Journal of Alzheimer’s Disease 88 (2022) 1263–1278 1263

DOI 10.3233/JAD-220243
IOS Press

Systematic Review

Exercise Training for Mild Cognitive

Impairment Adults Older Than 60:
A Systematic Review and Meta-Analysis
Hui Lia,b,c , Wenlong Sub , Hui Danga,b,c , Kaiyue Hanb , Haitao Lub ,
Shouwei Yuea,c,∗ and Hao Zhanga,b,c,∗
a Rehabilitation Center, Qilu Hospital, Cheeloo College of Medicine, Shandong University, Jinan, Shandong,
China
b China Rehabilitation Research Center, Beijing, China
c University of Health and Rehabilitation Sciences, Qingdao, Shandong, China

Handling Associate Editor: Jianping Jia

Accepted 31 May 2022

Pre-press 2 July 2022

Abstract.
Background: The prevalence of mild cognitive impairment (MCI) continues to increase due to population aging. Exercise
has been a supporting health strategy that may elicit beneficial effects on cognitive function and prevent dementia.
Objective: This study aimed to examine the effects of aerobic, resistance, and multimodal exercise training on cognition in
adults aged > 60 years with MCI.
Methods: We searched the Cochrane Library, PubMed, and Embase databases and ClinicalTrials.gov (https://clinicaltrials.
gov) up to November 2021, with no language restrictions. We included all published randomized controlled trials (RCTs)
comparing the effect of exercise programs on cognitive function with any other active intervention or no intervention in
participants with MCI aged > 60 years.
Results: Twelve RCTs were included in this review. Meta-analysis results revealed significant improvements in resistance
training on measures of executive function (p < 0.05) and attention (p < 0.05); no significant differences were observed between
aerobic exercise and controls on any of the cognitive comparisons.
Conclusion: Exercise training had a small beneficial effect on executive function and attention in older adults with MCI.
Larger studies are required to examine the effects of exercise and the possible moderators.

Keywords: Aged, exercise, meta-analysis, mild cognitive impairment, systematic review

INTRODUCTION

Alzheimer’s disease (AD) is a neurodegenerative

∗ Corresponding
disorder that is the most common cause of demen-
authors: Prof. Shouwei Yue and Prof. Hao
Zhang, Department of Rehabilitation, Qilu Hospital, Shandong
tia [1]. In the initial stage of AD, the decline in
University, Jinan, Shandong, China. Tel.: +86 010 87569345; cognitive function can be significantly subtle and is
E-mails: [email protected]; [email protected]. currently being clinically identified as degenerative

ISSN 1387-2877 © 2022 – The authors. Published by IOS Press. This is an Open Access article distributed under the terms
of the Creative Commons Attribution License (CC BY 4.0).
1264 H. Li et al. / Review of Exercise Treatment on MCI

mild cognitive impairment (MCI) [2]. There are with and without cognitive impairment [38–40] with
three clinical subtypes of MCI: amnestic MCI, MCI a general effect of exercise training [41–43]. For
involving various degrees of impairment in multi- instance, all levels of physical activity were reported
ple cognitive domains, and MCI involving a single to have a significant and consistent protective effects
non-memory domain. Patients with MCI normally against the occurrence of cognitive decline [44]. Nev-
have self- or informant-reported cognitive complaints ertheless, studies on the effects of physical exercise
[3] and documented objective impairment in one or in older adults with MCI are insufficient and vary in
more cognitive domains that is greater than would be efficacy owing to the large variability in exercise pro-
expected for the patient’s age or educational back- tocols, compliance, and complicated interpretation of
ground [4], but still preserved the activities of daily the results [43]. This review involved patients with
living without significant impairment in social or MCI aged > 60 years, when the global prevalence
occupational functions [5]. The prevalence of MCI rates of dementia are exacerbated continuously. We
increases with age (6.7% for ages 60–64 years, 65.2% compared the results with previous reviews by exam-
in older adults aged > 60 years), and patients with ining the effects of aerobic, resistance, mind–body,
MCI are at a higher risk of progressing to dementia and multimodal exercise training on cognitive per-
than age-matched controls (progression to dementia formance in patients with MCI. These results may
at a rate of approximately 14.9% in MCI aged > 65 be helpful in clarifying the efficacy of exercise inter-
years followed by 2 years) [6]. To date, there is no ventions on cognitive function in older adults with
pharmacologic management currently approved for MCI.
MCI, and behavioral training, such as cognitive ther-
apy and exercise training, has symptomatic cognitive
MATERIALS AND METHODS
benefits in MCI [7–9].
Considering the feasibility and acceptability of
This study was conducted according to the Pre-
intervention training, exercise seems to be an avail-
ferred Reporting Items for Systematic Reviews and
able and economical therapeutic intervention for
Meta-Analysis. The study protocol was registered
individuals at an earlier point of cognitive decline.
with the International Prospective Register of Sys-
Evidence indicates that exercise elicits cognition-
tematic Reviews (CRD42021259555).
and neurovascular oxygenation-promoting effects in
rodents by reducing amyloid pathology [10–12];
enhancing hippocampal neurogenesis [13], oxidative Search strategy
stress [14–16], and intracranial energy metabolism
[17]; and reducing neuroinflammation [18]. How- A librarian-led systematic search of Embase,
ever, findings are mixed in human studies. Evidence PubMed, and Cochrane databases was performed
suggests that exercise reduces vascular risk factors, using a strategy combining selected Medical Sub-
such as hypertension [19], dyslipidemia [20], type 2 ject Headings terms (exercise, exercise therapy,
diabetes [21], obesity [22], subclinical atherosclero- resistance training, sports, physical fitness, circuit-
sis [23], and arrhythmias [24], which are associated based exercise, endurance training, Tai Ji, bicycling,
with a greater risk of cognitive impairment and yoga, walking, jogging, running, swimming, muscle
dementia [25]. Moreover, exercise may reduces the stretching exercises, plyometric exercise, cognitive
risk of stroke [26]. Recent meta-analyses showed dysfunction, aged) and free-text terms (For full
that physical exercise may help preserve or even search strategy, see the Supplementary Material).
improve cognitive function [27–29] and periph- No language or other limitations were imposed dur-
eral brain-derived neurotrophic factor concentrations ing the search process. Additionally, we searched
[30, 31] in healthy older adults. Exercise increases ClinicalTrial.gov, reference lists of selected articles,
precentral cortical thickness and reduces neuroplas- and related review articles to screen for related
ticity [32]; augments gray and white matter volumes clinical trials being conducted. The search terms
[33], hippocampal volume [34–36], and systemic regarding random control study design (randomized
levels of glycosylphosphatidylinositol-specific phos- controlled trial [Publication Type] OR randomized
pholipase; and ameliorates age-related regenerative [Title/Abstract] OR placebo [Title/Abstract]) were
impairments [37]. obtained from McMaster University Health Infor-
Previous studies have reported a valid effect of mation’s website. The last search was performed on
exercise training on cognitive function in older adults November 5, 2021.
 H. Li et al. / Review of Exercise Treatment on MCI 1265

