The effects of high-intensity interval
exercise and moderate intensity
continuous exercise on inhibitory control,
hippocampus-dependent memory, and
brain-derived neurotrophic factor in
healthy sedentary adults: a pilot study
A.M. van Schaik
Student number:11209496
Psychobiology, University of Amsterdam
Bachelor’s Thesis
Supervisor: Prof. S. Higgs & M. Sardjoe MSc
Table of contents
Abstract...3 Introduction...4 Methods...7 Results...13 Discussion...17 References...20Abstract
Physical inactivity is associated with lower cognitive function. Previous studies have shown that exercise, even after one session, could enhance cognitive function. Cognitive functioning is associated with brain-derived neurotrophic factor (BDNF) levels in the brain, which can be increased through exercise. Therefore, BDNF is hypothesized to influence exercise-induced cognitive enhancement. This thesis focusses on acute high-intensity interval exercise (HIIE) and moderate intensity continuous exercise (MCE) and how they affect inhibitory control, hippocampus-dependent memory, and plasma BDNF in ten healthy sedentary adults. Contrary to previous literature, results suggested impairment in inhibitory control after HIIE and MCE, and in hippocampus-dependent memory after MCE. Additionally, plasma BDNF levels were suppressed after HIIE and MCE. Because of the low sample size and high variability in measures of inhibitory control and hippocampus-dependent memory, and in sample duplicates of the BDNF analysis, the findings should be carefully considered. Replication of the study should be considered to provide insight into these unusual findings.
Introduction
Despite the benefits, many people do not exercise regularly and lead a sedentary lifestyle, which can lead to increased risk of cardiovascular diseases [ CITATION For12 \l 2057 ] and lower cognitive function [ CITATION Whe17 \l 1043 ]. In contrast, aerobic exercise is able to increase cardiorespiratory fitness resulting in lowered risk of developing cardiovascular diseases [ CITATION Lav19 \l 1043 ] and could therefore benefit people with a sedentary lifestyle. The effects of exercise on brain function in sedentary individuals, however, are less well known. Hötting & Röder (2013) claim that regular exercise could facilitate functional and structural changes in the hippocampus and frontal regions of the brain, which would contribute to the maintenance of cognitive function [ CITATION Luc15 \l 1043 ].
Exercise protocols with just one session, acute exercise, have shown beneficial effects on cognition. A meta-analysis revealed that acute exercise could improve cognitive functioning [ CITATION Cha12 \l 2057 ]. They described moderator variables including exercise type, exercise intensity, and type of cognitive task. Additionally, the largest effects were found in tasks measuring executive function, an umbrella term including inhibitory control, cognitive flexibility, working memory, and planning [ CITATION Miy00 \l 2057 ]. Inhibitory control refers to the ability to focus on relevant information and inhibit responses driven by irrelevant information [CITATION Miy12 \t \l 1043 ]. Thirty minutes of acute aerobic exercise has also been reported to improve hippocampus-dependent memory in young adults [ CITATION Gri11 \l 2057 ], although the effects of different exercise routines on hippocampus-dependent associative memory have not been examined yet. Associative memory is the ability to learn and remember the relationship between unrelated items [ CITATION Suz05 \l 2057 ]. The hippocampus is involved in forming these new associations (Zeineh, Engel, Thompson, & Bookheimer, 2003, Brasted, Bussey, Murray, & Wise, 2003). The retrieval of existing associations, however, depends more on the prefrontal cortex as it plays a role in memory consolidation [ CITATION Pre13 \l 2057 ]. This thesis will examine the effects of exercise on both inhibitory control and associative memory.
Studies examining acute aerobic exercise often include one of two exercise regimes: moderate intensity continuous exercise (MCE) or high-intensity interval exercise (HIIE). While MCE maintains the same moderate intensity throughout the exercise program, HIIE is characterized by short bursts of high intensity exercise interspersed with periods of lower intensity exercise. HIIE routines are shorter than MCE protocols and are therefore more appealing to people with limited time. Both protocols have been able to improve cognitive function in multiple studies (Basso & Suzuki, 2017, Kao, Drollette, Ritondale, Khan, & Hillman, 2018), though HIIE seems to elicit larger effects. For instance, while MCE and HIIE both improved post-exercise inhibitory control, HIIE could prolong improvement in inhibitory control whereas the effects of MCE returned to normal 30 minutes after
exercise [ CITATION Tsu16 \l 2057 ]. Examining the different effects of acute HIIE and MCE on cognitive function, as not all cognitive domains have been studied yet, could provide much needed insight for developing the optimal exercise intervention to enhance brain health and function.
Research on the mechanisms underlying exercise-induced cognitive improvement has found evidence for a key role of brain-derived neurotrophic factor (BDNF) [ CITATION Ras09 \l 2057 ]. BDNF is a protein found in brain areas such as the hippocampus and cortex, but also remains in peripheral skeletal muscle cells, where it is associated with different functions. Cerebral BDNF elicits and increases long-term potentiation, neurogenesis, axonal growth, and synaptogenesis through a tyrosine kinase B receptor (TrkB) [ CITATION Jim18 \l 1043 ]. An increase in the amount of neurons in the brain and the efficiency of synaptic signalling are in turn related to enhanced cognitive function (Hötting & Röder, 2013), for example, increased learning and memory in rats [ CITATION Vos13 \l 2057 ]. Because BDNF is able to cross the blood-brain barrier, an increase in peripheral BDNF could be an indication of an increase in BDNF release in the brain [ CITATION Pan98 \l 1043 ].
In general, acute high-intensity protocols result in larger increases in peripheral BDNF levels compared to protocols with lower intensities [ CITATION Kna10 \l 2057 ]. Additionally, both acute HIIE and MCE were able to increase peripheral BDNF levels, although HIIE evoked larger BDNF concentrations compared to MCE [ CITATION Sau15 \l 2057 ], suggesting HIIE to be more effective in increasing BDNF levels.
