; 30 minutes) approximately 1 week apart. Participants were asked not to exercise and not to consume any stimulants (eg, caffeine and cigarettes) within 3 hours of the visit. Of the 87 participants, 86 (99%) complied and did not exercise within 3 hours of testing, and 75 (86%) did not consume any stimulants within 3 hours of testing. Initial analyses were computed that statistically controlled for whether participants complied with these stipulations, as well as analyses that excluded these participants, and the results were unchanged; therefore, these participants were included in the analyses. During both visits (for test-retest reliability), participants completed a variety of questionnaires, including an assessment of self-reported PA, sedentary behavior, and affect (see later herein for further details). Height and weight were measured during the first visit to calculate body mass index (BMI; calculated as the weight in kilograms divided by the height in meters squared). During the first visit, participants
XXX 2015;nn(n):1-11
n
http://dx.doi.org/10.1016/j.mayocp.2014.12.023 www.mayoclinicproceedings.org
EXERCISE AND COGNITION
TABLE 1. Characteristics of the Analyzed Sample Across Experimental Groupsa Mean/proportion (95% CI)
Variable
Entire sample (N¼87)
Control group (n¼19)
Moderate-intensity exercise group (n¼22)
Vigorous-intensity exercise group (n¼22)
(20.4-21.8)
20.8 (20.1-21.4)
22.3 (21.1-23.5)
(42.4-82.5) (17.4-57.5)
59.0 (37.7-80.4) 40.9 (19.5-62.2)
68.1 (47.9-88.3) 31.8 (11.6-52.0)
(87.5-100.0) (0.0-12.4)
100 0
100 0
Light-intensity exercise group (n¼24)
Age (y), mean 21.4 (20.9-21.8) 21.3 (20.4-22.2) 21.1 Sex (%) Male 58.6 (48.0-69.1) 42.1 (18.9-65.2) 62.5 Female 41.3 (30.8-51.9) 57.8 (34.7-81.0) 37.5 Race/ethnicity (%) Non-Hispanic white 97.7 (94.4-100.0) 94.7 (84.2-100.0) 95.8 Other 2.3 (0.0-3.4) 5.2 (0.0-15.7) 4.1 Education (%) In college 85.0 (77.4-92.6) 84.2 (67.1-100.0) 91.6 Undergraduate degree 10.3 (3.8-16.8) 10.5 (0.0-24.9) 4.1 Master’s degree or more 4.5 (0.1-9.0) 5.2 (0.0-15.7) 4.1 Body mass index, mean 24.1 (23.3-24.9) 25.2 (23.1-27.3) 24.2 Affectc 39.2 (37.3-41.1) 36.2 (33.9-38.5) 39.8 Accelerometry data Sedentary (min/d), meand 518.3 (497-539) 526.4 (477-575) 519.4 MVPA (min/d), meand 41.2 (34.5-47.8) 45.2 (24.7-65.7) 31.7 Accelerometer wear time 782.6 (763-801) 785.1 (742-827) 774.4 (min/d), meand No. of valid accelerometer days, meand 4.1 (3.6-4.5) 4.5 (3.3-5.7) 4.2 Estimated from survey VO2max (mL/kg/min), meane 47.1 (45.4-48.8) 42.9 (38.9-46.9) 47.6 421.2 (384-457) 435.0 (354-515) 434.3 Sedentary (min/d), meanf 5285 (4427-6143) 3449.1 (2372-4525) 6841.4 MET-min/wk, meang
P valueb .08 .37
.39
.86 (80.2-100.0) (0.0-12.4) (0.0-12.4) (22.3-26.0) (36.1-43.5)
86.3 9.1 4.5 23.1 42.7
(71.4-100.0) (0.0-21.5) (0.0-13.5) (22.2-23.9) (37.6-47.8)
77.2 18.1 4.5 24.0 37.5
(59.0-95.4) (1.4-34.9) (0.0-13.5) (22.7-25.2) (34.8-40.2)
.31 .09
(483-555) (22.0-41.3)
505.6 (457-553) 47.2 (34.5-59.9)
522.5 (483-562) 41.8 (31.2-51.7)
.91 .36
(744-804)
789.5 (748-830)
782.1 (740-823)
.95
4.0 (3.3-4.5)
3.8 (2.8-4.8)
.71
(3.2-5.0)