Criteria for considering studies for this review resolved through discussions with expert HZ (study
selection procedure, Fig. 1).
Types of studies
Human subject studies designed as randomized Quality assessment and data extraction
controlled trial (RCT) published in English were
included without year restriction. Nonrandomized We extracted types of data from the full-text
controlled studies, conference abstracts, reviews, pro- articles, including study characteristics of author;
tocols, and secondary analyses were excluded. published year; country; RCT design, including
grouping, sample size, sex ratio, mean age, MMSE
Types of participants score before the intervention, and diagnostic crite-
Participants diagnosed with MCI according to ria; characteristics of the exercise intervention group,
accepted criteria (Petersen, Winblad, Morris, Albert including type, intensity, frequency, duration, and
criteria) and aged > 60 years were included. We time; and outcome neuropsychological tool used for
excluded healthy aging elderly adults, patients with measuring cognitive function. If the full text or data
any form of dementia, or those diagnosed with cog- were incomplete, the author was contacted via email.
nitive impairment due to definite etiologies, such as The protocol of each included study was examined
trauma or vascular or psychiatric diseases. Any clin- to verify reporting bias. The quality of individual
ical subtype of MCI was considered eligible. MCI study was assessed using the Cochrane Risk of Bias
defined by a single scale was considered insufficiently Tool (Fig. 2) [45]. Risk was assessed to be “low,”
credible and unreliable. “high,” or “unclear.” Two independent authors (HL
and HD) assessed the risk of bias. Disagreements
Types of interventions were resolved by discussion with a third author.
We included supervised structured exercises of
any frequency, intensity, duration, or time directed Data synthesis and statistical analyses
at improving physical fitness. Combined exercise
interventions with two or more types were included Review Manager (RevMan) version 5.3 was used
as multimodal training; however, studies involving for all analyses. The summary statistics required for
exercise programs that combined cognitive training each assessment for continuous data were the post-
or other non-exercise activities (e.g., diets, drugs, intervention mean scores, standard deviation, and
and video games), not fully supervised or lasted < 4 number of participants in each group (several stud-
weeks, were excluded. ies did not provide change scores). The mean was
entered as negative when a higher score indicated
Types of control groups worse performance (e.g., reaction time, trail mak-
We included studies that involved comparators ing test). Cognitive function is normally assessed
who engaged in a strength or balance tone, a pro- using both global and individual cognitive tests.
gram of social or mental activities, no treatment, or There are general cognitive tests (MMSE, MoCA)
on the waiting list. and specific tests that cover cognitive domains,
including memory, attention, visuospatial processing,
Types of outcomes language, executive function, and social comport-
The primary outcome was cognitive function that ment. We classified the cognitive outcomes assessed
was evaluated using a measurable cognitive screen- using various neuropsychological tests into seven
ing instrument, whereas the secondary outcome was domains: (1) immediate memory, (2) working mem-
the dropout rate. Studies with no outcomes in spe- ory, (3) delayed memory, (4) processing speed, (5)
cific domains of cognitive fields or studies that attention, (6) executive function, and (7) recognition
only involved the Mini-Mental State Examination [27]. For each domain, we included only one test from
(MMSE) or Montreal Cognitive Assessment (MoCA) a single trial that was used more frequently or closer
were excluded. Two authors (WS and KH) initially to the domain. For dichotomous data, we extracted
excluded articles that did not meet the inclusion the number of participants in each group for each
criteria, based on their titles and abstracts. After dedu- trial. Statistical analyses were performed if more than
plication, the remaining articles were independently one study could be grouped into cognitive domains
screened for eligibility by full-text assessment based according to intervention (e.g., aerobic or resistance
on the defined inclusion criteria. Disagreements were exercise). When only one study involved a category,
1266 H. Li et al. / Review of Exercise Treatment on MCI

Fig. 1. PRISMA flow diagram of the literature selection process.

Fig. 2. Cochrane’s Quality Assessment: summary of risk of bias for each quality item for each included study.
 H. Li et al. / Review of Exercise Treatment on MCI 1267

the statistics from that single study were reported, protocol involved home-based exercise and lacked
but not analyzed. For continuous variables measured supervision [60–62]. No further articles were found
on different scales, standardized mean differences by searching the reference lists of reviews identified
(SMDs) and 95% confidence interval (CI) were cal- during the initial search or in the included articles.
culated; for dichotomous variables, risk ratios and
95% CIs were calculated. Risk of bias
The I2 statistic was calculated to describe the
proportion of true heterogeneity in the observed Studies without a statement of random sequence
variance across studies. Cochrane’s handbook rec- generation or allocation concealment in sufficient
ommends heterogeneity as not important (0–40%), detail were judged to have an “unclear” risk of bias
moderate heterogeneity (30–60%), substantial het- [45]. Most of the studies did not report protocols;
erogeneity (50–90%), and considerable heterogene- therefore, it was not possible to determine if there
ity (75–100%). Sensitivity analysis was performed was a selection bias. In all the assessed studies, it was
when I2 was > 50% [46]. For which data were not feasible to blind the participants to exercise train-
available, meta-analyses were performed using the ing interventions. Therefore, the included studies all
inverse-variance method. A random-effects model showed “high risk” of blinding for participants and
was used to assess the cognitive outcomes and personnel (performance bias). Half of the included
dropout rates. Given the small number of included studies stated that non-blind or without details of
studies, we did not perform a funnel plot visual test blinding of outcome assessors. It was judged non-
to investigate publication bias [9, 47]. blind as a “high-risk” bias and “not clear” for that
without details of the detection bias. Studies were
RESULTS assessed as having a high risk of bias for incomplete
outcome data (attrition bias) if they had a relatively
Included studies high and unbalanced missing rate between groups or
did not account for dropout. All other sectors had a
The initial database search identified 3,723 records low risk of bias (Fig. 2).
(PubMed, n = 886; Embase, n = 865; Cochrane,
n = 1,972) (Fig. 1). After duplicates were removed Aerobic exercise intervention versus control
and screened, the full texts of 286 articles were group
assessed for eligibility, of which 12 met the eligibil- Seven studies including 234 participants with
ity criteria (Tables 1–3). A total of 708 participants MCI relative to aerobic exercise intervention and
were included and assessed in this systematic review. any active or no intervention control groups were
Seven trials of 122 participants assessed the treat- pooled for meta-analyses, which contributed data
ment effects of aerobic exercise, including bicycle or to no less than one cognitive domain [48–54]. The
motor-driven treadmill training [48–51] (n = 66) and grouping of cognitive scales and studies over spe-
walking [52–54] (n = 56); six trials of 116 participants cific domains of cognitive function is shown in
reported on resistance exercise, including elastic band Table 4. We were able to perform meta-analyses
training [55–57] (n = 48), circuit exercise [51, 53] at all seven of our pre-specified cognitive domains
(n = 46), and dumbbell training group [58] (n = 22). (Supplementary Material, Outcome 1.1 to Outcome
Moreover, one trial researched on multimodal exer- 1.7). The results revealed no heterogeneity (I2 = 0%)
cise [59] (n = 147). In total, 112 participants with across studies investigating processing speed and
MCI in the control groups were compared to the aer- recognition, low heterogeneity in immediate mem-
obic training group, who received a waiting list, a ory (I2 = 24%), moderate heterogeneity in working
maintaining lifestyle routine, a strength or balance memory (I2 = 47%), delayed memory (I2 = 36%),
tone, muscle-stretching exercises, or social visits; 126 and attention (I2 = 42%). However, studies including
participants in the control groups were compared to executive function demonstrated substantial het-
the resistance training group, who received a regular erogeneity (I2 = 62%). A sensitivity analysis was
lifestyle routine, a strength or balance tone, stretching performed with the main analysis performed with
exercises, or lecture counseling classes; and 131 par- and without each study in investigating execu-
ticipants in the control groups were compared with tive function; the significance of the results did
the multimodal exercise group. Three mind–body not change (Supplementary Material, Outcome 1.6
training studies were excluded because their exercise and Outcome 1.6.1). The duration of each aerobic
 Table 1