Previous research examining the relationship between acute exercise, peripheral BDNF, and cognitive function have developed a theory that exercise-induced cognitive improvement could be mediated by BDNF [ CITATION Ras09 \l 2057 ]. An increase in levels of serum BDNF post-exercise was
associated with improved inhibitory control [ CITATION Fer07 \l 2057 ] and hippocampus-dependent memory [ CITATION Gri11 \l 2057 ]. However, a difference between the effects of HIIE and MCE on this relationship has not been as extensively researched, though HIIE is hypothesized to have additional beneficial effects on exercise-induced cognitive enhancement (Saucedo Marquez et al., 2015) which is also examined in this thesis. Slusher, Patterson, Schwartz, and Acevedo (2018) found improved executive function, assessed by the Wisconsin Card Sorting Task (WCST), and greater serum BDNF levels after HIIE. However, no other study has looked at the direct relation between HIIE and MCE, cognitive function, and BDNF and their mediating effects.
To our knowledge, this study is the first to examine the effects of high-intensity interval exercise and moderate intensity exercise on inhibitory control, hippocampus-dependent memory, and on levels of plasma BDNF in healthy sedentary adults. It is hypothesized that HIIE and MCE will improve inhibitory control and hippocampus-dependent memory, and that HIIE will be more effective than MCE. Furthermore, BDNF will have a mediating role in HIIE- and MCE-induced cognitive enhancement.
Healthy adults with a sedentary lifestyle will undergo a high-intensity interval exercise session, a moderate intensity exercise session, and a control rest session. In each session, cognitive function will be assessed after exercise. Inhibitory control is measured with the Stroop task and hippocampus-dependent memory with a video-word pair task. Additionally, to analyse exercise-induced changes in circulating BDNF levels and its relation to cognitive functioning, blood will be drawn three times per session: before exercise, after exercise, and after the cognitive tasks.
Because exercise is hypothesized to improve inhibitory control, Stroop interference scores are expected to be lower after a HIIE session and MCE session, where HIIE would have the lowest scores. Additionally, hippocampus-dependent memory is thought to be enhanced through exercise and therefore, accuracy on the unrelated video-word pairs of the video-word pair task is expected to be higher after HIIE and MCE than after rest. BDNF levels were hypothesized to be equal at baseline and increase after HIIE and MCE compared to rest. Lastly, as BDNF is hypothesized to play a mediating role, HIIE and MCE are expected to increase plasma BDNF levels, HIIE more so than MCE, and these increased levels are expected to correlate with decreased interference scores on the Stroop Test and increased accuracy on the video-word pair task.
Methods
Participants
Ten sedentary healthy adults, seven females, participated in an exploratory within-subject design pilot study examining the effects of exercise on cognition and its transfer effects on eating behaviour (hence the small sample size). Additional measures of eating behaviour (e.g. food choice task and questionnaires) and cognition (e.g. N-back task, Go/No-Go task, the Trail Making Test, and a delay discounting task) were included but are not discussed in this thesis. Participants were recruited through online and in-person advertising in the form of flyers, Facebook posts, and through a University of Birmingham participant recruitment website. Participants were able to participate if their age was between 18 and 35 years, their body mass index between 18 and 25 kg/m2, and if their
physical fitness level was low as this study examined people with a sedentary lifestyle. Physical inactivity was assessed during the screening by asking how many days and how long they participated in moderate- or vigorous physical activity in the past week. People were excluded if they participated in more than 150 minutes of moderate intensity or 75 minutes of vigorous intensity physical activity per week. Further exclusion criteria included smoking and/or dietary restraints because it was thought to affect eating behaviour in comparison with healthy people without diets. Written consent was given prior to participation and ethical approval was given by the Ethics Committee of the University of Birmingham.
International Physical Activity Questionnaire
The long, self-administered version of the International Physical Activity Questionnaire (IPAQ; Patterson, 2002) is a validated measure of physical activity [ CITATION Cra03 \l 2057 ]. The IPAQ was administered to determine participants’ physical activity levels in the seven days leading up to each test day. The questionnaire describes four domains of physical activity: work-related, transportation, housework/gardening, and leisure-time activity. For every domain, the number of days per week and time per day spent on both moderate and vigorous activity were recorded for the last 7 days. Walking time was also included for work-related, transportation, and leisure-time activity. Physical activity was expressed in MET (Metabolic Equivalent Task) and one MET was equal to energy expenditure during rest (approximately 3.5 mL O2 kg-1min-1 in adults). The questionnaire also
included questions about the time spent sitting per day as an indicator of sedentary behaviour. Outcome measures used were MET-minutes per week (see Patterson (2005) for scoring protocol) and number of minutes spent sitting per week.
Exercise procedures
The exercise procedure consisted of either a warm-up (5 minutes) following high-intensity interval exercise (26 minutes) or moderate intensity exercise (30 minutes), or a 35-minute resting session as
control and the order of the exercise condition was randomly assigned. The HIIE protocol, presented in Figure 1A, was based on the protocol of Tsukamoto, et al., 2016. The protocol alternated between intervals of moderate intensity (60% VO2max) and high intensity (90% VO2max). After the 5-minute warm-up (40% VO2max), the HIIE started off with five minutes of cycling at 60% VO2max,
followed by three four-minute bouts of 90% VO2max alternated with a three-minute interval of 60% VO2max, for a total exercise period of 31 minutes. Intensity was lowered during the high-intensity bouts to an intensity above 60% VO2max if participants deemed it too difficult due to physical pain. The MCE protocol (Figure 1B) was based on the protocol of Lowe, Kolev, and Hall (2016) and consisted of a 5-minute warm-up (40% VO2max) and 30 minutes of moderate intensity exercise (60% VO2max).
