Preventive Medicine Reports 23 (2021) 101490

Contents lists available at ScienceDirect

Preventive Medicine Reports

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

Mindful walking and cognition in older adults: A proof of concept study

using in-lab and ambulatory cognitive measures
Chih-Hsiang Yang a, *, Jonathan G. Hakun b, c, Nelson Roque d, Martin J. Sliwinski d, e,
David E. Conroy f
a
Department of Exercise Science, University of South Carolina, USA
b
Departments of Neurology, The Pennsylvania State University, USA
c
Departments of Psychology, The Pennsylvania State University, USA
d
Center for Healthy Aging, The Pennsylvania State University, USA
e
Department of Human Development and Family Studies, The Pennsylvania State University, USA
f
Department of Kinesiology, The Pennsylvania State University, USA

A R T I C L E I N F O A B S T R A C T

Keywords: Mindfulness practice and walking have been linked individually to sustain cognition in older adults. This early-
Walking meditation phase study aimed to establish proof-of-concept by evaluating whether an intervention that integrates light-
Executive function intensity walking with mindfulness practices shows promising signs of improving cognition in older adults.
Lifestyle activity
Participants (N = 25, Mage = 72.4 ± 6.45) were community-dwelling older adults who engaged in a supervised
Ambulatory cognitive assessment
Healthy aging
mindful walking program over one month (8 sessions total, 2 sessions per week, 30-minute slow walking con­
Light-intensity physical activity taining mindfulness skills). They completed performance-based and subjective ratings of cognitive measures in
field before and after two mindful walking bouts using a smartphone app. They also completed in-lab perfor­
mance-based and self-report cognitive measures at baseline and after the entire program. Controlling for de­
mographics, potential covariates, and time trends, short-term improvements in perceived cognition and
processing speed were observed from pre- to post-mindful walking sessions (i.e., 30 min) across multiple
ambulatory cognitive measures (Cohen’s ds range = 0.46–0.66). Longer-term improvements in processing speed
and executive function were observed between baseline and end of the program (i.e., one month) across various
performance-based cognitive measures (ds range = 0.43–1.28). No significant changes were observed for other
cognitive domains. This early-phase study (Phase IIa) provides preliminary support that mindful walking activity
is promising for sustaining cognition in older adults. Our promising findings form the building blocks of evidence
needed to advance this intervention to a fully powered randomized controlled trial that examines program ef­
ficacy with a comparator. Favorable outcomes will inform the development of this lifestyle behavioral strategy
for promoting healthy brain aging in late adulthood.

1. Introduction adulthood. Two promising strategies for these purposes include physical
activity and mindfulness practices (Erickson et al., 2019; Gard et al.,
Human aging is associated with normative alterations in cognition 2014 Jan; Malinowski and Shalamanova, 2017; Sofi et al., 2011). It is
and increased risks for neurodegenerative disease in late life. These viable to integrate mindfulness practice with walking as an intervention
diseases are the most expensive US annual health expenditure and place strategy (i.e., mindful walking) (Kabat-Zinn, 2017). This “active form”
a tremendous economic burden on society and families (Alzheimer’s of mindfulness practice has been implemented as part of the standard
Association, 2020). Cognitive impairments caused by these diseases also mindfulness-based programs (i.e., Mindfulness-Based Stress Reduction
exact a toll on the overall health, well-being, and quality of life among program) to enhance psychological well-being (Gotink et al., 2016; Teut
older adults. Preventive interventions are needed to help older adults et al., 2013). However, mindful walking has not been used as a major
reduce risks for these diseases and preserve functioning into late strategy to study cognitive outcomes.

* Corresponding author at: Department of Exercise Science, Technology Center to Promote Healthy Lifestyles, Arnold School of Public Health, University of South
Carolina, 915 Greene St, Columbia, SC 29201, USA.
E-mail address: cy11@mailbox.sc.edu (C.-H. Yang).

https://doi.org/10.1016/j.pmedr.2021.101490
Received 6 March 2021; Received in revised form 5 July 2021; Accepted 10 July 2021
Available online 14 July 2021
2211-3355/© 2021 Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
C.-H. Yang et al. Preventive Medicine Reports 23 (2021) 101490