1268
Characteristics of studies – aerobic exercise versus control group
Study RCT Country Group (N) Sex (male/ Mean age MMSE Diagnosis criteria Exercise Outcome
design female)
Tsai et al. [51] Parallel China Aerobic exercise 5/14 (AE) 66.00 (AE) 27.16 ± 1.26 (AE) Gauthier 2006, AE: Working memory: Digit span
(AE) n = 19 7/11 (RE) 65.44 (RE) 26.56 ± 1.34 (RE) Petersen 2004, Type: Bicycle ergometer or of Wechsler-IV Adult
Resistance 5/13 (CO) 65.17 (CO) 27.00 ± 1.65 (CO) Winblad 2004 motor-driven treadmill intelligence test
exercise (RE) Intensity: 70–75% of the target
n = 18 heart rate reserve (HRR)
Static stretching Frequency: 3 sessions/week
exercises (CO) Duration: 40 min
n = 18 Time: 16 weeks
CO:
Type: Static stretching exercises
Intensity: No load
Frequency, duration, and time:
same to AE group

H. Li et al. / Review of Exercise Treatment on MCI

Yogev-Seligmann Parallel Israel AE n = 13 8/5 (AE) 70.84 (AE) Not mentioned Albert 2011 AE: Immediate and delayed
et al. [50] Balance and tone 7/7 (BAT) 71.92 (BAT) Type: Stationary bicycles memory: The Rey Auditory
(BAT) n = 14 Intensity: 70% to 80% HRR; Verbal Learning Test
Frequency: 3 sessions/week Working memory: Digit span
Duration: 40 min component of the Wechsler
Time: 16 weeks memory test
BAT: Processing speed: The color
Type: balance, gross motor version of the trail making
coordination, and light toning test 1
exercises Execution function: The
Intensity: maintaining heart rate verbal fluency test
below 30% of HRR Attention: The color version of
Frequency, duration, and time: the trail making test 2
same to AE group Recognition: Faces/houses
recognition test
Scherder et al. Parallel America AE n = 15 2/13 (AE) 84 (AE) 9.73 ± 1.94 (AE) Petersen 1999 AE: Executive function: Retrieve
[52] Hand and face 1/14 (CO) 86 (CO) 9.87 ± 1.41 (CO) Type: Self-paced slow walking familiar information from
exercises with an aid semantic memory (Category
Control (CO) Frequency: 3 sessions/week Naming);
n = 15 Duration:30 mins Processing speed:
Time:6 weeks Trail-making A + B
CO: Working memory: Digit span
One subgroup eight subjects from the Wechsler Memory
received social visits as a Scale—Revised
‘treatment’, while in the other Memory: Visual Memory
subgroup the remaining seven Span, The Verbal Learning
subjects continued their and Memory Test, Direct
normal social activities. Recall; Rivermead
Behavioural Memory Test
(faces, pictures)
Delayed memory and
recognition: The Verbal
Learning and Memory Test,
delayed recall and
recognition.
Study RCT Country Group (N) Sex (male/ Mean age MMSE Diagnosis criteria Exercise Outcome
design female)
Nagamatsu et al. Parallel Canada AE n = 30 All female 75.6 (AE) 27.4 ± 1.5 (AE) Petersen2004 AE: Verbal memory and learning:
[53] RE n = 28 73.9 (RE) 27.0 ± 1.8 (RE) Type: Outdoor walking The Rey Auditory Verbal
BAT n = 28 75.1 (CO) 27.1 ± 1.7 (CO) Intensity: 70% to 80% of HRR, Learning Test, delayed recall;
13–15 on the RPE scale Spatial Memory: recall the
Frequency: 2 sessions/week spatial location of dots
Duration: 60 min presented on a screen;
Time: 26 weeks Processing speed: Choice
BAT: Reaction Time
Type: Stretching exercises,
range of motion exercises,
balance exercises functional
and relaxation techniques

H. Li et al. / Review of Exercise Treatment on MCI

Intensity: No loading
Duration and time: same with
RE group
Law et al. [48] Parallel China AE n = 16 8/8 77.94 (AE) Not mentioned Albert 2011 AE: Memory: Chinese Version
Waitlist control 5/9 75.14 (CO) Type: Structured whole body Verbal Learning Test;
group (CO) movement exercise, bicycle, Processing speed: Trail
n = 14 and arm ergometry Making Test A
Intensity: Moderate intensity Attention: Trail Making Test B
aerobic exercise, at 4–5/10 on
rate of perceived exertion
Frequency: 12 sessions
Duration: 60-min
Time: 8 weeks
Combourieu Parallel France AE n = 18 Not 77.1 (AE) 28.2 ± 0.43; Petersen 2004 AE: Executive function: The
Donnezan et al. Cognitive mentioned 79.2 (CO) 27.3 ± 0.5 Type: Bikes training Matrix
[49] training; Intensity: moderate intensity Reasoning test
Combined (i.e., 60% HR) Attention: The flexibility part
training; Frequency: 2 sessions/week of the Stroop Color Word test,
Control group Duration: 60 min The Digit Span Forward test
(CO) n = 14 Time: 12 weeks Working memory: The Digit
Span Backward test
Damirchi et al. Parallel Iran AE n = 11 All female 68.81 (AE) 23.18 ± 2.18; Petersen 2004 Type: Walking with muscular Processing speed: Reaction
[54] Mental training; 69.11 (CO) 23.44 ± 2.06 strength training time, Digit Symbol Coding
Exer- Intensity: Walking from 55% to test
cise + mental; 75% heart rate reserve, Attention: Modified Stroop
Control group muscle strength rating of color word test error number,
(CO) n = 9 perceived exertion within the Digit span forward subtests of
desired 13 to 15 range; the Wechsler Adult
Frequency: 3 sessions/week; Intelligence Scale
Duration: 60 min;
Time: 8 weeks

1269
 Table 2

1270
Characteristics of studies – resistance exercise versus control group
Study RCT Country Group (N) Sex (male/ Mean age MMSE Diagnosis Exercise Outcome
design female) criteria
Tsai et al. [51] Parallel China Resistance 7/11 (RE) 65.44 (RE) 26.56 ± 1.34 (RE) Gauthier 2006, RE: Working memory: Digit
exercise (RE) 5/13 (CO) 65.17 (CO) 27.00 ± 1.65 (CO) Petersen 2004, Type: Circuit-exercise use span of Wechsler-IV Adult
n = 18 Winblad 2004 free weights and intelligence test
Static stretching bodybuilding machines
exercises (CO) Intensity: 75% of their target
n = 18 1-RM
Frequency: 3 sessions/week
Duration: 40 min
Time: 16 weeks
CO:
Type: Static stretching

H. Li et al. / Review of Exercise Treatment on MCI

exercises
Intensity: No load
Frequency, duration, and
time: same to RE group
Hong et al. [55] Parallel Korea RE n = 10 3/7 (RE) Male 78.33 Not mentioned Petersen 1999 Type: Elastic band Execution function:
CO n = 12 3/9 (CO) Female 77.71 Frequency: 2 sessions/week Category/Semantic fluency
(RE) Intensity: 15-repetition test, Letter/Phonemic
Male 78.33 maximum (15RM, about fluency test;
Female 75.11 65% of maximum) Attention and working
(CO) Duration: 60 min memory: The digit span
Time: 12 weeks test; Stroop test;
CO: maintain current lifestyle Immediate Memory and
recognition: Short-term
and recognition memory,
Rey 15-Item Memory Test
Yoon et al. [57] Parallel Korea RE n = 22 6/14 (RE) 73.82 (RE) 24.23 ± 2.89 (RE) CDR, Kelaiditi RE: Memory: Rey 15-Item
CO n = 23 7/16 (CO) 74.03 (CO) 24.22 ± 1.86 (CO) 2013 Type: High-speed resistance memory test;
training Processing speed and
Intensity: blue elastic bands, attention: Trail Making
at a perceived exertion rate A&B Test, Working
of 12–13; 2–3 sets of 12–15 memory: Digit Span (both
repetitions forward and backward) test;
Frequency: 3 sessions/week Executive functions: Frontal
Duration: 60 min assessment battery
Time: 16 weeks
CO:
Type: Balance and resistance
band stretching
Frequency: 2 sessions/week
Duration and time: same with
RE group
Study RCT Country Group (N) Sex (male/ Mean age MMSE Diagnosis Exercise Outcome
design female) criteria
Nagamatsu et al. Parallel Canada Aerobic All female 75.6 (AE) 27.4 ± 1.5 (AE) Petersen 2004 Type: Keiser Pressurized Air Verbal memory and
[53] Training (AE) 73.9 (RE) 27.0 ± 1.8 (RE) system learning: The Rey
n = 30 75.1 (CO) 27.1 ± 1.7 (CO) Intensity: 6–8 repetitions Auditory Verbal Learning
RE n = 28 (two sets), 13–15 on the Test, delayed recall; Spatial
Balance and Borg’s Rating of Perceived Memory: recall the spatial
Tone (BAT) Exertion location of dots presented
n = 28 Frequency: 2 sessions/week on a screen;
Duration: 60 min Processing speed: Choice
Time: 26 weeks Reaction Time

