The effects of using an active workstation on executive function in Chinese college students
The effects of using an active workstation on executive function in Chinese college students
Zhanjia Zhang 1 2
Bing Zhang 0 2
Chunmei Cao 0 2
Weiyun Chen 1 2
0 Department of Sports Science and Physical Education, Tsinghua University , Beijing , China
1 School of Kinesiology, University of Michigan , Ann Arbor, Michigan , United States of America
2 Editor: Yih-Kuen Jan, University of Illinois at Urbana-Champaign , UNITED STATES
This study aimed to examine the effects of active workstation use on the executive function by measuring the three components of executive function (Inhibition, Updating, and Shifting) during sitting, standing, and walking at an active workstation with different speeds. Twentyfour college students completed a cognitive test battery while sitting, standing, walking on an active workstation with a self-selected speed (mean = 2.3 km/h) and a faster speed (mean = 3.5 km/h). The three components of executive function (Inhibition, Updating, and Shifting) were assessed by Stroop task, N-back task, More-odd shifting task, respectively. Performance of each task was determined by the response time and accuracy. Repeated measures ANOVAs were conducted with workstation condition and trial type as within-subjects factors. There were no significant main effects for workstation condition and no interaction between workstation condition × trial type in Stroop task and More-odd shifting task. There was a significant main effect for workstation condition (F (3, 69) = 4.029, p = 0.011) and interaction effect between workstation condition × trial type (F (6, 138) = 9.371, p < 0.001) in N-back task. Decomposition of the interaction showed that accuracy of 2-back task in self-paced walking was significantly lower than that in sitting condition (p = 0.017) and in standing condition (p < .001). But there was no difference in accuracy of 2-back task between self-paced walking condition and faster walking condition (p = 0.517). Our results suggest that using an active workstation may have a selective impact on three components of executive function, in which the Updating may be impaired to a certain extent while the Inhibition and Shifting remain unaffected.
Data Availability Statement: All relevant data are
within the paper and its Supporting Information
Funding: The authors received no specific funding
for this work.
Competing interests: The authors have declared
that no competing interests exist.
Sedentary lifestyle and physical inactivity have numerous adverse effects on health, such as
increased morbidity of cardiovascular diseases and higher mortality from all causes [
Unfortunately, the opportunities for physical activity have been largely eliminated due to the changes
in the way we work, commute, and spend leisure time [
]. It has been pointed out that the
rapid development of technology in the past fifty years has been making our environment
more inclined to result in physical inactivity and sedentary lifestyle [
]. According to a survey,
the percentage of US adults occupied in sedentary work has increased by 67%, and 46% people
spend almost all their work time sitting [
]. Since nearly half of our waking hours are spent at
work, the sedentary work environment largely accounts for daily physical inactivity. At
present, the major strategy to address the issue of physical inactivity due to the sedentary work
environment is to encourage employees to engage in more exercise after work such as by
providing some free fitness programs for employees or using financial incentives [
]. However, a
systematic review showed that such worksite physical activity promotion strategies were not
effective as they intended to be [
]. A study assessed the employee's attitude toward worksite
health promotion services and found that the main reported barriers to participating the
worksite health programs included no time during the workday and no time before or after work
]. In addition, the main reported incentives that would promote the employee's physical
activity include convenient time and convenient location. Given these barriers and incentives,
it might be effective to promote physical activity if the way of doing physical activity does not
occupy the time after work and occurs at convenient location, which was the very idea that the
active workstation was derived from.