Walking and mindfulness programs may individually contribute to could read and spoke English fluently, and without allergy to plants and
short- (e.g., after brief practice) and longer-term (e.g., after completing flowers. A subset of older adults (N = 25, Mean age = 72.4 ± 6.45, age
the entire program) cognitive improvements in older adults, albeit range = 66–89, 84% female, 84% White) opted to participate in the
variability of intervention design exists among available studies cognitive assessments designed in this proof-of-concept study. A more
(Scherder et al., 2014; Venturelli et al., 2011; Berk et al., 2017; Chiesa detailed description of recruitment with a CONSORT diagram is re­
et al., 2011). It is plausible that an integrated mindful walking program ported previously (Yang and Conroy, 2019). The primary goal of the
may likewise be associated with both short- and longer-term cognitive proof-of-concept study is to “determine if a treatment package can
benefits in older adults. This early phase proof-of-concept study (Phase achieve benefit on a clinically significant target in a small, select sam­
IIa) evaluated whether a multi-session mindful walking program pro­ ple” (Czajkowski et al., 2015). In this context, “within-subjects designs
vided signals consistent with short- and longer-term cognitive im­ where subjects act as their own controls in a pre-post comparison are
provements in older adults (Czajkowski et al., 2015 Oct). ideal…[and] the sample can be selected from acceptable subjects, rather
Walking is the most prevalent type of physical activity among older than representative, because this initial test will determine only whether
adults and the most preferred physical activity among cognitively- the treatment merits more rigorous testing” (p. 977). The present sample
impaired older adults (Williams et al., 2008; Dai et al., 2015). Current size is comparable to proof-of-concept studies evaluating the potential
evidence indicates that accruing physical activity at a lower and more benefits of behavioral and technology-based health interventions
achievable intensity (i.e., walking) improves cognitive health in both (Conroy and Heartphone, 2020; Liu-Ambrose and Eng, 2015; Månsson
active and inactive older adults, as well as older adults with cognitive et al., 2013; Conroy et al., 2020).
impairments (Prohaska et al., 2009; Spartano et al., 2019; Wang et al., At baseline, participants were not sufficiently active (based on the
2012). Mindfulness practice trains individuals to elevate their attention Physical Activity Guidelines for Americans) and 16% (n = 4) of them
and awareness in every present moment, and engage their present were overweight/obese. They reported no formal mindfulness training
experience in a non-judgmental manner (Kabat-Zinn, 1994; Kabat-Zinn, experiences, cognitive/memory complaints, or diagnosis of neuropsy­
2012). Practicing mindfulness is appealing to older adults, as evidenced chological diseases. The majority of the walking sessions were
by the high compliance rates, and initial evidence suggesting that daily completed on weekdays (n = 160, 80%) and before noon (n = 152, 76%)
mindfulness practice may improve their cognitive health (Gard et al., in October and December. Participants who completed all walking ses­
2014 Jan; Wong et al., 2017). Previous work suggests that short bouts of sions and assessments were eligible to win one of nine $25 gift cards in a
mindful-walking sessions are feasible to implement in older adults living raffle. All procedures followed were in accordance with the ethical
in the community (Yang and Conroy, 2019). No study to date has standards of the responsible committee on human subject research and
evaluated whether short bouts of mindful-walking practices produce with the Helsinki Declaration. Written informed consent was obtained
cognitive benefits in older adults. from all participants. The Institutional Review Board approved all study
This study applied both performance-based and subjective ratings of protocols.
cognition both in the lab and at the walking site to broadly assess
cognition in response to the mindful walking activity. To assess longer-
term outcomes, conventional methods including performance-based 2.2. Procedure
neuropsychological assessments, computerized experimental assess­
ments, and questionnaires were applied at baseline and at post mindful Participants first completed an initial lab visit and completed base­
walking program across one month. These lab-based measures are line cognitive assessments. These assessments included a computerized
relatively time-consuming, and they are not suitable for administration Stroop task, two sets of neuropsychological tests using paper–pencil
in the field (e.g., outdoor environments) to capture any acute changes format (see “Measures” below), and a survey of perceived cognition.
experienced following walking activity (Ladouce et al., 2017). To assess Participants then scheduled eight sessions of outdoor mindful walking
short-term outcomes, this study applied recently validated, ultra-brief, within the following month, with a maximum of scheduling two sessions
ambulatory cognitive assessments on smartphones to evaluate short- per week. After completing the mindful walking sessions, participants
term subjective and performance-based cognitive changes associated returned to the lab for post-program cognitive assessments that were
with brief 30-min mindful walking bouts (Sliwinski et al., 2018). identical to the formats used in the initial lab visit.
The purpose of this study was to establish proof-of-concept for using Each walking session consisted of a 30-minute individual slow
this mindful walking program to improve short- and longer-term walking along a flat, designated route in an arboretum. Participants
cognition among community-dwelling older adults. Proof-of-concept were instructed to walk at a slower pace of approximately one step per
study represents an early phase of intervention development in the second (i.e., light-intensity activity). Walking at a slower speed helped
Obesity-Related Behavioral Intervention Trials (ORBIT) framework. We participants elicit the state of mindfulness and elevate their awareness to
defined the meaningful change in cognitive outcomes as 0.20 standard the present moment experiences (Kabat-Zinn, 2017). The research staff
deviation, which is equivalent to 10 years of normative cognitive aging met with participants on the walking site to provide instructions on
documented in previous reviews (Salthouse, 1996 Jul; Salthouse, 2000). mindfulness skills and conducted pre-and post- walking assessments.
Available reviews on physical activity interventions on cognition also Three fundamental mindfulness skills were introduced and incorporated
reported small-to-moderate effect sizes (range = 0.20–0.48) among progressively in sequence starting from the second session to help par­
cognitively normal older adults (Erickson et al., 2019; Mj, et al., 2016). ticipants build up mindful walking skills. These fundamental skills
Evidence from a proof-of-concept study is not sufficient to draw con­ involve being attentive to the rhythm of their breathing (i.e., each inhale
clusions about efficacy, but it is essential for determining whether this and exhale), being attentive to the movement of their every step, and
intervention warrants investment in a rigorous trial to evaluate effects in mentally scanning the body to identify and accept sensations/feelings
relation to a comparator (Czajkowski et al., 2015 Oct; Freedland, 2020). that arise in every present moment (Kabat-Zinn, 1994). In the last two
mindful walking sessions (7th and 8th), participants practiced all three
2. Methods mindfulness skills in sequence throughout their 30-minute walk.
Immediately before and after the 7th and 8th sessions, participants
2.1. Participants completed subjective ratings of cognition and a short battery of
smartphone-based ambulatory cognitive assessments (see “Measures”
Participants were community-dwelling older adults who participated below). Participants overall reported increased state mindfulness (p <
in an 8-session slow walking program at a local arboretum. Eligible older .001, d = 0.84) using items from the State Mindfulness Scale across all
adults were at least 65 years old, could walk without other’s assistance, mindful walking sessions (Tanay and Bernstein, 2013).