A) B)
Figure 1: Exercise protocols of HIIE and MCE including the warm-up. A) The HIIE protocol, based on
Tsukamoto, et al., 2016, started off with a 5-minute warm-up (40% VO2max), followed by 5 minutes of moderate intensity exercise (60% VO2max). Afterwards, three sequences of a 4-minute high-intensity bout (90% VO2max) alternated with a 3-minute moderate intensity period were administered. B) The MCE protocol began with a 5-minute warm-up (40% warm-up) followed by 30 minutes of moderate intensity exercise (60% VO2max).
Cognitive tasks
Six cognitive tasks were administered in randomized order between 15 and 75 minutes after exercise. The Stroop Test and a video-word pair task are discussed in this thesis.
Stroop Test
The Stroop Test (Figure 2), adapted from Miyake et al. (2000) and Lowe et al. (2016) measured response inhibition in which participants had to call out the font colour of the presented stimulus and not read the word aloud. The words that were used were ‘red’, ‘purple’, ‘blue’, ‘orange’, ‘green’, and ‘yellow’ and these six colours were featured in the task. Additionally, they determined if the stimulus matched the font colour by clicking the left mouse-click. If the stimulus and font colour did not match, the right mouse-click was pressed. The paradigm was presented using E-Prime software (Psychology Software Tools, Inc.) and consisted of 144 trials divided over two blocks. The stimulus on the screen was either five coloured X’s (neutral condition; 72 trials), a coloured word where the
font colour matched the meaning of the word (congruent condition; 12 trials), or a coloured word where the font colour and the word did not match (incongruent condition; 60 trials). Participants were instructed to name the font colour of the stimulus out loud and simultaneously determine whether it is a match (congruent) or not (incongruent) with the word on the screen by mouse-click. For the neutral trials, participants were instructed to click left, as if it were a match. The stimuli remained on the screen until the participant responded, followed by presentation of a 1000 ms fixation cross. Percentage accuracy and reaction time were recorded, and Stroop interference scores were calculated for every participant: the reaction time of correct incongruent trials minus the reaction time of correct neutral trials. Lower Stroop interference scores indicated higher inhibitory control.
Figure 2: The Stroop Test. Congruent (e.g. “blue” in blue font), incongruent (e.g. “red” in green font), and
neutral (e.g. “XXXXX” in orange font) trials were displayed until the participant responded. A left mouse-click was given when the participant thought that the word and the font colour matched, and a right mouse-click when they did not match. Participants clicked left for the neutral trials. In between trials, a 1000 ms fixation cross was presented.
Video-word pair task
The video-word pair task (Griffiths, et al., 2019) is a paired-associates memory task which differentiates between recalling word-video pairs that are semantically related (cortex-dependent) or unrelated to the participant (hippocampus-dependent). During the learning phase, a video was shown for three seconds in each trial, followed by a word that was presented for two seconds. Participants had to remember that the video was paired with the word by making a vivid association between the two. Afterwards, they indicated whether the video and word were semantically related or not, by pressing the ‘f’ or ‘g’ key on the keyboard, respectively (Figure 3A). A fixation point was presented between trials for 1500 ms. Whether or not the video-word pair were semantically related was subjective. There were 48 trials using four different videos relating to scenery and 48 words in randomized order. Afterwards, participants completed several short math questions (e.g. 458 – 3) as a distractor task. Then, in the retrieval phase, participants were presented with every word that was shown in the learning phase and selected the associated video out of the four options using the “f”,
“g”, “h”, and “j” keys. (Figure 3B). Participants had three seconds to choose the right video; accuracy and reaction times were determined. Then, participants judged how confident they felt about their decision by selecting one of three options: ‘remember’, ‘know’, or ‘guess’. ‘Remember’ was chosen when the participant could clearly recreate the mental image. ‘Know’ related to being unable to recreate the mental image, but the participant thought they knew the associated video. ‘Guess’ was selected when the participant had no idea what the associated video was. A distinction was made between the trials where the video and word were found to be related or unrelated, to distinguish between cortex-dependent and hippocampus-dependent memory. Furthermore, two sub scales were made: ‘forgotten’ and ‘remembered’. A trial was ‘forgotten’ when the chosen video was incorrect, confidence was rated as ‘guess’, or when no response was recorded. The remaining trials were sorted into the ‘remembered’ scale. The amount of correct remembered unrelated words was divided by the total amount of unrelated words to determine accuracy of hippocampal-dependent memory.
A) B)
Figure 3: Video-word pair task. A) Learning phase. One trial consisted of a 1500 ms fixation point, a video
of three seconds, a word presented for two seconds, and a question if the video and the word were semantically related. Participants had three seconds to answer the question. B) Retrieval phase. One trial started off with a 1500 ms fixation point, followed by one of the previously seen words, presented for two seconds. Afterwards, a display with pictograms of the four videos were shown and the participant chose the video that they thought to be associated with the word and judged how confident they were in their decision. Participants had three seconds to respond.
Experimental procedures
During the screening, participants performed a VO2max test on the ergo cyclometer to calculate exercise parameters for the test days. The VO2max test started off with a 5-minute warm-up. Afterwards, the intensity of the ergo cyclometer was increased every 2.5 minutes to determine the maximum amount of oxygen the body can absorb during intense and prolonged exercise (VO2max). Participants cycled until exertion or until it was deemed unsafe by guidelines from Tsukamoto, et al. (2016).
This pilot study had a within-subject design where participants came in for three three-hour long test days in the morning with at least 7 days in between each test day. Each test day, illustrated in Figure 4, consisted of three blood draws spread out during the day, a standardized breakfast, an
exercise procedure, and cognitive tasks, and differed only in the type of exercise. The breakfast consisted of Nature Valley Crunch Oats & Honey breakfast bars and the standardized amount was 40% of their calculated basal metabolic rate (BMR), measured as suggested in Henry (2005).