The idea of active workstation was first proposed by Edelson and Danoffz in 1989. Edelson
and Danoffz [
] designed the active workstation by combining a treadmill and an office desk
together so that people were able to perform working tasks while walking. Since the active
workstation was designed to reduce sedentariness and promote physical activity at workplace,
researchers were concerned about two questions in its application: 1) Does the use of the active
workstation effectively increase physical activity at workplace? 2) Will the use of the active
workstation significantly influence work performance? Results from previous studies
consistently showed that the use of active workstation effectively improved physical activity (PA) and
increased energy consumption at workplace. For instance, Levine and Miller [
] found that
the energy consumption while walking at 1.1 mph at an active workstation was 191 kcal/h,
which was 119 kcal/h higher than that in sitting condition. Regarding the influence of the use
of active workstation on work conformance, results of the existing literature were inconsistent
due to the large variability of participants and measurement. For example, some early studies
adopted typing performance as outcomes reporting detrimental impact [
] while some
used subjective measurement such as supervisors' rating showing no impairment . More
recently, a growing literature emerged focusing on the effects of active workstation use on
cognitive functions. For instance, Ohlinger [
] administered Stroop Color Word test and
Auditory Consonant Trigram test and showed that walking on an active workstation with 1 mph
did not affect selective attention and short-term auditory verbal memory. Similarly, Alderman
et al. [
] also revealed that selective attention was not affected when walking on an active
workstation with self-selected speed. Ehhamn et al. [
] examined executive function and
found the executive function performance was relatively unaffected while walking on an active
workstation with self-selected speed. The effects of the use of active workstation on cognitive
functions might be the potential foundation for work performance while using the active
workstation. Therefore, investigations of the relationship between the use of active workstation
and cognitive functions may help us to understand what types of working tasks may be
influenced and what types may not, and thus to better facilitate the use the active workstation.
Among the cognitive functions, executive function is considered as a higher-order cognitive
function since it matures the latest and controls lower-order cognitive functions [
Executive function is a set of cognitive processes that modulate and coordinate various cognitive
processes to facilitate the attainment of chosen goals [
]. According to the theoretical model
proposed by Miyake et al. [
], executive function consists of three core
componentsÐÐinhibition of proponent responses (ªInhibitionº), information updating and monitoring
(ªUpdatingº), and mental set shifting (ªShiftingº). The Inhibition refers to one's ability to deliberately
inhibit dominant, automatic, or prepotent responses when necessary [
]. The Updating,
2 / 13
which is highly close to the notion of working memory, refers to one's ability to monitor and
code incoming information for relevance to the task at hand and replace old, no longer
relevant information with newer, more relevant information [
]. The last component of executive
function requires shifting back and forth between multiple tasks, operations, or mental sets
]. Given the importance of the executive function, it was of our interest to investigate
whether the use of active workstation has an impact on executive function. Additionally, since
the executive function consists of three components that are related to different cognitive
processes, we were also interested in the effects of the use of active workstation on each of the
three components of executive function.
Taken together, as a promising way to promote physical activity workplace, the influences
of the use of active workstation on cognitive functions have remained largely unexplored. The
purpose of this study was to examine the effects of active workstation use on the three
components of executive function. Specifically, we measured and compared the three components of
executive function when people were sitting, standing, and walking on an active workstation
with different speeds (one self-selected speed, and one faster speed). In other words,
participants were tested executive function during four different workstation conditions. We
hypothesized that the three components of executive function would not be influenced when
participants were walking with a self-selected speed but would be impaired when participants
were walking with a faster speed. We made these hypotheses based on the fact that walking
was a highly automated skill, and thus walking with a self-selected speed should require
minimum cognitive resources while walking with a faster speed may require increased cognitive
resources. However, our results indicated there was no difference in executive function
between walking with a self-selected and walking with a faster speed. In addition, we found
that the use of active workstation had a selective impact on the three components of executive
function. Detailed procedures of the experiment, results, and relevant discussions are
Materials and methodsParticipants
Twenty-four college students (12 men, mean age = 24.0 years, SD = 1.5 years; 12 women,
mean age = 22.1 years, SD = 1.5 years) were recruited by posted flyers. Potential participants
were excluded if they were from Psychology Department, have color blindness or weakness,
have neurological disorders, have a balance disorder, or have existing injuries that would
restrict walking. All participants completed the Physical Activity Readiness Questionnaire
] and provided written informed consent prior to the experiment. The research
protocol was approved by the Institutional Review Board of Tsinghua University.
