Mental Health and Physical Activity 10 (2016) 18e24

Contents lists available at ScienceDirect

Mental Health and Physical Activity

journal homepage: www.elsevier.com/locate/menpa

The effects of exercise frequency on executive function in individuals

with Parkinson's disease
Manuela C. Caciula a, *, Michael Horvat b, Phillip D. Tomporowski b, Joe Nocera c
a
School of Health and Kinesiology, Georgia Southern University, PO Box 8076, Statesboro, GA, USA
b
Department of Kinesiology, University of Georgia, Ramsey Center 330 River Rd., Athens, GA 30602, USA
c
Department of Neurology, Emory University, 201 Dowman Drive, Atlanta, GA 30322 USA

a r t i c l e i n f o a b s t r a c t

Article history: Introduction: Cognitive function impairments can be observed early in Parkinson's disease (PD).
Received 16 June 2015 Although exercise is considered a valuable tool in delaying the progression of disease related outcomes,
Received in revised form the optimal frequency of exercise interventions has not been identied. Our purpose was to identify the
5 April 2016
effects of different frequencies of chronic exercise training on aspects of executive function, specically
Accepted 14 April 2016
Available online 16 April 2016
cognitive exibility and working memory in PD.
Methods: Forty-three participants (M age 68.5 (SD 11.3), 26 males), with idiopathic PD completed a
category switching task and an N-back task at baseline and after 12 weeks of multimodal exercise
Keywords:
Parkinson's disease
training. The participants were divided into two training frequency groups: high-frequency: 4e5 times
Exercise frequency each week (N 23, M age 68.6 (SD 5.8), 16 males), and low-frequency: 3 times or less each week
Executive function (N 20, M age 67.6 (SD 4.5), 10 males).
Results: Both frequency groups improved in executive function, showing improved global switch-cost
accuracy (F (1, 41) 5.08, p<.05, h_p^2 0.11) and N-back accuracy (F (1, 41) 17.37, p < 0.001,
h_p^2 0.29). The highfrequency group displayed signicantly greater reductions in global switch costs
(F(1, 41) 5.53, p < 0.05., h_p^2 0.09), and working memory response time (F(1,41) 14.96, p<0.001,
h_p^2 0.26), than the low-frequency group.
Conclusions: A high frequency, multimodal exercise program involving aerobic, strength, and exibility
exercises has a better impact on executive functioning in PD than a low-frequency exercise intervention.
2016 Elsevier Ltd. All rights reserved.

1. Introduction tasks involving planning, problem solving, set-elaboration, set

Parkinson's disease (PD) is identied as a movement disorder
and is characterized by a progressive neurodegeneration of the
basal ganglia which affects the motor control of planned and un-
planned movements (Agid, 1991). Cognitive impairments are
increasingly recognized as associated features of PD. In this context,
indices of executive function, especially working memory (WM),
are among the risk factors for the progression of dementia in PD
(Marder, Tang, Cote, Stern, & Mayeux, 1995; Stern, Marder, Tang, &
Mayeux, 1993; Woods & Troster, 2003). Other indicators of the
impaired executive function may include poor performance in
* Corresponding author. School of Health and Kinesiology, Georgia Southern
University, 62 Georgia Ave., Hollis Room 2117 A, Statesboro, GA 30458-8076, USA.
E-mail addresses: mbarna@georgiasouthern.edu (M.C. Caciula), mhorvat@uga.
edu (M. Horvat), ptomporo@uga.edu (P.D. Tomporowski), joenocera@emory.edu
(J. Nocera).
shifting, and set maintenance (Horstink, Berger, van Spaendonck,
van den Bercken, & Cools, 1990; Pagonabarraga & Kulisevsky, 2012).
The hallmark of WM is the ability to both maintain information
in a transient short-term store of limited capacity and simulta-
neously manipulate and transform information. Alteration of WM
could be partly responsible for the executive function impairments
encountered in individuals with PD (Beato et al., 2008). As the
disease progresses WM tends to deteriorate in PD. This decreases
the psychomotor speed at which individuals can plan, initiate, and
perform multi-tasking (Koerts, Van Beilen, Tucha, Leenders, &
Brouwer, 2011).
Presently, the body of literature suggests benets of exercise on
cognitive function for individuals with PD (Ahlskog, 2011; Cruise
et al., 2011; Murray, Sacheli, Eng, & Stoessl, 2014; Nocera,
Altmann, Sapienza, Okun, & Hass, 2010; Ridgel, Kim, Fickes,
Muller, & Alberts, 2011). The benets of chronic exercise in-
terventions on cognitive performance, particularly in older adults,

http://dx.doi.org/10.1016/j.mhpa.2016.04.001
1755-2966/ 2016 Elsevier Ltd. All rights reserved.
 M.C. Caciula et al. / Mental Health and Physical Activity 10 (2016) 18e24
typically have been explained in two ways. Neurologically-based
hypotheses have been advanced that target adaptations in brain
structure and function; e.g., angiogenesis, synaptogenesis, and in-
creases in brain-derived neurotropic factors, insulin-like growth
factors, and vascular endothelial growth factor (Prakash, Voss,
Erickson, & Kramer, 2015). The data amassed from studies con-
ducted over the past two decades that examine the effects of
chronic exercise on cognition implicate the importance of frontal
lobes and networks that underlie executive functioning.
