Young Et Al-2015-The Cochrane Library

Cochrane Database of Systematic Reviews
Aerobic exercise to improve cognitive function in older people

without known cognitive impairment (Review)
Young J, Angevaren M, Rusted J, Tabet N
Young J, Angevaren M, Rusted J, Tabet N.

Aerobic exercise to improve cognitive function in older people without known cognitive impairment.
Cochrane Database of Systematic Reviews 2015, Issue 4. Art. No.: CD005381.
DOI: 10.1002/14651858.CD005381.pub4.
www.cochranelibrary.com
Aerobic exercise to improve cognitive function in older people without known cognitive impairment (Review)
Copyright © 2015 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
TABLE OF CONTENTS
HEADER . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
ABSTRACT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
PLAIN LANGUAGE SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
BACKGROUND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
OBJECTIVES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
METHODS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Figure 1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Figure 2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Figure 3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Figure 4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Figure 5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Figure 6. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Figure 7. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Figure 8. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Figure 9. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Figure 10. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
Figure 11. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
Figure 12. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
Figure 13. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
Figure 14. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
Figure 15. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
Figure 16. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Figure 17. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Figure 18. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
Figure 19. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
Figure 20. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Figure 21. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Figure 22. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
Figure 23. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
Figure 24. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Figure 25. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
DISCUSSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
AUTHORS’ CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
ACKNOWLEDGEMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
CHARACTERISTICS OF STUDIES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
DATA AND ANALYSES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
Analysis 1.1. Comparison 1 Aerobic exercise versus any active intervention, Outcome 1 Cognitive speed. . . . . 66
Analysis 1.2. Comparison 1 Aerobic exercise versus any active intervention, Outcome 2 Verbal memory functions
(immediate). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
Analysis 1.3. Comparison 1 Aerobic exercise versus any active intervention, Outcome 3 Visual memory functions
(immediate). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
Analysis 1.4. Comparison 1 Aerobic exercise versus any active intervention, Outcome 4 Working memory. . . . . 69
Analysis 1.5. Comparison 1 Aerobic exercise versus any active intervention, Outcome 5 Memory functions (delayed). 70
Analysis 1.6. Comparison 1 Aerobic exercise versus any active intervention, Outcome 6 Executive functions. . . . 71
Analysis 1.7. Comparison 1 Aerobic exercise versus any active intervention, Outcome 7 Perception. . . . . . . 72
Analysis 1.8. Comparison 1 Aerobic exercise versus any active intervention, Outcome 8 Cognitive inhibition. . . . 73
Analysis 1.9. Comparison 1 Aerobic exercise versus any active intervention, Outcome 9 Visual attention. . . . . 74
Analysis 1.10. Comparison 1 Aerobic exercise versus any active intervention, Outcome 10 Auditory attention. . . . 75
Analysis 1.11. Comparison 1 Aerobic exercise versus any active intervention, Outcome 11 Motor function. . . . . 76
Aerobic exercise to improve cognitive function in older people without known cognitive impairment (Review) i
Analysis 1.12. Comparison 1 Aerobic exercise versus any active intervention, Outcome 12 Drop-out. . . . . . . 77
Analysis 2.1. Comparison 2 Aerobic exercise versus no intervention, Outcome 1 Cognitive speed. . . . . . . . 78
Analysis 2.2. Comparison 2 Aerobic exercise versus no intervention, Outcome 2 Verbal memory functions (immediate). 79
Analysis 2.3. Comparison 2 Aerobic exercise versus no intervention, Outcome 3 Visual memory functions (immediate). 80
Analysis 2.4. Comparison 2 Aerobic exercise versus no intervention, Outcome 4 Working memory. . . . . . . 81
Analysis 2.5. Comparison 2 Aerobic exercise versus no intervention, Outcome 5 Memory functions (delayed). . . . 82
Analysis 2.6. Comparison 2 Aerobic exercise versus no intervention, Outcome 6 Executive functions. . . . . . . 83
Analysis 2.7. Comparison 2 Aerobic exercise versus no intervention, Outcome 7 Cognitive inhibition. . . . . . 84
Analysis 2.8. Comparison 2 Aerobic exercise versus no intervention, Outcome 8 Visual attention. . . . . . . . 85
Analysis 2.9. Comparison 2 Aerobic exercise versus no intervention, Outcome 9 Auditory attention. . . . . . . 86
Analysis 2.10. Comparison 2 Aerobic exercise versus no intervention, Outcome 10 Motor function. . . . . . . 87
Analysis 2.11. Comparison 2 Aerobic exercise versus no intervention, Outcome 11 Drop-out. . . . . . . . . 88
Analysis 3.1. Comparison 3 Aerobic exercise versus flexibility/balance programme, Outcome 1 Cognitive speed. . . 89
Analysis 3.2. Comparison 3 Aerobic exercise versus flexibility/balance programme, Outcome 2 Verbal memory functions
(immediate). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
Analysis 3.3. Comparison 3 Aerobic exercise versus flexibility/balance programme, Outcome 3 Visual memory functions
(immediate). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
Analysis 3.4. Comparison 3 Aerobic exercise versus flexibility/balance programme, Outcome 4 Working memory. . 92
Analysis 3.5. Comparison 3 Aerobic exercise versus flexibility/balance programme, Outcome 5 Memory functions
(delayed). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
Analysis 3.6. Comparison 3 Aerobic exercise versus flexibility/balance programme, Outcome 6 Executive functions. . 94
Analysis 3.7. Comparison 3 Aerobic exercise versus flexibility/balance programme, Outcome 7 Perception. . . . . 95
Analysis 3.8. Comparison 3 Aerobic exercise versus flexibility/balance programme, Outcome 8 Cognitive inhibition. 96
Analysis 3.9. Comparison 3 Aerobic exercise versus flexibility/balance programme, Outcome 9 Visual attention. . . 97
Analysis 3.10. Comparison 3 Aerobic exercise versus flexibility/balance programme, Outcome 10 Auditory attention. 98
Analysis 3.11. Comparison 3 Aerobic exercise versus flexibility/balance programme, Outcome 11 Motor function. . 99
Analysis 3.12. Comparison 3 Aerobic exercise versus flexibility/balance programme, Outcome 12 Drop-out. . . . 100
Analysis 4.1. Comparison 4 Aerobic exercise versus strength programme, Outcome 1 Verbal memory functions
(immediate). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Analysis 4.2. Comparison 4 Aerobic exercise versus strength programme, Outcome 2 Executive functions. . . . . 102
Analysis 4.3. Comparison 4 Aerobic exercise versus strength programme, Outcome 3 Perception. . . . . . . . 103
Analysis 4.4. Comparison 4 Aerobic exercise versus strength programme, Outcome 4 Cognitive speed. . . . . . 104
Analysis 5.1. Comparison 5 Fitness Improved: aerobic exercise versus any active intervention, Outcome 1 Cognitive
speed. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
Analysis 5.2. Comparison 5 Fitness Improved: aerobic exercise versus any active intervention, Outcome 2 Verbal memory
functions (immediate). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
Analysis 5.3. Comparison 5 Fitness Improved: aerobic exercise versus any active intervention, Outcome 3 Visual memory
functions (immediate). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
Analysis 5.4. Comparison 5 Fitness Improved: aerobic exercise versus any active intervention, Outcome 4 Working
memory. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
Analysis 5.5. Comparison 5 Fitness Improved: aerobic exercise versus any active intervention, Outcome 5 Memory
functions (delayed). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
Analysis 5.6. Comparison 5 Fitness Improved: aerobic exercise versus any active intervention, Outcome 6 Executive
functions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
Analysis 5.7. Comparison 5 Fitness Improved: aerobic exercise versus any active intervention, Outcome 7 Perception. 111
Analysis 5.8. Comparison 5 Fitness Improved: aerobic exercise versus any active intervention, Outcome 8 Cognitive
inhibition. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
Analysis 5.9. Comparison 5 Fitness Improved: aerobic exercise versus any active intervention, Outcome 9 Visual
attention. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
Analysis 5.10. Comparison 5 Fitness Improved: aerobic exercise versus any active intervention, Outcome 10 Auditory
attention. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
Analysis 5.11. Comparison 5 Fitness Improved: aerobic exercise versus any active intervention, Outcome 11 Motor
function. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
Aerobic exercise to improve cognitive function in older people without known cognitive impairment (Review) ii
Analysis 5.12. Comparison 5 Fitness Improved: aerobic exercise versus any active intervention, Outcome 12 Drop-out. 116
Analysis 6.1. Comparison 6 Fitness improved: aerobic exercise versus no intervention, Outcome 1 Cognitive speed. . 117
Analysis 6.2. Comparison 6 Fitness improved: aerobic exercise versus no intervention, Outcome 2 Verbal memory functions
(immediate). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
Analysis 6.3. Comparison 6 Fitness improved: aerobic exercise versus no intervention, Outcome 3 Visual memory functions
(immediate). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
Analysis 6.4. Comparison 6 Fitness improved: aerobic exercise versus no intervention, Outcome 4 Working memory. 120
Analysis 6.5. Comparison 6 Fitness improved: aerobic exercise versus no intervention, Outcome 5 Memory functions
(delayed). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
Analysis 6.6. Comparison 6 Fitness improved: aerobic exercise versus no intervention, Outcome 6 Executive functions. 122
Analysis 6.7. Comparison 6 Fitness improved: aerobic exercise versus no intervention, Outcome 7 Cognitive inhibition. 123
Analysis 6.8. Comparison 6 Fitness improved: aerobic exercise versus no intervention, Outcome 8 Visual attention. . 124
Analysis 6.9. Comparison 6 Fitness improved: aerobic exercise versus no intervention, Outcome 9 Auditory attention. 125
Analysis 6.10. Comparison 6 Fitness improved: aerobic exercise versus no intervention, Outcome 10 Motor function. 126
Analysis 6.11. Comparison 6 Fitness improved: aerobic exercise versus no intervention, Outcome 11 Drop-out. . . 127
ADDITIONAL TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
APPENDICES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
WHAT’S NEW . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
HISTORY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
CONTRIBUTIONS OF AUTHORS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
DECLARATIONS OF INTEREST . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
SOURCES OF SUPPORT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
INDEX TERMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
Aerobic exercise to improve cognitive function in older people without known cognitive impairment (Review) iii
[Intervention Review]
Aerobic exercise to improve cognitive function in older people

without known cognitive impairment
Jeremy Young1 , Maaike Angevaren2 , Jennifer Rusted1 , Naji Tabet3

1 School of Psychology, University of Sussex, Brighton, UK. 2 Research Group Lifestyle and Health, University of Applied Sciences,
Utrecht, Netherlands. 3 Postgraduate Medicine, Brighton and Sussex Medical School, Brighton, UK
Contact address: Jeremy Young, School of Psychology, University of Sussex, Brighton, BN1 9QH, UK. J.Young@sussex.ac.uk.
Editorial group: Cochrane Dementia and Cognitive Improvement Group.