H. Li et al. / Review of Exercise Treatment on MCI

BAT:
Type: Stretching exercises,
range of motion exercises,
balance exercises functional
and relaxation techniques
Intensity: No load
Duration and time: same with
RE group
Lee et al. [56] Parallel Korea RE n = 18 7/11 73.7 (RE) 23.8 ± 2.9; CDR Morris RE: Executive function: Frontal
Control: lecture 9/13 74.2 (CO) 23.4 ± 1.3 1993 Type: High speed elastic band Assessment Battery
consultation Frequency:3 sessions/week
(CO) n = 22 Intensity:10–12 repetitions of
two to three sets, perceived
exertion index ranging from
12 to 13
Duration: 50 min
Time: 8 weeks
Lü et al. [58] Parallel China RE n = 22 6/16 69.00 (RE) 27.23 ± 1.63; Petersen, 1999 RE: Attention: The Trail Making
CO n = 23 7/23 70.43 (CO) 26.43 ± 2.00 Type: Momentum based Test B; The Digit Span Test
dumbbell-training forward
Intensity: Each individual Working memory: The Digit
spinning exercise lasted 1–2 Span Test backward
minutes with repetitions set
at 4–5 minutes.
Frequency: 3 sessions/week
Duration: 60 min
Time:12-week

1271
1272 H. Li et al. / Review of Exercise Treatment on MCI

exercise intervention in these studies lasted from 30

learning delayed recall test

Episodic memory: The list
to 60 min, with 1–3 sessions/week, ranging from 6

Category verbal fluency

Memory Inventory for

Cognitive complaints:
to 26 weeks. There was no significant difference

Executive function:
between the aerobic exercise and control groups

test; Subjective
in immediate memory (p = 0.79), executive func-
tion (p = 0.1), working memory (p = 0.49), processing

Chinese
Outcome

speed (p = 0.93), delayed memory (p = 0.96), atten-

tion (p = 0.39), or recognition (p = 0.56). One study
observed the maintenance effects of aerobic exer-

Attending in social activities

cise after treatment termination, reporting that the
Frequency: 3 sessions/week
toning exercise, one mind

aerobic exercise session

treatment effects declined during the treatment-free

body exercise and one
Type: One stretching &

period [52]. There was no difference between the

Duration: 60 min
Time: 12 months

aerobic exercise and control groups in dropout rates

(odds ratio [OR], 2.47; 95% CI, 0.94–6.49; seven
trials, 251 participants; Supplementary Material,
Exercise

Outcome 1.8). Two studies reported no adverse events

AE:

S:

in any group during the course of the intervention

Characteristics of studies –Multimodal exercise versus control group

[48, 50]. One study reported the presence of adverse

Winblad 2004

events, including episodes of shortness of breath that

Diagnosis

resolved with rest and non-injurious falls [53], with

criteria

no significant differences between groups (p = 0.54).

The other studies did not mention any adverse events.
25.8 ± 2.3 (ME);

Resistance exercise intervention versus control

25.6 ± 2.4 (S)

group
Table 3

Six studies including 242 participants with MCI

MMSE

relative to resistance exercise intervention and any

active or no intervention control groups were pooled
for meta-analyses, which contributed data to at least
75.5 (ME)
Sex (male/ Mean age

one cognitive domain [51, 53, 55–58]. We performed

75.4 (S)

meta-analyses on five cognitive domains because

only one study was involved in the domain of delayed
memory [53] and recognition [55] each (Figs. 3
29/101 (S)
female)

and 4; Supplementary Material, Outcome 2.1 to

34/113
(ME)

Outcome 2.4, Outcome 2.7). The results revealed

no heterogeneity (I2 = 0%) across studies investigat-
motion exercise

ing immediate memory and executive function and

Cognitive group;

Social group (S)

(ME) n = 147;

low heterogeneity in attention (I2 = 13%); however,

cognitive-
Multimodal
Group (N)

physical

studies including working memory (I2 = 32%) and

Integrated

n = 131
Group;

processing speed (I2 = 42%) demonstrated moderate

heterogeneity.
The duration of each resistance exercise interven-
Country

China

tion in these studies lasted from 40 to 60 minutes,

with 2–3 sessions/week, ranging from 8 to 26 weeks.
Meta-analysis on pooled data revealed significant
Parallel
design

improvement differences for the resistance exer-

RCT

cise intervention versus control groups on measures

of attention (0.48 [0.06–0.89], p < 0.05) and exec-
Lam et al. [59]

utive function (0.44 [0.05–0.83], p < 0.05), but not

on immediate memory (p = 0.79), working memory
(p = 0.19), processing speed (p = 0.92), delayed mem-
Study

ory (p = 0.82), or recognition (p = 0.77). There was

 H. Li et al. / Review of Exercise Treatment on MCI 1273

Table 4
Grouping of cognitive tests and studies over cognitive functions
Cognitive function Cognitive tests Trial
Immediate memory The Verbal Learning and Memory Test Direct recall Scherder et al. [52]
Chinese Version Verbal Learning Test, total free recall Law et al. [48]
The Rey Auditory Verbal Learning Test Nagamatsu et al. [53], Yogev-Seligmann et al. [50]
Rey 15-Item Memory Test Hong et al. [55], Yoon et al. [57]
Working memory Digit span backward Lu et al. [58], Combourieu Donnezan et al. [49], Hong
et al. [55]
Digit span component of the Wechsler memory test Scherder et al. [52], Yoon et al. [57], Tsai et al. [51],
Yogev-Seligmann et al. [50]
Delayed memory Rey auditory verbal learning test delayed recall trail Nagamatsu et al. [53], Yogev-Seligmann et al. [50]
The Verbal Learning and Memory Test, delayed recall Scherder et al. [52]
Chinese Version Verbal Learning Test, delayed recall Law et al. [48]
The list learning delayed recall test Lam et al. [59]
Processing speed Reaction time of modified Stroop color-word test Damirchi et al. [54]
Reaction time of spatial memory Nagamatsu et al. [53]
Trail making test A Yoon et al. [57], Law et al. [48], Yogev-Seligmann et al.
[50]
Trail-making A + B Scherder et al. [52]
Attention Trail making test B Lu et al. [58], Yoon et al. [57], Law et al. [48],
Yogev-Seligmann et al. [50]
Digit span (forward) Combourieu Donnezan et al. [49], Damirchi et al. [54],
Hong et al. [55]
Executive function Category verbal fluency Scherder et al. [52], Lam et al. [59]
Verbal fluency, semantic Hong et al. [55], Yogev-Seligmann et al. [50]
Frontal Assessment Battery Yoon et al. [57], Lee et al. [56]
Matrix Reasoning test Combourieu Donnezan et al. [49]
Recognition Faces/houses recognition Yogev-Seligmann et al. [50]
The Verbal Learning and Memory Test, recognition Scherder et al. [52]
Rey 15-Item Memory Test, recognition Hong et al. [55]