The main independent variable of the current study was the workstation conditions, and the
main dependent variable was the executive function performance, reflected by response time
and accuracy. This study used a within-subject design, in which each participant performed a
test battery of executive function under each of four conditions, including sitting, standing,
walking at an active workstation with self-selected speed (self-paced walking), and walking at
an active workstation with 1.5 times the self-selected speed (faster walking).
In April and May 2015, each participant visited the lab during four days with one day apart
between two consecutive visiting days. In each day, participants performed the executive
function test under one of the four experimental conditions (sitting, standing, self-paced walking,
3 / 13
Note. Subjects were randomly divided into four groups (A, B, C, and D). Each row of the table showed the sequence of the test conditions of each group.
faster walking). In order to eliminate the order effects within the repeated measures design, the
order in which participants performed the experimental conditions was counterbalanced
(Table 1). During the sitting condition, participants performed the executive function test
sitting at an office desk. During the standing condition and two walking conditions, participants
performed the executive function test standing and walking on an active workstation,
respectively (Fig 1). The active workstation used in this study (Lifespan, Salt Lake City, UT, USA)
Fig 1. A participant was performing executive function test on an active workstation.
4 / 13
consisted of a treadmill and a height-adjustable desk. Prior to the first time using the active
workstation, each participant had 10 minutes to practice walking on the active workstation.
After participants got used to walking on the active workstation, self-selected speed was
determined as their most comfortable walking speed on the active workstation.
Participants visited the lab in the morning at two hours after breakfast and were not allowed
to have exercised prior to the experiment that morning. When participants were performing
the executive function test, a heart rate monitor (Polar Electro, Finland) was applied to identify
the aerobic intensity of the active workstation condition. The executive function test started
immediately after the participant reached a stable status of heart rate under each workstation
condition. The entire test session lasted about 30 minutes.
The test battery of executive function consisted of three cognitive tasks to test the three
components of the executive function respectively. There were 2-minute intervals between each task.
The total duration of the executive function test was about 25 minutes. The program of the test
battery was written using the Psychtoolbox package within Matlab (version 2014a).
Stroop task. Stroop Color Word task was used to measure the Inhibition function, in
which participants were required to name a color word. There were congruent and
incongruent for the trials. In the congruent trials, the name of the color word was same as the ink of the
color word (e.g., the word ªredº printed in red ink), while in the incongruent condition, the
name of the color word was different as the ink of the color word (e.g., the word ªredº printed
in blue ink). In this study, there were six kinds of the trials, 1) the word ªredº printed in red
ink, 2)the word ªblueº printed in blue ink, and 3) the word ªgreenº printed in green ink, which
were regarded as congruent condition, and 4) the word ªredº printed in blue or green ink, 5)
the word ªblueº printed in red or green ink, and 6) the word ªgreenº printed in red or blue
ink, which were regarded as incongruent condition. There were 96 trials in which 48 trials
were congruent and 48 trials were incongruent. Each stimulus was presented 2000ms and
between two stimuli was 2 to 8 seconds interval with sign ª+º presented on the screen. The
stimuli were presented in a random order and participants were required to tell the color name
of the words rather than the color of ink by pressing corresponding buttons on the keyboard.
N-back task. N-back task was used to measure the Updating function, which required
participants to monitor a series of letters shown on the screen and match the current stimulus
with the one that presented N steps earlier in the sequence. The parameter N in this study
included 0, 1, 2, with increased number N associated with increased complexity of the task. In
the 0-back task, the participants were only required to identify a pre-specified letter ªXº (e.g.,
in a series letters ªABSXISXSDº, the underline letters were the target stimuli). In the 1-back
task, the participants were required to identify the letter that was same as the last presented
letter (e.g., in a series letters ªASXDDARRSIEEº, the underline letters were the target stimuli). In
the 2-back task, the participants were required to identify the letter that was same as the letter
presented prior to the last letter (e.g., in a series letters ªADFDJSJISISOº, the underline letters
were the target stimuli). There were two blocks for each task and 18 trials in each block, in
which 6 trials were target stimuli. Each stimulus was presented 500ms followed by 1500ms
sign ª+º presented at the center of the screen.