Publication status and date: New search for studies and content updated (no change to conclusions), published in Issue 4, 2015.
Review content assessed as up-to-date: 24 August 2013.
Citation: Young J, Angevaren M, Rusted J, Tabet N. Aerobic exercise to improve cognitive function in older people
without known cognitive impairment. Cochrane Database of Systematic Reviews 2015, Issue 4. Art. No.: CD005381. DOI:
10.1002/14651858.CD005381.pub4.
ABSTRACT
Background
There is increasing evidence that physical activity supports healthy ageing. Exercise is helpful for cardiovascular, respiratory and
musculoskeletal systems, among others. Aerobic activity, in particular, improves cardiovascular fitness and, based on recently reported
findings, may also have beneficial effects on cognition among older people.
Objectives
To assess the effect of aerobic physical activity, aimed at improving cardiorespiratory fitness, on cognitive function in older people
without known cognitive impairment.
Search methods
We searched ALOIS - the Cochrane Dementia and Cognitive Improvement Group’s Specialized Register, the Cochrane Controlled
Trials Register (CENTRAL) (all years to Issue 2 of 4, 2013), MEDLINE (Ovid SP 1946 to August 2013), EMBASE (Ovid SP
1974 to August 2013), PEDro, SPORTDiscus, Web of Science, PsycINFO (Ovid SP 1806 to August 2013), CINAHL (all dates to
August 2013), LILACS (all dates to August 2013), World Health Organization (WHO) International Clinical Trials Registry Platform
(ICTRP) (http://apps.who.int/trialsearch), ClinicalTrials.gov (https://clinicaltrials.gov) and Dissertation Abstracts International (DAI)
up to 24 August 2013, with no language restrictions.
Selection criteria
We included all published randomised controlled trials (RCTs) comparing the effect on cognitive function of aerobic physical activity
programmes with any other active intervention, or no intervention, in cognitively healthy participants aged over 55 years.
Data collection and analysis
Two review authors independently extracted the data from included trials. We grouped cognitive outcome measures into eleven categories
covering attention, memory, perception, executive functions, cognitive inhibition, cognitive speed and motor function. We used the
mean difference (or standardised mean difference) between groups as the measure of the treatment effect and synthesised data using a
random-effects model. We conducted separate analyses to compare aerobic exercise interventions with no intervention and with other
exercise, social or cognitive interventions. Also, we performed analyses including only trials in which an increase in the cardiovascular
fitness of participants had been demonstrated.
Aerobic exercise to improve cognitive function in older people without known cognitive impairment (Review) 1
Main results
Twelve trials including 754 participants met our inclusion criteria. Trials were from eight to 26 weeks in duration.
We judged all trials to be at moderate or high risk of bias in at least some domains. Reporting of some risk of bias domains was poor.
Our analyses comparing aerobic exercise to any active intervention showed no evidence of benefit from aerobic exercise in any cognitive
domain. This was also true of our analyses comparing aerobic exercise to no intervention. Analysing only the subgroup of trials in which
cardiorespiratory fitness improved in the aerobic exercise group showed that this improvement did not coincide with improvements
in any cognitive domains assessed. Our subgroup analyses of aerobic exercise versus flexibility or balance interventions also showed no
benefit of aerobic exercise in any cognitive domain.
Dropout rates did not differ between aerobic exercise and control groups. No trial reported on adverse effects.
Overall none of our analyses showed a cognitive benefit from aerobic exercise even when the intervention was shown to lead to improved
cardiorespiratory fitness.
Authors’ conclusions
We found no evidence in the available data from RCTs that aerobic physical activities, including those which successfully improve
cardiorespiratory fitness, have any cognitive benefit in cognitively healthy older adults. Larger studies examining possible moderators
are needed to confirm whether or not aerobic training improves cognition.
PLAIN LANGUAGE SUMMARY
Aerobic exercise to improve cognitive function in older people without known cognitive impairment
Aerobic exercise is beneficial for healthy ageing. It has been suggested that the increased fitness brought about by aerobic exercise may
help to maintain good cognitive function in older age. We looked for randomised controlled trials of aerobic exercise programmes for
people over the age of 55 years, without pre-existing cognitive problems, which measured effects on both fitness and cognition. The
aerobic exercise programmes could be compared with no intervention (e.g. being on a waiting list for the exercise group) or with other
kinds of activity (including non-aerobic exercises such as strength or balance exercises, or social activities).
In this Cochrane Review, 12 trials including 754 participants met our inclusion criteria. Eight of the 12 trials reported that the aerobic
exercise interventions resulted in increased fitness of the trained group. However, when we combined results across the trials, we did
not find any significant benefits of aerobic exercise or increased fitness on any aspect of cognition. Many included trials had problems
with their methods or reporting which reduced our confidence in the findings.
We did not find evidence that aerobic exercise or increased fitness improves cognitive function in older people. However, it remains
possible that it may be helpful for particular subgroups of people, or that more intense exercise programmes could be beneficial.
Therefore further research in this area is necessary.
BACKGROUND crease to over 1.4 billion people. Subjective complaints about cog-
nitive capacities increase with (older) age (Martin 2003; Newson
2006) and an objective decline in cognitive performance accel-
erates around the age of 50 (Salthouse 2003; Verhaeghen 1997),
Description of the condition
with the exception of cognitive skills with a large crystallised in-
In 2005, there were over 925 million people worldwide aged 55 telligence component. Research has shown that a regular exercise
years or older according to the population database of the United programme can slow down or prevent functional decline associ-
Nations (WPP 2006). It is predicted that in 10 years this will in-
ated with ageing and improve health in this age group. The physi- Sturman 2005; van Gelder 2004). However, results from train-
cal health benefits for older people who regularly participate in en- ing studies by Hill 1993 and Blumenthal 1991 failed to correlate
durance, balance and resistance training programmes are well es- changes in aerobic power (VO2 max) with changes in cognitive
tablished. Such health benefits include improved muscle mass, ar- measures. At the same time, trials seldom report combinations of
terial compliance, energy metabolism, cardiovascular fitness, mus- activity, fitness and cognition in a single trial.
cle strength and overall functional capacity (Lemura 2000). It is
suspected that physical activity may also enhance cognitive func-
tion (Colcombe 2003). Why it is important to do this review
Previous meta-analyses have reported a robust effect of physical
activity on cognitive function in older adults (Colcombe 2003;
Description of the intervention Etnier 1997b; Heyn 2004; Smith 2010), but it remains unclear
In this Cochrane Review we included the interventions of exercise whether improvement in cardiovascular fitness (as reflected by car-
programmes for older people which aimed to improve cardiorespi- diovascular parameters such as VO2 max) accounts for the effects
ratory fitness, the ability of the circulatory and respiratory to sup- of physical activity on cognitive capacity. Physiological or psycho-
ply oxygen to muscles during sustained physical activity, through logical mechanisms other than aerobic fitness might still account
for example walking, running or cycling. We compared their ef- for the effects found in these meta-analyses. This Cochrane Re-
fects with a variety of control interventions: either no intervention view intends to investigate a hypothesised link between physical
or exercise interventions which would not be expected to enhance activity specifically aimed at the improvement of cardiorespiratory
cardiorespiratory fitness, such as strength or balance programmes, fitness and cognitive function. Such information will be useful in
or social or mental activities. Cardiorespiratory fitness may be as- the quest to identify interventions that may be helpful for healthy
sessed in a variety of ways. A common method is to measure VO2 ageing and protective against the development of neurodegenera-
max, which is the maximal oxygen uptake measured during exer- tive disorders such as Alzheimer’s disease.
cise on a treadmill or cycle, although other physiological measures
or walk times may also be used.
OBJECTIVES
How the intervention might work To assess the effectiveness of physical activity, aimed at improving
Research using animal models has provided insight into the pos- cardiorespiratory fitness, on cognitive function in older people
sible cellular and molecular mechanisms that could underlie an without known cognitive impairment.
effect of physical activity on cognitive function. Increased aerobic
fitness increases oxygen extraction, glucose utilisation and cerebral
blood flow (Churchill 2002). Cerebral blood flow meets metabolic METHODS
needs of the brain and removes waste (Lojovich 2010). Increased
aerobic fitness also increases Brain-Derived Neurotrophic Fac-
tor (BDNF) and other growth factors which mediate structural
Criteria for considering studies for this review
changes (Cotman 2002; Cotman 2007). For example, BDNF is
implicated in neurogenesis, synaptogenesis, dendritic branching
and neuroprotection (Lojovich 2010). A preliminary survey of
the literature on human research points towards the same possi- Types of studies
ble physiological mechanisms that could explain the association We only included randomised controlled clinical trials (RCTs).
between physical activity and cognitive vitality (Aleman 2000; Blinding of outcome assessors was not required for inclusion in
Brown 2008; Colcombe 2006; Davenport 2012; Erickson 2009; this review. We did not apply any language restrictions but trials
McAuley 2004; Prins 2002). Hence it is hypothesised that im- must have been published in peer-reviewed journals.
provements in cardiovascular (aerobic) fitness mediate the benefits
of physical activity on cognitive capacity (Etnier 2007; McAuley
2004). Therefore this cardiovascular fitness hypothesis implies that Types of participants
changes in cognitive function are preceded by changes in aerobic Participants were aged 55 or older and not objectively cogni-
fitness. The evidence for this hypothetical link between physical tively impaired in any way greater than that expected from age
activity, cardiovascular fitness and cognitive function in older peo- alone. Hence, we excluded patients with mild cognitive impair-
ple comes from several longitudinal studies (Abbott 2004; Barnes ment (MCI) or any form of dementia and patients with other con-
2003; Etgen 2010; Laurin 2001; Middleton 2011; Richards 2003; ditions likely to be associated with cognitive impairment, such as
stroke and depression. However, we included trials of participants Search methods for identification of studies
with age-related illnesses (e.g. osteoporosis, arthrosis) or specific
disorders (e.g. chronic obstructive pulmonary disease (COPD),
heart failure). Electronic searches
We searched ALOIS - the Cochrane Dementia and Cognitive Im-
provement Group’s Specialized Register, Cochrane Central Reg-
Types of interventions ister of Controlled Trials (CENTRAL), MEDLINE (1946 to
August 2013), EMBASE (Ovid SP 1974 to August 2013), PE-
We included the physical activity interventions of any programme Dro, SPORTDiscus, Web of Science (Web of Science platform),
of exercise of any intensity, duration or frequency which was aimed PsycINFO (Ovid SP 1806 to August 2013), CINAHL (EBSCO-
at improving cardiorespiratory fitness. Therefore, trials must have host), LILACS (BIREME), World Health Organization (WHO)
reported at least one objective measure of cardiorespiratory fit- International Clinical Trials Registry Platform (ICTRP) (http:
ness. Acceptable comparator interventions were: no treatment; a //apps.who.int/trialsearch), ClinicalTrials.gov (https://clinicaltri-
strength or balance programme; or a programme of social activities als.gov) and Dissertation Abstracts International (DAI) up to 24
or mental activities. Trials which had both an active comparator August 2013 with no language restrictions.
group and a no treatment group could contribute data to the ’aer- We used a combination of MeSH and free text terms to find records
obic exercise vs. any active intervention’ meta-analyses and to the of physical activity, including: exercise*, motor activit*, leisure
’aerobic exercise vs. no intervention’ meta-analyses. activit*, physical fitness, physical endurance, exercise tolerance,
exercise test, aerobic, aerobic capacity, physical activity, physical
capacity, physical performance, training. We have listed the search
strategy details in Appendix 1.
Types of outcome measures
We performed a further search update up to November 2014.
Trials had to report an objective measure of cardiorespiratory fit- We have inserted the search results into the Studies awaiting
ness. Acceptable measures included, but were not limited to: VO2 classification section and will fully incorporate these trials in the
max, Graded Exercise Test (GXT) rate-pressure product, heart next review update.
rate and blood pressure during modified step test, the Six-Minute
Walk Test (6MWT), 400-metre walk time, and ¼ mile walk time.
Where trials measured more than one fitness parameter, we pre- Searching other resources
ferred the measure that we considered to be the purest measure of We checked reference lists of the included trials and in reviews of
cardiorespiratory fitness, or was previously show to be correlated the literature screened for relevant trials. Also we contacted experts
with VO2 max, or both. in this area and relevant associations.
Data collection and analysis

Primary outcomes
The primary outcome measurement was cognitive function, tested

with a neuropsychological test (sensitive to changes in cognitive Selection of studies
function in adults) or test battery (a combination of several neu- The Cochrane Trials Search Coordinator (ANS) assessed the titles
ropsychological tests). and available abstracts of all trials identified by the initial search
and excluded irrelevant trials. Two review authors (JY and NT; or
MA and GA previously) independently assessed full paper copies
Secondary outcomes of reports of potentially relevant trials. We resolved any disagree-
ments on inclusion by discussion and through arbitration by a
Other outcome measures were drop-out, as a measure of accept- third review author (JR). Details of the study selection process can
ability, and adverse events. be found in Figure 1.
Figure 1. Study flow diagram for the August 2013 update search
quence generation, allocation concealment, blinding, incomplete
Data extraction and management
outcome data, selective outcome reporting and other issues). For
Two review authors (JY and NT) independently extracted data each trial the six domains are analysed, described as reported in
from the published reports and JY entered them into RevMan the trial and a final judgment on the likelihood of bias is pro-
2014, with full agreement of the second review author. The sum- vided. This is achieved by answering a pre-specified question about
mary statistics required for each trial and each outcome for con- the adequacy of the trial in relation to each domain, such that a
tinuous data were the mean (or mean change from baseline), the judgement of “yes” indicates low risk of bias, “no” indicates high
standard deviation (SD) and the number of participants for each risk of bias, and “unclear” indicates unclear or unknown risk of
treatment group at each assessment. For cognitive data in which a bias. To make these judgments we used the criteria indicated by
higher score denotes worse performance (e.g. reaction times, digit the Cochrane Handbook for Systematic Reviews of Interventions (see
vigilance, trail making part A, trail making part B, Stroop inter- Higgins 2011 for a detailed description) and their applicability
ference data and error rates), we entered the mean as a negative on the addiction field. We assessed the included trials using the
variable. If only the standard error of the mean was reported, we criteria and the method indicated in Higgins 2011.
calculated the SD using SD = SE x sqrt(N). For dichotomous data,
we extracted the number of participants with each outcome in
each group. Measures of treatment effect
The included articles measured cognitive function using various For continuous outcome data, we used the weighted mean dif-
rating scales. We grouped neuropsychological tests measuring ap- ference (WMD) if trials used the same cognitive tests and if the
proximately the same construct in a total of eleven categories (see outcome measurements were on the same scale. We calculated the
Table 1; Kessels 2000; Lezak 2004). For each trial, only a single standardised mean difference (SMD) in all other cases. For di-
test was admitted to each category. Where a trial used more than chotomous data, such as drop-out, we used the odds ratio (OR).
one test within a category, then first we chose the one which was
used most frequently in the included trials; if not, then the one
that had been found to load onto the category in previous factor Dealing with missing data
analysis (Salthouse 1996) or which we considered closer to the To allow an intention-to-treat (ITT) analysis, we sought data on
core construct of the category. We chose all included tests prior to every participant randomised irrespective of compliance, whether
extraction of results. or not the participant was subsequently deemed ineligible, or oth-
One trial (Blumenthal 1989) reported results for men and women erwise excluded from treatment or follow-up. If ITT data were
separately in the same paper. In this case, we calculated pooled unavailable in the publications, we sought “on-treatment” data or
means and SDs by combining results for both genders. the data of those who completed the trial, where indicated.
Assessment of risk of bias in included studies Data synthesis