Fig. 3. Outcome 2.5 Impact of resistance exercise training on attention. CI, confidence interval; IV, inverse variance; SD standard deviation;
Std, standardized.

no difference in the dropout rates between the resis- intervention [56, 58]. One study reported no signifi-
tance exercise and control groups (OR, 1.21; 95% cant differences of adverse event occurrence between
CI, 0.59–2.50; six trials; Supplementary Material, groups (p = 0.54) [53]. The other studies did not men-
Outcome 2.8). One study reported the maintenance tion any adverse events.
effects of resistance exercise treatment on partici-
pants with MCI [49], and the outcome of the matrix Multimodal exercise intervention versus control
reasoning test (p < 0.01) and digit span forward group
(p < 0.05) still improved for the 6-month post-test Lam et al. assessed the effect of multimodal
relative to the pre-test. Two studies reported no exercise intervention on cognitive function in 114
adverse events in any group during the course of the individuals with MCI [59]. These cognitive domains
1274 H. Li et al. / Review of Exercise Treatment on MCI

Fig. 4. Outcome 2.6 Impact of resistance exercise training on executive function. CI, confidence interval; IV, inverse variance; SD standard
deviation; Std, standardized.

involved outcomes of pre- and post-training measures of each measure. We observed that half of the stud-
of delayed recall and executive function (category ies reported significant improvements in composite
verbal fluency test). The trial had a fair method- measures of cognitive function for exercise versus
ological quality. The author found a significant controls, including executive function [49, 52, 57],
improvement favoring multimodal exercise interven- working memory [55], and processing speed [50,
tion over the control group on measures of both 52, 53, 57]. Despite this, the majority of compar-
the category verbal fluency test and delayed recall isons yielded no significant differences. Additionally,
after 12 months of intervention (multilevel model, the results of this meta-analysis could have been
p < 0.001). There was no difference between the mul- affected by the quality of the included trials. Across
timodal exercise and control groups in the dropout the included papers, approximately half did not state
rates. This study reported no adverse events associ- the methods of randomization or blinding of outcome
ated with this intervention. No differences in age and assessors. For all but one trial, allocation was poorly
educational level were observed between completers reported. No trial evaluated contamination bias that
and defaulters. could have interfered with the outcomes. Only two tri-
als reported protocols [57, 59]; for most trials, it was
DISCUSSION not feasible to determine whether there was selective
reporting of results.
Summary of main results Some researchers have suggested that improve-
ments in cognitive function could be attributed to
We examined the effects of aerobic, resistance, and improvements in cardiovascular fitness due to exer-
multimodal exercise training on cognitive function in cise [63]. Changes in grip strength and walking
adults aged > 60 years with MCI. One multimodal speed have been correlated with mental state and
exercise intervention study was presented insuffi- fluid cognition [64]. However, most trials involved
ciently. Meta-analysis results revealed that, compared in this review did not report any objective measures
to the control group, resistance exercise significantly of cardiorespiratory fitness. In addition, studies have
improved performance on measures of attention and reported that short session duration and higher fre-
executive function. No significant differences were quency might predict higher effect sizes [65, 66].
observed in any of the remaining cognitive domains. However, owing to the inconsistency of exercise
We included 12 studies after screening because modalities in this review, dose-response tests were
of a more qualified inclusion criterion. Most stud- not performed in this study.
ies involved small sample sizes. There was significant
heterogeneity in the methodologies across the studies Agreements and disagreements with other studies
involved, including sample size, exercise schemes, or reviews
participant compliance, and study quality. Although
the review analyzed the fine-sorted subgroups of cog- Seven meta-analyses published data based on sim-
nitive function, there could be variations in the results ilar hypotheses yet failed to find comparable results.
 H. Li et al. / Review of Exercise Treatment on MCI 1275

Gates et al. assessed the effects of exercise on insulin resistance. We excluded those who were not
cognitive function in 14 RCT trials with 1,695 par- purely cognitively impaired but had other diseases,
ticipants [43]. Their eligibility criteria differed from which might disturb the exercise effect and cognitive
this study in that they included participants with function. They reported improved global cognitive
probable MCI, and subjective memory decline who function, executive function, and delayed recall after
were aged > 65. We both set the criteria for exercise the exercise.
intervention up longer than 4 weeks and prescribed Similarly, Zhu et al. reviewed and analyzed aerobic
them specifically. However, we only included trials dance in participants with MCI aged > 50 years. This
that reported exercise training effectively supervised meta-analysis involved five studies with 842 patients
by centers, physicians, or professionals; those self- [70]. They found that aerobic dance, including Tai
reported, at home, or without reported supervision Chi, improved global cognitive function and memory
were excluded. The authors concluded that there is executive function in older adults with MCI.
limited evidence to prove that any exercise modality A meta-analysis by Lin et al. assessed the effects of
improves cognitive function in individuals with MCI. Tai Chi on cognitive performance. Except for exercise
De Souto Barreto et al. reviewed five papers on intervention, the criteria differed from this review in
long-term exercise (12 months or longer) in older including younger participants with MCI [71]. This
adults with MCI, onset of dementia, or clinically meta-analysis indicated that Tai Chi has positive clin-
meaningful cognitive decline [67]. Apart from the ical effects on cognitive function (global cognitive
inclusion criteria, another difference is that they con- function, memory and learning, executive function)
sidered the change in MMSE score as a main outcome in older adults with MCI.
indicator and analyzed the overall effect of exercise,
whereas we discussed the different exercise modali- Implications for practice and research
ties separately. Their review observed no significant
effects of exercise, either individually or in combina- We found that resistance exercise improved exec-
tion, on reducing the risk of cognitive decline. utive function and attention performance in patients
Zhang et al. published a meta-analytic review aged > 60 years. However, there might be some con-
of traditional Chinese exercise in older adults with founding factors in the outcomes. First, there are
MCI [68]. They found that four Tai Chi studies and more studies and diversified scales involved in aer-
one Liuzijue study met the inclusion criteria. We obic exercise analyses, which might lead to higher
excluded trials on participants with MCI with a def- heterogeneity, and the large variability in training
inite etiology, whereas this review included medical paradigms and study quality might also influence the
or neurological disorders, such as AD, dementia, outcome. Further investigations into the parameters
or Parkinson’s disease. The authors concluded that and populations most associated with the efficacy of
exercise improved the visuospatial function for indi- physical training for cognitive function are neces-
viduals with MCI. sary. Cognition is a complex and multi-class function;
The meta-analysis presented by Lee et al. has a however, there is insufficient evidence to illustrate the
similar aim (reporting the effects of exercise interven- relationship between exercise and brain-processed
tion for older adults with MCI) and exclusion criteria cognitive function. Certainly, clear differences exist
(cross-sectional, protocol, and review studies) [69]. in the effects of exercise on the cognition subgroups.
In contrast to this review, our study subdivided cog- Within the seven disparate intrinsic connectivity net-
nitive function into detailed items instead of MMSE. works in the brain, the beneficial effects exerted by
In addition, this review finally involved three eligible aerobic exercise might be mediated by the greatest
studies, including five exercise interventions (aerobic association with the executive and dorsal attention
or resistance exercise) combined for data analysis, networks [72]. Exercise activates the hub region of
and only one study was classified into a functional the executive network (e.g., attention, working mem-
category. This review reported that cognitive function ory, cognitive control) in the brain [11, 73], and the
was significantly increased in the exercise group. duration of exercise may be positively correlated with
Biazus-Sehn et al. reviewed the global effects of its volume [74]. Exercise-induced brain functional
various exercise modalities on MCI. Their search homogeneity variation is likely to improve executive
identified 27 studies that reported data [42]. The dif- control behavior and predict attention behavior [75].
ference is that they involved participants with a mean This evidence may help to explain the results. Thus,
age of ≥60 years who had other diseases, such as to elucidate the exact relationship between exercise
1276 H. Li et al. / Review of Exercise Treatment on MCI