2
C.-H. Yang et al. Preventive Medicine Reports 23 (2021) 101490

2.3. Measures 2.3.7. Perceived sleep quality (in field)

One item was used to assess participants’ overall sleep quality in the
2.3.1. Demographics (lab-based) previous night at the beginning of each walking session. Participants
Participants’ basic demographic variables, including gender, age, answer one question, “What was your overall quality of sleep last
race/ethnicity, socioeconomic status, and educational level, were night?” on a 7-point scale ranging from 1 (very bad) to 7 (very good). This
collected by self-report surveys during the initial lab visit. item was included to account for the impact of the previous day sleep
quality on cognition on the next day (Nebes et al., 2009).
2.3.2. Neuropsychological tests (lab-based)
The paper-and-pencil format of Trail Making Tests (forms A and B) 2.4. Data analysis
and Porteus Maze Tests (forms Adult-I and Adult-II) were used to assess
older adults’ various domains of cognition during the two lab visits For lab-based cognitive tests, paired t-tests and within-subjects effect
before and after the entire program (Reitan, 1986; Porteus and Peters, sizes were used to examine the preliminary magnitude of change be­
1947). The outcome variables in these in-lab tests include the task tween baseline and post-program measures. The standard 2 × 2 repeated
completion time and the number of errors. Each participant followed the measures ANOVA was used for Stroop Task to test the main effects and
instruction by trained staff to complete the tests individually during the the occasion by condition interaction (Bugg et al., 2008). For ambula­
two lab visits. tory cognitive assessments, the mixed-effects linear models were used to
test within-person differences between their paired pre- and post-walk
2.3.3. Perceived cognition (lab-based) scores from the 7th and the 8th walking sessions. The four cognitive
Four subscales were selected and slightly modified from the measures were coded (0 = pre-walks, 1 = post-walks) to test whether
Everyday Cognition Scale to assess subjective ratings of cognition: post-walking cognitive scores significantly differed from pre-walk scores
everyday memory (8 items), everyday planning (5 items), everyday after controlling for covariates. These models adjusted for demographics
organization (6 items), and everyday divided attention (4 items) (Farias (age, sex) and time-varying temporal and contextual factors that may
et al., 2008). Participants reported each question on a 6-point Likert impact the outcomes. Temporal factors included day of the week, time of
scale ranging from 1 (almost always) to 6 (almost never). The average day (to adjust for diurnal influences), and number of walking session to
score in each subscale was calculated to represent the general level in account for any session-to-session trends associated with retest im­
the specific cognitive domain of everyday life, with a higher score provements that account for main sources of practice effect. Contextual
indicating better cognition (Marshall et al., 2014). factors included previous night sleep quality and mean daytime tem­
perature. Separate models were tested for each outcome variable.
2.3.4. Stroop task (lab-based) Cohen’s d was calculated using the Satterthwaite approximations to
Procedures for the computerized Stroop task are described in greater calculate the degrees of freedom to estimate the effect sizes fixed effects
detail elsewhere (Kim et al., 2014). In brief, participants were instructed in each model (Valliant and Rust, 2010).
to, as quickly and as accurately as possible, select the response option
from the bottom of the screen (color words written in white font) that 3. Results
matched the font color of the target stimulus presented centrally on a
black background. The meaning and font color of the target stimulus All participants completed the baseline and the post-program in-lab
either matched (“congruent”) or mismatched (“incongruent”) with a cognitive assessments.
50% probability across all trials. During congruent trials, the incorrect
response option was selected randomly from the remaining five color 3.1. Longer-term cognitive change (in-lab)
word options. During incongruent trials, the incorrect response matched
the orthography of the target stimulus. Participants completed 80 total Table 1 summarizes descriptive statistics and the paired t-test results
trials (40 trials per condition: congruent/incongruent). Primary out­ for the laboratory-based cognitive assessments. At baseline, relatively
comes for the Stroop task included mean accuracy and response time high levels of everyday cognition on all four domains of the Everyday
during each condition. Cognition Scale, including Memory, Planning, Organization, and
Divided Attention (mean scores ≥ 4.83 on a 1–6 scale). There were no
2.3.5. Ambulatory cognitive tests (in field) differences in scores on any domain of the Everyday Cognition Scale
Three ultra-brief ambulatory cognitive tasks described in detail in after exposure to the mindful walking program.
Sliwinski et al. (2018) were used to assess processing speed, working Results of paired t-test revealed that mean completion times on the
memory, and executive function: Symbol Search, Dot Memory, and N- Trail Making Test were faster post-program compared to baseline (forms
Back. These tasks were administered using a custom java-based mobile A and B, ps < 0.05, ds = 0.44 and 0.43). No change in difference scores
application loaded onto Samsung Galaxy S5 Android smartphones. for forms B-A was observed. Similarly, the completion time for both test
These three cognitive tasks were performed during the 7th and 8th forms on the Porteus Maze Test was faster post-program compared to
walking sessions where participants carried out mindful walking skills baseline (ps < 0.05, ds = 0.49 and 0.45). No changes in error rate on the
throughout the 30-minute walk, with two pre-walk tests and two post- Trail Making Test and Porteus Maze Test were observed between
walk tests. Outcome variables included mean response time (Symbol occasions.
Search/N-Back/Dot Memory) and mean accuracy (mean of trial-level Results of a 2 (occasion: baseline/post-program) × 2 (condition:
binary correct/incorrect for Symbol Search/N-Back; mean distance of congruent/incongruent) repeated measures ANOVA on Stroop Task
dot locations between actual and recall arrays for Dot Memory). response times revealed significant main effects of measurement occa­
sion (F(1,23) = 9.65, MSE = 14521.95, p < .01) and condition (F(1,23) =
2.3.6. Momentary rating of cognition (in field) 101.11, MSE = 6365.74, p < .001). No significant condition × occasion
One item adapted from the PROMIS Applied Cognitive Abilities interaction was observed (p = .88). Post-hoc analyses revealed that
Short Form (v1.0) was used to assess perceived cognition immediately overall response times were faster post-program compared with baseline
before and after the walking session (Fries et al., 2005). Participants (d = 1.28, p < .01) and during congruent trials compared with incon­
responded to the question - “My mind is sharper than usual now” - on a 1 gruent trials (d = 3.88, p < .001).
(Strongly Disagree) to 7 scale (Strongly Agree). Results of a 2 × 2 repeated measures ANOVA on Stroop Task accu­
racy rate revealed significant main effects of measurement occasion
(F(1,23) = 10.86, MSE = 0.11, p < .01) and condition (F(1,23) = 46.16,