Two review authors (JY, NT) independently evaluated the For each cognitive outcome category, we synthesised the data using
methodological quality of the selected articles using two different a random-effects model. We analysed the possible effects of aer-
methods. We used the criteria list for quality assessment of non- obic exercise versus any active comparator (strength programme,
pharmaceutical trials (CLEAR NPT) developed using consensus flexibility or balance programme, social or mental programme)
(Boutron 2005). This checklist includes information on sampling and versus no intervention (usual care or waiting list).
method, measurement, intervention and reporting of biases and
limitations (see Table 2). We performed a small pilot exercise to
clarify the method with some articles that we already excluded Subgroup analysis and investigation of heterogeneity
from the review process. We calculated Cohen’s kappa (K) as a mea- Heterogeneity was low across all domains in all meta-analyses,
sure of inter-observer agreement, and we relied on Landis 1977’s therefore we did not subgroup analyses to explore heterogeneity.
benchmarks for assessing the relative strength of agreement. We In order to explore further the potential effects the different forms
resolved any discordance in assessment through a single round of of exercise, we conducted subgroup analyses which compared aer-
discussion and arbitration by a third review author (JR). obic exercise with (a) flexibility or balance interventions and (b)
We also used the recommended approach for assessing risk of strength training. We further explored our hypothesis by perform-
bias in trials included in Cochrane Reviews, which is based on ing analyses of only those trials in which an increase in fitness was
the evaluation of six specific methodological domains (namely, se- demonstrated.
As an extension to subgroup analyses, a meta-regression would groups (we included this pooled group in the ’exercise versus any
allow the effect of cardiovascular fitness (VO2 max or any other intervention’ analyses). This trial included tests for cognitive speed
measure of the degree of aerobic fitness) on cognitive outcomes to and auditory attention.
be investigated. However, we did not consider meta-regression in Fabre 2002 presented data from 32 participants randomly assigned
this Cochrane Review due to the small number of included trials to an aerobic exercise programme, a mental training programme,
(< eight trials) in all meta-analyses. a combined aerobic/mental programme or a social activity group.
We did not use data from the combined aerobic exercise/mental
training group in this review. There was a significant increase in
VO2 max in the aerobic training group but no change in the other
RESULTS two groups. The trial included tests for verbal and visual memory,
perception and executive functions.
Description of studies Kramer 2001 recruited a total of 174 participants and randomly
assigned participants to an aerobic walking group or a stretch-
ing and toning group. The aerobic walking group improved their
VO2 max measures while the stretching and toning group de-
Results of the search
creased their VO2 max measures. The trial authors assessed cog-
The August 2013 search identified 352 promising abstracts (see nitive speed, verbal and visual memory, perception, executive and
PRIMSA flow diagram). We identified seven potentially relevant motor functions as well as cognitive inhibition, visual and auditory
theses but these had no associated peer-reviewed publications. We attention with various cognitive tests. Mean results of the subtests
asked the authors of the theses to provide information on published of the pursuit rotor task, Rey’s auditory verbal learning test, spatial
data, but none were provided. We examined the full texts of 82 attention and visual search task were summed and divided by the
articles. We identified 2 new trials for inclusion bringing the total number of tasks. SD values of these subtests were pooled.
number of trials included to 12 trials involving 754 participants. Langlois 2012 randomly assigned 83 participants, ensuring gen-
der ratio equivalence, to a 12-week exercise training group or a
Included studies control group that maintained their previous activity levels. Par-
ticipants in the exercise training group improved in physical fit-
We have listed the details of the methods, participants, interven-
ness, as measured by the 6MWT, significantly more than controls.
tions and outcomes for each included trial in the Characteristics of
Outcomes included tests of cognitive speed, verbal and working
included studies table. Also, we have summarized the intervention
memory, executive functions and inhibition.
types in each trial in Table 2.
Legault 2011 published a pilot RCT of 73 participants randomly
Bakken 2001 conducted a small RCT (N = 15) comparing an aer-
assigned to a physical activity training group, a cognitive training
obic exercise group to a waiting list control group for eight weeks.
group, a combined intervention group or a ’healthy aging’ control
Both groups showed slight improvement in a measure of aerobic
group, which we considered an active intervention. We did not
fitness over the course of the trial. The only cognitive outcome
use data from the combined intervention group in this review.
parameter was the accuracy index - a test of visual attention.
The physical activity training group improved in a fitness measure
Blumenthal 1989 randomised 101 participants to aerobic exer-
while the cognitive training and control group did not. Cognitive
cise training, a yoga/flexibility programme or a waiting list control
speed, verbal memory, working memory, executive function and
group over 16 weeks. Participants in the aerobic training group
cognitive inhibition were tested in the participants.
only experienced a significant increase in their VO2 max. Out-
Moul 1995 recruited 30 participants and randomly assigned them
comes included tests of cognitive speed, verbal, visual and work-
to a walking condition, weight training or control condition,
ing memory, executive functions, cognitive inhibition, visual and
which we considered to be a flexibility intervention, for 16 weeks.
auditory attention and motor function.
VO2 max significantly increased in the walking group but not in
Madden 1989 reported different cognitive outcomes for a subset
the weight training or control conditions. The Ross Information
of the participants from Blumenthal 1989. We did not included
Processing Assessment was used to evaluate changes in cognitive
any of the data from this paper in the analyses because Blumenthal
function.
1989 reported data for the same outcome categories.
Oken 2006 randomised 135 participants into an aerobic group, a
Emery 1990a assigned 48 participants to an exercise programme,
yoga group or a waiting list control group for six months. There
a social activity group or a waiting list control group for 12 weeks.
were no significant differences between the groups in their fitness
No effect of the exercise programme on cardiovascular function
measure. Cognitive speed, delayed memory functions, executive
was demonstrated. As attrition from the social group was compa-
functions, visual attention and cognitive inhibition were assessed
rable to that of the control group, and attendance for the social
in order to test for effects on cognition.
group was poor overall (ranged from 10% to 94%), the trial au-
Panton 1990 included data on 49 participants randomly assigned
thors pooled data from the social activity and waiting list control
to a walk/jog group, a strength group or a no intervention control kappa (K) as a measure of inter-observer reliability after the initial
condition for 26 weeks. VO2 max significantly improved for the screening and reached 0.84, almost perfect according to Landis
walk/jog group while there was no significant change for strength 1977.
as well as the control groups. Tests for cognitive speed were per- We have presented the results of our ’Risk of bias’ assessment in
formed to analyse cognitive function. the Characteristics of included studies tables and in Figure 2. We
Whitehurst 1991 recruited 14 participants and randomly assigned only considered one trial to be at low risk of bias for sequence
them to an exercise programme or a no intervention control con- generation (Oken 2006). We judged the remaining 11 trials to
dition for eight weeks. Participants in the exercise group signif- be at unclear risk of bias for sequence generation. Procedures to
icantly increased their VO2 max scores, whereas participants in ensure allocation concealment were not described in the included
the control group did not. Choice reaction times were tested for papers; all 12 papers were judged to be at unclear risk of bias in
evaluation of cognitive function. this domain. In all 12 included trials blinding of participants and
trainers was not feasible. This was unlikely to introduce bias in
trainers, so we considered all 12 trials to be at low risk of bias for
Excluded studies blinding trainers. This may have introduced bias in participants,
We have listed details of excluded trials in the Characteristics of so all 12 trials were judged to be at high risk of blinding of the
excluded studies table. We excluded trials because they were not participants. We judged five trials (Bakken 2001; Legault 2011;
RCTs (19), did not use a cognitively normal older population (11), Oken 2006; Panton 1990; Whitehurst 1991) to be at low risk
did not meet other inclusion criteria (1: Kharti 2001 included de- of bias for blinding of the assessors for the cognitive outcomes
pressed participants), did not have objective aerobic fitness param- because assessment of cognition was by means of computerised
eters (16), did not have objective cognitive outcomes (5), assessed tests. We considered the other seven trials to be at unclear risk of
cognition during exercise (3), did not have pre- to post- interven- bias for this item. Four trials (Fabre 2002; Legault 2011; Moul
tion data (4), did not have a non-aerobic control group (2), had 1995; Whitehurst 1991) were judged to be at low risk of bias
not been published (7), the data was published in an already in- for addressing incomplete data. Besides Legault 2011, in all cases
cluded trial (2), or for other reasons: objective cognitive measures this was due to the fact that there were no drop-outs from these
were not analysed by group (Emery 1990b) or the control group trials. Legault 2011 reported drop-outs per group and analysed
was exercising but not given a formal program (Etnier 2001). using ITT principles. All other eight trials were judged being at
high risk of bias for this item since they reported drop-outs but
either lacked information on the group assignment of these drop-
outs (Panton 1990) or lacked ITT analysis, or both. We judged
Risk of bias in included studies all trials, except Blumenthal 1989, to be at unclear risk of bias
We have presented the results of the quality assessment of non- for selective reporting since there was insufficient information to
pharmaceutical trials (CLEAR NPT) (Boutron 2005) in Table 3. permit a judgment. Blumenthal 1989 was judged being at high
The overall methodological quality score of the included trials risk for this item since data on one pre-specified primary cognitive
ranged from 24 to 39 (minimum possible score of 14 points, outcome was missing. We considered all trials to be at low risk of
maximum possible score of 48 points; lower scores denote a better bias for other potential threats to validity. However, we could not
methodological quality). For most trials, the blinding treatment rule out risk of contamination bias, where the control group, on
providers and participants was scored “no, because blinding is finding out the purpose of a trial, could have increased their levels
not feasible”. Two review authors (JY, NT) calculated Cohen’s of aerobic exercise as well.
Figure 2. Methodological quality summary: review authors’ judgements about each methodological quality
item for each included study.
Effects of interventions
Aerobic exercise versus any active intervention

Eight trials including 506 participants contributed data on at least
one cognitive domain. Duration of the intervention in these trials
ranged from eight weeks to 26.07 weeks. In six trials, trial authors
showed an increase in aerobic fitness in the active intervention but
not the comparison group. We were able to conduct meta-analyses
for all 11 of our pre-specified cognitive domains (Analysis 1.1 to
Analysis 1.11; Figure 3; Figure 4; Figure 5; Figure 6; Figure 7;
Figure 8; Figure 9; Figure 10; Figure 11; Figure 12; Figure 13).
There was no evidence of benefit of the aerobic exercise interven-
tion in any cognitive domain.
Figure 3. Forest plot of comparison: 1 Aerobic exercise versus any active intervention, outcome: 1.1
Cognitive speed.
Figure 4. Forest plot of comparison: 1 Aerobic exercise versus any active intervention, outcome: 1.2 Verbal
memory functions (immediate).
Figure 5. Forest plot of comparison: 1 Aerobic exercise versus any active intervention, outcome: 1.3 Visual
Working memory.
Memory functions (delayed).
Executive functions.
Perception.
Cognitive inhibition.
Figure 11. Forest plot of comparison: 1 Aerobic exercise versus any active intervention, outcome: 1.9 Visual
attention.
Auditory attention.
Motor function.
There was no difference in dropout rates between the aerobic

exercise intervention and comparison groups (OR 0.96, 95% CI
0.44 to 2.10; seven trials, 469 participants; Analysis 1.12; Figure
14).
Drop-out.
Aerobic exercise versus no intervention
Six trials including 296 participants contributed data on at least
one cognitive domain. The duration of the intervention in these
trials ranged from eight to 26.07 weeks. In four trials, trial authors
showed an increase in aerobic fitness in the active intervention
but not the comparison group. We were able to conduct meta-
analyses for 10 of our 11 pre-specified cognitive domains, besides
perception (Analysis 2.1 to Analysis 2.10; Figure 15; Figure 16;
Figure 17; Figure 18; Figure 19; Figure 20; Figure 21; Figure 22;
Figure 23; Figure 24). There was no evidence of benefit of the
aerobic exercise intervention in any cognitive domain.
Figure 15. Forest plot of comparison: 2 Aerobic exercise versus no intervention, outcome: 2.1 Cognitive
speed.
Figure 16. Forest plot of comparison: 2 Aerobic exercise versus no intervention, outcome: 2.2 Verbal
Figure 17. Forest plot of comparison: 2 Aerobic exercise versus no intervention, outcome: 2.3 Visual
Figure 18. Forest plot of comparison: 2 Aerobic exercise versus no intervention, outcome: 2.4 Working
memory.
Figure 19. Forest plot of comparison: 2 Aerobic exercise versus no intervention, outcome: 2.5 Memory
functions (delayed).
Figure 20. Forest plot of comparison: 2 Aerobic exercise versus no intervention, outcome: 2.6 Executive
functions.
Figure 21. Forest plot of comparison: 2 Aerobic exercise versus no intervention, outcome: 2.7 Cognitive
inhibition.
Figure 22. Forest plot of comparison: 2 Aerobic exercise versus no intervention, outcome: 2.8 Visual
attention.
Figure 23. Forest plot of comparison: 2 Aerobic exercise versus no intervention, outcome: 2.9 Auditory
attention.
Figure 24. Forest plot of comparison: 2 Aerobic exercise versus no intervention, outcome: 2.10 Motor
function.