(44.7-50.5) 49.7 (46.8-52.5) 47.8 (44.1-51.4) (347-521) 409.1 (349-468) 407.1 (341-472) (4514-9168) 5120.0 (3863-6376) 5325.6 (4068-6583)
.04 .91 .05
MET-min/wk ¼ metabolic equivalent of task minutes per week; MVPA ¼ moderate-to-vigorous physical activity; VO2max ¼ volume of maximum oxygen consumption. One-way analysis of variance was used to examine whether there were any differences in continuous variables (eg, age) across the 4 groups; c2 analysis was used to test differences for categorical variables (eg, sex). c The summed affect score from the depression, anxiety, and fatigue subscales of the Profile of Mood States questionnaire. d Estimates are from the accelerometry data (n¼78); a valid accelerometer day of monitoring was defined as at least 10 h/d of wear time. e Estimated from the prediction equation.37 f Estimated from the adult Sedentary Behavior Questionnaire (average of the 2 assessments). g Estimated from the International Physical Activity Questionnaire (average of the 2 assessments). a
b
were given a GT1M accelerometer (ActiGraph LLC) to wear for the following 7 days; they returned it during their second visit to the laboratory. Using a counterbalanced, crossover, randomized controlled design (Figure), cognitive function was assessed after an acute bout of treadmill exercise (30 minutes), with participants randomized to 1 of 4 exercise groups: no exercise or light-, moderate-, or vigorousintensity exercise. To minimize potential carryover effects in the cognitive assessments, an approximately 1-week washout period elapsed between the 2 visits. Carryover effect analyses (Koch analyses) were used by ranking the sum scores for each cognitive test across both Mayo Clin Proc. n XXX 2015;nn(n):1-11 www.mayoclinicproceedings.org
n
sequences (eg, AB/BA) and using nonparametric analyses to examine differences in the ranks. No carryover effects were observed for any of the 10 cognitive assessments (P>.05 for all the tests). Allocation concealment was used by not informing the participants of which tests (eg, cognitive testing only or cognitive testing after treadmill exercise) would take place at the visit. Randomization for the group assignment (ie, control or light-, moderate-, or vigorousintensity exercise), randomization for the crossover design (ie, at which visit they completed the treadmill exercise; AB/BA), and randomization for the order of the cognitive tests was conducted using the Excel random list (RAND) function (Microsoft Corp).
http://dx.doi.org/10.1016/j.mayocp.2014.12.023
3
MAYO CLINIC PROCEEDINGS
Control
No PA
PA
Cognition
No PA
Cognition
PA
Cognition
Cognition
Washout
Light-intensity exercise No PA
Cognition No PA
PA
Cognition
Cognition PA
Cognition
Washout Random allocation
Moderate-intensity exercise No PA
Cognition
PA
Cognition
No PA PA
Cognition Cognition
Washout
Vigorous-intensity exercise No PA
Cognition No PA
Cognition
1-week accelerometry assessment
FIGURE. Schematic of crossover study design. Physical activity (PA) / cognition ¼ acute 30-minute exercise followed by cognitive assessment; No PA / cognition ¼ no exercise before cognitive assessment. Washout was 1 week.