More-odd shifting task. More-odd shifting task was used to measure the Shifting
function, which required participants to switch between different mental tasks and make
corresponding responses. A series of number (1to 4, and 6 to 9) were presented at the center of the
screen with two conditions: 1) if the color of the number was red, participants were required
to identify whether the number is larger or smaller than 5 by pressing corresponding buttons
5 / 13
on keyboard; and 2) if the color of the number was green, participants were required to
identify whether the number is odd or even by pressing corresponding buttons on the keyboard. In
the block ªAº, all trials were red numbers, known as the ªMore-trialsº. In the block ªBº, all
trials were green numbers, known as the ªOdd-trialsº. In the block ªCº, trials were mixed
conditions with red numbers and green numbers presented in a random order, known as the
ªMixed-trialsº. There were each 16 trials in block ªAº and ªBº, and 32 trials in block ªCº. The
order of the blocks was ªABCCBAº with each block presented twice. Each stimulus was
presented 2000ms and between two stimuli was 3 seconds interval with sign ª+º presented on the
In order to examine the effects of active workstation use on executive function, a series of
analysis of variance (ANOVA) with repeated measures was conducted with workstation condition
(sitting, standing, self-paced walking, faster walking) and trial type (congruence vs.
incongruence for Stroop task; 0-, 1-, 2-back for N-back task; More-, Odd-, Mixed-trial for Shifting task)
as within-subjects factors. Bonferroni's post hoc procedure was used for post hoc comparisons
if ANOVAs reported a significant main effect. Skewness and kurtosis of the data were checked
for normality according to Kline's criteria [
] prior to performing ANOVAs. The results
indicated that most variables were normal except for the accuracy of 1-back task under sitting
condition and the accuracy of congruent Stroop task under self-paced walking condition. Given
that the ANOVA produces valid results even when the normality is violated [
], and given
that the repeated measures design removes the individual differences, we tolerated the
nonnormality in these two variables, but nevertheless we considered it as a limitation. The main
dependent variables in this study were the three components of executive function: Inhibition,
Updating, and Shifting. The three components were determined by the average response time
and accuracy of Stroop task, N-back task, and More-odd shifting task, respectively. Shorter
response time and higher accuracy represented better performance. Statistical significant level
for all analyses was set at p < 0.05. All data were collated and analyzed using IBM SPSS version
time in incongruent trials than in congruent trials. However, there was no significant main
effect for workstation condition and no interaction between workstation condition × trial type
in both accuracy and response time (all p values > 0.05), suggesting that the Stroop task
performance did not vary across four workstation conditions.
Regarding the N-back task, there was a significant main effect for trial type in accuracy
(F (2, 46) = 677.319, p < 0.001). Post hoc comparisons indicated participants showed lower
accuracy in 2-back task than in 0-back task (p < 0.001) and 1-back task (p < 0.001), but there
was no difference in accuracy between 0-back task and 1-back task (p = 0.075). There was also
a significant main effect for workstation condition in accuracy (F (3, 69) = 4.029, p = 0.011) but
it was superseded by interaction effect between workstation condition × trial type (F (6, 138) =
9.371, p < 0.001). The decomposition of the interaction indicated the accuracy of 2-back task
at self-paced walking (74.9%) was significantly lower than sitting (78.1%, p = 0.017) and
standing (83.0%, p < 0.001). Accuracy of 2-back task at faster walking (74.0%) was also significantly
lower than sitting (p < 0.015) and standing (p < 0.001). In addition, accuracy of 2-back task at
sitting was significantly lower than standing (p = 0.017), but there was no significant difference
between self-paced walking and faster walking (p = 0.517). No difference in accuracy across
four workstation condition was found within the 0-back task and 1-back task.
7 / 13
Fig 2. Comparisons of accuracy and response time among workstation conditions. Note: error bars denote
standard error of the mean; asterisks denote significant pairwise comparisons.
There was also a significant main effect for trial type (F (2, 46) = 26.963, p < 0.001) but no
main effect for workstation condition (p = 0.739) and no interaction between workstation
condition × trial type (p = 0.377) in response time. Post hoc comparisons indicated
participants spent more time in 2-back task than in 1-back (p < 0.001) and 0-back task (p < 0.001).