0.79 to 4.29; five trials, 267 participants; Analysis 2.11; Figure
25).
Figure 25. Forest plot of comparison: 2 Aerobic exercise versus no intervention, outcome: 2.11 Drop-out.
Aerobic exercise versus strength training intervention

Aerobic exercise versus flexibility/balance
intervention Subgroup analyses of aerobic exercise intervention compared to
strength training controls was not possible since we could only
Analysing only the subgroup of trials in which the aerobic exer- include one trial in these analyses.
cise intervention was compared to flexibility or balance control
groups, four trials (351 participants) contributed data on at least
one cognitive domain (Blumenthal 1989; Kramer 2001; Moul Fitness improved: aerobic exercise versus any active
1995; Oken 2006). Intervention duration in these trials ranged intervention
from 16 to 26.07 weeks. We were able to conduct meta-analyses Analysing only the subgroup of trials in which the aerobic exercise
on all 11 of our pre-specified cognitive domains (Analysis 3.1 to intervention was shown to enhance fitness relative to any active
Analysis 3.11). There was no evidence of benefit of the aerobic intervention control groups, six trials including 367 participants
exercise intervention in any cognitive domain. contributed data on at least one cognitive domain (Blumenthal
There was no difference in dropout rates between the aerobic 1989; Fabre 2002; Kramer 2001; Legault 2011; Moul 1995;
exercise intervention and comparison groups (OR 0.99, 95% CI Panton 1990). The duration of the intervention in these trials
0.58 to 1.72; four trials, 351 participants; Analysis 3.12). ranged from eight to 26.07 weeks. We were able to conduct meta-
analyses for all 11 of our pre-specified cognitive domains (Analysis already points to exercise having a beneficial effect on cardiorespi-
5.1 to Analysis 5.11). There was no evidence of benefit of the aer- ratory fitness. However, this was not accompanied by any impact
obic exercise intervention in any cognitive domain. on cognitive function. Several issues need further consideration.
There was no difference in dropout rates between the aerobic Firstly, the quality of the included trials could have also affected
exercise intervention and comparison groups (OR 1.22, 95% CI our results. Reporting of methods in the included papers was gen-
0.66 to 2.25; five trials, 330 participants; Analysis 5.12). erally quite poor. For all but one trial, the randomisation methods
were unclear. It was not feasible to blind participants and trainers,
but for most trials it was also unclear if outcome assessors were
Fitness improved: aerobic exercise versus no blinded, raising the risk of detection bias. Attrition was poorly
intervention reported. No trials had published protocols so it was not possi-
Analysing only the subgroup of trials in which the aerobic ex- ble to tell if there was selective reporting of results. Of note, no
ercise intervention was shown to significantly improve fitness included trials assessed for contamination bias which could have
relative to no intervention control groups, four trials involving worked against finding group differences. Secondly, with healthy
183 participants contributed data on at least one cognitive do- older populations, it is possible that “ceiling effects” prevented de-
main. Intervention duration in these trials ranged from eight tection of cognitive improvement. The risk of this will depend on
to 26 weeks (Blumenthal 1989; Langlois 2012; Panton 1990; the task used and what is being measured. In the included papers,
Whitehurst 1991). We were able to conduct meta-analyses for no trial author discussed any potential impact of a ceiling effect
10 of our 11 pre-specified cognitive domains, besides perception on the variables measured. However, there was much variation in
(Analysis 6.1 to Analysis 6.10). There was no evidence of benefit each measure included in our analyses which makes ceiling effects
of the aerobic exercise intervention in any cognitive domain. unlikely.
0.50 to 4.50; three trials, Analysis 6.11). Agreements and disagreements with other
All analyses showed no difference on cognitive test scores between studies or reviews
aerobic exercise groups and either active comparator or no treat- Five meta-analytic studies and one systematic review published
ment groups (controls or waiting list groups). In terms of dropout data based on very similar hypotheses yet failed to find comparable
(without Panton 1990, which did not include dropouts by group), results:
there were no differences between aerobic exercise and any of our • Etnier 1997b included 134 articles in their review. Their
other intervention groups. Also, no trial included adverse events aim was to give a comprehensive overview of all literature
as an outcome and none of the trial reports made any mention of available with sufficient information to calculate effect sizes.
adverse events. Therefore, apart from RCTs, the review included several cross-
sectional studies. It reported data on the acute effects of exercise
and data on strength and flexibility regimens as well as results for
younger age groups and cognitively impaired individuals. The
DISCUSSION authors concluded that exercise has a small positive effect on
cognition and with the effect size depending on the exercise
paradigm, the quality of the trial, the participants and the
Summary of main results cognitive tests used as outcome measures.
This Cochrane Review examined the effect of physical activity • van Uffelen 2008 set out to systematically review the effect
aimed at improving cardiorespiratory fitness on cognitive func- of exercise on cognitive performance in older adults with and
tion in healthy older people without known cognitive impair- without dementia. They found 23 papers that met their
ment. The hypothesis being tested is that physical activity brings inclusion criteria. They included strength exercise interventions,
about improvements in cognition which are mediated by increased trials which did not assess any fitness parameters and a trial
cardiovascular (aerobic) fitness (Colcombe 2004; Kramer 1999; where both groups received aerobic training, while this review
McAuley 2004). If true, this would imply that a physically active did not. Their review observed exercise programmes in healthy
lifestyle resulting in enhanced fitness could positively affect peo- older adults improved memory, information processing abilities
ple’s cognitive abilities as they age and may even prevent, or at least and executive function.
delay, the onset of neurodegenerative disorders such as Alzheimer’s • Smith 2010 meta-analytic review assessed the effects of
disease. aerobic exercise on cognitive performance. Their criteria differed
Nine of the 12 included trials reported that aerobic exercise in- from this Cochrane Review in including participants with MCI,
terventions resulted in increased cardiorespiratory fitness of the younger participants and trials which did not assess
intervention group. This is not unexpected as significant evidence cardiorespiratory fitness. They also included some unpublished
trials. The authors concluded that aerobic exercise is significantly which may have implications for the effectiveness of some of the
and positively related to modest improvements in attention and training programmes in the included RCTs. However, Smith 2010
processing speed, executive function and memory. did not find any relationship between intensity of physical activity
• The meta-analysis presented by Colcombe 2003 included and change in cognitive function.
18 studies. Their aim (“to examine the hypothesis that aerobic
fitness training enhances the cognitive vitality of healthy but
sedentary older adults”) and exclusion criteria (cross-sectional
design, no random assignment, unsupervised exercise AUTHORS’ CONCLUSIONS
programme, training lacking in fitness component and an
average age below 55) were similar to ours. The reviews differed Implications for practice
in that we excluded trials in which allocation was clearly quasi-
randomised or did not present any fitness parameter. We also We found no evidence that improving cardiorespiratory fitness
excluded interventions that were not purely exercise and which necessarily results in improvements in cognitive performance in
included participants who were cognitively impaired or suffered healthy older adults without known cognitive impairment.
from depression. Colcombe 2003 concluded that physical
activity is beneficial for all analysed cognitive functions. Implications for research
• Etnier 2006 published a meta-analytic review on the We consider that larger studies with robust methodology exploring
relationship between aerobic fitness and cognitive performance. possible moderators are still required to confirm whether or not
Their primary goal was “to provide a statistically powerful test of aerobic training improves cognition in this population.
the viability of the cardiovascular fitness hypothesis by examining
the dose-response relationship between aerobic fitness and We wish to emphasise two important points:
cognition”. Their search identified 30 studies which reported
data on cross-sectional comparisons, pre-post comparisons and 1. Our review includes results from as many as 40 different
RCTs. Etnier 2006 included only those studies which assessed cognitive tests. This is already a smaller sample of tests than the
aerobic fitness by maximal, submaximal or a composite measure absolute total reported in the included trials (tests were lost from
of fitness which included VO2 max, whereas we included all analyses in order to avoid double representation of trials within
measures of aerobic fitness. We imposed a lower age limit and cognitive categories). A broad battery of tests can give insight
did not include trials on depressed participants whereas Etnier into the specificity of physical activity effects. At the same time,
2006 included all ages and at least one trial on depressed too great a number of cognitive tests can be confusing and
subjects. Etnier 2006 included unpublished master theses and obscure overall effects. We would recommend that researchers in
doctoral dissertations, whereas we only included data published the field seek agreement on a smaller battery of cognitive tests to
in peer reviewed journals. Post-test comparisons showed no use in order to increase comparability between trials. This
significant relationships between aerobic fitness and cognitive smaller core-set of cognitive tests should incorporate measures of
performance. For the exercise groups, increased fitness was key cognitive domains which are important both scientifically
associated with worse cognitive function. Age interacted with and clinically.
fitness and was a significant negative predictor of cognitive 2. Any intervention that is to be effective against age-related
performance for older adults. cognitive decline should be assessed over a significant period of
time. A limitation of the included RCTs is the lack of long-term
follow-up (with an average duration of 15.62 weeks). Longer-
Although we did not identify any relationship between physi- term intervention trials would be very valuable in the future.
cal activity or cardiorespiratory fitness and cognitive function,
it is possible that certain subgroups of the population, such as
those starting from a lower baseline of fitness, could react dif-
ferently to aerobic training. Other factors which might influ-
ACKNOWLEDGEMENTS
ence the relationship include: age, frequency of cognitive activi-
ties (Christensen 1993; Hultsch 1993; Hultsch 1999; Lachman We thank Jenny McCleery, Co-ordinating Editor of the Cochrane
2010; Marquine 2012; Wilson 1999; Wilson 2005), social net- Dementia and Cognitive Improvement Group (CDCIG), for as-
work (Crooks 2008; Seeman 2001), and adherence to a Mediter- sistance. We are grateful to Anna Noel-Storr, Trials Search Coor-
ranean diet (Panagiotakos 2007; Tangney 2011). The search for dinator, for her initial assessment of trials identified by searches in
possible subgroups has provided some promising results (examples this iteration. We thank Sue Marcus, Managing Editor of CDCIG,
in Etnier 2007; Podewils 2005; Schuit 2001). for assistance. Also, we thank Geert Aufdemkampe, HJJ Verhaar,
It is possible that the intensity of physical activities is important A Aleman and Luc Vanhees for their help with a previous version
(Angevaren 2007; Brown 2012; Tierney 2010; van Gelder 2004) of this manuscript.
REFERENCES
References to studies included in this review in healthy seniors: effects on cognition and quality of life.
Alternative Therapies in Health and Medicine 2006;12(1):
Bakken 2001 {published data only} 40–7.
Bakken RC, Carey JR, Di Fabio RP, Erlandson TJ, Hake Panton 1990 {published data only}
JL, Intihar TW. Effect of aerobic exercise on tracking Panton LB, Graves JE, Pollock ML, Hagberg JM, Chen W.
performance in elderly people: a pilot study. Physical Effect of resistance training on fractionated reaction time
Therapy 2001;81(12):1870–9. and speed of movement. Journal of Gerontology 1990;45(1):
Blumenthal 1989 {published data only} M26–31.
Blumenthal JA, Emery CF, Madden DJ, George LK, Whitehurst 1991 {published data only}
Coleman RE, Riddle MW, et al. Cardiovascular and Whitehurst M. Reaction time unchanged in older women
behavioral effects of aerobic exercise training in healthy following aerobic training. Perception and Motor Skills 1991;
older men and women. Journal of Gerontology 1989;44(5): 72(1):251–6.
M147–57.
References to studies excluded from this review
Emery 1990a {published data only}
Emery CF, Gatz M. Psychological and cognitive effects of Alessi 1999 {published data only}
an exercise program for community-dwelling older adults. Alessi CA, Yoon EJ, Schnelle JF, Al-Samarrai NR, Cruise
Gerontologist 1990;30(2):184–8. PA. A randomized trial of a combined physical activity and
environmental intervention in nursing home residents:
Fabre 2002 {published data only}
do sleep and agitation improve?. Journal of the American
Fabre C, Chamari K, Mucci P, Massé-Birron J, Préfaut
Geriatrics Society 1999;47(7):784–91.
C. Improvement of cognitive function by mental and/or
individualized aerobic training in healthy elderly subjects. Alves 2012 {published data only}
International Journal of Sports Medicine 2002;23(6):415–21. Alves CRR, Gualano B, Takao PP, Avakian P, Fernandes
RM, Morine D, et al. Effects of acute physical exercise on
Kramer 2001 {published data only}
∗ executive functions: a comparison between aerobic and
Kramer AF, Hahn S, McAuley E, Cohen NJ, Banich MT,
strength exercise. Journal of Sport & Exercise Psychology
Harrison C, et al. Exercise, aging and cognition: Healthy
2012;34(4):539–49.
body, healthy mind?. In: Fisk AD, Rogers W editor(s).
Human Factors Interventions for the Health Care of Older Barry 1966 {published data only}
Adults. Hillsdale, NJ: Erlbaum, 2001:91–120. Barry AJ, Steinmetz JR, Page HF, Rodahl K. The effects
of physical conditioning on older individuals. II. Motor
Langlois 2012 {published data only} performance and cognitive function. Journal of Gerontology
Langlois F, Vu TTM, Chassé K, Dupuis G, Kergoat 1966;21(2):192–9.
MJ, Bherer L. Benefits of physical exercise training on
Blumenthal 1988 {published data only}
cognition and quality of life in frail older adults. Journals of
Blumenthal JA, Madden DJ. Effects of aerobic exercise
Gerontology Series B: Psychological Sciences and Social Sciences
training, age, and physical fitness on memory-search
2012;68(3):400–4. [DOI: doi:10.1093/geronb/gbs069]
performance. Psychology and Aging 1988;3(3):280–5.
Legault 2011 {published data only}
Blumenthal 1991 {published data only}
Legault C, Jennings JM, Katula JA, Dagenbach D, Gaussoin
Blumenthal JA, Emery CF, Madden DJ, Schniebolk S,
SA, Sink KM, et al. Designing clinical trials for assessing
Walsh-Riddle M, George LK, et al. Long-term effects of
the effects of cognitive training and physical activity
exercise on psychological functioning in older men and
interventions on cognitive outcomes: the Seniors Health
women. Journal of Gerontology 1991;46(6):352–61.
and Activity Research Program Pilot (SHARP-P) study, a
randomized controlled trial. BMC Geriatrics 2011;11:27. Bream 1996 {unpublished data only}
Bream JH. Interventions to aging: Immunologic and
Madden 1989 {published data only}
cognitive responses to 16 weeks of low-intensity exercise
Madden DJ, Blumenthal JA, Allen PA, Emery CF.
training in older adults. Dissertation Abstracts International
Improving aerobic capacity in healthy older adults does
1996. [DA9716196]
not necessarily lead to improved cognitive performance.
Carles 2007 {published data only}
Psychology and Aging 1989;4(3):307–20.
Carles S Jr, Curnier D, Pathak A, Roncalli J, Bousquet M,
Moul 1995 {published data only} Garcia JL, et al. Effects of short-term exercise and exercise
Moul JL, Goldman B, Warren B. Physical activity and training on cognitive function among patients with cardiac