Figure 4: Schematic overview of the test days. A test day consisted of the following parts: the first blood
sample (t=0), breakfast (t=10), an exercise or rest session (t=65), the second blood sample (t=100), cognitive tasks (t=115), and a third blood sample (t=175). Time indicated by dotted lines consisted of questionnaires and other tasks, which were not discussed except for the IPAQ.
Blood sampling and BDNF analysis
All subjects were required to fast for 12 hours before each test day to collect a baseline measure, unaffected by nutritional factors, of circulating BDNF levels. Approximately 4 ml blood was drawn through venepuncture at three different time points: before breakfast (baseline), immediately after exercise, and immediately after cognitive task administration. Whole blood samples were collected into a BD Vacutainer containing K2EDTA as anticoagulant (Franklin Lakes, NJ, USA) and
centrifuged at 1000 g for 15 min at 4 °C to collect plasma supernatant. Because platelets store and release BDNF into circulating blood [ CITATION Fuj02 \l 2057 ], plasma supernatant was additionally centrifuged at 10.000 g for 10 min at 4 °C to obtain platelet-free plasma containing BDNF [ CITATION Par15 \l 2057 ].
Circulating BDNF levels were determined using a Quantikine ELISA quantitative enzyme immunoassay kit with a minimum detectable dose of <80 pg/mL (R&D Systems Europe, Abingdon, UK). Two 96-well polystyrene microplates were pre-coated with anti-BDNF monoclonal antibody and eight working standards were prepared from undiluted Human Free BDNF Stock Standard (400 ng/mL). 100 µL of standard, control, or 100 times diluted sample was added to a well, followed by 2.5 hours of incubation at room temperature. Wells were then washed three times with 300 µL of 1X Wash Solution. Afterwards, 100 µL of 1X Biotinylated anti-Human BDNF Detector Antibody was added to the wells and incubated for one hour at room temperature, followed by the same washing protocol. 100 µL of HRP-conjugated streptavidin was then added to the wells for 45 minutes. The wells were again washed and 100 µL of TMB substrate solution was added to the wells to add blue colour to the wells in proportion to the amount of bound BDNF. Colour development was stopped after 45 minutes and changed from blue to yellow with 50 µL of Stop Solution. Optical density was measured at 450 nm using a microplate reader. A standard curve for optical density (OD) and BDNF concentrations in ng/mL was made to estimate BDNF levels in the samples.
Data analysis
First, the relation between exercise conditions (HIIE, MCE, and rest) and cognition was assessed. Therefore, within-subject effects of exercise on Stroop Test accuracy (dependent variable; DV) for the congruent, incongruent, and neutral trials were assessed in a two-way repeated measures ANOVA. Exercise-induced changes in Stroop Test’s interference scores (DV) were assessed in a one-way repeated measures ANOVA. Furthermore, within-subject effects of exercise on accuracy on unrelated video-word pairs (DV) of the video-word pair task were determined with a one-way repeated measures ANOVA.
Afterwards, the within-subject effects of exercise on BDNF levels on three time points (DV) were determined with a two-way repeated measures ANOVA. All repeated measures ANOVAs included gender as a between-subjects factor and were, if statistically significant, followed by Bonferroni corrected post hoc paired t-tests. Effects of gender were only reported if they were significant.
The relation between exercise, cognition, and BDNF was assessed through multiple mediation analyses (MEMORE pre-release version 3.0, Montoya & Hayes, 2017) at which the dependent measures of the Stroop Test and the video-word pair task were examined separately. Outliers were removed if the data point was above or below three standard deviations from the group mean. Repeated measures ANOVA’s assumption of sphericity was determined with the Mauchly test and the Greenhouse-Geisser correction was applied if the assumption was violated. Because the repeated measures ANOVA is quite robust against changes in normality [ CITATION Bla17 \l 2057 ], normality was not corrected for. Results were statistically significant with p-values lower than 0.05. Statistical analyses were conducted using SPSS (Version 26, IBM Corp).
Results
Participants’ characteristics
Initially, ten participants, seven females, aged between 19 and 29 [mean ± SD; 23 ± 3.559], were examined on their BMI, physical activity level, and the amount of time spent sitting (sedentary behaviour) leading up to the test days to control for confounding factors (Table 1). No differences between test days were found.
Table 1. Participants’ characteristics. Descriptive statistics of BMI, physical activity levels, and
amount of sedentary behaviour.
Characteristics Screening Rest MCE HIIE
BMI (kg m-2) 23 ± 0.92 23.2 ±
0.95
23.1 ± 0.94
23.2 ± 0.96 Physical activity level
(MET-minutes/week)
453 ± 129 622 ± 238 505 ± 145
Sedentary behaviour (min) 401 ± 48.1 348 ± 62.8 440 ± 52
Notes. Means ± SEM.
Effects of exercise on cognitive function
Stroop Test accuracy
The repeated measures ANOVA revealed no main effect of exercise [Figure 5A, F(2, 16) = 0.75, p = 0.49, ƞp2 = 0.09] nor any interaction effect between exercise and Stroop condition on accuracy [F(4,
32) = 0.8, p = 0.54, ƞp2 = 0.09]. However, accuracy was affected by Stroop condition [F(2, 16) = 4.27,
p = 0.03, ƞp2 = 0.35] but a post hoc test found no significant differences. Stroop interference
The ANOVA did not reveal an effect of exercise on Stroop interference scores [Figure 5B; F(2, 16) = 0.74, p = 0.49, ƞp2 = 0.08]. The data was also visually inspected to examine individual differences
which could explain the findings. On group level, the interference scores seemed to be lower after rest compared to HIIE and MCE. However, on individual level, a high variance in individual’s interference scores was seen, especially in the rest and MCE conditions (Figure 5C), making it difficult to determine a clear pattern.