and cognitive function in patients with MCI, more cognitive function in older adults with mild cognitive
large-sample RCTs and animal studies are required impairment or dementia: A meta-analysis. Ageing Res Rev
40, 75-83.
in the future. It also raises our wider understanding [8] Gavelin HM, Dong C, Minkov R, Bahar-Fuchs A, Ellis
of how to improve exercise compliance and to use a KA, Lautenschlager NT, Mellow ML, Wade AT, Smith AE,
smaller battery of cognitive scales for the uniformity Finke C, Krohn S, Lampit A (2021) Combined physical
and comparability of results. and cognitive training for older adults with and without
cognitive impairment: A systematic review and network
meta-analysis of randomized controlled trials. Ageing Res
ACKNOWLEDGMENTS Rev 66, 101232.
[9] Meng Q, Yin H, Wang S, Shang B, Meng X, Yan M, Li
G, Chu J, Chen L (2021) The effect of combined cogni-
This work was supported by grants from the Cap- tive intervention and physical exercise on cognitive function
ital Health Research and Development of Special in older adults with mild cognitive impairment: A meta-
(No.2020-1-6011) to HZ. We thank the authors who analysis of randomized controlled trials. Aging Clin Exp
provided the original article and information during Res 34, 261-276.
[10] Ryan SM, Kelly ÁM (2016) Exercise as a pro-cognitive,
the screening and analysis. pro-neurogenic and anti-inflammatory intervention in trans-
Authors’ disclosures available online (https:// genic mouse models of Alzheimer’s disease. Ageing Res Rev
www.j-alz.com/manuscript-disclosures/22-0243r2). 27, 77-92.
[11] Lu X, Moeini M, Li B, de Montgolfier O, Lu Y, Bélanger
S, Thorin É, Lesage F (2020) Voluntary exercise increases
SUPPLEMENTARY MATERIAL brain tissue oxygenation and spatially homogenizes oxygen
delivery in a mouse model of Alzheimer’s disease. Neuro-
The supplementary material is available in the biol Aging 88, 11-23.
[12] Choi DH, Kwon KC, Hwang DJ, Koo JH, Um HS, Song HS,
electronic version of this article: https://dx.doi.org/ Kim JS, Jang Y, Cho JY (2021) Treadmill exercise alleviates
10.3233/JAD-220243. brain iron dyshomeostasis accelerating neuronal amyloid-␤
production, neuronal cell death, and cognitive impairment
in transgenic mice model of Alzheimer’s disease. Mol Neu-
REFERENCES robiol 58, 3208-3223.
[13] Liu Y, Chu JMT, Yan T, Zhang Y, Chen Y, Chang
[1] Fazio S, Pace D, Maslow K, Zimmerman S, Kallmyer B RCC, Wong GTC (2020) Short-term resistance exercise
(2018) Alzheimer’s Association dementia care practice rec- inhibits neuroinflammation and attenuates neuropatho-
ommendations. Gerontologist 58, S1-S9. logical changes in 3xTg Alzheimer’s disease mice. J
[2] Petersen RC (2004) Mild cognitive impairment as a diag- Neuroinflammation 17, 4.
nostic entity. J Intern Med 256, 183-194. [14] Nakanishi K, Sakakima H, Norimatsu K, Otsuka S, Takada
[3] Petersen RC, Caracciolo B, Brayne C, Gauthier S, Jelic V, S, Tani A, Kikuchi K (2021) Effect of low-intensity motor
Fratiglioni L (2014) Mild cognitive impairment: A concept balance and coordination exercise on cognitive functions,
in evolution. J Intern Med 275, 214-228. hippocampal A␤ deposition, neuronal loss, neuroinflamma-
[4] Albert MS, DeKosky ST, Dickson D, Dubois B, Feldman tion, and oxidative stress in a mouse model of Alzheimer’s
HH, Fox NC, Gamst A, Holtzman DM, Jagust WJ, Petersen disease. Exp Neurol 337, 113590.
RC, Snyder PJ, Carrillo MC, Thies B, Phelps CH (2011) The [15] Kim TW, Park SS, Park JY, Park HS (2020) Infusion of
diagnosis of mild cognitive impairment due to Alzheimer’s plasma from exercised mice ameliorates cognitive dys-
disease: Recommendations from the National Institute on function by increasing hippocampal neuroplasticity and
Aging-Alzheimer’s Association workgroups on diagnostic mitochondrial functions in 3xTg-AD mice. Int J Mol Sci
guidelines for Alzheimer’s disease. Alzheimers Dement 7, 21, 3291.
270-279. [16] Garcı́a-Mesa Y, Colie S, Corpas R, Cristòfol R, Comellas F,
[5] Liss JL, Seleri Assuncao S, Cummings J, Atri A, Geld- Nebreda AR, Giménez-Llort L, Sanfeliu C (2016) Oxidative
macher DS, Candela SF, Devanand DP, Fillit HM, Susman stress is a central target for physical exercise neuroprotection
J, Mintzer J, Bittner T, Brunton SA, Kerwin DR, Jackson against pathological brain aging. J Gerontol A Biol Sci Med
WC, Small GW, Grossberg GT, Clevenger CK, Cotter V, Sci 71, 40-49.
Stefanacci R, Wise-Brown A, Sabbagh MN (2021) Practical [17] Pang R, Wang X, Pei F, Zhang W, Shen J, Gao X, Chang
recommendations for timely, accurate diagnosis of symp- C (2019) Regular exercise enhances cognitive function
tomatic Alzheimer’s disease (MCI and dementia) in primary and intracephalic GLUT expression in Alzheimer’s disease
care: A review and synthesis. J Intern Med 290, 310-334. model mice. J Alzheimers Dis 72, 83-96.
[6] Petersen RC, Lopez O, Armstrong MJ, Getchius TSD, Gan- [18] Binda KH, Lillethorup TP, Real CC, Bærentzen SL, Nielsen
guli M, Gloss D, Gronseth GS, Marson D, Pringsheim T, MN, Orlowski D, Brooks DJ, Chacur M, Landau AM (2021)
Day GS, Sager M, Stevens J, Rae-Grant A (2018) Prac- Exercise protects synaptic density in a rat model of Parkin-
tice guideline update summary: Mild cognitive impairment: son’s disease. Exp Neurol 342, 113741.
Report of the Guideline Development, Dissemination, and [19] Borjesson M, Onerup A, Lundqvist S, Dahlof B (2016)
Implementation Subcommittee of the American Academy Physical activity and exercise lower blood pressure in indi-
of Neurology. Neurology 90, 126-135. viduals with hypertension: Narrative review of 27 RCTs. Br
[7] Karssemeijer EGA, Aaronson JA, Bossers WJ, Smits T, J Sports Med 50, 356-361.
Olde Rikkert MGM, Kessels RPC (2017) Positive effects [20] O’Malley PG, Arnold MJ, Kelley C, Spacek L, Buelt A,
of combined cognitive and physical exercise training on Natarajan S, Donahue MP, Vagichev E, Ballard-Hernandez
 H. Li et al. / Review of Exercise Treatment on MCI 1277