cognitive performance in the older population. Journal of disease. Journal of Cardiopulmonary Rehabilitation and
Aging and Physical Activity 1995;3:135–45. Prevention 2007;27(6):395–9.
Oken 2006 {published data only} Cassilhas 2007 {published data only}
Oken BS, Zajdel D, Kishiyama S, Flegal K, Dehen C, Haas Cassilhas RC, Viana VA, Grassmann V, Santos RT, Santos
M, et al. Randomized, controlled, six-month trial of yoga RF, Tufik S, et al. The impact of resistance exercise on the
cognitive function of the elderly. Medicine and Science in Fabre 1999 {published data only}
Sports and Exercise 2007;39(8):1401–7. Fabre C, Massé-Biron J, Chamari K, Varray A, Mucci P,
Préfaut C. Evaluation of quality of life in elderly healthy
Colcombe 2006 {published data only}
subjects after aerobic and/or mental training. Archives of
Colcombe SJ, Erickson KI, Scalf PE, Kim JS, Prakash R,
Gerontology and Geriatrics 1998;28(1):9–22.
McAuley E, et al. Aerobic exercise training increases brain
volume in aging humans. Journals of Gerontology Series A- Gates 2011 {published data only}
Biological Sciences & Medical Sciences 2006;61(11):1166–70. Gates NJ, Valenzuela M, Sachdev PS, Singh NA, Baune BT,
Brodaty H, et al. Study of Mental Activity and Regular
Deary 2006 {published data only}
Training (SMART) in at risk individuals: A randomised
Deary IJ, Whalley LJ, Batty GD, Starr JM. Physical fitness
double blind, sham controlled, longitudinal trial. BMC
and lifetime cognitive change. Neurology 2006;67(7):
Geriatrics 2011;11:19.
1195–200.
Dietrich 2004 {published data only} Glisky 1997 {unpublished data only}
Dietrich A, Sparling PB. Endurance exercise selectively Glisky ML. Interventions for cognitive and psychosocial
impairs prefrontal-dependent cognition. Brain and functioning in older adults: A comparison of aerobic
Cognition 2004;55(3):516–24. exercise and cognitive training. Dissertation Abstracts
International 1997. [DA9806830]
Dorner 2007 {published data only}
Dorner T, Kranz A, Zettl-Wiedner K, Ludwig C, Rieder Hassmén 1992 {published data only}
A, Gisinger C. The effect of structured strength and Hassmén P, Ceci R, Bäckman L. Exercise for older women: a
balance training on cognitive function in frail, cognitive training method and its influences on physical and cognitive
impaired elderly long-term care residents. Aging Clinical performance. European Journal of Applied Physiology and
and Experimental Research 2007;19(5):400–5. Occupational Physiology 1992;64(5):460–6.
Dustman 1984 {published data only} Hassmén 1997 {published data only}
Dustman RE, Ruhling RO, Russell EM, Shearer DE, Hassmén P, Koivula N. Mood, physical working capacity
Bonekat HW, Shigeoka JW, et al. Aerobic exercise training and cognitive performance in the elderly as related to
and improved neuropsychological function of older physical activity. Aging 1997;9(1-2):136–42.
individuals. Neurobiology of Aging 1984;5(1):35–42. Hawkins 1992 {published data only}
Emery 1990b {published data only} Hawkins HL, Kramer AF, Capaldi D. Aging, exercise, and
Emery CF, Blumenthal JA. Percieved change among attention. Psychology and Aging 1992;7(4):643–53.
participants in an exercise program for older adults. Hill 1993 {published data only}
Gerontologist 1990;30(4):516–21. Hill RD, Storandt M, Malley M. The impact of long-term
Emery 1998 {published data only} exercise training on psychological function in older adults.
Emery CF, Schein RL, Hauck ER, MacIntyre NR. Journal of Gerontology 1993;48(1):P12–7.
Psychological and cognitive outcomes of a randomized
Ijuin 2013 {published data only}
trial of exercise among patients with chronic obstructive Ijuin M, Sugiyama M, Sakuma N, Inagaki H, Miyamae
pulmonary disease. Health Psychology 1998;17(3):232–40.
F, Ito K, et al. Walking exercise and cognitive functions
Emery 2003 {published data only} in community-dwelling older adults: preliminary results
Emery CF, Hsiao ET, Hill SM, Frid DJ. Short-term effects of a randomized controlled trial. International Journal of
of exercise and music on cognitive performance among Geriatric Psychiatry 2013;28(1):109–10.
participants in a cardiac rehabilitation program. Heart &
Jacobson 2007 {unpublished data only}
Lung: the Journal of Critical Care 2003;32(6):368–73.
Jacobson A. Specificity of Exercise on Enhancing Cognitive
Erickson 2007 {published data only} Abilities: Argentine Tango and Walking. Montreal, Canada:
Erickson KI, Colcombe SJ, Elavsky S, McAuley E, Korol McGill University, 2007.
DL, Scalf PE, et al. Interactive effects of fitness and
Jedrziewski 2007 {published data only}
hormone treatment on brain health in postmenopausal
Jedrziewski MK, Lee VM-Y, Trojanowski JQ. Physical
women. Neurobiology of Aging 2007;28(2):179–85.
activity and cognitive health. Alzheimer’s and Dementia
Etnier 1997a {published data only} 2007;3(2):98–108.
Etnier JL, Landers DM. The influence of age and fitness on
Kerschan 2002 {published data only}
performance and learning. Journal of Aging and Physical
Kerschan-Schindl K, Wiesinger G, Zauner-Dungl A,
Activity 1997;5:175–89.
Kollmitzer J, Fialka-Moser V, Quittan M. Step aerobic
Etnier 2001 {published data only} vs. cycle ergometer training: Effects on aerobic capacity,
Etnier JL, Berry M. Fluid intelligence in an older COPD coordinative tasks, and pleasure in untrained adults - A
sample after short- or long-term exercise. Medicine and randomized controlled trial. Wiener Klinische Wochenschrift
Science in Sports and Exercise 2001;33(10):1620–8. 2002;114(23-24):992–8.
Kharti 2001 {published data only} students with leaning disabilities and deficit disorder.
Khatri P, Blumenthal JA, Babyak MA, Craighead WE, Dissertation Abstracts International 2000. [DA9972924]
Herman S, Baldewicz T, et al. Effects of exercise training Molloy 1988 {published data only}
on cognitive functioning among depressed older men and Molloy DW, Richardson LD, Crilly RG. The effects of a
women. Journal of Aging and Physical Activity 2001;9(1): three-month exercise programme on neuropsychological
43–57. function in elderly institutionalized women: a randomized
Kramer 1999 {published data only} controlled trial. Age and Ageing 1988;17(5):303–10.
Kramer AF, Hahn S, Cohen NJ, Banich MT, McAuley E,
Mortimer 2012 {published data only}
Harrison CR, et al. Ageing, fitness and neurocognitive
Mortimer JA, Ding D, Borenstein AR, DeCarli C, Guo Q,
function. Nature 1999;400(6743):418–9.
Wu Y, et al. Changes in brain volume and cognition in
Kramer 2007 {published data only}
a randomized trial of exercise and social interaction in a
Kramer AF, Erickson KI. Capitalizing on cortical plasticity:
community-based sample of non-demented Chinese elders.
influence of physical activity on cognition and brain
Journal of Alzheimer’s Disease 2012;30(4):757–66.
function. Trends in Cognitive Sciences 2007;11(8):342–8.
Munguía-Izquierdo 2007 {published data only}
Larson 2006 {published data only}
Munguía-Izquierdo D, Legaz-Arrese A. Exercise in warm
Larson EB, Wang L, Bowen JD, McCormick WC, Teri L,
water decreases pain and improves cognitive function
Crane P, et al. Exercise is associated with reduced risk for
in middle-aged women with fibromyalgia. Clinical and
incident dementia among persons 65 years of age and older.
Experimental Rheumatology 2007;25(6):823–30.
Annals of Internal Medicine 2006;144(2):73–81.
Lautenschlager 2008 {published data only} Netz 2007 {published data only}
Lautenschlager NT, Cox KL, Flicker L, Foster JK, van Netz Y, Tomer R, Axelrad S, Argov E, Inbar O. The effect
Bockxmeer FM, Xiao J, et al. Effect of physical activity of a single aerobic training session on cognitive flexibility
on cognitive function in older adults at risk for Alzheimer in late middle-aged adults. International Journal of Sports
disease: a randomized trial. JAMA 2008;300(9):1027–37. Medicine 2007;28(1):82–7.
Leinonen 2007 {published data only} O’Dwyer 2007 {published data only}
Leinonen R, Heikkinen E, Hirvensalo M, Lintunen T, O’Dwyer ST, Burton NW, Pachana NA, Brown WJ
Rasinaho M, Sakari-Rantala R, et al. Customer-oriented . Protocol for Fit Bodies, Fine Minds: a randomized
counseling for physical activity in older people: study controlled trial on the affect of exercise and cognitive
protocol and selected baseline results of a randomized- training on cognitive functioning in older adults. BMC
controlled trial (ISRCTN 07330512). Scandinavian Journal Geriatrics 2007;7:23.
of Medicine & Science in Sports 2007;17(2):156–64. Oken 2004 {published data only}
Littbrand 2006 {published data only} Oken BS, Kishiyama S, Zajdel D, Bourdette D, Carlsen
Littbrand H, Rosendahl E, Lindelöf N, Lundin-Olsson J, Haas M, et al. Randomized controlled trial of yoga
L, Gustafson Y, Nyberg L. A high-intensity functional and exercise in multiple sclerosis. Neurology 2004;62(11):
weight-bearing exercise program for older people dependent 2058–64.
in activities of daily living and living in residential care Okumiya 1996 {published data only}
facilities: evaluation of the applicability with focus on Okumiya K, Matsubayashi K, Wada T, Kimura S, Doi Y,
cognitive function. Physical Therapy 2006;86(4):489–98. Ozawa T. Effects of exercise on neurobehavioral function
Liu-Ambrose 2010 {published data only} in community-dwelling older people more than 75 years of
Liu-Ambrose T, Davis JC, Nagamatsu LS, Hsu CL, age. Journal of the American Geriatrics Society 1996;44(5):
Katarynych LA, Khan KM. Changes in executive functions 569–72.
and self-efficacy are independently associated with improved Palleschi 1996 {published data only}
usual gait speed in older women. BMC Geriatrics 2010;10: Palleschi L, Vetta F, De Gennaro E, Idone G, Sottosanti G,
25. Gianni W, et al. Effect of aerobic training on the cognitive
Masley 2009 {published data only} performance of elderly patients with senile dementia of
Masley S, Roetzheim R, Gualtieri T. Aerobic exercise Alzheimer type. Archives of Gerontology and Geriatrics 1996;
enhances cognitive flexibility. Journal of Clinical Psychology 22(Suppl 1):47–50.
in Medical Settings 2009;16(2):186–93. Palmer 1995 {unpublished data only}
McAuley 2006 {published data only} Palmer AC. The effects of aerobic exercise on cognitive
McAuley E, Konopack JF, Motl RW, Morris KS, Doerksen ability and creativity in senior citizens. Dissertation
SE, Rosengren KR. Physical activity and quality of life in Abstracts International 1995. [DA9530364]
older adults: influence of health status and self-efficacy. Perri 1984 {published data only}
Annals of Behavioral Medicine 2006;31(1):99–103. Perri S 2nd, Templer DI. The effects of an aerobic
McKenzie 2000 {unpublished data only} exercise program on psychological variables in older adults.
McKenzie DL. An investigation of the relationship between International Journal of Aging and Human Development
exercise and the cognitive function of attention in adult 1984;20(3):167–72.
Pierce 1993 {published data only} healthy older adults. Frontiers in Aging Neuroscience 2013;5:
Pierce TW, Madden DJ, Siegel WC, Blumenthal JA. 8.
Effects of aerobic exercise on cognitive and psychosocial Sibley 2007 {published data only}
functioning in patients with mild hypertension. Health Sibley BA, Beilock SL. Exercise and working memory: an
Psychology 1993;12(4):286–91. individual differences investigation. Journal of Sport and
Plati 2006 {published data only} Exercise Psychology 2007;29(6):783–91.
Plati MC, Covre P, Lukasova K, de Macedo EC. Depressive
Small 2006 {published data only}
symptoms and cognitive performance of the elderly:
Small GW, Silverman DH, Siddarth P, Ercoli LM, Miller
relationship between institutionalization and activity
KJ, Lavretsky H, et al. Effects of a 14-day healthy longevity
programs. Revista Brasileira de Psiquiatria 2006;28(2):
lifestyle program on cognition and brain function. American
118–21.
Journal of Geriatric Psychiatry 2006;14(6):538–45.
Powell 1975 {published data only}
Powell RR. Effects of exercise on mental functioning. Smiley-Oyen 2008 {published data only}
Journal of sports medicine and physical fitness 1975;15(2): Smiley-Oyen AL, Lowry KA, Francois SJ, Kohut ML,
125–31. Ekkekakis P. Exercise, fitness, and neurocognitive
function in older adults: the “selective improvement” and
Powers 2007 {unpublished data only}
“cardiovascular fitness” hypotheses. Annals of Behavioral
Powers M. High-intensity resistance training in older adults:
Medicine 2008;36(3):280–91.
Impact on physical and cognitive function. Dissertation
Abstracts International 2007;68:2308. Stevenson 1990 {published data only}
Predovan 2012 {published data only} Stevenson JS, Topp R. Effects of moderate and low intensity
Predovan D, Fraser SA, Renaud M, Bherer L. The Effect of long-term exercise by older adults. Research in Nursing &
Three Months of Aerobic Training on Stroop Performance Health 1990;13(4):209–18.
in Older Adults. Journal of Aging Research 2012;2012: van Uffelen 2007 {published data only}
269815. [DOI: 10.1155/2012/269815] van Uffelen JG, Chin A Paw MJ, Hopman-Rock M,
Prohaska 2007 {published data only} van Mechelen W. The effect of walking and vitamin B
Prohaska TR, Peters KE. Physical activity and cognitive supplementation on quality of life in community-dwelling
functioning: translating research to practice with a public adults with mild cognitive impairment: a randomized,
health approach. Alzheimer’s & Dementia 2007;3(2 Suppl): controlled trial. Quality of Life Research 2007;16(7):
S58–64. 1137–46.
Querry 1998 {unpublished data only} Verghese 2006 {published data only}
Querry PA. Effects of aerobic exercise training and Verghese J, LeValley A, Derby C, Kuslansky G, Katz M,
cognitive function in aging adults. Dissertation Abstracts Hall C, et al. Leisure activities and the risk of amnestic mild
International 1998. [DA9901203] cognitive impairment in the elderly. Neurology 2006;66(6):
Rikli 1991 {published data only} 821–7.
Rikli RE, Edwards DJ. Effects of a three-year exercise Wallman 2004 {published data only}
program on motor function and cognitive processing speed Wallman KE, Morton AR, Goodman C, Grove R, Guilfoyle
in older women. Research Quarterly for Exercise and Sport AM. Randomised controlled trial of graded exercise in
1991;62(1):61–7. chronic fatigue syndrome. Medical Journal of Australia
Rosendahl 2006 {published data only} 2004;180(9):444–8.
Rosendahl E, Lindelöf N, Littbrand H, Yifter-Lindgren Wilbur 2005 {published data only}
E, Lundin-Olsson L, Håglin L, et al. High-intensity Wilbur J, Miller AM, McDevitt J, Wang E, Miller J.
functional exercise program and protein-enriched energy Menopausal status, moderate-intensity walking, and
supplement for older persons dependent in activities of daily symptoms in midlife women. Research and Theory for
living: a randomised controlled trial. Australian Journal of Nursing Practice 2005;19(2):163–80.
Physiotherapy 2006;52(2):105–13.
Williams 1997 {published data only}
Russell 1984 {unpublished data only} Williams P, Lord SR. Effects of group exercise on cognitive
Sato 2007 {published data only} functioning and mood in older women. Australian and New
Sato D, Kaneda K, Wakabayashi H, Nomura T. The water Zealand Journal of Public Health 1997;21(1):45–52.
exercise improves health-related quality of life of frail elderly Williamson 2009 {published data only}
people at day service facility. Quality of Life Research 2007; Williamson JD, Espeland M, Kritchevsky SB, Newman AB,
16(10):1577–85. King AC, Pahor M, et al. Changes in cognitive function
Shatil 2013 {published data only} in a randomized trial of physical activity: results of the
Shatil E. Does combined cognitive training and physical lifestyle interventions and independence for elders pilot
activity training enhance cognitive abilities more than either study. Journals of Gerontology. Series A: Biological sciences
alone? A four-condition randomized controlled trial among and medical sciences 2009;64(6):688–94.
Winter 2007 {published data only} fitness and cerebral blood flow on cognitive outcomes in
Winter B, Breitenstein C, Mooren FC, Voelker K, Fobker older women. Neurobiology of Aging 2008;31(12):2047–57.
M, Lechtermann A, et al. High impact running improves
Brown 2012
learning. Neurobiology of Learning & Memory 2007;87(4):
Brown BM, Peiffer JJ, Sohrabi HR, Mondal A, Gupta
597–609.
VB, Rainey-Smith SR, et al. Intense physical activity
Zlomanczuk 2006 {published data only} is associated with cognitive performance in the elderly.
Zlomanczuk P, Milczarek B, Dmitruk K, Sikorski W, Translational Psychiatry 2012;2:e191.
Adamczyk W, Zegarski T, et al. Improvement in the
face/name association performance after three months of Christensen 1993
physical training in elderly women. Journal of Physiology & Christensen H, Mackinnon A. The association between
Pharmacology 2006;57(Suppl 4):417–24. mental, social and physical activity and cognitive
performance in young and old subjects. Age and Ageing
References to studies awaiting assessment 1993;22(3):175–82.
Chapman 2013 {published data only} Churchill 2002