However, there was no difference in response time across four workstation condition within
More-odd shifting task
ANOVA revealed a significant main effect for trial type in both accuracy (F (2, 46) = 9.912,
p < 0.001) and response time (F (2, 46) = 142.132, p < 0.001) of More-odd shifting task. Post
hoc comparisons found participants showed lower accuracy and longer response time in
Mixed-trials than that in More-trials and Odd-trials (all p values < 0.05). The accuracy in
Odd-trials was significantly lower than in More-trials (p = 0.006), and the response time of
Odd-trials was significantly longer than that of More-trials (p < 0.001). There was no
significant main effect for workstation condition and no interaction between workstation condition ×
trial type in both accuracy and response time (all p values > 0.05), indicating the performance
of More-odd shifting task did not differ across four workstation conditions.
8 / 13
According to a widely accepted theoretical model [
], this study tested the three core
components of executive function under different workstation conditions. The results indicated that
active workstation use had a selective impact on executive function.
Walking on the active workstation did not affect the Inhibition, a core component of the
executive function, whether with a self-selected speed or with a faster speed. Inhibition was
measured by Stroop Color Word task, in which participants were required to identify the color
name of the word in two conditions, congruent or incongruent. For both congruent and
incongruent trials, there was no significant difference in accuracy and response time across the
four workstation condition (sitting, standing, self-paced walking, and faster walking). Our
results were consistent with the findings by John et al. [
]and Alderman et al. [
], both of
whom employed Stroop task as a measurement of selective attention. John et al. [
walking at an active workstation at 1 mph did not affect the performance of Stroop compared
to sitting. Alderman et al. [
]also found no significant difference in response speed and
accuracy of Stroop task between walking and sitting conditions. The walking speed in Alderman
et al.'s [
] study was self-selected and the average walking speed was 2.45km/h, which was
very close to the average self-selected walking speed in the current study (2.3km/h). However,
the current study found that walking with an even faster speed also did not influence the
results of Stroop task compared to sitting.
Results in this study showed that the performance in both in 0-back and 1-back task did not
differ across the four workstation conditions. But the performance in 2-back task was
diminished during self-paced walking and faster walking compared to sitting, suggesting that the
active workstation use might impair the Updating to a certain extent and such impairment
was dependent on the working memory load. Updating, which is another component of the
executive function, was assessed by N-back task, in which participants were required to match
the current information with previously presented information, so the working memory was
largely involved in the N-back task. The increase of the parameter N (from 0 to 2 in the current
study) was associated with increasing complexity of N-back task because of the elevating
working memory load, which could be indicated from the decreasing accuracy and increasing
response time from 0-back task to 2-back task.
Shifting ability was not affected by the active workstation use. Shifting was evaluated by the
More-odd shifting task, in which participants were required to quickly switch between two
different tasks, one to judge the size of a number and the other to judge the odd-even of a
number. This result is in line with Ehmann et al. study [
], which adopted Wisconsin Card
Sorting test as a measure of cognitive flexibility and found the cognitive flexibility was
unaffected during walking on an active workstation with self-preferred speed compared to sitting.
Overall, our study found that active workstation use had a selective impact on the three
components of the executive function. Active workstation use did not affect the Inhibition and
Shifting but caused the impairment of Updating to some extent. Since the Updating
component of executive function is highly related to working memory, our findings suggest that
working memory might be more affected than other cognitive processes when using an active
workstation. Previous studies regarding the effects of the use of active workstation on executive
function mainly focused on the Inhibition component of the executive function [
current study extended the extant scholarship by examining the effects of the use of active
workstation on the three components of executive function separately. Our findings showing a
selective impact of the use of active workstation also indirectly support the Miyake et al.'s 
proposition that the three components of executive function are distinguishable and should be
assessed separately [
9 / 13
There might be two potential mechanisms to explain the change of cognitive function
under active workstation condition. First, according to the arousal theory [
], the increased
level of arousal induced by exercise intensity, which is usually measured by heart rate, oxygen
uptake, or perceived exertion, has an inverted-U impact on the performance of cognitive tasks.