Video-word pair task
Additionally, accuracy on unrelated video-word pairs of the video-word pair task were not affected by exercise condition [Figure 6A; F(2, 16) = 0.46, p = 0.64, ƞp2 = 0.05]. The pattern of the data suggested
a small decrease in accuracy after MCE compared to HIIE and rest. Individual performance is plotted in Figure 6B and showed high variance and an unclear pattern where some individuals seemed to improve through exercise and others showed the opposite effect, which agrees with the overall lack of effect of exercise shown on group level.
Because none of the measures of the cognitive tasks were affected by exercise, mediation analyses were not conducted.
A) B)
C)
Figure 5: Accuracy on all Stroop conditions and interference scores on the Stroop Test were not significantly affected by exercise condition. A) Accuracy (%) on the congruent (blue), incongruent
(orange), and neutral (grey) conditions of the Stroop Test did not differ between exercise conditions. B) Interference score (RT incongruent – RT neutral) on the Stroop Test did not differ between exercise conditions. C) Interference scores were plotted for each participant.
A)
Figure 6: Accuracy on the unrelated video-word pairs in the video-word pair task were not significantly affected by exercise condition. A) Accuracy (%) on unrelated word pairs of the
video-word pair task was not affected by exercise. B) Accuracy on the unrelated video-video-word pairs in the video-video-word pair task were plotted for each participant.
Effects of exercise on plasma levels of brain-derived neurotrophic factor
The effect of exercise condition on the plasma levels of BDNF on all three time points was examined (Table 2). The BDNF levels of one female participant (P5) was not included due to technical difficulties. A two-way repeated measures ANOVA with exercise condition and time point as within-subject variables revealed a main effect of exercise [F(2, 14) = 9.1, p = 0.003, ƞp2 = 0.57], time point
[F(2, 14) = 8.38, p = 0.004, ƞp2 = 0.55], and an interaction effect between exercise and time point [F(4,
28) = 3.4, p = 0.022, ƞp2 = 0.33]. An interaction effect between time point and gender was also found
[F(2, 14) = 5.28, p = 0.02, ƞp2 = 0.43] suggesting that men and women had had different levels of
BDNF in their blood at certain time points, P2, P8, and P9 being men. Post hoc paired t-tests showed an effect of time only in the rest condition with time point 3 (T3) being significantly higher than time point 1 (T1) [t(8) = 2.53, p = 0.035, d = 0.84]. Furthermore, BDNF levels during the rest condition were higher than HIIE at T1 [t(8) = 2.95, p = 0.018, d = 0.98], T2 [t(8) = 5.88, p < 0.001, d = 1.96], and T3 [t(8) = 2.96, p = 0.018, d = 0.99]. Levels at T1 [t(8) = 2.9, p = 0.02, d = 0.97] and T2 [t(8) = 2.73, p = 0.026, d = 0.91] were also higher in the MCE condition compared to the HIIE test days. At time point 3, BDNF levels were higher after rest than after MCE [t(8) = 2.71, p = 0.027, d = 0.9].
Individual BDNF concentrations at all time points for each exercise condition were plotted in Figure 7. The pattern on T1 was in line with the findings in Table 2: BDNF concentrations were lowest at HIIE and highest at rest. This was largely similar on T2, though some individuals showed increased levels at MCE which did not fit the pattern. High variance in BDNF levels were seen on T3, especially for rest and HIIE. P8 and P9, which are two men, showed a deep V-curve with a big drop in BDNF concentration after MCE.
Figure 7: Individual differences in BDNF levels across exercise conditions and time. The levels of BDNF
(ng/ml) were plotted for each time point and exercise condition for nine participants. P2, P8, and P9 were male participants.
Table 2. Levels of BDNF across time for rest, MCE, and HIIE descriptive statistics.
Variable Rest MCE HIIE
BDNF levels (ng/ml)
Time point 1 (baseline) 1.23 ± 0.15 1.21 ± 0.17 1.02 ± 0.14a,b
Time point 2 (post exercise) 1.3 ± 0.19 1.16 ± 0.25 0.98 ± 0.13a,b
Time point 3 (post cognitive tests)
1.49 ± 0.27c 1.14 ± 0.16a 1.19 ± 0.28a Notes. Means ± Standard Deviations. a Indicates a significant difference relative to rest values
within the time point (p < 0.05). b Indicates a significant difference relative to MCE values within
the time point (p < 0.05). c Indicates a significant difference relative to time point 1 value within
Discussion
The aim of this thesis was to examine the effects of high-intensity interval exercise and moderate intensity exercise on inhibitory control, hippocampus-dependent memory, and on plasma levels of brain-derived neurotrophic factor in healthy sedentary adults. Our results showed no changes in Stroop interference scores or accuracy on unrelated trials of the video-word pair task after HIIE or MCE compared to rest. The pattern of the results suggested that Stroop interferences scores were lower after rest than after HIIE or MCE and that accuracy on the video-word pair task was decreased after MCE compared to HIIE and rest. Additionally, BDNF concentrations were unaffected by HIIE or MCE compared to baseline, whereas levels were increased after a resting period.
To answer the research question, high-intensity interval exercise and moderate intensity exercise impaired inhibitory control or hippocampus-dependent memory in sedentary adults and plasma BDNF was depressed after both exercise regimes compared to the control, which is inconsistent with the hypotheses. Because exercise did not lead to cognitive improvement, we were unable to examine a possible mediating effect of BDNF.