Acute Exercise Testing Participants were randomized to 1 of 4 acute exercise groups based on guidelines proposed by Mayo Clinic16: the control group (no exercise), the light-intensity exercise group (40%50% of predicted [220-age] maximum HR [HRmax]; calculation of HRmax was determined by subtracting the participant’s age from 220), the moderate-intensity exercise group (51%70% of predicted HRmax), or the vigorousintensity exercise group (71%-85% of predicted HRmax). During the 30-minute bout of exercise on the treadmill, HR (HR monitor; Polar Electro Inc) and rating of perceived exertion (10-point Borg scale) were assessed every 5 minutes. Based on expected mean and SD values from previous studies,17 using a 2-tailed test with an effect size d of 0.62 and alpha error probability of 0.05, 22 participants in each acute exercise group were needed to have an achieved power (1ebeta error probability) of 0.80 to detect differences in the evaluated cognitive function parameters. Of the 87 participants, 19 (22%) were randomized to the control group, 24 (28%) to the light-intensity group, 22 (25%) to the moderate-intensity group, 4
Mayo Clin Proc.
n
and 22 (25%) to the vigorous-intensity group. Cognitive Tests. During the visit at which the participant completed the cognitive assessments after the acute bout of exercise, the cognitive assessments started when the postexercise HR was within 10% of the baseline HR or 15 minutes after the acute exercise ended, whichever came first. Participants completed several cognitive tests that assessed different areas of brain function and varied in task complexity; some evidence suggests that cognitive task complexity may moderate the relationship between exercise and cognition, and different areas of the brain (eg, frontal lobe and temporal lobe) may be differentially influenced by exercise.18 Ten cognitive function tests were administered, including 2 paper-and-pencil tests (Trail Making Test A and B) assessing cognitive visual attention and task switching.19 The remaining 8 tests were administered using electronic software: Spatial Span20 and Paired Associates21,22 (to assess memory), Grammatical Reasoning and Odd
XXX 2015;nn(n):1-11
n
http://dx.doi.org/10.1016/j.mayocp.2014.12.023 www.mayoclinicproceedings.org
EXERCISE AND COGNITION
One Out (to assess reasoning),23 Feature Match24,25 and Polygon26 (to assess concentration), and Spatial Search27 and Spatial Slider28 (to assess planning). The cognitive tests were administered in a randomized order. Affect. The Profile of Mood States questionnaire was used to assess affect, with the depression/dejection (13 items), anger/hostility (11 items), and fatigue/inertia (7 items) subscales used for the present study. The Profile of Mood States questionnaire has demonstrated reasonable internal consistency on the subscales (eg, a¼0.96 for depression).29 In the present sample, internal consistencies, as measured by the Cronbach a, for the subscales were as follows (nontreadmill and treadmill visits): depression/dejection (a¼0.84 and a¼0.86), anger/hostility (a¼0.86 and a¼0.75), and fatigue/inertia (a¼0.80 and a¼0.81). In the present sample, 1-week test-retest reliability for the depression/dejection subscale was intraclass correlation coefficient (ICC)¼0.74 (P
Exercise and Cognitive Function: A Randomized Controlled Trial Examining Acute Exercise and Free-Living Physical Activity and Sedentary Effects Paul D. Loprinzi, PhD, and Christy J. Kane, PhD Abstract Objective: To simultaneously examine the effects of acute exercise intensity and free-living physical activity and sedentary behavior on cognitive function in young, healthy adults. Patients and Methods: Using a counterbalanced, crossover, randomized controlled design, 87 young adults (mean age, 21.4 years) completed various cognitive assessments with and without an acute bout of exercise preceding the assessment. Participants were randomized into 1 of 4 groups to complete a 30-minute bout of acute exercise: control (no exercise), light intensity (40%-50% of predicted maximum heart rate [HRmax]), moderate intensity (51%-70% of predicted HRmax), or vigorous intensity (71%-85% of predicted HRmax). Subjectively and objectively determined (accelerometry) physical activity and sedentary behavior were assessed to examine the association between these free-living behaviors and cognitive function. The study duration was August 26, 2013, to September 11, 2014. Results: Concentration-related cognition (mean SD Feature Match test score) was significantly higher after a 30-minute acute bout of moderate-intensity exercise (145.126.9) compared with cognitive assessment without exercise (121.319.2; P¼.004). Furthermore, questionnaire-determined sedentary behavior was inversely associated with visual attention and task switching (Trail Making Test A score) (b¼e0.23; P¼.04). Last, estimated cardiorespiratory fitness (volume of maximum oxygen consumption) was positively associated with reasoning-related cognitive function (Odd One Out test score) (b¼0.49; P¼.05); when adding metabolic equivalent of task minutes per week to this model, the results were not significant (b¼0.47; P¼.07). Conclusion: These findings provide some support for acute moderate-intensity exercise, sedentary behavior, and cardiorespiratory fitness being associated with executive functioningerelated cognitive function in young, healthy adults. ª 2015 Mayo Foundation for Medical Education and Research