3
C.-H. Yang et al. Preventive Medicine Reports 23 (2021) 101490

Table 1
Descriptives of in-lab cognitive assessments and the within-group differences between baseline and post mindful walking program.
Variable Baseline mean(SD) Post-program mean(SD) Mean difference(SD) 95%CI of mean difference t Pre-post Correlation
a
Everyday Cognition Scale
Memory 4.83 (0.39) 4.80 (0.56) 0.03 (0.57) [− 0.20 , 0.25] 0.21 0.31
Planning 5.54 (0.44) 5.61 (0.47) − 0.07 (0.51) [− 0.27 , 0.13] − 0.69 0.38
Organization 4.92 (0.85) 5.05 (0.76) − 0.13 (0.55) [− 0.34 , 0.09] − 1.17 0.77***
Divided attention 4.96 (0.79) 4.83 (0.80) 0.13 (0.72) [− 0.15 , 0.41] 0.94 0.60**
Trail Making Testa
Trail A completion time (sec) 26.62 (7.50) 24.50 (7.10) 2.12 (4.90) [0.14 , 4.10] 2.20* 0.78***
Trail B completion time (sec) 59.11 (20.88) 51.08 (19.14) 8.03 (18.75) [0.46 , 15.60] 2.18* 0.56**
Trail B-A time difference (sec) 32.49 (18.45) 26.58 (16.31) 5.91 (19.80) [− 2.08 , 13.91] 1.52 0.36
Trail A Errors 0.26 (0.66) 0.37 (0.63) − 0.11 (0.80) [− 0.43 , 0.21] − 0.72 0.22
Trail B Errors 1.19 (1.62) 0.59 (0.89) 0.59 (1.80) [− 0.12 , 1.31] 1.71 0.06
Porteus Maze Testa
Maze I completion time (sec) 60.81 (34.96) 44.45 (28.72) 16.35 (33.65) [2.76 , 29.94] 2.48* 0.46*
Maze II completion time (sec) 103.72 (72.02) 72.49 (41.17) 31.23 (69.19) [2.67 , 59.79] 2.26* 0.35
Maze I Errors 2.85 (1.98) 2.04 (2.26) 0.82 (2.42) [− 0.14 , 1.78] 1.75 0.35
Maze II Errors 2.70 (1.88) 2.30 (1.44) 0.41 (2.12) [− 0.43 , 1.25] 1.00 0.21
Stroop Testb
Congruence reaction time (ms) 1120.26 (141.83) 1047.46 (125.23) − 72.80 (132.47) [− 127.48 , − 18.11] − 2.75* 0.51**
Incongruence reaction time (ms) 1282.78 (136.85) 1205.85 (139.95) − 76.93 (141.35) [− 135.27 , − 18.58] − 2.72* 0.48*
Congruence accuracy rate (%) 96.30 (7.94) 98.80 (3.16) 2.50 (7.97) [− 0.79 , 5.79] 1.57 0.19
Incongruence accuracy rate (%) 80.62 (17.10) 91.56 (6.79) 10.94 (13.29) [5.33 , 16.55] 4.03** 0.70***

Note: Number of participants = 25; a paper-and-pencil format; b

computer-based test; sec = second, ms = millisecond.
*p < .05, **p < .01, ***p < .001.