Psychologically-based hypotheses grounded in psychological the-
ories of self-control emphasize the importance of the acquisition of
self-regulation skills that are learned during exercise training
(Audiffren & Andre , 2015; Tomporowski, Davis, Miller, & Naglieri,
2015). From this perspective, improvements in self-regulation are
realized via regular practice on tasks that require executive control.
Further, self-regulation skills acquired from repeated bouts of
physically and mentally challenging activities generalize to non-
exercise domains. The exercise-training intervention developed
for the present study was designed specically to promote mental
engagement and problem solving. The cognitive tests used in the
study were selected for their sensitivity to change. Tasks that
measure the task-switching, working memory, and inhibition have
been observed to reect the inuence of chronic exercise training
(Diamond, 2013).
Moreover, recent results suggested that exercise selectively
benets cognitive functioning in people with PD, by improving
frontal lobe based executive function (Cruise et al., 2011). Larson
et al. (2006) also indicated that exercising three or more times
per week may improve higher e order cognitive function, and
delay the onset of dementia in people with neurodegenerative
disease. Studies on animal models with PD exposed to an intensive
and progressive protocol of training on a treadmill, for 30 days at a
frequency of 2 times a day, showed signicant improvement in
motor performance of animals (running duration and speed)
(Fisher et al., 2004). The same protocol was used by Petzinger et al.
(2013) and the results indicated similar motor gains in treadmill
performance and an increase in dopamine release within the
motor basal ganglia. These changes could potentially explain the
neuroprotective effect of exercise. Another study looked at the
effects of 8 weeks of treadmill running on the cognitive function of
animal models, and compared four different exercise frequencies.
The results indicated that the frequency and intensity of exercise
have a profound impact on cognitive functions mainly in elderly
(Costa et al., 2003).
Although the latest ndings support the benets of exercise for
cognitive function in individuals with PD, lacking is information
concerning the optimal frequency of exercise training need to
promote cognitive benets. This study examined the effects of
frequency of multimodal exercise on selective aspects of executive
function, including WM and cognitive exibility, in individuals with
idiopathic PD without dementia. We hypothesized changes of ex-
ecutive functioning in individuals with PD following exercise
intervention, with highefrequency group (4e5 times/week) dis-
playing greater improvements than a lowerefrequency group (3
times/week). The cut-off frequencies were selected based on the
recommendations for physical activity in older adults formulated in
the WHO Guidelines Global Recommendations on Physical Activ-
ity for Health (Organization, 2010), and the recommendations
published by the American College of Sport Medicine (ACSM) po-
sition stand (Chodzko-Zajko et al., 2009). A key message is that at
least 150 min of moderate-intensity aerobic activity, or at least
75 min of vigorous-intensity aerobic activity, or an equivalent
combination is required for health benet in older adults. For
additional health benets, up to 300 min of moderate-intensity or
150 min of vigorous-intensity aerobic activity, or an equivalent
19
combination is recommended (Organization, 2010). Also, in-
dividuals with poor mobility should do balance exercise to prevent
falls on 3 or more days, and muscle-strengthening activities should
be done on two or more days (Organization, 2010).
2. Methods
2.1. Participants
Initially, 90 individuals with PD attended a screening visit; based
on inclusion and exclusion criteria 46 individuals were recruited
(n 46; 26 males and 20 females) from the Atlanta Metropolitan
area through the PD Support Group meetings. After the recruit-
ment, 3 females dropped out from the study. Individuals were
included based on the following criteria: 1) diagnosis of PD; 2) a
Montreal Cognitive Assessment (MoCA) scores  22 (a score lower
than 21 indicates increased odds of dementia) (Nasreddine et al.,
2005); 3) age between 50 yr. and 80 yr. and 4) medical clearance
for exercise and absence of other health related problems that
could interfere with safe participation in an exercise training pro-
gram. Participants were excluded from the study if they were
experiencing any neurological severe motoric impairment that
impacted their mobility, had a cardiovascular disease or metabolic
disorder, or had undergone deep brain stimulation surgery. A
summary of the clinical and demographic features of the partici-
pants is presented in Table 1.