The positive effect of arousal on cognitive function is also found to relate to the activation of
central nervous system linked to the level of catecholamine [
]. A previous study indicated
that the optimal zone of acute exercise intensity for cognitive performance ranged from 40%
to 60% of maximal oxygen uptake [
]. Although the current study did not measure the
oxygen uptake directly, we monitored the heart rate of participants during each workstation
condition. It showed that the heart rate increased significantly from the sitting and standing
conditions to active workstation conditions, indicating an elevating level of arousal. However,
it is possible that such increase in arousal had not reached a level that was needed for altering
the cognitive performance.
Second, under the active workstation condition, the cognitive tests were performed during
low-intensity exercise, which could be regarded as a dual-task scenario. Participants were
confronting one physical task and one cognitive task. According to the classic and temporary
theories of attention, attention has a limited capacity and the cognitive performance might be
compromised if there is competition for the attention resources . Studies indicate that
walking, although highly practiced, is not an entirely automated motor skill and need certain
attention resources [
]. Therefore, allocation of attention resources plays an important
role in cognitive performance during active workstation condition. In addition, walking speed
matters in the attention allocation when simultaneously performing a cognitive task and a
physical task. Walking at a higher speed increases the instability of body and thus imposes a
greater attentional demand on participants to maintain balance. A previous study found typing
while walking at a lower speed (1.6 km/h) changed the gait kinematics and decreased the
postural stability . Funk et al. [
] compared the typing performance under four conditions
including sitting, walking at 1.3 km/h, 2.25 km/h, and 3.2 km/h, and found that typing
performance diminished when walking at 1.3 km/h and 3.2 km/h compared to sitting while there
was no difference in typing performance between walking at 2.25 km/h and sitting, indicating
there might exist an optimal zone of walking speed in active workstation use. Our study
employed two different walking speeds including a self-selected speed and a faster speed which
was 1.5 times the self-selected speed. However, we did not find any difference in cognitive test
outcomes between two different walking speeds. It was possible that although the speeds in
two walking conditions differed greatly, the exercise intensity did not differ much, which was
indicated by the close heart rate under two active workstation conditions. It has been indicated
that the attentional demand could be strongly related to the energy demand of the task, with
greater energy demand associated with more attentional demand for controlling the
]. Furthermore, the type of cognitive task could also impact the allocation of
attention resources during dual-task scenario. Easier tasks might require fewer attention resources
than more complex ones. It was supported in our study by the results of N-back task. While
performing N-back task during walking, participants only showed decreased performance in
2-back task but not in 0-back task and 1-back task, in which the former one requires more
working memory resources than the latter two.
Several study limitations should be noted. First, the study was limited to a narrow population
of sample. The participants in our study were college students and not actually employed office
workers, which might prevent the generalization of results to other population, such as middle
10 / 13
age or elderly people, who might also be potential active workstation users. Second, we did not
examine the role of physical fitness in the relationship between cognitive performance and
active workstation use. It is possible that people with higher physical fitness level might feel
more comfortable in using active workstation. A third limitation of the study was the short
period of the assessment. It could be possible that prolonged use of active workstation might
cause fatigue and thus impair cognitive function. In addition, effects of long-term use of active
workstation on cognitive function were not examined in the current study. Long-term use
might get people more accustomed to the active workstation and thus eliminate its potential
negative effects. Future research is also suggested to investigate the delayed effects of active
workstation use on cognitive functions as well as work performance.
Despite these limitations, our study has strengths in its counterbalanced design and testing
different components of executive function. We employed self-selected walking speed rather
than fixed walking speed for every participant since the preferred walking speed was known to
differ among individuals based on various factors. In addition, the current study compared the
effects of two different walking speeds in active workstation use on cognitive performance.