MSE = 0.32, p < .001). Post-hoc analyses revealed that overall accuracy 1994; Salthouse, 1996 Jul; Salthouse, 2000). Previous studies of mindful
rate was higher post-program compared with baseline (d = 1.37, p < walking have focused on mental health (Mj, et al., 2016; Peavy et al.,
.01) and during congruent trials compared with incongruent trials (d = 2012). This study extended the literature by modifying key domains of
2.78, p < .001). Additionally, there was a condition × occasion inter­ cognition in response to a multi-session mindful walking program for
action in predicting overall accuracy rate (F(1,23) = 22.57, MSE = 0.01, p older adults.
< .001). During incongruent trials, participants significantly increased The current study identified longer-term within-person improve­
overall accuracy rate post-program compared with baseline (d = 1.96, p ments in processing speed and executive function across paper–pencil
< .001). There was no difference in accuracy rate during congruent trials and computerized assessments. Performance improvements on the
between baseline and post-program measures. Stroop task appeared to be specifically associated with incongruent
condition accuracy. This finding may imply improvements in inhibitory
control, selective attention, and overall executive function (Scarpina,
3.2. Short-term cognitive change (in field) 2017). However, we caution that accuracy during the congruent con­
dition was overall very high at baseline, and thus, the observed inter­
Table 2 summarizes the mixed-effects model results for the perceived action effect may be driven by either the changes in executive function
and objective ambulatory cognitive assessments from the mobile (incongruent condition-only) or general task performance improve­
cognitive assessment protocol. Four participants had missing records in ments (in both conditions) that were masked by baseline ceiling per­
their 7th walking session due to malfunction identified in one of the formance in the congruent condition. A potential ceiling effect may also
study smartphones, resulting in a total of 92 measurement occasions. explain the no difference in subjective measures of everyday cognition.
Controlling for contextual and time-based factors (main sources of Participants in this study were not cognitively impaired; their ability to
practice effect), subjective ratings of cognition were better at post- carry out daily cognitive tasks should be similar before and after the
walking session compared to pre-walking session (d = 1.15, p < .001). program.
Further, participants’ response time was generally faster post- compared Mirroring the longer-term cognitive improvements, short-term im­
to pre-walking sessions across objective ambulatory cognitive assess­ provements in processing speed were observed during performance of a
ments. Significant faster post-walking response time was observed task with instructions that stressed speeded performance (Symbol
during two of the three ambulatory cognitive tasks, including Symbol Match) and another that stressed accuracy (N-Back), indicated that a
Search (d = 0.46, p < .05) and the N-Back task (d = 0.66, p < .01). The general impact on cognition may exist from practicing mindful walking.
mean reduction in response time observed in the Dot Memory task was It is possible that these short-term changes of mindful walking on pro­
not significant (p = .61). No significant changes in mean accuracy were cessing speed are the mechanisms by which longer-term advantages are
observed from pre- to post-walking sessions among cognitive tasks. conferred (e.g., improvements are immediate and incremental). Pro­
cessing speed is a central marker of neurocognitive function that
4. Discussion changes with age, and is altered significantly by the presence of
neurodegenerative disease (Salthouse, 2000; Finkel et al., 2007). Slower
Overall, the observed within-person changes of cognitive outcomes processing speed can have a widespread influence on other higher-order
in both short- and longer-term exceeded the meaningful benchmarks cognitive processes that unfold over time and require coordination of
that were given (i.e., d ≥ 0.20) for concluding that there was a favorable lower-level processes (e.g., working memory) (Kail, 2000). This proof-
signal on sustaining cognition from mindful walking. These results of-concept study controlled for potential practice effect, but an effi­
indicate that mindful walking warrants progression in the intervention cacy trial is needed to evaluate if mindful walking practice contributes to
development pipeline (Phase IIb/III) described in the ORBIT model. The improvements of processing speed (Duff et al., 2007). Further, a short-
benefits of both acute (from 30-min bout) and accumulated practice term improvement in subjective cognition from pre-to post-walking
(from multiple sessions) of mindful walking appear to be conferred to session was observed using a single self-report item. This single item is a
information processing speed, which holds implications for a wide range global measure of cognition that does not represent a specific cognitive
of cognitive processes affected by cognitive aging (Kail and Salthouse,

4
C.-H. Yang et al. Preventive Medicine Reports 23 (2021) 101490

domain. Thus, finding from this self-report may be different from those

Perceivedcognitivefunction
measured by Everyday Cognition Scale in which daily cognitive aspects
were targeted.
The mindfulness skills practiced in this study emphasized cultivating
heightened awareness and attention to every present moment and
movement. These basic skills may enhance older adults’ ability to focus
5.05(1.12)

− 1.52***
5.92*** on timed or speeded tasks that lead to the observed improvement in
0.74***

− 0.65
− 0.02

− 0.10
processing speed. A review on mindfulness trainings suggested that the
0.02

0.02

0.07

0.85
0.79
development of focused attention is linked to improved executive
function and selective attention (Chiesa et al., 2011). The simultaneous
rhythmic walking activity carried out in this program may also evoke
Dotmemorymean error dist.

the mindfulness state that facilitates attention and cognition (Spartano

et al., 2019; Christie et al., 2017). Light-intensity physical activity also
influences lipid and glucose metabolism, and both markers may regulate
risks for neurodegenerative diseases in older adults (Sato and Morishita,
2.15(1.44)

2015; Füzéki et al., 2017). Future trials can collect blood drops to un­
− 0.38
− 0.08

− 0.01

derstand whether those bio-physiological markers change as a function

1.36
0.11
0.10
0.03

0.01
0.09

0.89
1.18

of mindful walking practice, and, in turn, explain cognitive improve­

ments. The null findings in the memory aspect did not support previous
findings from moderate-to-vigorous physical activity engagement
DotmemorymeanRT

8138.66(4704.54)

(Erickson et al., 2019). One possible explanation may be that brain re­
gion governing memory capacities (i.e., amygdala, hippocampus, cere­
bellum, prefrontal cortex) are less likely to be engaged by light-intensity
394.02**
− 321.21
3671.83

210.63

114.22

375.86
130.65

movement targeted in the current program (Tulving and Markowitsch,

13.11

14.46

2788
3581

1997). This hypothesis warrants future intervention studies using neu­

roimaging measures to detect brain structural differences in those areas.
This proof-of-concept study combines objective and subjective
2-Backpropaccuracy

cognitive measures administered both in-lab and in the field to broadly

access key domains of cognition in response to a mindful walking pro­
2. Analyses were based on the 7th and the 8th walking sessions in which 30 min of mindfulness practice was incorporated.

gram. It demonstrates readiness for developing a well-powered ran­

0.77(0.13)

3. Perceived cognitive function was measured using self-report; all other cognitive outcomes were measured objectively.
<-0.0.01

domized controlled trial to draw causal inferences by ruling out other

1.05***

<-0.01

<-0.01
− 0.01

− 0.02

− 0.02
− 0.01
Results of the within-person changes in mobile-based cognitive outcomes from pre to post mindful walking session.