Because this was a pilot study, and therefore not fully powered, the lack of statistically significant results was not unexpected. However, the calculated effect sizes show a medium-to-large effect of exercise condition on inhibitory control (ƞp2 = 0.08) and a small-to-medium effect of exercise on
hippocampus-dependent memory (ƞp2 = 0.05), suggesting that the acquired results could show a
significant effect if the study would be fully powered. The patterns of the data indicate that exercise impaired both inhibitory control and hippocampus-dependent memory, which are in the opposite direction of what was expected and inconsistent with previous research which showed either improvement in or no effect on cognition. The present results could partially be explained by the high variability in Stroop interference scores and accuracy of hippocampus-dependent memory due to individual differences, which explains why there was no effect at the group level. For example, Stroop interference scores were affected by exercise in some people, while others were unaffected. This view is supported by Schwarck et al. (2019), who also reported inter-individual differences in cognition as a result of acute exercise. The causes of individual differences in cognition are not well understood as of yet, but recent literature suggests that neural correlates, heritability, environmental influences, and personality play a role [ CITATION Boo18 \l 1043 ].The high variability in the data and low sample size also demand caution in the interpretation because they make the pattern unclear. Even with favourable effect sizes, it is unclear whether the findings are representative.
BDNF concentrations were significantly influenced by exercise with large effect sizes. Surprisingly, BDNF levels were suppressed after exercise, which is inconsistent with the previous literature. However, some notable methodological factors may explain the results. Specifically, high variance amongst the sample duplicates was found, indicating possible inaccuracies in the assay
analyses. Therefore, the presented BDNF values could be different from the legitimate BDNF levels in the blood. Furthermore, BDNF samples from the HIIE and half of the MCE condition were analysed on a second BDNF assay which gave overall lower values than the first BDNF assay. This could be due to the inter-assay variance of the kits (12%) and could explain the difference in baseline BDNF for the HIIE condition. In the future, it could be beneficial to put the samples of every participant on the same assay to diminish some of the variance in sample duplicates.
Additionally, methodological differences between this study and other literature in the field could explain our findings. First, as mentioned before, relatively little research has investigated HIIE compared to other types of exercise. There is therefore less known about HIIE-induced responses on cognition and BDNF. Most studies also choose to focus on serum BDNF (Ferris, et al., 2007, Griffin, et al., 2011, Hötting & Röder, 2013, Hwang, et al., 2016), while we focussed on plasma BDNF for practical reasons: blood plasma can legally be stored in the United Kingdom for 3 months and blood serum only a few days. The effects of exercise on plasma BDNF are inconsistent across studies with exercise either increasing (see Dinoff, Herrmann, Swardfager, & Lanctôt, 2017 for meta-analysis) or not affecting plasma BDNF (Slusher et al., 2018). Studies that found a decrease in BDNF concentrations detected it in serum BDNF after several weeks of aerobic training in overweight individuals (Glud, Christiansen, Larsen, Richelsen, & Bruun, 2019) and individuals with metabolic syndrome (Damirchi, Tehrani, Alamdari, & Babaei, 2014), though those effects could be due to differences in metabolic processes between lean and overweight participants and those with metabolic syndrome. Also, most studies examining the relationship between exercise and BDNF choose to focus on men alone to exclude confounding variables due to the menstrual cycle [ CITATION Beg07 \l 1043 ]. Therefore, little is known about exercise-induced BDNF responses in women. A study that looked into gender differences in BDNF found higher baseline levels of serum BDNF in women compared to men but the same amount of exercise-induced BDNF response to exercise (Glud et al., 2019) However, a meta-analysis suggested that BDNF levels increase after exercise in men, but not in women [ CITATION Szu15 \l 1043 ]. These findings suggest that circulating BDNF could have been influenced by sex and that studies with equally sized groups of men and women should be examined to provide more insight into the potential effect.
Some strengths of this study include having a within-subject design to rule out between-subject factors. Participants went through two different exercise routines which could then be directly compared to determine which program would be most influential on cognitive function and BDNF levels. Additionally, including a variety of cognitive tasks and the ability to measure BDNF values over time and in response to exercise in this one study made it possible to directly examine a possible mediating effect of BDNF on exercise-induced cognitive improvement. However, the limited number of participants and unequally sized gender groups limited the ability to examine these relations. The small number of men also made it difficult to make assumptions about influence of sex. Therefore, a
fully powered replication study is necessary to provide more insight into the unusual findings of exercise-induced cognitive impairment and suppressed BDNF. In this replication study, it would be beneficial to have an equal number of men and women to inspect sex differences. Furthermore, a future study with other tasks that measure inhibitory control and hippocampus-dependent memory, for example the Go/No-Go task and the face-name matching task, respectively, would help determine if the present effects were task-specific or not. The Go/No-Go task was administered in this pilot study, but the data could not be collected due to software errors and we were therefore unable to compare the Go/No-Go task and the Stroop task.
This thesis contributes to research into the influences of exercise on cognitive function and BDNF. To further inspect these relations, future studies could focus on the effects of multiple sessions of HIIE and MCE in sedentary adults, rather than acute exercise, and see if the exercise routines differ as sedentary people are at risk for developing diseases and exercise could reduce those risks. Additionally, developing an easy and fast exercise protocol that has many cognitive benefits compared to time consuming exercise protocols could potentially motivate people to start exercising.
In summary, the findings suggest that acute HIIE and MCE tended towards impairing inhibitory control and hippocampus-dependent memory in sedentary healthy adults. Additionally, peripheral BDNF was suppressed after HIIE and MCE compared to rest. However, the high variability in measures of cognitive function and BDNF sample duplicates could have influenced the findings. A powered replication study is necessary to provide more insight.