Chapman SB, Aslan S, Spence JS, Defina LF, Keebler Churchill JD, Galvez R, Colcombe S, Swain RA, Kramer
MW, Didehbani N, et al. Shorter term aerobic exercise AF, Greenough WT. Exercise, experience and the aging
improves brain, cognition, and cardiovascular fitness in brain. Neurobiology of Aging 2002;23(5):941–55.
aging. Frontiers in Aging Neuroscience 2013;5:75. Colcombe 2003
Linde 2014 {published data only} Colcombe S, Kramer AF. Fitness effects on the cognitive
Linde K, Alfermann D. Single versus combined cognitive function of older adults; a meta-analytic study. Psychological
and physical activity effects on fluid cognitive abilities of Science 2003;14(2):125–30.
healthy older adults: a 4-month randomized controlled trial
Colcombe 2004
with follow-up. Journal of Aging and Physical Activity 2014;
Colcombe SJ, Kramer AF, Erickson KI, Scalf P, McAuley E,
22(3):302–13.
Cohen NJ, et al. Cardiovascular fitness, cortical plasticity,
Additional references and aging. Proceedings of the National Academy of Sciences of
the United States of America 2004;101(9):3316–21.
Abbott 2004
Abbott RD, White LR, Ross GW, Masaki KH, Curb JD, Cotman 2002
Petrovitch H. Walking and dementia in physically capable Cotman CW, Berchtold NC. Exercise: a behavioral
elderly men. JAMA 2004;292(12):1447–53. intervention to enhance brain health and plasticity. Trends
in Neurosciences 2002;25(6):295–301.
Aleman 2000
Aleman A, de Haan EHF, Verhaar HJJ, Samson MM, de Cotman 2007
Vries WR, Koppeschaar HP. Relationship between cognitive Cotman CW, Berchtold NC, Christie LA. Exercise builds
and physical function in healthy older men; a role for brain health: key roles of growth factor cascades and
aerobic power?. Journal of the American Geriatrics Society inflammation. Trends in Neurosciences 2007;30(9):464–72.
2000;48(1):104–5.
Crooks 2008
Angevaren 2007 Crooks V, Lubben J, Petitti D, Little D, Chiu V. Social
Angevaren M, Vanhees L, Wendel-Vos W, Verhaar HJJ, network, cognitive function, and dementia incidence
Aufdemkampe G, Aleman A, et al. Intensity, but not among elderly women. American Journal of Public Health
duration, of physical activities is related to cognitive 2008;98(8):1221–7.
function. European Journal of Cardiovascular Prevention and
Rehabilitation 2007;14(6):825–30. Davenport 2012
Davenport MH, Hogan DB, Eskes GA, Longman RS,
Barnes 2003
Poulin MJ . Cerebrovascular reserve: the link between
Barnes DE, Yaffe K, Satariano WA, Tager IB. A longitudinal
fitness and cognitive function?. Exercise and Sport Sciences
study of cardiorespiratory fitness and cognitive function
Reviews 2012;40(3):153–8.
in healthy older adults. Journal of the American Geriatrics
Society 2003;51(4):459–65. Erickson 2009
Boutron 2005 Erickson KI, Prakash RS, Voss MW, Chaddock L, Hu
Boutron I, Moher D, Tugwell P, Giraudeau B, Poiraudeau L, Morris KS, et al. Aerobic fitness is associated with
S, Nizard R, et al. A checklist to evaluate a report of a hippocampal volume in elderly humans. Hippocampus
nonpharmaceutical trial (CLEAR NPT) was developed 2009;19(10):1030–9.
using consensus. Journal of Clinical Epidemiology 2005;58 Etgen 2010
(12):1233–40. Etgen T, Sander D, Huntgeburth U, Poppert H, Förstl
Brown 2008 H, Bickel H. Physical activity and incident cognitive
Brown AD, McMorris CA, Longman RS, Leigh R, Hill impairment in elderly persons: the INVADE study. Archives
MD, Friedenreich CM, et al. Effects of cardiorespiratory of Internal Medicine 2010;170(2):186–93.
Etnier 1997b 90: a meta-analysis. Journal of Sports Medicine and Physical
Etnier JL, Salazar W, Landers DM, Petruzzello SJ, Han M, Fitness 2000;40(1):1–10.
Nowell P. The influence of physical fitness and exercise upon
Lezak 2004
cognitive functioning; a meta-analysis. Journal of Sports and
Lezak MD, Howieson DB, Loring DW. Neuropsychological
Exercise Psychology 1997;19:249–77.
Assessment. New York: Oxford University Press, 2004.
Etnier 2006
Etnier JL, Nowell PM, Landers DM, Sibley BA. A meta- Lojovich 2010
regression to examine the relationship between aerobic Lojovich JM. The relationship between aerobic exercise
fitness and cognitive performance. Brain Research Reviews and cognition: is movement medicinal?. Journal of Head
2006;52(1):119–30. Trauma Rehabilitation 2010;25(3):184–92.
Etnier 2007 Marquine 2012
Etnier JL, Caselli RJ, Reiman EM, Aleander GE, Sibley BA, Marquine MJ, Segawa E, Wilson RS, Bennett DA, Barnes
Tessier D, et al. Cognitive performance in older women LL. Association between cognitive activity and cognitive
relative to ApoE-e4 genotype and aerobic fitness. Medicine function in older Hispanics. Journal of the International
and Science in Sports and Exercise 2007;39(1):199–207. Neuropsychological Society 2012;18(6):1041–51.
Heyn 2004 Martin 2003
Heyn P, Abreu BC, Ottenbacher KJ. The effects of exercise Martin M, Zimprich D. Are changes in cognitive
training on elderly persons with cognitive impairment and functioning in older adults related to changes in subjective
dementia; a meta-analysis. Archives of Physical Medicine and complaints?. Experimental Aging Research 2003;29(3):
Rehabilitation 2004;85(10):1694–704. 335–52.
Higgins 2011
McAuley 2004
Higgins JPT, Green S (editors) [Cochrane Handbook
McAuley E, Kramer AF, Colcombe SJ. Cardiovascular
for Systematic Reviews of Interventions, Version 5.1.0
fitness and neurocognitive function in older adults: a brief
[updated March 2011]. The Cochrane Collaboration,
review. Brain, Behavior and Immunity 2004;18(3):214–20.
2011]. Available at www.cochrane-handbook.org.
Hultsch 1993 Middleton 2011
Hultsch DF, Hammer M, Small BJ. Age differences in Middleton LE, Manini TM, Simonsick EM, Harris TB,
cognitive performance in later life: relationships to self- Barnes DE, Tylavsky F, et al. Activity energy expenditure
reported health and activity life style. Journal of Gerontology and incident cognitive impairment in older adults. Archives
1993;48(1):P1–11. of Internal Medicine 2011;171(14):1251–7.
Hultsch 1999 Newson 2006
Hultsch DF, Hertzog C, Small BJ, Dixon RA. Use it or lose Newson RS, Kemps EB. The nature of subjective cognitive
it: engaged lifestyle as a buffer of cognitive decline in aging? complaints of older adults. International Journal of Aging
. Psychology and Aging 1999;14(2):245–63. and Human Development 2006;63(2):139–51.
Kessels 2000 Panagiotakos 2007
Kessels RPC, Aleman A, Verhagen WIM, Luijtelaar EL. Panagiotakos DB, Pitsavos C, Arvaniti F, Stefanadis C.
Cognitive functioning after whiplash injury: a meta- Adherence to the Mediterranean food pattern predicts the
analysis. Journal of the International Neuropsychological prevalence of hypertension, hypercholesterolemia, diabetes
Society 2000;6(3):271–8. and obesity, among healthy adults; the accuracy of the
Lachman 2010 MedDietScore. Preventive Medicine 2007;44(4):335–40.
Lachman ME, Agrigoroaei S, Murphy C, Tun PA. Frequent
Podewils 2005
cognitive activity compensates for education differences in
Podewils LJ, Guallar E, Kuller LH, Fried LP, Lopez OL,
episodic memory. American Journal of Geriatric Psychiatry
Carlson M, et al. Physical activity, APOE genotype, and
2010;18(1):4–10.
dementia risk: findings from the Cardiovascular Health
Landis 1977 Cognition Study. American Journal of Epidemiology 2005;
Landis JR, Koch GG. The measurement of observer 161(7):639–51.
agreement for categorical data. Biometrics 1977;33(1):
159–74. Prins 2002
Laurin 2001 Prins ND, Den Heijer T, Hofman A, Koudstaal PJ, Jolles J,
Laurin D, Verreault R, Lindsay J, MacPherson K, Rockwood Clarke R, et al. Homocysteine and cognitive function in
K. Physical activity and risk of cognitive impairment and the elderly: the Rotterdam Scan Study. Neurology 2002;59
dementia in elderly persons. Archives of Neurology 2001;58 (9):1375–80.
(3):498–504. RevMan 2014 [Computer program]
Lemura 2000 The Nordic Cochrane Centre, The Cochrane Collaboration.
Lemura LM, von Duvillard SP, Mookerjee S. The effects of Review Manager (RevMan). Copenhagen: The Nordic
physical training of functional capacity in adults. Ages 46 to Cochrane Centre, The Cochrane Collaboration, 2014.
Richards 2003 cognitive decline in elderly men: the FINE Study. Neurology
Richards M, Hardy R, Wadsworth ME. Does active leisure 2004;63(12):2316–21.
protect cognition? Evidence from a national birth cohort.
van Uffelen 2008
Social Science & Medicine 2003;56(4):785–92.
van Uffelen JGZ, Chin A Paw MJM, Hopman-Rock M, van
Salthouse 1996 Mechelen W. The effects of exercise on cognition in older
Salthouse TA. General and specific speed mediation of adult adults with and without cognitive decline: a systematic
age differences in memory. The Journals of Gerontology. review. Clinical Journal of Sport Medicine 2008;18(6):
Series B, Psychological Sciences and Social Sciences 1996;51 486–500.
(1):30–42.
Verhaeghen 1997
Salthouse 2003
Verhaeghen P, Salthouse TA. Meta-analyses of age-cognition
Salthouse TA. Memory aging from 18 to 80. Alzheimer’s
relations in adulthood: estimates of linear and nonlinear age
Disease and Associated Disorders 2003;17(3):162–7.
effects and structural models. Psychological Bulletin 1997;
Schuit 2001 122(3):231–49.
Schuit AJ, Feskens EJ, Launer LJ, Kromhout D. Physical
activity and cognitive decline, the role of the apolipoprotein Wilson 1999
e4 allele. Medical Science and Sports Exercise 2001;33(5): Wilson RS, Bennett DA, Beckett LA, Morris MC, Gilley
772–7. DW, Bienias JL, et al. Cognitive activity in older persons
Seeman 2001 from a geographically defined population. Journals of
Seeman TE, Lusignolo TM, Albert M, Berkman L. Social Gerontology. Series B, Psychological Sciences and Social
relationships, social support, and patterns of cognitive Sciences 1999;54(3):P155–60.
aging in healthy, high-functioning older adults: MacArthur Wilson 2005
studies of successful aging. Health Psychology 2001;20(4): Wilson RS, Barnes LL, Krueger KR, Hoganson G, Bienias
243–55. JL, Bennett DA. Early and late life cognitive activity and
Smith 2010 cognitive systems in old age. Journal of the International
Smith PJ, Blumenthal JA, Hoffman BM, Cooper H, Neuropsychological Society 2005;11(4):400–7.
Strauman TA, Welsh-Bohmer K, et al. Aerobic exercise
and neurocognitive performance: a meta-analytic review of WPP 2006
randomized controlled trials. Psychosomatic Medicine 2010; World Population Prospects, The 2006 Revision Population
72(3):239–52. Database. United Nations. http://www.un.org/esa/
Sturman 2005 population/publications/wpp2006/wpp2006.htm (accessed
Sturman MT, Morris MC, Mendes de Leon CF, Bienias JL, 15 November 2007).
Wilson RS, Evans DA. Physical activity, cognitive activity,
and cognitive decline in a biracial community population.
References to other published versions of this review
Archives of Neurology 2005;62(11):1750–4.
Tangney 2011 Angevaren 2008a
Tangney CC, Kwasny MJ, Li H, Wilson RS, Evans DA, Angevaren M, Aufdemkampe G, Verhaar HJ, Aleman A,
Morris MC. Adherence to a Mediterranean-type dietary Vanhees L. Physical activity and enhanced fitness to improve
pattern and cognitive decline in a community population. cognitive function in older people without known cognitive
American Journal of Clinical Nutrition 2011;93(3):601–7. impairment. Cochrane Database of Systematic Reviews 2008,
Issue 4. [DOI: 10.1002/14651858.CD005381.pub2]
Tierney 2010
Tierney MC, Moineddin R, Morra A, Manson J, Blake J. Angevaren 2008b
Intensity of recreational physical activity throughout life Angevaren M, Aufdemkampe G, Verhaar HJ, Aleman A,
and later life cognitive functioning in women. Journal of Vanhees L. Physical activity and enhanced fitness to improve
Alzheimer’s Disease 2010;22(4):1331–8. cognitive function in older people without known cognitive
van Gelder 2004 impairment. Cochrane Database of Systematic Reviews 2008,
van Gelder BM, Tijhuis MAR, Kalmijn S, Giampaoli S, Issue 7. [DOI: 10.1002/14651858.CD005381.pub3]
∗
Nissinen A, Kromhout D. Physical activity in relation to Indicates the major publication for the study
CHARACTERISTICS OF STUDIES
Characteristics of included studies [ordered by study ID]
Bakken 2001
Methods Parallel-group RCT: 1 intervention group and 1 control group.