0.02

0.08
0.10

confounders with a comparator. We suggest that future efficacy trials

should establish criteria to determine meaningful magnitude of the
within-person cognitive changes using ambulatory assessments. Device-
2005.21(719.11)

based activity measures, such as accelerometers, should be applied

2-BackmeanRT

during the intervention to control for potential extra practices. Multiple

Note: 1. Number of measurement occasions = 92; number of participants = 25; RT = response time.
− 266.43**

occasions of post-program measures could be applied in future studies to

1230.69

− 10.36
187.85

180.61
50.39*

23.72

63.03
35.32

444.6
496.7

reveal the timing and duration of cognitive improvements in response to

the intervention. Given recent evidence that exposure to nature may
enhance short-term cognition (Berman et al., 2008; Bratman et al., 2015
Jun), future trials should also consider different mindful walking con­
Symbolsearchpropaccuracy

texts (e.g., indoor track, treadmill) to examine whether cognitive im­

provements are consistent across settings.
The walking and mindfulness training were delivered as an inte­
grated intervention package in this study. A muli-arm randomized
0.96(0.77)

controlled trial can be conducted to understand the relative cognitive

0.85***

<-0.01

<-0.01

effect between walking, mindful practice, and the integrated mindful

<0.01
− 0.03

− 0.04
0.01

0.01

0.01

0.05
0.06

walking conditions. Lastly, it will be valuable to investigate if older

adults who already have lower neurocognitive performance or with mild
cognitive impairment may also benefit from practicing mindful walking.
Symbol searchmeanRT

Validating the immediate and longer-term efficacy of brief mindful

3746.60(1131.09)

walking activity in future efficacy trials can contribute to designing

scalable and sustainable behavioral interventions to promote healthy
− 266.88*

aging in everyday life. Favorable results in future efficacy trials may

3032.62*

157.11*
− 61.07

− 17.06
280.00
80.64*

warrant the dissemination of mindful walking as part of a lifestyle

17.38

13.04

765.7
695.3

4. *p < .05; ** p < .01; *** p < .001.

strategy to sustaining healthy cognitive aging in late adulthood.

CRediT authorship contribution statement

Mean temperature (centered)
Day of the week (Mon = 0)
Post walking session (=1)
Walking session number

Chih-Hsiang Yang: Conceptualization, Methodology, Investigation,

Perceived sleep quality
Time of the day (hour)

Writing - original draft. Jonathan G. Hakun: Software, Visualization,

Writing - review & editing. Nelson Roque: Data curation, Writing -
Random Effect
Sex (male = 1)
Age (centered)

Intercept (SD)
Residual (SD)
Fixed Effect

review & editing. Martin J. Sliwinski: Resources, Writing - review &

(Intercept)
Mean(SD)

editing. David E. Conroy: Supervision, Conceptualization, Writing -

Table 2

Model

review & editing.