References
Basso, J. C., & Suzuki, W. A. (2017). The Effects of Acute Exercise on Mood, Cognition,
Neurophysiology, and Neurochemical Pathways: A Review. Brain Plasticity, 2(2), 127-152. doi:10.3233/BPL-160040
Begliuomini,, S., Casarosa, E., Pluchino, N., Lenzi, E., Centofanti, M., Freschi, L., . . . Genazzani, A. R. (2007). Influence of endogenous and exogenous sex hormones on plasma brain-derived neurotrophic factor. Human Reproduction, 22(4), 995–1002. doi:10.1093/humrep/del479 Blanca, M. J., Alarcón, R., Arnau, J., Bono, R., & Bendayan, R. (2017). Non-normal data: Is ANOVA
still a valid option? Psicothema, 29(4), 552-557. doi:10.7334/psicothema2016.383 Boogert, N. J., Madden, J. R., Morand-Ferron, J., & Thornton, A. (2018). Measuring and
understanding individual differences in cognition. Philosophical Transactions of the Royal Society of London, Series B, Biological sciences, 373(1756). doi:10.1098/rstb.2017.0280 Brasted, P. J., Bussey, T. J., Murray, E. A., & Wise, S. P. (2003). Role of the hippocampal system in
associative learning beyond the spatial domain. Brain: A Journal of Neurology, 126(5), 1202– 1223. doi:10.1093/brain/awg103
Chang, Y. K., Labban, J. D., Gapin, J. I., & Etnier, J. L. (2012). The effects of acute exercise on cognitive performance: a meta-analysis. Brain Research, 1453, 87-101.
doi:10.1016/j.brainres.2012.02.068
Craig, C. L., Marshall, A. L., Sjöström, M., Bauman, A. E., Booth, M. L., Ainsworth, B. E., . . . Oja, P. (2003). International Physical Activity Questionnaire: 12-Country Reliability and Validity. Medicine & Science in Sports & Exercise, 35(8), 1381-1395.
doi:10.1249/01.MSS.0000078924.61453.FB
Damirchi, A., Tehrani, B. S., Alamdari, K. A., & Babaei, P. (2014). Influence of Aerobic Training and Detraining on Serum BDNF, Insulin Resistance, and Metabolic Risk Factors in Middle-Aged Men Diagnosed With Metabolic Syndrome. 24(6), 513-518.
Dinoff, A., Herrmann, N., Swardfager, W., & Lanctôt, K. L. (2017). The effect of acute exercise on blood concentrations of brain-derived neurotrophic factor in healthy adults: a meta-analysis. The European Journal of Neuroscience, 46(1), 1635-1646. doi:10.1111/ejn.13603
Ferris, L. T., Williams, J. S., & Shen, C. (2007). The Effect of Acute Exercise on Serum Brain-Derived Neurotrophic Factor Levels and Cognitive Function. Medicine & Science in Sports & Exercise, 39(4), 728-734. doi:10.1249/mss.0b013e31802f04c7
Ford, E. S., & Caspersen, C. J. (2012). Sedentary behaviour and cardiovascular disease: a review of prospective studies. INTERNATIONAL JOURNAL OF EPIDEMIOLOGY, 41(5), 1338-1353. doi:10.1093/ije/dys078
Fujimura, H., Altar, C. A., Chen, R., Nakamura, T., Nakahashi, T., Kambayashi, J., . . . Tandon, N. N. (2002). Brain-derived neurotrophic factor is stored in human platelets and released by agonist stimulation. Thrombosis and Haemostasis, 87(4), 728-734. doi:10.1055/s-0037-1613072 Glud, M., Christiansen, T., Larsen, L. H., Richelsen, B., & Bruun, J. M. (2019). Changes in
Circulating BDNF in relation to Sex, Diet, and Exercise: A 12-Week Randomized Controlled Study in Overweight and Obese Participants. Journal of Obesity. doi:10.1155/2019/4537274
Griffin, É. W., Mullally, S., Foley, C., Warmington, S. A., O'Mara, S. M., & Kelly, Á. M. (2011). Aerobic exercise improves hippocampal function and increases BDNF in the serum of young adult males. Physiology & Behavior, 104, 934-941. doi:10.1016/j.physbeh.2011.06.005 Griffiths, B. J., Parish, G., Roux, F., Michelmann, S., van der Plas, M., Kolibius, L. D., . . .
Hanslmayr, S. (2019). Directional coupling of slow and fast hippocampal gamma with neocortical alpha/beta oscillations in human episodic memory. Proceedings of the National Academy of Sciences of the United States of America. doi:10.1073/pnas.1914180116 Henry, C. J. (2005). Basal metabolic rate studies in humans: measurement and development of new
equations. Public Health Nutrition, 8(7A), 1133-1152. doi:10.1079/PHN2005801 Hötting, K., & Röder, B. (2013). Beneficial effects of physical exercise on neuroplasticity and
cognition. Neuroscience & Biobehavioral Reviews, 37(9), 2243-2257. doi:10.1016/j.neubiorev.2013.04.005
Hwang, J., Brothers, R. M., Castelli, D. M., Glowacki, E. M., Chen, Y. T., Salinas, M. M., . . . Calvert, H. G. (2016). Acute high-intensity exercise-induced cognitive enhancement and brain-derived neurotrophic factor in young, healthy adults. Neuroscience Letters, 630, 247-253. doi:10.1016/j.neulet.2016.07.033
Jiménez-Maldonado, A., Rentería, I., García-Suárez, P. C., Moncada-Jiménez, J., & Freire-Royes, L. F. (2018). The Impact of High-Intensity Interval Training on Brain Derived Neurotrophic Factor in Brain: A Mini-Review. Frontiers in Neuroscience, 12, 839.
doi:10.3389/fnins.2018.00839
Kao, S., Drollette, E. S., Ritondale, J. P., Khan, N., & Hillman, C. H. (2018). The acute effects of high-intensity interval training and moderate-intensity continuous exercise on declarative memory and inhibitory control. Psychology of Sport & Exercise, 38, 90-99.
doi:10.1016/j.psychsport.2018.05.011
Knaepen, K., Goekint, M., Heyman, E. M., & Meeusen, R. (2010). Neuroplasticity - Exercise-Induced Response of Peripheral Brain-Derived Neurotrophic Factor. Sports Medicine, 40(9), 765-801. doi:10.2165/11534530-000000000-00000
Lavie, C. J., Ozemek, C., Carbone, S., Katzmarzyk, P. T., & Blair, S. N. (2019). Sedentary Behavior, Exercise, and Cardiovascular Health. Circulation Research, 124(5), 799–815.