At randomisation 15 enrolled; 8 in the aerobic exercise group, 7 in the control group.
Follow-up: 8 weeks
Participants 10 participants (4 males, 6 females) in the age range of 72 to 91 years from a senior
housing complex in Minneapolis, Minnesota.
Inclusion criteria: > 65 years of age with no history of pulmonary disease, recurring falls,
orthopaedic limitations or acute arthritis in the hands
Interventions Aerobic exercise: 1 hour sessions for 3 sessions per week for 8 consecutive weeks. 10
minutes of warming up, aerobic conditioning period that increased in duration and
intensity (callisthenics, walking and cycling) systematically each week, 10 minutes of
cooling down. Subjects heart rates did not exceed the upper limit of their THRR*.
Control: continued their normal everyday routine, which did not include any aerobic
exercise according to the subjects report
Outcomes AI (Accuracy Index by finger movement)

Resting heart rate
Resting systolic BP
Resting rate-pressure product
GXT test heart rate**
GXT test systolic BP
GXT test RPP (Rate-Pressure product)
Notes Testing took place at the University of Minnesota.

*THRR: (Karvonen) training HR = resting HR + [0.60-0.75 (HRR)].
HRR = age-predicted max HR - resting HR
**GXT: submaximal graded exercise tolerance test. Stage 1; stepping back and forth on
the ground at a frequency of 20 mounts per minute for 3 minutes. Stage 2: stepping up
and down a 10.16 cm high step. Stage 3: stepping up and down a 20.32 cm high step.
Stage 4: stepping up and down a 30.48 cm high step.
RPP; rate-pressure product = systolic BP multiplied by heart rate. A decrease in RPP is
a quantitative measure of aerobic training
Both groups showed slight increases in RPP from pre-test to post-test
Risk of bias
Bias Authors’ judgement Support for judgement
Random sequence generation (selection Unclear risk No information provided.

bias)
Allocation concealment (selection bias) Unclear risk No information provided.
Bakken 2001 (Continued)
Blinding (performance bias and detection High risk Participants were not blinded to their group
bias) assignment, but it was not feasible to do so
Participants
Blinding (performance bias and detection Low risk It is not feasible that the trainers were
bias) blinded to the condition but this non-
Trainers blinding was unlikely to introduce bias
Blinding (performance bias and detection Low risk Cognition was assessed with a computer
bias) and therefore adequately blinded. At the
Outcome assessors same time the researchers where unaware
of the group assignment of the participants
Incomplete outcome data (attrition bias) High risk 3/8 participants were lost from the exer-
All outcomes cise condition, 2/7 were lost in the control
group. Main outcomes were not analysed
according to the ITT principle
Selective reporting (reporting bias) Unclear risk Insufficient information.
Other bias Low risk However, we cannot rule out contamina-

tion bias.
Blumenthal 1989
Methods Parallel-group RCT: 2 intervention groups and 1 control group. 101 participants (50
males and 51 females) were randomised either to aerobic exercise (N = 33), yoga/flexibility
(N = 34) or control (N = 34)
Follow-up: 16 weeks
Participants 101 participants aged 60 to 83 years

Inclusion criteria: free from clinical manifestations of coronary disease assessed by medical
history, physical examination, bicycle ergometry exercise testing.
Exclusion criteria: positive ECG during exercise testing, evidence of coronary artery
disease, asthma, pulmonectomy, uncontrolled hypertension, beta-blocker therapy
Interventions Aerobic exercise: 3 supervised sessions per week for 16 weeks. Training based on 70%
of max heart rate achieved on exercise test. 10 minutes of warming up, 30 minutes of
bicycle ergometry, 15 minutes of brisk walking/jogging and arm ergometry, 5 minutes
of cooling down.
Yoga/flexibility: 2 supervised sessions a week for 60 minutes over 16 weeks.
Controls: not to change their physical activity habits and especially not to engage in any
aerobic exercise for the trial period
Outcomes Tapping (dominant/non-dominant)

Digit span (forward / backward)
Benton Revised Visual Retention test (correct/error)
Story Recall of the Randt Memory test (immediate) - data on the delayed Story Recall
Blumenthal 1989 (Continued)
of the Randt Memory test could not be traced by the authors.

Selective reminding test (total/intrusions)
Trail making (part B)
Digit Symbol substitution
2 & 7 test (digits/letters)
Stroop colour word
Stroop interference
Verbal fluency
Non-verbal fluency
VO2 max
AT
Notes Testing took place at the Duke University Medical Center.

A summary combination of the scores on both 2&7 test (letters and digits) was calculated
and SDs were pooled and used in analysis
Subjects in the aerobic training group experienced a significant 11.6% increase in their
VO2 max (from 19.4 to 21.4 mL/kg/min), whereas the participants in the yoga/flexibility
and control groups experienced a 1 to 2% decrease in VO2 max (from 18.8 to 18.7 mL/
kg/min and 18.5 to 17.9 mL/kg/min, respectively)
Risk of bias

bias)
Participants
Blinding (performance bias and detection Unclear risk Insufficient information is provided to
bias) judge the blinding of the cognitive out-
Outcome assessors comes
Incomplete outcome data (attrition bias) High risk 2/33 participants were lost from both the
All outcomes aerobic group and yoga/flexibility group
and 2/34 the control group. Main out-
comes were not analysed according to the
ITT principle
Blumenthal 1989 (Continued)
Selective reporting (reporting bias) High risk The methods section describes assessment
of the Story Recall of the Randt Memory
test after 30 minutes delay. Data on this
subtest could not be traced by the authors
Other bias Low risk Although contamination bias could not be

ruled out.
Emery 1990a
Methods Parallel-group RCT: 2 intervention groups and 1 control group. 48 subjects (8 males
and 40 females) were randomly assigned to an aerobic exercise programme (N = 15),
social activity group (N = 15) or a control group (N = 18)
Follow-up: 12 weeks
Participants 48 participants aged 61 to 86 years from a metropolitan inner-city community.