In summary, the present study extended the extant knowledge on active workstation by
examining the effects of active workstation use on the three components of executive function. The
results showed that walking at an active workstation had a selective impact on the three
components of executive function, in which Updating was impaired to a certain extent while
Inhibition and Shifting were not affected. Since Updating is highly correlated to the working
memory, it is indicated that active workstation use might be more compatible with
non-working memory-intensive tasks. In conjunction with its ability to increase energy consumption
and daily physical activity, which has been well demonstrated in previous studies, active
workstation might be a feasible solution to eliminate sedentariness in work environment.
S1 Dataset. Table containing the raw data for all executive function tests.
We would like to thank all the participants in this study. We thank Keying Zhang, Tianqing Li,
and Huiyuan Jia for their help during the experiment.
Conceptualization: Bing Zhang.
Formal analysis: Zhanjia Zhang.
Investigation: Zhanjia Zhang.
Methodology: Chunmei Cao.
Resources: Bing Zhang.
Supervision: Bing Zhang.
Writing ± original draft: Zhanjia Zhang.
Writing ± review & editing: Chunmei Cao, Weiyun Chen.
11 / 13
12 / 13
1. Katzmarzyk PT , Church TS , Craig CL , Bouchard C . Sitting time and mortality from all causes, cardiovascular disease, and cancer . Med Sci Sports Exerc . 2009 ; 41 ( 5 ): 998 ± 1005 . https://doi.org/10.1249/ MSS.0b013e3181930355 PMID: 19346988
2. Hallal PC , Andersen LB , Bull FC , Guthold R , Haskell W , Ekelund U , et al. Global physical activity levels: surveillance progress, pitfalls, and prospects . The lancet . 2012 ; 380 ( 9838 ): 247 ± 57 .
3. Brownson RC , Boehmer TK , Luke DA . Declining rates of physical activity in the United States: what are the contributors ? Annu Rev Public Health . 2005 ; 26 : 421 ± 43 . https://doi.org/10.1146/annurev. publhealth. 26 .021304.144437 PMID: 15760296
4. Sparling PB . Worksite health promotion: principles, resources, and challenges . Prev Chronic Dis . 2010 ; 7(1):A25 . PMID: 20040240
5. Dishman RK , Oldenburg B , O'Neal H , Shephard RJ . Worksite physical activity interventions . American journal of preventive medicine . 1998 ; 15 ( 4 ): 344 ± 61 . PMID: 9838977
6. Kruger J , Yore MM , Bauer DR , Kohl HW III. Selected barriers and incentives for worksite health promotion services and policies . American Journal of Health Promotion . 2007 ; 21 ( 5 ): 439 ± 47 . https://doi.org/ 10.4278/ 0890 -1171- 21 .5.439 PMID: 17515009
7. Edelson N , Danoffz J. Walking on an electric treadmill while performing VDT office work . ACM SIGCHI Bulletin . 1989 ; 21 ( 1 ): 72 ± 7 .
8. Levine JA , Miller JM . The energy expenditure of using a ªwalk-and-workº desk for office workers with obesity . British journal of sports medicine . 2007 ; 41 ( 9 ): 558 ± 61 . https://doi.org/10.1136/bjsm. 2006 . 032755 PMID: 17504789
9. Straker L , Levine J , Campbell A . The effects of walking and cycling computer workstations on keyboard and mouse performance . Human factors . 2009 ; 51 ( 6 ): 831 ± 44 . https://doi.org/10.1177/ 0018720810362079 PMID: 20415158
10. Thompson WG , Levine JA . Productivity of transcriptionists using a treadmill desk . Work . 2011 ; 40 ( 4 ): 473 ±7. https://doi.org/10.3233/WOR-2011 -1258 PMID: 22130064
11. Koepp GA , Manohar CU , McCrady-Spitzer SK , Ben-Ner A , Hamann DJ , Runge CF , et al. Treadmill desks: A 1-year prospective trial . Obesity . 2013 ; 21 ( 4 ): 705 ± 11 . https://doi.org/10.1002/oby.20121 PMID: 23417995
12. Ohlinger CM . The effect of active workstation use on measures of cognition, attention and motor skill: Miami University; 2009 .