5
C.-H. Yang et al. Preventive Medicine Reports 23 (2021) 101490

Declaration of Competing Interest Ladouce, S., Donaldson, D.I., Dudchenko, P.A., Ietswaart, M., 2017. Understanding
minds in real-world environments: toward a mobile cognition approach. Front. Hum.
Neurosci. 10, 694.
The authors declare that they have no known competing financial Liu-Ambrose, Teresa, Eng, Janice J., 2015. Exercise training and recreational activities to
interests or personal relationships that could have appeared to influence promote executive functions in chronic stroke: A proof-of-concept study. J. Stroke
the work reported in this paper. Cerebrovasc. Dis. 24 (1), 130–137.
Malinowski, Peter, Shalamanova, Liliana, 2017. Meditation and cognitive ageing: The
role of mindfulness meditation in building cognitive reserve. J. Cogn. Enhanc. 1 (2),
References 96–106.
Månsson, Kristoffer NT, Skagius Ruiz, Erica, Gervind, Elisabet, Dahlin, Mats,
Alzheimer’s Association. 2020. Alzheimer’s disease facts and figures. Alzheimers Dement Andersson, Gerhard, 2013. D Development and initial evaluation of an internet-
J. Alzheimers Assoc. based support system for face-to-face cognitive behavior therapy: A proof of concept
Berk, Lotte, van Boxtel, Martin, van Os, Jim, 2017. Can mindfulness-based interventions study. J. Med. Internet Res. 15 (12), e280. https://doi.org/10.2196/jmir.3031.
influence cognitive functioning in older adults? A review and considerations for Marshall, Gad A., Zoller, Amy S., Kelly, Kathleen E., Amariglio, Rebecca E.,
future research. Aging Ment Health. 21 (11), 1113–1120. Locascio, Joseph J., Johnson, Keith A., Sperling, Reisa A., Rentz, Dorene M., 2014.
Berman, Marc G., Jonides, John, Kaplan, Stephen, 2008. The cognitive benefits of Everyday cognition scale items that best discriminate between and predict
interacting with nature. Psychol. Sci. 19 (12), 1207–1212. progression from clinically normal to mild cognitive impairment. Curr. Alzheimer
Bratman, G.N., Daily, G.C., Levy, B.J., Gross, J.J., 2015 Jun. The benefits of nature Res. 11 (9), 853–861.
experience: Improved affect and cognition. Landsc. Urban Plan. 1 (138), 41–50. Mj, K., Ca, D., C W, Mj, S, A E, Me, Z., et al. 2016. Influence of perceived stress on
Bugg, J.M., Jacoby, L.L., Toth, J.P., 2008. Multiple levels of control in the Stroop task. incident amnestic mild cognitive impairment: Results from the Einstein Aging Study.
Mem. Cognit. 36 (8), 1484–1494. Alzheimer Dis. Assoc. Disord. 1;30(2):93–98.
Chiesa, Alberto, Calati, Raffaella, Serretti, Alessandro, 2011. Does mindfulness training Nebes, R.D., Buysse, D.J., Halligan, E.M., Houck, P.R., Monk, T.H., 2009. Self-reported
improve cognitive abilities? A systematic review of neuropsychological findings. sleep quality predicts poor cognitive performance in healthy older adults.
Clin. Psychol. Rev. 31 (3), 449–464. J. Gerontol. Ser. B. 64B (2), 180–187.
Christie, G.J., Hamilton, T., Manor, B.D., Farb, N.A.S., Farzan, F., Sixsmith, A., Peavy, G.M., Jacobson, M.W., Salmon, D.P., Gamst, A.C., Patterson, T.L., Goldman, S.,
Temprado, J.-J., Moreno, S., 2017. Do lifestyle activities protect against cognitive et al., 2012. The influence of chronic stress on dementia-related diagnostic change in
decline in aging? A review. Front. Aging Neurosci. 9, 381. https://doi.org/10.3389/ older adults. Alzheimer Dis. Assoc. Disord. 26 (3), 260–266.
fnagi.2017.00381. Porteus, S.D., Peters, H.N. 1947. Maze test validation and psychosurgery. Genet Psychol
Conroy, D.E., Heartphone, Kim I., 2020. Mobile evaluative conditioning to enhance Monogr. 36:86–86.
affective processes and promote physical activity. Health Psychol. Off. J. Div. Health. Prohaska, T.R., Eisenstein, A.R., Satariano, W.A., Hunter, R., Bayles, C.M., Kurtovich, E.,
Psychol. Am. Psychol. Assoc. Jun 11. et al., 2009. Walking and the preservation of cognitive function in older populations.
Conroy, D.E., West, A.B., Brunke-Reese, D., Thomaz, E., Streeper, N.M., 2020. Just-in- Gerontologist 49 (S1), S86–93.
time adaptive intervention to promote fluid consumption in patients with kidney Reitan, R.M. 1986. Trail Making Test: Manual for Administration and Scoring. Reitan
stones. Health Psychol. 39 (12), 1062–1069. Neuropsychology Laboratory. book.
Czajkowski, Susan M., Powell, Lynda H., Adler, Nancy, Naar-King, Sylvie, Reynolds, Kim Salthouse, T.A., 1996 Jul. The processing-speed theory of adult age differences in
D., Hunter, Christine M., Laraia, Barbara, Olster, Deborah H., Perna, Frank M., cognition. Psychol. Rev. 103 (3), 403–428.
Peterson, Janey C., Epel, Elissa, Boyington, Josephine E., Charlson, Mary E., 2015 Salthouse, Timothy A., 2000. Aging and measures of processing speed. Biol. Psychol. 54
Oct. From ideas to efficacy: The ORBIT Model for developing behavioral treatments (1–3), 35–54.
for chronic diseases. Health Psychol. Off. J. Div. Health Psychol. Am. Psychol. Assoc. Sato, N., Morishita, R., 2015. The roles of lipid and glucose metabolism in modulation of
34 (10), 971–982. β-amyloid, tau, and neurodegeneration in the pathogenesis of Alzheimer disease.
Dai, S., Carroll, D.D., Watson, K.B., Paul, P., Carlson, S.A., Fulton, J.E., 2015. Front. Aging Neurosci. 7, 199.
Participation in types of physical activities among US adults—National health and Scarpina, F., Tagini, S. 2017. The Stroop Color and Word Test. Front Psychol [Internet].
nutrition examination survey 1999–2006. J Phys. Act Health 12 (s1), S128–S140. [cited 2020 Apr 26];8. Available from: https://www.ncbi.nlm.nih.gov/pmc/artic