merging research is starting to find neuroprotective effects of physical activity (PA) on the brain,1 particularly in youth2 and older adults,3,4 with fewer investigations in young, healthy adults.5 There is also some evidence indicating that some aspects of cognitive function may start to decline in the early adult years (ie, 20s).6 The underlying mechanisms through which PA may improve cognition are likely a result of PA-induced changes at the systemic, molecular, and cellular levels.1,3 For example, PA may influence neural systems (eg, improved information processing and memory encoding) involved in attention, learning, and memory7; increase molecular mediators (eg, brain-derived neurotrophic factor) by which
E
n
Mayo Clin Proc. 2015;nn(n):1-11
PA affects cognition; and promote a cellular environment that enhances cognition through PA-induced neurogenesis and vascular function.3,8 In addition to the need for further studies in young, healthy adults, delineating the effects of different exercise intensity levels on cognition is important; some evidence suggests a differential effect of exercise intensity on cognition. For example, studies have reported that lower-intensity exercise may be more beneficial regarding brain protection and restoration9,10 compared with vigorousintensity exercise, which may result in much larger increases in catecholamine levels, ultimately inducing neural noise and
Mayo Clin Proc. n XXX 2015;nn(n):1-11 n http://dx.doi.org/10.1016/j.mayocp.2014.12.023 www.mayoclinicproceedings.org n ª 2015 Mayo Foundation for Medical Education and Research
From the Department of Health, Exercise Science, and Recreation Management, School of Applied Sciences, Center for Health Behavior Research, The University of Mississippi, University (P.D.L.); and Department of Respiratory Therapy, Donna and Allan Lansing School of Nursing and Health Sciences, Bellarmine University, Louisville, KY (C.J.K.).
1
MAYO CLINIC PROCEEDINGS
inhibiting exercise-induced cognitive changes.11 Furthermore, high-intensity exercise may require greater activation of the premotor and supplementary motor areas of the brain, which may result in less cognitive adaption as these areas of the brain will be activated at the expense of the prefrontal cortex. Moderateintensity exercise should, in theory, facilitate cognition because this intensity should be high enough to elicit changes in brain neurotransmitters but low enough to prevent large increases in catecholamine levels that may be induced during vigorous-intensity PA. In addition to the potential differential effects of exercise intensity on cognition, forced vs voluntary/habitual exercise may have differential effects on brain function. Forced exercise is often defined as exercise that is augmented mechanically to assist the individual in achieving or maintaining a given heart rate (HR).12 Forced exercise has been found to have neurotransmitter-sparing effects.13 However, there is also some evidence suggesting that habitual free-living PA, compared with forced exercise, has a greater ability to elicit higher concentrations of brain-derived neurotrophic factor and to induce less of a corticosterone stress response.14 In contrast, other studies15 have found that forced exercise, compared with habitual PA, may be more beneficial in increasing cerebral blood flow and cerebral glycolysis. Most of the human intervention studies examining the effect of exercise on cognition have used a forced exercise model, often consisting of acute exercise followed by cognitive testing. We are aware of no studies that have simultaneously considered acute exercise and freeliving PA and their potential relationships with cognition. To address these gaps in the literature, the primary purposes of this study were to examine the effects of exercise on cognitive function in young, healthy adults and, specifically, to address the role of exercise intensity and acute vs habitual exercise on cognition. Furthermore, we examined whether exercise is associated with cognition independent of mood and anxiety because it is unclear whether the effects of exercise on cognition are independent of changes in mood and anxiety. That is, it is unclear whether exerciseinduced effects on cognition are due to 2
Mayo Clin Proc.