2.2. Procedures
All participants were informed on the components of the study
and signed the University of Georgia Institutional Review Board
consent forms prior to beginning testing and exercising. The off-
times, which are an invariable consequence of levodopa and
dopamine agonists treatment in patients with PD, were controlled
and diminished by assessing and exercising the participants after
medicine ingestion, when clinically - dened best on state (Ouchi,
2001). Off-time refers to periods of the day when the medica-
tion is not working appropriately, causing worsening of motor and
non-motor PD symptoms. Levodopa administration has been
shown to improve motor function and subjective alertness without
compromising either reaction time or accuracy (Molloy et al.,
2006). Prolonged levodopa usage leads to a decline in efcacy
which creates uctuations in the motor performance of voluntary
movements, but only secondary, non-signicant uctuations in
cognitive, autonomic and sensory functions (Colosimo & De
Michele, 1999). Medication prescribed for control of PD was not
adjusted or withdrawn for any participants during the study. Par-
ticipants completed two testing sessions conducted one week apart
prior to beginning the exercise intervention, and in the rst 48 h
after the conclusion of 12 weeks of exercise training. During the
rst session, participants completed a demographic and medical
history questionnaire, took a screening test (MoCA), administered
by a PD specialist, and were trained to perform two executive
function (EF) tests: an auditory switch task and an auditory N-back
task. Session two took place after approximately 5 days (2 days)
and included retraining on the EF tasks followed by the baseline
administration of the two EF tests, counterbalanced across partic-
ipants, at 10 min after retraining. Retraining included one short trial
of the executive test and was given because discrete tasks, espe-
cially those with a relatively large cognitive component, are
forgotten relatively quickly. Also, considering that the initial
training was given approximately one week prior to the baseline
testing, we wanted to control for any retention decit that could
have occurred due to the loss of activity set. The learning effects on
the cognitive tests were diminished by training participants to a
20 M.C. Caciula et al. / Mental Health and Physical Activity 10 (2016) 18e24
Table 1
Demographic and clinical features of 43 patients with Parkinson's disease (PD) pre-intervention.
Variable
Number of patients used to describe how the Male/Female
Age (years)
Age of PD onset (years)
PD duration (years)
Education (years)
Exercise habits (times/week)
MoCA score
UPDRS motor score (on-state)
Hoehn and Yahr (on-state)
Notes: Values are expressed as mean (SD). HF high-frequency; LD low-frequency; MoCA Montreal Cognitive Assessment; UPDRS Unied Parkinson's Disease Rating
Scale; Hoehn and Yahr a system symptoms of PD progress, from stage 1 (the least debilitating), to stage 5 (the most debilitating); on-state periods when the PD
medication is working and symptoms are controlled.
performance criterion (<5% errors for a block of trials). All training
and testing trials were administered individually in a quiet room,
and the assessors were blinded to the exercise groups. Participants
sat in front of a laptop computer, wore headphones and pressed a
regular size optical mouse. Participants were instructed to respond
as quickly and accurately as possible.
Following session two, participants were assigned into one of
two exerciseefrequency groups, based on their preference: a
highefrequency group (exercise 4e5 times/week, for 30e45 min/
bout), and a lowefrequency exercise group (exercise dose  3
times/week, 30e45 min/bout). More precisely, participants chose
the number of exercise sessions they wanted to attend, and were
assigned to groups accordingly. Following treatments, all partici-
pants completed a replication of session 2, performed with alter-
nate versions of the switch task and the N- back, but maintaining
the same structure of the baseline testing session.
2.2.1. Auditory switch test
The auditory switch test is designed to measure cognitive ex-
ibility (Miyake et al., 2000), and shifting attention, an aspect of EF
known to deteriorate in PD (Horstink et al., 1990; Rustamov et al.,
2014). In the switch test, computer-generated letters or numbers
were presented binaurally via headphones via a commercial soft-
ware program, Cedrus-SuperLab. The letters consisted of four
vowels (A, E, I, and O) and four randomly selected consonants (B, D,
L, and C). The numbers consisted four even numbers (2, 4, 6. and 8)
and four odd numbers (1, 3, 5, and 7). The participant was required
to respond to each stimulus by pressing a specied key on the
mouse (even number-left key; odd number-right key; vowel letter-
left key; consonant letter-right key). Each key press was followed
100 ms later by the presentation of the next stimulus. Two types of
stimulus blocks of trials were used: in homogenous blocks, par-
ticipants were instructed to respond to letters only or numbers
only; in mixed, or heterogeneous blocks, participants were
instructed to respond alternatively to letters and numbers. In
mixed blocks, letters or numbers were presented in series lengths
of either two, three, or four stimuli. All participants completed one
200 numbers-only test, one 200 letters-only test, and one 220
mixed-condition test. Global switch cost was calculated based on
Response times (RT) for each trial and response. Accuracy was
scored as the percentage of correct responses. This protocol has
been utilized previously and is demonstrated to be reliable
(Cepeda, Kramer, & Gonzalez de Sather, 2001; Okumura, Cooper,
Ferrara, & Tomporowski, 2013).
2.2.2. N-back task
The N-back task is a continuous performance task of working
HF exercise LF exercise p
23 20
16/7 10/10
mean (SD) mean (SD)
68.6 (5.8) 67.65 (4.5) 0.55
57.5 (12.9) 58 (13.5) 0.62
10.6 (5.2) 9.8 (4.9) 0.83
18.3 (2.9) 17.9 (2.7) 0.81
2.5 (1.1) 2.2 (0.6) 0.24
25.5 (2.7) 25.4 (2.5) 0.88
26.4 (4.3) 26.5 (5.0) 0.95
median median
3 3
memory, and has been previously used in individuals with PD
(Beato et al., 2008; Costa et al., 2003). Participants performed an N-
back task based on the protocol developed by Perlstein and his
colleagues (Perlstein, Dixit, Carter, Noll, & Cohen, 2003). In the 0-
back condition, the target was any letter that matched a pre-
specied letter (i.e., c), which was a condition that required
sustained attention but no working memory demand. In the 1-back
condition, the target was any letter identical to the letter imme-
diately preceding it (i.e., the letter presented one trial back). In the
2-back condition, the target was any letter that was identical to the
one presented two trials back. In the 3-back condition, the target
was any letter that was identical to the one presented three trials
back. Stimuli were random consonants presented binaurally to
headphones via a commercial software program (Cedrus.