J, Logan A, Thomas L, Ritter J, Neubauer BE, Downs increases hippocampal subfield volumes in younger adults
JR (2020) Management of dyslipidemia for cardiovascu- and prevents volume decline in the elderly. Brain Imaging
lar disease risk reduction: Synopsis of the 2020 updated Behav 14, 1577-1587.
U.S. Department of Veterans Affairs and U.S. Department [35] Erickson KI, Voss MW, Prakash RS, Basak C, Szabo A,
of Defense Clinical Practice Guideline. Ann Intern Med 173, Chaddock L, Kim JS, Heo S, Alves H, White SM, Wojcicki
822-829. TR, Mailey E, Vieira VJ, Martin SA, Pence BD, Woods JA,
[21] Balducci S, Sacchetti M, Haxhi J, Orlando G, D’Errico V, McAuley E, Kramer AF (2011) Exercise training increases
Fallucca S, Menini S, Pugliese G (2014) Physical exercise size of hippocampus and improves memory. Proc Natl Acad
as therapy for type 2 diabetes mellitus. Diabetes Metab Res Sci U S A 108, 3017-3022.
Rev 30, 13-23. [36] Thomas AG, Dennis A, Rawlings NB, Stagg CJ, Matthews
[22] Yarizadeh H, Eftekhar R, Anjom-Shoae J, Speakman JR, L, Morris M, Kolind SH, Foxley S, Jenkinson M, Nichols
Djafarian K (2021) The effect of aerobic and resistance TE, Dawes H, Bandettini PA, Johansen-Berg H (2016)
training and combined exercise modalities on subcutaneous Multi-modal characterization of rapid anterior hippocampal
abdominal fat: A systematic review and meta-analysis of volume increase associated with aerobic exercise. Neuroim-
randomized clinical trials. Adv Nutr 12, 179-196. age 131, 162-170.
[23] Aengevaeren VL, Mosterd A, Sharma S, Prakken NHJ, [37] Horowitz AM, Fan X, Bieri G, Smith LK, Sanchez-Diaz CI,
Möhlenkamp S, Thompson PD, Velthuis BK, Eijsvo- Schroer AB, Gontier G, Casaletto KB, Kramer JH, Williams
gels TMH (2020) Exercise and coronary atherosclerosis: KE, Villeda SA (2020) Blood factors transfer beneficial
Observations, explanations, relevance, clinical manage- effects of exercise on neurogenesis and cognition to the aged
ment. Circulation 141, 1338-1350. brain. Science 369, 167-173.
[24] Grabs V, Peres T, Zelger O, Haller B, Pressler A, Braun S,
[38] Wayne PM, Walsh JN, Taylor-Piliae RE, Wells RE, Papp
Halle M, Scherr J (2015) Decreased prevalence of cardiac
KV, Donovan NJ, Yeh GY (2014) Effect of tai chi on cog-
arrhythmias during and after vigorous and prolonged exer-
nitive performance in older adults: Systematic review and
cise in healthy male marathon runners. Am Heart J 170,
meta-analysis. J Am Geriatr Soc 62, 25-39.
149-155.
[39] Wu C, Yi Q, Zheng X, Cui S, Chen B, Lu L, Tang C
[25] Kalaria RN, Maestre GE, Arizaga R, Friedland RP, Galasko
(2019) Effects of mind-body exercises on cognitive func-
D, Hall K, Luchsinger JA, Ogunniyi A, Perry EK, Potoc-
tion in older adults: A meta-analysis. J Am Geriatri Soc 67,
nik F, Prince M, Stewart R, Wimo A, Zhang ZX, Antuono
749-758.
P; World Federation of Neurology Dementia Research
Group (2008) Alzheimer’s disease and vascular dementia in [40] Bhattacharyya KK, Andel R, Small BJ (2021) Effects of
developing countries: Prevalence, management, risk factors. yoga-related mind-body therapies on cognitive function in
Lancet Neurol 7, 812-826. older adults: A systematic review with meta-analysis. Arch
[26] Niewada M, Michel P (2016) Lifestyle modification for Gerontol Geriatr 93, 104319.
stroke prevention: Facts and fiction. Curr Opin Neurol 29, [41] Zhou XL, Wang LN, Wang J, Zhou L, Shen XH (2020)
9-13. Effects of exercise interventions for specific cognitive
[27] Kelly ME, Loughrey D, Lawlor BA, Robertson IH, Walsh domains in old adults with mild cognitive impairment: A
C, Brennan S (2014) The impact of exercise on the cognitive meta-analysis and subgroup analysis of randomized con-
functioning of healthy older adults: A systematic review and trolled trials. Medicine 99, e20105.
meta-analysis. Ageing Res Rev 16, 12-31. [42] Biazus-Sehn LF, Schuch FB, Firth J, Stigger FdS (2020)
[28] Chen FT, Etnier JL, Chan KH, Chiu PK, Hung TM, Chang Effects of physical exercise on cognitive function of
YK (2020) Effects of exercise training interventions on older adults with mild cognitive impairment: A system-
executive function in older adults: A systematic review and atic review and meta-analysis. Arch Gerontol Geriatr 89,
meta-analysis. Sports Med 50, 1451-1467. 104048.
[29] Northey JM, Cherbuin N, Pumpa KL, Smee DJ, Rat- [43] Gates N, Fiatarone Singh MA, Sachdev PS, Valenzuela M
tray B (2018) Exercise interventions for cognitive function (2013) The effect of exercise training on cognitive func-
in adults older than 50: A systematic review with meta- tion in older adults with mild cognitive impairment: A
analysis. Br J Sports Med 52, 154-160. meta-analysis of randomized controlled trials. Am J Geriatr
[30] Marinus N, Hansen D, Feys P, Meesen R, Timmermans A, Psychiatry 21, 1086-1097.
Spildooren J (2019) The impact of different types of exercise [44] Sofi F, Valecchi D, Bacci D, Abbate R, Gensini GF, Casini
training on peripheral blood brain-derived neurotrophic fac- A, Macchi C (2011) Physical activity and risk of cognitive
tor concentrations in older adults: A meta-analysis. Sports decline: A meta-analysis of prospective studies. J Intern
Med 49, 1529-1546. Med 269, 107-117.
[31] Wang R, Holsinger RMD (2018) Exercise-induced [45] Higgins J, Green SR (2011) Cochrane Collaboration.
brain-derived neurotrophic factor expression: Therapeutic Cochrane Handbook for Systematic Reviews of Interven-
implications for Alzheimer’s dementia. Ageing Res Rev 48, tions. Chichester, England.
109-121. [46] Higgins JP, Green S (2008) Cochrane Handbook for System-
[32] Rogge AK, Roder B, Zech A, Hotting K (2018) Exercise- atic Reviews of Interventions. The Cochrane Collaboration.
induced neuroplasticity: Balance training increases cortical [47] Cumpston M, Li T, Page MJ, Chandler J, Welch VA, Higgins
thickness in visual and vestibular cortical regions. Neuroim- JP, Thomas J (2019) Updated guidance for trusted system-
age 179, 471-479. atic reviews: A new edition of the Cochrane Handbook for
[33] Colcombe SJ, Erickson KI, Scalf PE, Kim JS, Prakash R, Systematic Reviews of Interventions. Cochrane Database
McAuley E, Elavsky S, Marquez DX, Hu L, Kramer AF Syst Rev 10, ED000142.
(2006) Aerobic exercise training increases brain volume in [48] Law LLF, Mok VCT, Yau MMK (2019) Effects of functional
aging humans. J Gerontol A Biol Sci Med Sci 61, 1166-1170. tasks exercise on cognitive functions of older adults with
[34] Frodl T, Strehl K, Carballedo A, Tozzi L, Doyle M, Amico F, mild cognitive impairment: A randomized controlled pilot
Gormley J, Lavelle G, O’Keane V (2020) Aerobic exercise trial. Alzheimers Res Ther 11, 98.
1278 H. Li et al. / Review of Exercise Treatment on MCI