n
exercise-induced changes in affect or whether separate neural systems mediate these effects. METHODS Sample and Design Participants were eligible for the study if they were 35 years or younger, were “ready” to engage in PA as determined by the Physical Activity Readiness Questionnaire, spoke English, and provided written informed consent. Participants were excluded from the study if they perceived having any difficulty completing all the tests or had a current illness. No recruited participants were excluded. The study was conducted between August 26, 2013, and September 11, 2014. Eighty-seven young (mean SD age, 21.42.1 years) (Table 1), healthy (as determined by the Physical Activity Readiness Questionnaire) adults were recruited to participate, with all the participants providing informed consent; all the study procedures were approved by Bellarmine University’s institutional review board. Participants were recruited by the student researchers using a nonprobability convenience sampling approach at Bellarmine University (ie, student researchers proposed the study to students enrolled in university courses). Participants completed 2 visits (at approximately the same time of day; 30 minutes) approximately 1 week apart. Participants were asked not to exercise and not to consume any stimulants (eg, caffeine and cigarettes) within 3 hours of the visit. Of the 87 participants, 86 (99%) complied and did not exercise within 3 hours of testing, and 75 (86%) did not consume any stimulants within 3 hours of testing. Initial analyses were computed that statistically controlled for whether participants complied with these stipulations, as well as analyses that excluded these participants, and the results were unchanged; therefore, these participants were included in the analyses. During both visits (for test-retest reliability), participants completed a variety of questionnaires, including an assessment of self-reported PA, sedentary behavior, and affect (see later herein for further details). Height and weight were measured during the first visit to calculate body mass index (BMI; calculated as the weight in kilograms divided by the height in meters squared). During the first visit, participants
XXX 2015;nn(n):1-11
n
http://dx.doi.org/10.1016/j.mayocp.2014.12.023 www.mayoclinicproceedings.org
EXERCISE AND COGNITION
TABLE 1. Characteristics of the Analyzed Sample Across Experimental Groupsa Mean/proportion (95% CI)
Variable
Entire sample (N¼87)
Control group (n¼19)
Moderate-intensity exercise group (n¼22)
Vigorous-intensity exercise group (n¼22)
(20.4-21.8)
20.8 (20.1-21.4)
22.3 (21.1-23.5)
(42.4-82.5) (17.4-57.5)
59.0 (37.7-80.4) 40.9 (19.5-62.2)
68.1 (47.9-88.3) 31.8 (11.6-52.0)
(87.5-100.0) (0.0-12.4)
100 0
100 0
Light-intensity exercise group (n¼24)
Age (y), mean 21.4 (20.9-21.8) 21.3 (20.4-22.2) 21.1 Sex (%) Male 58.6 (48.0-69.1) 42.1 (18.9-65.2) 62.5 Female 41.3 (30.8-51.9) 57.8 (34.7-81.0) 37.5 Race/ethnicity (%) Non-Hispanic white 97.7 (94.4-100.0) 94.7 (84.2-100.0) 95.8 Other 2.3 (0.0-3.4) 5.2 (0.0-15.7) 4.1 Education (%) In college 85.0 (77.4-92.6) 84.2 (67.1-100.0) 91.6 Undergraduate degree 10.3 (3.8-16.8) 10.5 (0.0-24.9) 4.1 Master’s degree or more 4.5 (0.1-9.0) 5.2 (0.0-15.7) 4.1 Body mass index, mean 24.1 (23.3-24.9) 25.2 (23.1-27.3) 24.2 Affectc 39.2 (37.3-41.1) 36.2 (33.9-38.5) 39.8 Accelerometry data Sedentary (min/d), meand 518.3 (497-539) 526.4 (477-575) 519.4 MVPA (min/d), meand 41.2 (34.5-47.8) 45.2 (24.7-65.7) 31.7 Accelerometer wear time 782.6 (763-801) 785.1 (742-827) 774.4 (min/d), meand No. of valid accelerometer days, meand 4.1 (3.6-4.5) 4.5 (3.3-5.7) 4.2 Estimated from survey VO2max (mL/kg/min), meane 47.1 (45.4-48.8) 42.9 (38.9-46.9) 47.6 421.2 (384-457) 435.0 (354-515) 434.3 Sedentary (min/d), meanf 5285 (4427-6143) 3449.1 (2372-4525) 6841.4 MET-min/wk, meang