SuperLab) for a 500-ms duration with a 2500-ms inter-stimulus
interval. Participants completed 12 blocks of trials (three blocks
of each of the four conditions) with each block consisting of 25
trials. The rst three trials of each block were never targets and of
the remaining trials 30% were targets. The participant controlled
the duration of a short break (5e20 s) provided between blocks.
Response time and accuracy measures (% correct) were obtained for
each trial.
2.3. Exercise protocol
All participants engaged in group exercise classes organized
under the auspice of American Parkinson's Disease Association
(APDA). The activities were consistent at all ve centers; they
included aerobic activities such as walking, running, stationary
bike, water aerobics; resistance exercises with bands and dumb-
bells, and movement activities such as Zumba and Tai Chi. Training
was conducted by instructors who were certied by APDA. The
exercise intervention was goal oriented and focused on aspects of
functional ability specic to people with PD. The weekly exercise
frequency, the types of exercises, the duration and perceived in-
tensity, as well as adverse events were documented through indi-
vidual daily exercise logs reported by each participant in the study.
The intensity and type of the exercise intervention was matched for
all the participants in the study, and the frequency was varied by
group: high-frequency group exercised 4e5 times/week, and low
frequency 3 times/week. A paired sample t-test revealed that the
high-frequency group (M 2.3, SD 0.92) and the low-frequency
exercise group (M 2.15, SD 0.74) were not signicantly different
in exercise participation prior to the intervention, t (19) 0.64,
p 0.527. The perceived intensity of exercise, reported by partici-
pants in the study, was on average moderate-high, with no signif-
icant differences between the two groups.
 M.C. Caciula et al. / Mental Health and Physical Activity 10 (2016) 18e24
2.4. Statistical analyses
Separate switch-cost scores were calculated for tests composed
for the numbers and letter as well as the mixed condition. Global
switch cost scores were calculated using the procedure recom-
mended by Wasylyshyn, Verhaeghen, & Sliwinski (2011). The
average RT of the two pure, within-category conditions (number
and letter discrimination) was subtracted from the average RT of
switch and non-switch trials in the mixed, between-category
condition. The frequency of the correct responses made during
the mixed, between-category tests of 220 trials were converted to
percentage of total responses. The global switch cost scores and
accuracy scores are presented in Table 3. For both the global switch
cost scores and percentage accuracy scores an initial 2 (exercise
frequency group: high-frequency, low-frequency) x 2 (time: pre-
exercise, post-exercise) mixed measures factorial ANOVA, with
time as the repeated factor was conducted to assess the effects of
exercise frequency and training.
Means and standard errors of RTs for correct responses were
computed for each N-back condition. RT greater or less than three
standard deviations, calculated per participant, per condition were
excluded from further analyses. Excluded trials accounted for only
2.4% of the total number of trials. N-back accuracy was calculated
with the following algorithm: [1  (number of
commissions number of omissions)/(total possible
correct)]  100. Mixed measures factorial ANOVA were used to
examine group, time, and load differences on these measures.
Analyses were conducted using SPSS 22 software (SPSS IBM).
The p < 0.05 rejection level was used in all analyses. Where
appropriate, signicant effects were decomposed through post-hoc
t-tests. Levene's test indicated that the error variance of the
dependent variable is equal across groups F (1, 41) 0.961, p > 0.05.
3. Results
All participants included in the study completed the 12 weeks
exercise intervention at the self-selected frequency. We assessed
habitual physical activity through questionnaires at the beginning
of the study, and through daily logs during the intervention. Most of
the participants were sedentary prior to the initiation of the study
and changes in physical activity after the intervention were sig-
nicant. There were no adverse events among the participants in
the study. Adverse events were dened as undesirable outcomes
that may be directly or indirectly related to the intervention, such
as muscle strain or joint pain.
3.1. Auditory switch task
3.1.1. Global switch cost
Means and standard errors for the auditory switch task are
Table 2
Auditory Switch Task performance by group and time.
Condition High - frequency Low - frequency
Pre-exercise
Switch-cost 139.48 (10.8) 140.15 (11.6)
Accuracy (% correct) 93.86 94.01
Post-exercise
Switch-cost 110.81 (10.6) 143.72 (11.3)
Accuracy (% correct) 94.93 94.33
Notes: Values are expressed as mean (SE). Global switch cost scores are in milli-
seconds, and accuracy scores are percentage correct. Values in bold indicate a sig-
nicant group difference (p < 0.05). Switch-cost scores showed a signicant
interaction between time and exercise frequency. Accuracy data showed a main
effect of time on both exercise frequencies.