Duff, K., Beglinger, L.J., Schultz, S.K., Moser, D.J., McCaffrey, R.J., Haase, R.F., et al. les/PMC5388755/.
2007. Practice effects in the prediction of long-term cognitive outcome in three Scherder, Erik, Scherder, Rogier, Verburgh, Lot, Königs, Marsh, Blom, Marco,
patient samples: A novel prognostic index. Arch. Clin. Neuropsychol. Off. J. Natl. Kramer, Arthur F., Eggermont, Laura, 2014. Executive functions of sedentary elderly
Acad. Neuropsychol. 2007;22(1):15–24. may benefit from walking: A systematic review and meta-analysis. Am. J. Geriatr.
Erickson, K.I., Hillman, C., Stillman, C.M., Ballard, R.M., Bloodgood, B., Conroy, D.E., Psychiatry. 22 (8), 782–791.
et al. 2019. Physical activity, cognition, and brain outcomes: A review of the 2018 Sliwinski, Martin J., Mogle, Jacqueline A., Hyun, Jinshil, Munoz, Elizabeth,
physical activity guidelines. Med. Sci. Sports Exerc. 51(6):1242–1251. Smyth, Joshua M., Lipton, Richard B., 2018. Reliability and validity of ambulatory
Farias, Sarah Tomaszewski, Mungas, Dan, Reed, Bruce R., Cahn-Weiner, Deborah, cognitive assessments. Assessment 25 (1), 14–30.
Jagust, William, Baynes, Kathleen, DeCarli, Charles, 2008. The measurement of Sofi, F., Valecchi, D., Bacci, D., Abbate, R., Gensini, G.F., Casini, A., et al., 2011. Physical
everyday cognition (ECog): Scale development and psychometric properties. activity and risk of cognitive decline: a meta-analysis of prospective studies.
Neuropsychology 22 (4), 531–544. J. Intern. Med. 269 (1), 107–117.
Finkel, D., Reynolds, C.A., McArdle, J.J., Pedersen, N.L. 2007. Age changes in processing Spartano, Nicole L., Davis-Plourde, Kendra L., Himali, Jayandra J., Andersson, Charlotte,
speed as a leading indicator of cognitive aging. Psychol. Aging. 22(3):558–568. Pase, Matthew P., Maillard, Pauline, DeCarli, Charles, Murabito, Joanne M.,
Freedland, Kenneth E., 2020. Pilot trials in health-related behavioral intervention Beiser, Alexa S., Vasan, Ramachandran S., Seshadri, Sudha, 2019. Association of
research: Problems, solutions, and recommendations. Health Psychol. 39 (10), accelerometer-measured light-intensity physical activity with brain volume: The
851–862. Framingham Heart Study. JAMA Netw. Open 2 (4), e192745. https://doi.org/
Fries, J.F., Bruce, B., Cella, D., 2005 Oct. The promise of PROMIS: using item response 10.1001/jamanetworkopen.2019.2745.
theory to improve assessment of patient-reported outcomes. Clin. Exp. Rheumatol. Tanay, G., Bernstein, A., 2013. State Mindfulness Scale (SMS): Development and initial
23 (5 Suppl 39), S53–S57. validation. Psychol. Assess. 25 (4), 1286–1299.
Füzéki, Eszter, Engeroff, Tobias, Banzer, Winfried, 2017. Health benefits of light- Teut, M., Roesner, E.J., Ortiz, M., Reese, F., Binting, S., Roll, S., Fischer, H.F.,
intensity physical activity: A systematic review of accelerometer data of the National Michalsen, A., Willich, S.N., Brinkhaus, B., 2013. Mindful walking in psychologically
Health and Nutrition Examination Survey (NHANES). Sports Med. 47 (9), distressed individuals: A randomized controlled trial. Evid. Based Complement
1769–1793. Alternat. Med. 2013, 1–7.
Gard, T., Hölzel, B.K., Lazar, S.W., 2014 Jan. The potential effects of meditation on age- Tulving, Endel, Markowitsch, Hans J, 1997. Memory beyond the hippocampus. Curr.
related cognitive decline: A systematic review. Ann. N.Y. Acad. Sci. 1307, 89–103. Opin. Neurobiol. 7 (2), 209–216.
Gotink, Rinske A., Hermans, Karlijn S.F.M., Geschwind, Nicole, De Nooij, Reinier, De Valliant, R., Rust, K.F., 2010. Degrees of freedom approximations and rules-of-thumb.
Groot, Wouter T., Speckens, Anne E.M., 2016. Mindfulness and mood stimulate each J. Off. Stat. 26 (4), 585–602.
other in an upward spiral: A mindful walking intervention using experience Venturelli, Massimo, Scarsini, Renato, Schena, Federico, 2011. Six-month walking
sampling. Mindfulness 7 (5), 1114–1122. program changes cognitive and ADL performance in patients with Alzheimer. Am. J.
Kabat-Zinn, J., 1994. Wherever You Go, There You Are: Mindfulness Meditation in Alzheimers Dis. Dementiasr. 26 (5), 381–388.
Everyday Life. Hyperion, New York. Wang, Hui-Xin, Xu, Weili, Pei, Jin-Jing, 2012. Leisure activities, cognition and dementia.
Kabat-Zinn, Jon, 2017. Walking meditations. Mindfulness 8 (1), 249–250. Biochim. Biophys. Acta. BBA Mol. Basis Dis. 1822 (3), 482–491.
Kabat-Zinn, J. 2012. Mindfulness for Beginners: Reclaiming the Present Moment—and Williams, D.M., Matthews, C., Rutt, C., Napolitano, M.A., Marcus, B.H., 2008.
Your Life. Sounds True. Interventions to increase walking behavior. Med. Sci. Sports Exerc. 40 (7 Suppl),
Kail, Robert, 2000. Speed of information processing: Developmental change and links to S567–S573.
intelligence. J. Sch. Psychol. 38 (1), 51–61. Wong, W.P., Coles, J., Chambers, R., Wu, D.B.-C., Hassed, C., 2017. The effects of
Kail, Robert, Salthouse, Timothy A., 1994. Processing speed as a mental capacity. Acta mindfulness on older adults with mild cognitive impairment. J. Alzheimers Dis. Rep.
Psychol. Amst. 86 (2–3), 199–225. 1 (1), 181–193.
Kim, C., Johnson, N.F., Gold, B.T., 2014. Conflict adaptation in prefrontal cortex: Now Yang, Chih-Hsiang, Conroy, David E., 2019. Feasibility of an outdoor mindful walking
you see it, now you don’t. Cortex 1 (50), 76–85. program for reducing negative affect in older adults. J. Aging Phys. Act. 27 (1),
18–27.