P valueb .08 .37
.39
.86 (80.2-100.0) (0.0-12.4) (0.0-12.4) (22.3-26.0) (36.1-43.5)
86.3 9.1 4.5 23.1 42.7
(71.4-100.0) (0.0-21.5) (0.0-13.5) (22.2-23.9) (37.6-47.8)
77.2 18.1 4.5 24.0 37.5
(59.0-95.4) (1.4-34.9) (0.0-13.5) (22.7-25.2) (34.8-40.2)
.31 .09
(483-555) (22.0-41.3)
505.6 (457-553) 47.2 (34.5-59.9)
522.5 (483-562) 41.8 (31.2-51.7)
.91 .36
(744-804)
789.5 (748-830)
782.1 (740-823)
.95
4.0 (3.3-4.5)
3.8 (2.8-4.8)
.71
(3.2-5.0)
(44.7-50.5) 49.7 (46.8-52.5) 47.8 (44.1-51.4) (347-521) 409.1 (349-468) 407.1 (341-472) (4514-9168) 5120.0 (3863-6376) 5325.6 (4068-6583)
.04 .91 .05
MET-min/wk ¼ metabolic equivalent of task minutes per week; MVPA ¼ moderate-to-vigorous physical activity; VO2max ¼ volume of maximum oxygen consumption. One-way analysis of variance was used to examine whether there were any differences in continuous variables (eg, age) across the 4 groups; c2 analysis was used to test differences for categorical variables (eg, sex). c The summed affect score from the depression, anxiety, and fatigue subscales of the Profile of Mood States questionnaire. d Estimates are from the accelerometry data (n¼78); a valid accelerometer day of monitoring was defined as at least 10 h/d of wear time. e Estimated from the prediction equation.37 f Estimated from the adult Sedentary Behavior Questionnaire (average of the 2 assessments). g Estimated from the International Physical Activity Questionnaire (average of the 2 assessments). a
b
were given a GT1M accelerometer (ActiGraph LLC) to wear for the following 7 days; they returned it during their second visit to the laboratory. Using a counterbalanced, crossover, randomized controlled design (Figure), cognitive function was assessed after an acute bout of treadmill exercise (30 minutes), with participants randomized to 1 of 4 exercise groups: no exercise or light-, moderate-, or vigorousintensity exercise. To minimize potential carryover effects in the cognitive assessments, an approximately 1-week washout period elapsed between the 2 visits. Carryover effect analyses (Koch analyses) were used by ranking the sum scores for each cognitive test across both Mayo Clin Proc. n XXX 2015;nn(n):1-11 www.mayoclinicproceedings.org
n
sequences (eg, AB/BA) and using nonparametric analyses to examine differences in the ranks. No carryover effects were observed for any of the 10 cognitive assessments (P>.05 for all the tests). Allocation concealment was used by not informing the participants of which tests (eg, cognitive testing only or cognitive testing after treadmill exercise) would take place at the visit. Randomization for the group assignment (ie, control or light-, moderate-, or vigorousintensity exercise), randomization for the crossover design (ie, at which visit they completed the treadmill exercise; AB/BA), and randomization for the order of the cognitive tests was conducted using the Excel random list (RAND) function (Microsoft Corp).