21
presented (Table 2). A mixed factorial ANOVA of the switch cost
scores yielded a signicant interaction between time and exercise
frequency, F (1, 41) 5.53, p > 0.05, h2p 0:09: Post-hoc t-tests
indicated that the two exercise frequency groups were not signi-
cantly different prior to the exercise intervention, t 0.04,
p 0.96, but their performance was signicantly different after the
12 weeks of training, t 2.11, p < 0.05 with the high frequency
group demonstrating a better performance on global switch cost.
These results indicate a better performance in the measure of
cognitive exibility for the high e frequency exercise group after
the intervention.
A mixed model factorial ANOVA revealed a signicant main
effect of time on the response accuracy, F (1, 41) 5.08, p < 0.05,
h2p 0:11, indicating that regardless of the frequency, exercise
might benet shifting accuracy in individuals with PD. The group
by time interaction was non-signicant F (1, 41) 1.64, p 0.20,
h2p 0:39:
3.2. N-back task
3.2.1. Response time (RT)
Means and standard errors for the N-back are presented in
Table 3. Analysis of RTs revealed a signicant interaction between
time and group, F (1, 41) 14.96, p < 0.001, h2p 0:26; indicating a
better performance in working memory response time for the high
e frequency exercise group after the intervention (Table 3). Post-
hoc t-tests revealed that regardless of the WM load there was no
signicant difference between the two groups prior to exercise, but
their performance was signicantly different in the 2-back condi-
tion, t 2.61, p < 0.05, and 3-back condition, t 2.84, p < 0.05
after the exercise intervention, with the high-frequency group
performing better on working memory tasks than the low-
frequency group. Also, analysis revealed a signicant main effect
of load, F (3, 123) 468.94, p < 0.001, h2p 0:92. Post-hoc t-tests
indicated that RTs for each load differed signicantly from RTs for
all other loads, both prior to exercise and at the end of the
intervention.
3.2.2. Response accuracy (% correct)
Analysis of response accuracy revealed a signicant main effect
of time, F (1, 41) 17.37, p < 0.001, h2p 0:29 suggesting that both
groups performed better in working memory accuracy after the
exercise intervention. Also, analyses revealed a main effect of load,
F (3, 123) 613.96, p < 0.001, h2p 0:93. Post-hoc t-tests indicated
that accuracy for each load differed signicantly from accuracy for
all other loads, both prior to exercise and at the end of the inter-
vention (Table 3).
4. Discussion
The intent of our study was to determine the effects of exercise
frequencies on select components of executive function in in-
dividuals with PD, after 12 weeks of multimodal exercise training. A
paired sample t-test revealed that the high-frequency group
(M 2.3, SD 0.92) and the low-frequency exercise group
(M 2.15, SD 0.74) were not signicantly different in exercise
participation prior to the intervention, t (19) 0.64, p 0.527. It
was apparent that exercise benetted executive function in all in-
dividuals with PD. More importantly the higher frequency exercise
group attained greater benets on cognitive exibility and working
memory than those individuals in the lower frequency exercise
group.
Cognitive exibility was assessed via an auditory switch task
and working memory performances via an auditory N-back task,
which were administered before and after the exercise
22 M.C. Caciula et al. / Mental Health and Physical Activity 10 (2016) 18e24
Table 3
N-back performance by group and time.
Group
High-frequency
Pre-exercise
0-back
Response time 692.02 (23.1)
Accuracy 92.78
1-back
Response time 789.06 (28.3)
Accuracy 83.08
2-back
Response time 997.09 (38.8)
Accuracy 74.04
3-back
Response time 1164.16 (40.8)
Accuracy 69.69
Notes: Values are expressed as mean (SE). Response times are in milliseconds and accuracy scores are percentage correct. Values in bold indicate a signicant group difference
(p < 05). Accuracy data showed a main effect of time. Response time data showed a signicant interaction between time and exercise frequency.
intervention. Analyses of the global switch costs data revealed that
higher frequency of exercise are more benecial for cognitive
exibility for individuals with PD than lower frequency of exercise.
The switch cost index reects the changes in the attention-
demanding mental processes required of participants to abandon
one response set and to recongure a different response set
(Monsell & Driver, 2000; Rogers & Monsell, 1995). Also, exercise,
regardless of its frequency, had a positive impact on response ac-
curacy. The lack of a differential effect of exercise frequency on
accuracy could be limited by the ceiling effect, with most partici-
pants' response accuracy between 90% and 100%. However, a
speedeaccuracy tradeoff between the switch cost scores and
response accuracy that often occurs when individuals attempt to
maintain or reduce RT but at the cost of increasing errors was not
observed, which provides support for exercise interventions on the
shifting aspects of executive function in individuals with PD.