[49] Combourieu Donnezan L, Perrot A, Belleville S, Bloch F, (Tai Chi) with stretching and toning exercise on cognitive
Kemoun G (2018) Effects of simultaneous aerobic and cog- function in older Chinese adults at risk of cognitive decline.
nitive training on executive functions, cardiovascular fitness J Am Med Dir Assoc 13, 568.e15-568.e20.
and functional abilities in older adults with mild cognitive [63] Angevaren M, Aufdemkampe G, Verhaar HJJ, Aleman
impairment. Ment Health Phys Act 15, 78-87. A, Vanhees L (2008) Physical activity and enhanced fit-
[50] Yogev-Seligmann G, Eisenstein T, Ash E, Giladi N, Sharon ness to improve cognitive function in older people without
H, Nachman S, Bregman N, Kodesh E, Hendler T, Lerner known cognitive impairment. Cochrane Database Syst Rev,
Y (2021) Neurocognitive plasticity is associated with CD005381.
cardiorespiratory fitness following physical exercise in [64] Clouston SA, Brewster P, Kuh D, Richards M, Cooper
older adults with amnestic mild cognitive impairment. J R, Hardy R, Rubin MS, Hofer SM (2013) The dynamic
Alzheimers Dis 81, 91-112. relationship between physical function and cognition in lon-
[51] Tsai CL, Pai MC, Ukropec J, Ukropcová B (2019) Dis- gitudinal aging cohorts. Epidemiol Rev 35, 33-50.
tinctive effects of aerobic and resistance exercise modes [65] Chen ML, Wotiz SB, Banks SM, Connors SA, Shi Y (2021)
on neurocognitive and biochemical changes in individuals Dose-response association of Tai Chi and cognition among
with mild cognitive impairment. Curr Alzheimer Res 16, community-dwelling older adults: A systematic review and
316-332. meta-analysis. Int J Environ Res Public Health 18, 3179.
[52] Scherder EJ, Van Paasschen J, Deijen JB, Van Der Knokke [66] Sanders LMJ, Hortobágyi T, la Bastide-van Gemert S, van
S, Orlebeke JF, Burgers I, Devriese PP, Swaab DF, Sergeant der Zee EA, van Heuvelen MJG (2019) Dose-response rela-
JA (2005) Physical activity and executive functions in the tionship between exercise and cognitive function in older
elderly with mild cognitive impairment. Aging Ment Health adults with and without cognitive impairment: A systematic
9, 272-280. review and meta-analysis. PloS One 14, e0210036.
[53] Nagamatsu LS, Chan A, Davis JC, Beattie BL, Graf P, Voss [67] de Souto Barreto P, Demougeot L, Vellas B, Rolland Y
MW, Sharma D, Liu-Ambrose T (2013) Physical activity (2018) Exercise training for preventing dementia, mild
improves verbal and spatial memory in older adults with cognitive impairment, and clinically meaningful cognitive
probable mild cognitive impairment: A 6-month random- decline: A systematic review and meta-analysis. J Gerontol
ized controlled trial. J Aging Res 2013, 861893. A Biol Sci Med Sci 73, 1504-1511.
[54] Damirchi A, Hosseini F, Babaei P (2018) Mental training [68] Zhang Q, Hu J, Wei L, Cao R, Ma R, Song H, Jin Y (2019)
enhances cognitive function and BDNF more than either Effects of traditional Chinese exercise on cognitive and
physical or combined training in elderly women with MCI: psychological outcomes in older adults with mild cogni-
A small-scale study. Am J Alzheimers Dis Other Demen 33, tive impairment: A systematic review and meta-analysis.
20-29. Medicine 98, e14581.
[55] Hong SG, Kim JH, Jun TW (2018) Effects of 12-week [69] Lee J (2020) Effects of aerobic and resistance exercise inter-
resistance exercise on electroencephalogram patterns and ventions on cognitive and physiologic adaptations for older
cognitive function in the elderly with mild cognitive impair- adults with mild cognitive impairment: A systematic review
ment: A randomized controlled trial. Clin J Sport Med 28, and meta-analysis of randomized control trials. Int J Environ
500-508. Res Public Health 17, 9216.
[56] Lee DW, Yoon DH, Lee JY, Panday SB, Park J, Song W [70] Zhou HX, Chen X, Shen YQ, Li L, Chen NX, Zhu ZC,
(2020) Effects of high-speed power training on neuromus- Castellanos FX, Yan CG (2020) Rumination and the default
cular and gait functions in frail elderly with mild cognitive mode network: Meta-analysis of brain imaging studies and
impairment despite blunted executive functions: A random- implications for depression. Neuroimage 206, 116287.
ized controlled trial. J Frailty Aging 9, 179-184. [71] Lin R, Cui S, Yang J, Yang H, Feng Z, Wahner-Roedler DL,
[57] Yoon DH, Lee JY, Song W (2018) Effects of resistance exer- Zhou X, Salinas M, Mallory MJ, Do A, Bublitz SE, Chon
cise training on cognitive function and physical performance TY, Tang C, Bauer BA, Xu M (2021) Effects of Tai Chi
in cognitive frailty: A randomized controlled trial. J Nutr on patients with mild cognitive impairment: A systematic
Health Aging 22, 944-951. review and meta-analysis of randomized controlled trials.
[58] Lu J, Sun M, Liang L, Feng Y, Pan X, Liu Y (2016) Effects BioMed Res Int 2021, 5530149.
of momentum-based dumbbell training on cognitive func- [72] Talukdar T, Nikolaidis A, Zwilling CE, Paul EJ, Hillman
tion in older adults with mild cognitive impairment: A pilot CH, Cohen NJ, Kramer AF, Barbey AK (2018) Aerobic fit-
randomized controlled trial. Clin Interv Aging 11, 9-16. ness explains individual differences in the functional brain
[59] Lam LC, Chan WC, Leung T, Fung AW, Leung EM (2015) connectome of healthy young adults. Cereb Cortex 28,
Would older adults with mild cognitive impairment adhere 3600-3609.
to and benefit from a structured lifestyle activity intervention [73] Kujach S, Byun K, Hyodo K, Suwabe K, Fukuie T,
to enhance cognition? A cluster randomized controlled trial. Laskowski R, Dan I, Soya H (2018) A transferable
PLoS One 10, e0118173. high-intensity intermittent exercise improves executive
[60] Kasai JY, Busse AL, Magaldi RM, Soci MA, Rosa Pde performance in association with dorsolateral prefrontal acti-
M, Curiati JA, Jacob Filho W (2010) Effects of Tai Chi vation in young adults. Neuroimage 169, 117-125.
Chuan on cognition of elderly women with mild cognitive [74] Northey JM, Rattray B, Pumpa KL, Pryor DJ, Fraser MA,
impairment. Einstein (Sao Paulo) 8, 40-45. Shaw ME, Anstey KJ, Cherbuin N (2020) Objectively
[61] Sungkarat S, Boripuntakul S, Chattipakorn N, Watcharasak- measured physical activity is associated with dorsolateral
silp K, Lord SR (2017) Effects of Tai Chi on cognition and prefrontal cortex volume in older adults. Neuroimage 221,
fall risk in older adults with mild cognitive impairment: A 117150.
randomized controlled trial. J Am Geriatr Soc 65, 721-727. [75] Wei GX, Dong HM, Yang Z, Luo J, Zuo XN (2014) Tai Chi
[62] Lam LC, Chau RC, Wong BM, Fung AW, Tam CW, Leung Chuan optimizes the functional organization of the intrin-
GT, Kwok TC, Leung TY, Ng SP, Chan WM (2012) A 1-year sic human brain architecture in older adults. Front Aging
randomized controlled trial comparing mind body exercise Neurosci 6, 74.