http://dx.doi.org/10.1016/j.mayocp.2014.12.023
3
MAYO CLINIC PROCEEDINGS
Control
No PA
PA
Cognition
No PA
Cognition
PA
Cognition
Cognition
Washout
Light-intensity exercise No PA
Cognition No PA
PA
Cognition
Cognition PA
Cognition
Washout Random allocation
Moderate-intensity exercise No PA
Cognition
PA
Cognition
No PA PA
Cognition Cognition
Washout
Vigorous-intensity exercise No PA
Cognition No PA
Cognition
1-week accelerometry assessment
FIGURE. Schematic of crossover study design. Physical activity (PA) / cognition ¼ acute 30-minute exercise followed by cognitive assessment; No PA / cognition ¼ no exercise before cognitive assessment. Washout was 1 week.
Acute Exercise Testing Participants were randomized to 1 of 4 acute exercise groups based on guidelines proposed by Mayo Clinic16: the control group (no exercise), the light-intensity exercise group (40%50% of predicted [220-age] maximum HR [HRmax]; calculation of HRmax was determined by subtracting the participant’s age from 220), the moderate-intensity exercise group (51%70% of predicted HRmax), or the vigorousintensity exercise group (71%-85% of predicted HRmax). During the 30-minute bout of exercise on the treadmill, HR (HR monitor; Polar Electro Inc) and rating of perceived exertion (10-point Borg scale) were assessed every 5 minutes. Based on expected mean and SD values from previous studies,17 using a 2-tailed test with an effect size d of 0.62 and alpha error probability of 0.05, 22 participants in each acute exercise group were needed to have an achieved power (1ebeta error probability) of 0.80 to detect differences in the evaluated cognitive function parameters. Of the 87 participants, 19 (22%) were randomized to the control group, 24 (28%) to the light-intensity group, 22 (25%) to the moderate-intensity group, 4
Mayo Clin Proc.
n
and 22 (25%) to the vigorous-intensity group. Cognitive Tests. During the visit at which the participant completed the cognitive assessments after the acute bout of exercise, the cognitive assessments started when the postexercise HR was within 10% of the baseline HR or 15 minutes after the acute exercise ended, whichever came first. Participants completed several cognitive tests that assessed different areas of brain function and varied in task complexity; some evidence suggests that cognitive task complexity may moderate the relationship between exercise and cognition, and different areas of the brain (eg, frontal lobe and temporal lobe) may be differentially influenced by exercise.18 Ten cognitive function tests were administered, including 2 paper-and-pencil tests (Trail Making Test A and B) assessing cognitive visual attention and task switching.19 The remaining 8 tests were administered using electronic software: Spatial Span20 and Paired Associates21,22 (to assess memory), Grammatical Reasoning and Odd
XXX 2015;nn(n):1-11
n
http://dx.doi.org/10.1016/j.mayocp.2014.12.023 www.mayoclinicproceedings.org
EXERCISE AND COGNITION
One Out (to assess reasoning),23 Feature Match24,25 and Polygon26 (to assess concentration), and Spatial Search27 and Spatial Slider28 (to assess planning). The cognitive tests were administered in a randomized order. Affect. The Profile of Mood States questionnaire was used to assess affect, with the depression/dejection (13 items), anger/hostility (11 items), and fatigue/inertia (7 items) subscales used for the present study. The Profile of Mood States questionnaire has demonstrated reasonable internal consistency on the subscales (eg, a¼0.96 for depression).29 In the present sample, internal consistencies, as measured by the Cronbach a, for the subscales were as follows (nontreadmill and treadmill visits): depression/dejection (a¼0.84 and a¼0.86), anger/hostility (a¼0.86 and a¼0.75), and fatigue/inertia (a¼0.80 and a¼0.81). In the present sample, 1-week test-retest reliability for the depression/dejection subscale was intraclass correlation coefficient (ICC)¼0.74 (P