Exercise also benetted participants' working memory pro-
cesses. For example, the N-back test, participants who exercised
more frequently evidence more rapid response times than those
who exercised less frequently. The lack of a speed-accuracy tradeoff
for the N-back task provides verication of benets of multimodal
exercise on selective aspects of executive functions. Our results are
similar to those of Tanaka et al. (2009), Ahlskog (2011), and Cruise
et al. (2011). They suggest that physical exercise benets executive
function in older individuals with PD, and that vigorous exercise
and physical tness may have a neuroprotective effect in PD
(Ahlskog, 2011; Cruise et al., 2011; Tanaka et al., 2009). The im-
provements in cognitive function as a result of exercise may be due
to enhanced neuroplasticity, exercise-related protection from
dopaminergic neurotoxins, or increased neurotrophic factor
expression (Hillman, Erickson, & Kramer, 2008). Benjamin & Sibley
(2007) suggest that people with generally lower cognitive perfor-
mances tend to benet more from exercise interventions, which
could be the case of our participants since their MoCA screening
scores were on the lower end of non-demented status. Similarly,
our ndings are in line with the work by Kramer et al. (1999), who
propose that exercise benets executive function during the aging
process.
Previous clinical studies have shown that various types of ex-
ercise, and more specically, exercise that incorporates goal-based
training and aerobic activity have the potential to improve both
cognitive and automatic components of motor control in in-
dividuals with mild to moderate disease through experience-
dependent neuroplasticity (Ahlskog, 2011; Petzinger et al., 2013;
Tanaka et al, 2009). In our study, all participants were exposed to
group exercise classes specically designed for people with PD,
Low-frequency
Post-exercise Pre-exercise Post-exercise
649.02 (20.04) 700.54 (24.8) 694.62 (21.4)
94.56 92.05 93.1
741.43 (24.95) 821.24 (30.41) 806.09 (26.7)
85.04 83.8 84.05
954.13 (34.05) 1028.73 (41.6) 1022.67 (36.5)
76.17 74.25 75.15
1083.46 (35.6) 1177.63 (43.8) 1165.1 (38.2)
71.52 69.05 70.1
thus the exercise training was goal-based and helped the partici-
pants to learn through instructions, feedback (reinforcement), and
encouragement to perform beyond self-perceived capability.
Through practice and learning of new movements and skills, in-
dividuals with PD are thought to become more cognitively engaged,
providing another potential explanation for the changes observed
in selective aspects of executive function following multimodal
exercise (Petzinger et al., 2013). The observed dose-response rela-
tion between exercise frequency and enhanced cognitive function
can be explained in terms of neuro-physiological adaptations,
enhancement of psychological self-regulation skills, or their com-
bination. It will be important for researchers to determine whether
the type of exercise intervention differentially inuences cognitive
outcomes. If it is the case that executive functions improve due
solely to changes in physical tness, the maintenance of the
cognitive benets derived from exercise will require individuals to
maintain their exercise regimens indenitely. However, should
changes in cognitive performance be due to the acquisition of self-
regulation and response strategies, levels of physical tness might
decline, but unless they were associated with a disease state it
would be difcult to see how reductions in physical functions
would compromise mental functioning. Clearly, identifying the
mechanisms that underlie changes in cognitive functions will be
important for establishing clinical interventions for individuals
with PD. The theory recently presented by Audiffren & Andre 
(2015) provides valuable heuristics for future research by reconsi-
dering the strength model of self-control. The new application of
the model explains the positive effects of chronic exercise on ex-
ecutive function through strengthening the self-control capacity.
Thus, exercise provides not only a physiological stimulation that
leads to changes in cardiovascular tness and muscle strength, but
also a psychological stimulation that leads to improvements of
executive function and strengthens self-control (Audiffren & Andre ,
2015).
Although we have demonstrated the benecial aspects of
cognitive function in PD, this study was limited to a highly educated
sample of participants, which may not represent the typical PD
patient. In addition, all participants were tested during on stages
of dopaminergic medication, which may alter task performance
(Costa et al., 2003). The lack of a non-exercise control group is a
limitation in this study. The eld-based nature of the exercise
intervention challenged attempts to recruit PD individuals who
would agree to be assigned to a condition that postponed partici-
pation in a health-promoting activity. Considerable research
demonstrated benets of chronic exercise on mental functioning,
particularly for older adults (Prakash et al., 2015). Less information
 M.C. Caciula et al. / Mental Health and Physical Activity 10 (2016) 18e24
is available that addresses exercise frequency and its relation to
gains in mental processing. The results of the present study provide
preliminary evidence for the importance of identifying the exercise
dose that confers optimal cognitive benets.
Also, the results obtained in the present study argue for the need
for randomized trials that can evaluate the role of such potential
moderators as the social environment experienced by participants.
Sauleau, Eusebio, Vandenberghe, Nuttin, and Brown (2009) sug-
gested that long-term usage of dopaminergic medications can
cause a decrease in motivation to participate in general activities.
Since the participants in this study chose their weekly exercise
frequency, motivation to participate in these activities might be a
variable inuencing the study ndings.
Noteworthy aspects of the present study suggest that multi-
modal exercises performed consistently, at least four times/week at
moderate to high intensities can benet selective aspects of exec-
utive function, such as WM and set shifting, in individuals with PD
without dementia. Because of the availability of services in a large
metropolitan area, individuals were able to access programs
designed specically for PD. Although these services may not be
available in less populated areas, the value of exercise in treatment
of PD should become a primary factor in the treatment of PD.