13. Alderman BL , Olson RL , Mattina DM . Cognitive function during low intensity walking: a test of the treadmill workstation . J Phys Act Health . 2013 ; 11 : 752 ± 8 .
14. Ehmann PJ , Brush CJ , Olson RL , Bhatt SN , Banu AH , Alderman BL . Active Workstations Do Not Impair Executive Function in Young and Middle-Age Adults. Medicine and science in sports and exercise . 2017 ; 49 ( 5 ): 965 ± 74 . https://doi.org/10.1249/MSS.0000000000001189 PMID: 28009785
15. Romine CB , Reynolds CR . Sequential memory: a developmental perspective on its relation to frontal lobe functioning . Neuropsychology review . 2004 ; 14 ( 1 ): 43 ± 64 . PMID: 15260138
16. Funahashi S. Neuronal mechanisms of executive control by the prefrontal cortex . Neuroscience research . 2001 ; 39 ( 2 ): 147 ± 65 . PMID: 11223461
17. Miyake A , Friedman NP , Emerson MJ , Witzki AH , Howerter A , Wager TD . The unity and diversity of executive functions and their contributions to complex ªfrontal lobeº tasks: A latent variable analysis . Cognitive psychology . 2000 ; 41 ( 1 ): 49 ± 100 . https://doi.org/10.1006/cogp. 1999 .0734 PMID: 10945922
18. Thomas S , Reading J , Shephard RJ . Revision of the physical activity readiness questionnaire (PAR-Q) . Canadian journal of sport sciences . 1992 .
19. Kline R. Methodology in the social sciences. Principles and practice of structural equation modeling , 2nd ed. New York: Guilford Press; 2005 .
20. Glass GV , Peckham PD , Sanders JR . Consequences of failure to meet assumptions underlying the fixed effects analyses of variance and covariance . Review of educational research . 1972 ; 42 ( 3 ): 237 ± 88 .
21. John D , Bassett D , Thompson D , Fairbrother J , Baldwin D . Effect of using a treadmill workstation on performance of simulated office work tasks . Journal of physical activity & health . 2009 ; 6 ( 5 ): 617 ± 24 .
22. Yerkes RM , Dodson JD . The relation of strength of stimulus to rapidity of habit-formation . Journal of comparative neurology and psychology . 1908 ; 18 ( 5 ): 459 ± 82 .
23. Chmura J , Nazar K , Kaciuba-Uścilko H . Choice reaction time during graded exercise in relation to blood lactate and plasma catecholamine thresholds . International journal of sports medicine . 1994 ; 15 ( 04 ): 172 ± 6 .
24. Reilly T , Smith D . Effect of work intensity on performance in a psychomotor task during exercise . Ergonomics . 1986 ; 29 ( 4 ): 601 ±6. https://doi.org/10.1080/00140138608968294 PMID: 3709512 Wickens CD . Processing resources and attention . Multiple-task performance . 1991 ; 1991 : 3 ± 34 .
26. Regnaux JP , Robertson J , Smail DB , Daniel O , Bussel B . Human treadmill walking needs attention . Journal of neuroengineering and rehabilitation . 2006 ; 3 ( 1 ): 1 .
27. Abernethy B , Hanna A , Plooy A . The attentional demands of preferred and non-preferred gait patterns . Gait & Posture . 2002 ; 15 ( 3 ): 256 ± 65 .
28. Funk RE . Using the Active Workstation: Effects on Typing Speed and Walking Mechanics: Miami University; 2009 .
29. Funk RE , Taylor ML , Creekmur CC , Ohlinger CM , Cox RH , Berg WP . Effect of walking speed on typing performance using an active workstation . Perceptual and motor skills . 2012 ; 115 ( 1 ): 309 ± 18 . https://doi. org/10.2466/06.23.26.PMS. 115 .4. 309 -318 PMID: 23033765
30. Brisswalter J , Collardeau M , ReneÂ A . Effects of acute physical exercise characteristics on cognitive performance . Sports Medicine . 2002 ; 32 ( 9 ): 555 ± 66 . PMID: 12096929