doi:10.1161/CIRCRESAHA.118.312669
Lowe, C. J., Kolev, D., & Hall, P. A. (2016). An exploration of exercise-induced cognitive
enhancement and transfer effects to dietary self-control. Brain and Cognition, 110, 102-111. doi:10.1016/j.bandc.2016.04.008
Lucas, S. J., Cotter, J. D., Brassard, P., & Bailey, D. M. (2015). High-intensity interval exercise and cerebrovascular health: curiosity, cause, and consequence. Journal of Cerebral Blood Flow & Metabolism, 35, 902-911. doi:10.1038/jcbfm.2015.49
Miyake, A., & Friedman, N. P. (2012). The nature and organization of individual differences in executive functions: Four general conclusions. Current Directions in Psychological Science, 21(1), 8-14. doi:10.1177/0963721411429458
Miyake, A., Friedman, N. P., Emerson, M. J., Witzki, A. H., Howerter, A., & Wager, T. D. (2000). The unity and diversity of executive functions and their conributions to complex "Frontal lobe" tasks: a latent variable analysis. Cognitive Psychology, 41(1), 49-100.
Montoya, A. K., & Hayes, A. F. (2017). Two condition within-participant statistical mediation analysis: A path-analytic framework. Psychological Methods, 22(1), 6-27.
doi:10.1037/met0000086
Pan, W., Banks, W. A., Fasold, M. B., Bluth, J., & Kastin, A. J. (1998). Transport of brain-derived neurotrophic factor across the blood-brain barrier. Neuropharmacology, 37(12), 1553-1561. doi:10.1016/s0028-3908(98)00141-5
Pareja-Galeano, H., Alis, R., Sanchis-Gomar, F., Cabo, H., Cortell-Ballester, J., Carmen Gomez-Cabrera, M., . . . Viña, J. (2015). Methodological considerations to determine the effect of exercise on brain-derived neurotrophic factor levels. Clinical Biochemistry, 48(3), 162-166. doi:10.1016/j.clinbiochem.2014.11.013
Patterson, E. (2002, October). Downloadable questionnaires. Retrieved from The International Physical Activity Questionnaire: https://sites.google.com/site/theipaq/questionnaire_links Patterson, E. (2005, November). IPAQ scoring protocol. Retrieved from The International Physical
Activity Questionnaire: https://sites.google.com/site/theipaq/scoring-protocol
Preston, A. R., & Eichenbaum, H. (2013). Interplay of hippocampus and prefrontal cortex in memory. Current Biology, 23(17), R764–R773. doi:10.1016/j.cub.2013.05.041
Rasmussen, P., Brassard, P., Adser, H., Pedersen, M. V., Leick, L., Hart, E., . . . Pilegaard, H. (2009). Evidence for release of brain-derived neurotrophic factor from the brain during exercise. Experimental Physiology, 94(10), 1062-1069. doi:10.1113/expphysiol.2009.048512
Saucedo Marquez, C. M., Vanaudenaerde, B., Troosters, T., & Wenderoth, N. (2015). High-intensity interval training evokes larger serum BDNF levels compared with intense continuous exercise. Journal of Applied Physiology, 119, 1363-1373.
doi:10.1152/japplphysiol.00126.2015
Schwarck, S., Schmicker, M., Dordevic, M., Rehfeld, K., Müller, N., & Müller, P. (2019). Inter-Individual Differences in Cognitive Response to a Single Bout of Physical Exercise—A Randomized Controlled Cross-Over Study. Journal of Clinical Medicine, 8(8), 1101. doi:10.3390/jcm8081101
Slusher, A. L., Patterson, V. T., Schwartz, C. S., & Acevedo, E. O. (2018). Impact of high intensity interval exercise on executive function and brain derived neurotrophic factor in healthy college aged males. Physiology & Behavior, 191, 116-122.
doi:10.1016/j.physbeh.2018.04.018
Suzuki, W. A. (2005, February). Associative Learning and the Hippocampus. Retrieved from American Psychological Association: https://www.apa.org/science/about/psa/2005/02/suzuki Szuhany, K. L., Bugatti, M., & Otto, M. W. (2015). A meta-analytic review of the effects of exercise
on brainderived neurotrophic factor. Journal of Psychiatric Research, 60, 56-64. doi:10.1016/j.jpsychires.2014.10.003.
Tsukamoto, H., Suga, T., Takenaka, S., Tanaka, D., Takeuchi, T., Hamaoka, T., . . . Hashimoto, T. (2016). Greater impact of acute high-intensity interval exercise on post-exercise executive function compared to moderate-intensity continuous exercise. Physiology & Behaviour, 155, 224-230. doi:10.1016/j.physbeh.2015.12.021
Voss, M. W., Vivar, C., Kramer, A. F., & van Praag, H. (2013). Bridging animal and human models of exercise-induced brain plasticity. Trends in Cognitive Sciences, 17(10), 525-544.
Wheeler, M. J., Dempsey, P. C., Grace, M. S., Ellis, K. A., Gardiner, P. A., Green, D. J., & Dunstan, D. W. (2017). Sedentary behavior as a risk factor for cognitive decline? A focus on the influence of glycemic control in brain health. Alzheimer's & Dementia: Translational Research & Clinical Interventions, 3(3), 291-300. doi:10.1016/j.trci.2017.04.001 Zeineh, M. M., Engel, S. A., Thompson, P. M., & Bookheimer, S. Y. (2003). Dynamics of the
hippocampus during encoding and retrieval of face-name pairs. Science, 299(5606), 577-58-. doi:10.1126/science.1077775