Additional intervention opportunities such as home e based pro-
grams that were used by Nocera, Horvat, & Ray (2009) should be
encouraged to supplement our ndings, and to help further identify
the optimal delivery content of exercise for PD.
Financial disclosure
To the best of our knowledge, no nancial conict of interest
exists and there is no outside funds that have been given that effect
this submission.
Acknowledgement
I gratefully acknowledge the support of Dr. Patricia Creel from PT
Solutions Atlanta, and the generosity of Dr. David E. Jones without
which the present study could not have been completed.
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.mhpa.2016.04.001.
References
Agid, Y. (1991). Parkinson's disease: pathophysiology. Lancet Neurology, 337,
1321e1323.
Ahlskog, J. (2011). Does vigorous exercise have a neuroprotective effect in Parkinson
disease? Neurology, 77, 288e294.
Audiffren, M., & Andre , N. (2015). The strength model of self-control revisited:
linking acute and chronic effects of exercise on executive functions. Journal of
Sport and Health Science, 4(1), 30e46.
Beato, R., Levy, R., Pillon, B., Vidal, C., Tezenas du Montcel, S., Deweer, B., et al.
(2008). Working memory in Parkinson's disease patients: clinical features and
response to levodopa. Arquivos de Neuro-psiquiatria, 66(2A), 147e151.
Benjamin, A., & Sibley, S. L. B. (2007). Exercise and working memory: an individual
differences investigation. Journal of Sport & Exercise Psychology, 29, 783e791.
Cedrus. SuperLab., ed., Editor., Cedrus Corporation: San Pedro (CA).
Cepeda, N. J., Kramer, A. F., & Gonzalez de Sather, J. C. (2001). Changes in executive
control across the life span: examination of task-switching performance.
Developmental Psychology, 37(5), 715e730.
Chodzko-Zajko, W. J., Proctor, D. N., Fiatarone-Singh, M. A., Minson, C. T., Nigg, C. R.,
Salem, G. J., et al. (2009). American College of Sports Medicine position stand.
Exercise and physical activity for older adults. Medicine and Science in Sports and
Exercise, 41(7), 1510e1530.
Colosimo, C., & De Michele, M. (1999). Motor uctuations in Parkinson's disease:
pathophysiology and treatment. European Journal of Neurology, 6(1), 1e21.
Costa, A., Peppe, A., Dell'Agnello, G., Carlesimo, G. A., Murri, L., Bonuccelli, U., et al.
(2003). Dopaminergic modulation of visual-spatial working memory in
23
Parkinson's disease. Dementia and Geriatric Cognitive Disorders, 15(2), 55e66.
Cruise, K. E., Bucks, R. S., Loftus, A. M., Newton, R. U., Pegoraro, R., & Thomas, M. G.
(2011). Exercise and Parkinson's: benets for cognition and quality of life. Acta
Neurologica Scandinavica, 123, 13e19.
Diamond, A. (2013). Executive functions. Annual Review of Psychology, 64(1),
135e168.
Fisher, B. E., Petzinger, G. M., Nixon, K., Hogg, E., Bremmer, S., Meshul, C. K., et al.
(2004). Exercise-induced behavioral recovery and neuroplasticity in the 1-
methyl-4-phenyl-1,2,3,6-tetrahydropyridine-lesioned mouse basal ganglia.
Journal of Neuroscience Research, 77(3), 378e390.
Hillman, C. H., Erickson, K. I., & Kramer, A. F. (2008). Be smart, exercise your heart:
exercise effects on brain and cognition. Nature Reviews Neuroscience, 9(1),
58e65.
Horstink, M. W., Berger, H. J., van Spaendonck, K. P., van den Bercken, J. H., &
Cools, A. R. (1990). Bimanual simultaneous motor performance and impaired
ability to shift attention in Parkinson's disease. Journal of Neurology, Neurosur-
gery & Psychiatry, 53(8), 685e690.
Koerts, J., Van Beilen, M., Tucha, O., Leenders, K. L., & Brouwer, W. H. (2011). Ex-
ecutive functioning in daily life in Parkinson's disease: initiative, planning and
multi-task performance. PLoS One, 6(12), e29254.
Kramer, A. F., Hahn, S., Cohen, N. J., Banich, M. T., McAuley, E., Harrison, C. R., et al.
(1999). Ageing, tness and neurocognitive function. Nature, 400, 418e419.
Larson, E. B., Wang, L., Bowen, J. D., McCormick, W. C., Teri, L., Crane, P., et al. (2006).
Exercise is associated with reduced risk for incident dementia among persons
65 years of age and older. Annals of Internal Medicine, 144(2), 73e81.
Marder, K., Tang, M. X., Cote, L., Stern, Y., & Mayeux, R. (1995). The frequency and
associated risk factors for dementia in patients with Parkinson's disease. Ar-
chives of Neurology, 52(7), 695e701.
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 contributions to
complex frontal lobe tasks: a latent variable analysis. Cognitive Psychology,
41(1), 49e100.
Molloy, S. A., Rowan, E. N., O'Brien, J. T., McKeith, I. G., Wesnes, K., & Burn, D. J.
(2006). Effect of levodopa on cognitive function in Parkinson's disease with and
without dementia and dementia with Lewy bodies. Journal of Neurology,
Neurosurgery, and Psychiatry, 77(12), 1323e1328.
Monsell, S., & Driver, J. (2000). Control of cognitive processes: Attention and perfor-
mance XVIII. Cambridge (MA): The MIT Press.
Murray, D. K., Sacheli, M. A., Eng, J. J., & Stoessl, A. J. (2014). The effects of exercise on
cognition in Parkinson's disease: a systematic review. Translational Neuro-
degeneration, 3(1), 5.
Nasreddine, Z. S., Phillips, N. A., Be dirian, V., Charbonneau, S., Whitehead, V.,
Collin, I., et al. (2005). The montreal cognitive assessment, MoCA: a brief
screening tool for mild cognitive impairment. Journal of the American Geriatrics
Society, 53, 695e699.
Nocera, J. R., Altmann, L. J., Sapienza, C., Okun, M. S., & Hass, C. J. (2010). Can exercise
improve language and cognition in Parkinson's disease? A case report. Neuro-
case, 16(4), 301e306.
Nocera, J., Horvat, M., & Ray, C. T. (2009). Effects of home-based exercise on postural
control and sensory organization in individuals with Parkinson disease.
Parkinsonism & Related Disorders, 15(10), 742e745.
Okumura, M. S., Cooper, S. L., Ferrara, M. S., & Tomporowski, P. D. (2013). Global
switch cost as an index for concussion assessment: reliability and stability.
Medicine and Science in Sports and Exercise, 45(6), 1038e1042.
Organization, W. H. (2010). In W. H. Organization (Ed.), Global recommendations on
physical activity for health. Geneva.
Ouchi, Y. (2001). Changes in dopamine availability in the nigrostriatal and meso-
cortical dopaminergic systems by gait in Parkinson's disease. Brain, 124,
784e792.
Pagonabarraga, J., & Kulisevsky, J. (2012). Cognitive impairment and dementia in
Parkinson's disease. Neurobiology of Disease, 46(3), 590e596.
Perlstein, W. M., Dixit, N. K., Carter, C. S., Noll, D. C., & Cohen, J. D. (2003). Prefrontal
cortex dysfunction mediates decits in working memory and prepotent
responding in schizophrenia. Biological Psychiatry, 53(1), 25e38.
Petzinger, G. M., Fisher, B. E., McEwen, S., Beeler, J. A., Walsh, J. P., & Jakowec, M. W.
(2013). Exercise-enhanced neuroplasticity targeting motor and cognitive cir-
cuitry in Parkinson's disease. Lancet Neurology, 12(7), 716e726.
Prakash, R. S., Voss, M. W., Erickson, K. I., & Kramer, A. F. (2015). Physical activity and
cognitive vitality. Annual Review of Psychology, 66(1), 769e797.
Ridgel, A. L., Kim, C. H., Fickes, E. J., Muller, M. D., & Alberts, J. L. (2011). Changes in
executive function after acute bouts of passive cycling in Parkinson's disease.
Journal Aging and Physical Activity, 19, 87e98.
Rogers, D. R., & Monsell, S. (1995). Costs of a predictable switch between simple
tasks. Journal of Experimental Psychology: General, 124, 207e231.
Rustamov, N., Rodriguez-Raecke, R., Timm, L., Agrawal, D., Dressler, D., Schrader, C.,
et al. (2014). Attention shifting in Parkinson's disease: an analysis of behavioral
and cortical responses. Neuropsychology, 28(6), 929e944.
Sauleau, P., Eusebio, A., Vandenberghe, W., Nuttin, B., & Brown, P. (2009). Deep brain
stimulation modulates effects of motivation in Parkinson's disease. NeuroReport,
20(6), 622e626.
SPSS IBM, New York, U. S. A.
Stern, Y., Marder, K., Tang, M. X., & Mayeux, R. (1993). Antecedent clinical features
associated with dementia in Parkinson's disease. Neurology, 43(9), 1690e1692.
Tanaka, K., Quadros, A. C., Jr, Santos, R. F., Stella, F., Gobbi, L. T., & Gobbi, S. (2009).
Benets of physical exercise on executive functions in older people with
24 M.C. Caciula et al. / Mental Health and Physical Activity 10 (2016) 18e24

Parkinson's disease. Brain and Cognition, 69, 435e441.
Tomporowski, P. D., Davis, C. L., Miller, P. H., & Naglieri, J. A. (2015). Exercise and
children's cognition: the role of exercise characteristics and a place for meta-
cognition. Journal of Sport and Health Science, 4(1), 47e55.
Wasylyshyn, C., Verhaeghen, P., & Sliwinski, M. J. (2011). Aging and task switching: a
meta-analysis. Psychology and Aging, 26(1), 15e20.
Woods, S. P., & Troster, A. I. (2003). Prodromal frontal/executive dysfunction pre-
dicts incident dementia in Parkinson's disease. Journal of the International
Neuropsychological Society, 9(1), 17e24.