Inclusion criteria not described.
Exclusion criteria not described.
Interventions Exercise: 3 sessions per week for approximately 60 minutes. 10 to 15 minutes of stretching
exercises followed by 20 to 25 minutes of aerobic exercise (at 70% of age-adjusted max
= 220-age), including rapid walking as well as rhythmic muscle strengthening exercises
(e.g. repeatedly standing up and sitting down). 5 minutes of cooling down with dancing
and light exercises.
Social activity: 3 sessions per week for 60 minutes. Participation in non-physical activities
(card games, art projects, political discussion groups, watching films).
Controls: not described.
Outcomes Digit Symbol substitution

Digit Span
Copying Words
Copying Numbers (digit/sec)
Weight (kg)
Resting HR
Resting blood pressure (syst/diast)
BP during modified step test
HR during modified step test
Sit-and-reach test

Resting heart rate, maximum heart rate and systolic/diastolic blood pressure indicated
no significant differences between the groups. Both groups showed a significant time
main effect decrease in diastolic blood pressure, other measures indicated no significant
effects
Risk of bias
Emery 1990a (Continued)

bias)
Participants
Blinding (performance bias and detection Unclear risk Insufficient information provided to assess
bias) the blinding of the cognitive outcome mea-
Outcome assessors sures
Incomplete outcome data (attrition bias) High risk 1/15 participants was lost from the aero-
All outcomes bic group, 4/15 from the social group and
4/18 from the control group. Main out-
comes were not analysed according to the
ITT principle
Selective reporting (reporting bias) Unclear risk Insufficient information provided.

ruled out.
Fabre 2002
Methods Parallel-group RCT: 3 intervention groups and 1 control group. 32 participants (5 males
and 27 females) at randomisation; each group (physical training, memory training,
combined physical/memory training and controls) contained 8 subjects
Follow-up: 8 weeks
Participants 32 participants in the age range of 60 to 76 years “from clubs”.

Inclusion criteria are not described. Exclusion criteria: positive ECG during exercise
testing, present depression, could not breathe through the tube during exercise testing,
various other reasons such as disease during training
Interventions Physical training: two supervised 1 hour exercise sessions per week for 8 weeks: walking
and running to maintain target heart rate (target heart rate corresponded to the ventila-
tory threshold). 5 minutes of warming up, 45 minutes of walking/running, 10 minutes
of cooling down
Memory training: 90 minutes of sessions once a week for 8 weeks. 15 minutes of ex-
plaining, Israel’s method in core
Combined physical training and memory training.
Controls: no training whatsoever.
Fabre 2002 (Continued)
Outcomes Memory quotient (= total score of al WAIS subtests)

Paired associates learning
Digit span forward
Logical memory immediate recall
Orientation
General information
Mental control
Visual reproductions
VO2 max
VO2 max at Vth
Max O2 pulse
Max O2 pulse at Vth
Notes Testing took place at the University of Montpellier.

The physical training resulted in an average significant increase in VO2 max of 12%
(from 1350 to 1630 mL/min) and 11% (from 1510 to 1625 mL/min) in the aerobic
training group and the combined aerobic/mental group, respectively. The VO2 max
scores of the participants in the other two groups were unchanged compared to initial
values (mental training group from 1060 to 999 mL/min and controls from 1256 to
1265 mL/min)
Risk of bias

bias)
Participants
Incomplete outcome data (attrition bias) Low risk All participants completed the trial.
All outcomes
Fabre 2002 (Continued)

ruled out.
Kramer 2001
Methods Parallel-group RCT: 1 aerobic walking intervention group and 1 stretching/toning con-
trol group.174 participants at randomisation. The trial was completed by 124 individ-
uals; 58 (13 men) in the aerobic group and 66 (20 men) in the stretching and toning
group
Follow-up period: 6 months.
Participants 124 participants aged 60 to 75 years of age.

Inclusion criteria: aged 60 to 75 years, sedentary (no physical activity in the preced-
ing 6 months), capable of performing exercise, physicians examination and consent to
participate, successful completion of graded exercise test without evidence of cardiac
abnormalities, initial depression score on the GDS below clinical level, no history of
neurologic disorders, corrected (near & far) acuity of 20/40 or better, fewer than three
errors on the Pfeiffer Mental Status questionnaire.
Exclusion criteria: younger than 60 years, self-reported activity on a regular basis (2 times
a week) in the preceding 6 months, any physical disability that prohibits mobility, non-
consent of physician, evidence of abnormal cardiac responses during graded exercise test-
ing, depression score on the GDS indicative of clinical depression, history of neurologic
disorders, corrected (near & far) acuity greater than 20/40, more than three errors on
the Pfeiffer questionnaire
Interventions Aerobic walking exercise: 3 supervised sessions per week for 6 months. Warming up, 40
minutes of brisk walking (gradually beginning at 10 to 15 minutes up to 40 minutes),
cooling down. Initial exercise was performed at 50 to 55% of VO2 max and increased
to 65 to 70% of VO2 max.
Stretching and toning: 3 times a week supervised sessions for 6 months. The programme
emphasized stretches for all the large muscle group of the upper and lower extremities.
Each stretch was held for 20 to 30 seconds and repeated 5 to 10 times. Each session was
proceeded and followed by 10 minutes of warm-up and cooling down
Outcomes Visual search task

Response compatibility task
Task switching paradigm
Stopping paradigm
Spatial attention task
Rey auditory verbal learning test
Pursuit rotor task
Self-ordered pointing task
Spatial working memory
Verbal working memory
Face recognition task
Digit-digit and digit-symbol tests
Forward and backward digit span
VO2 max (mL/kg/min)
Kramer 2001 (Continued)
Time on treadmill (min)

Rockport 1-mile walk (min)
Notes Testing took place at the University of Illinois at Urbana-Champaign

The physical training resulted in improvements of 5.1% on VO2 max measures (from
21.5 to 22.6 mL/kg/min). The toning group showed a 2.8% decrease in VO2 max scores
(from 21.8 to 21.2 mL/kg/min).
Risk of bias

bias)
Participants
Incomplete outcome data (attrition bias) High risk 25/83 subjects from walking group and 25/
All outcomes 91 subjects from stretching/toning group
were dropped from the trial because of
withdrawal from the training protocol or
incomplete data. These participants did
not differ in demographic characteristics
from those who completed the trial. Main
outcomes were not analysed according to
the ITT principle

ruled out.
Langlois 2012
Methods Parallel-group RCT: 1 one exercise training intervention group and 1 waiting list control
group. 83 participants at randomisation, randomised ensuring gender ratio equivalence:
43 in the intervention group, 40 in the control group
Follow-up: 3 months
Participants 72 participants aged 61 to 89.

Inclusion criteria: assessed via complete geriatric assessment as able to perform exercise
programme at low risk.
Exclusion criteria: limitations to undertaking exercise programme, MMSE < 25, or GDS
> 10, or both
Interventions Physical exercise training group: 12 weeks of 1 hour exercise 3 days/week conducted in
supervised 3 to 5 participant subgroups. 10 mins stretching and balancing warm up, 10
to 30 mins aerobic workout, 10 mins strength training, 10 mins cool down. Intensity and
duration of aerobic workout increased individually using modified Borg RPE reaching
moderate to hard intensity.
Control group: maintain level of activity during period and were offered physical training
programme after trial
Outcomes MMSE
WAIS-III Similarities
WAIS-III Digit-Symbol Coding
Trailmaking part A
Trailmaking part B
modified Stroop Colour-Word Test
WAIS-III Letter-Number Sequencing
Digit Span Backwards
Rey Auditory Verbal Learning Task
6MWT
modified Physical Performance Test
Timed Up and Go Test
Gait speed (comfortable and maximum)
Notes Testing took place at the Université du Québec à Montréal.

There was a significantly larger improvement in the exercise training group in comparison
to the control group in the 6MWT
Risk of bias

bias)
Participants
Langlois 2012 (Continued)
Incomplete outcome data (attrition bias) High risk 7/43 participants was lost from the inter-
All outcomes vention group, 4/40 from the from the con-
trol group. Main outcomes were not anal-
ysed according to the ITT principle

ruled out.
Legault 2011
Methods Parallel-group RCT: 3 intervention groups and 1 control group. 73 participants at ran-
domisation, 18 (10 female) were put into the physical activity training group, 18 (8 fe-
male) into the cognitive training, 19 (12 female) into the combined intervention group,
and 18 (7 female) into the healthy aging control group
Follow-up period: 4 months.
Participants 73 participants, aged 70 to 85 years of age that were community-dwelling

Inclusion criteria: aged 70 to 85, identified as “individuals who were appropriate candi-
dates for physical activity and cognitive training and who appeared likely to adhere to
interventions and data collection protocols” as detailed in a previous paper
Exclusion criteria: Related to Physical Activity: Severe rheumatologic or orthopedic dis-
eases, severe pulmonary heart disease, actively participating in a formal exercise pro-
gramme within the past month (> 30 min/week), severe cardiac disease, clinically signif-
icant aortic stenosis, history of cardiac arrest which required resuscitation, use of cardiac
defibrillator or uncontrolled angina. Other significant co-morbid disease that would im-
pair ability to participate in the exercise based intervention. Receiving physical therapy
for gait, balance or other lower extremity training. Serious conduction disorder, un-
controlled arrhythmia. Pulmonary embolism or deep venous thrombosis within past 6
months. Hip fracture, hip or knee replacement, or spinal surgery within past 4 months.
Severe hypertension
Related to Cognition: Neurological disease, stroke that required hospitalisation, Parkin-
son’s, multiple sclerosis, Amyotrophic Lateral Sclerosis, or MCI. Telephone interview for
cognitive status ≤ 31. Current use of cognitive enhancing prescription or investigational
medications. History of participation in a cognitive training programme in the last two
years. 3MSE score < 88 (< 80 for ≤ 8 years education). Scores ≥ 2 SDs below normal on
memory or non-memory domain tests (speed of processing and verbal fluency). Other
significant factors that may affect the ability for cognitive training, including a history of
Legault 2011 (Continued)
head trauma resulting in a loss of consciousness, current use of benzodiazepines, hypnotic

or anticholinergic agents. Stroke within past 4 months. Baseline Geriatric Depression
Scale score ≥ 8
Related to trial design or adherence: Age < 70 or > 85 years. Unwillingness to be randomized
to any of the four intervention conditions. Failure to provide the name of a personal
physician. Living in a nursing home. Terminal illness with life expectancy less than 8
months. Unable to communicate because of severe hearing loss or speech disorder. Severe
visual impairment. Excessive alcohol use (> 14 drinks per week). Member of household
is already enrolled. Lives distant from the trial site or is planning to move out of the
area in the next year or leave the area for more than one month during the next year.
Other temporary intervening events, such as sick spouse, bereavement or recent move.
Participation in another intervention trial. Inability to commit to intervention schedule
requirements. Failure to provide informed consent
Interventions Physical activity training: centre-based and home-based sessions aimed at aerobic and
flexibility training targeting duration of 150 minutes/week. Two centre-based sessions
per week for four months, focus on walking (or other endurance activity if contraindi-
cated) with explicit intent of improving cardiovascular fitness. Centre-based sessions ap-
proximately 60 minutes - 40 minutes walking, 20 minutes flexibility. Tailored home-
based walking 1 to 2 sessions per week for first month and encouraged to slowly increase
duration, speed and frequency to achieve 150 min/week goal
Cognitive training: four consecutive 10 to 12 min sessions per day, administered two
times per week for two months, then one time per week for two additional months at
centre via computer, where participants studied a list of 30 words, then were given a
recognition test consisting of the 30 studied words and 30 new words with each new
word repeated once, and asked to respond “yes” to trial words or “no” to new words.
Intervals between the first and second presentation of new words increased as participants
reached accuracy thresholds
Combined physical activity and cognitive training: received both, cognitive was delivered
prior to physical activity to avoid impact of fatigue
Controls: weekly lectures based on health education, topics such as medications, foot
care, travelling and nutrition
Outcomes Trailmaking part A

Trailmaking part B
Hopkins Verbal Learning Test
2-Back
Flanker Task
400-metre walk time
Notes Testing took place at Wake Forest University.

400-metre walk times for the physical activity training group decreased by 5.31 seconds
and were not different from the combined intervention group. Walk times for the cog-
nitive training group and the ’healthy ageing’ group did not improve
Risk of bias
Legault 2011 (Continued)

bias)
Participants
Blinding (performance bias and detection Low risk Trial was “single-blinded”, since it is not
bias) possible to blind participants, outcome as-
Outcome assessors sessors must have been blinded
Incomplete outcome data (attrition bias) Low risk All analyses conducted according to ITT
All outcomes principles. 2/18 participants in the physical
activity group were excluded because they
did not return for the 4-month visit. 1/18
in the physical activity group and 1/19 in
the combined intervention group were ex-
cluded for not attending any of the centre-
based training sessions

ruled out.
Madden 1989
males and 41 females) at randomisation; 28 in the aerobic group, 30 in the yoga group
and 27 served as controls
Follow-up: 16 weeks.
Participants Participants were 60 to 83 years of age.

Inclusion criteria: free of medical conditions that would preclude a programme of either
aerobic exercise or yoga.
Exclusion criteria: uncontrolled hypertension, diabetes, or coronary heart disease, use of
beta-blockers or psychotropic medication
Interventions Aerobic exercise: 3 supervised sessions per week for 16 weeks. 10 minutes of warming
up, 30 minutes of cycling, 15 minutes of brisk walking or jogging or both, 5 minutes of
cooling down. All exercise was performed in target (training) heart range (70% of max
during initial exercise test)
Madden 1989 (Continued)
Yoga: 2 times a week 60 minutes of supervised yoga sessions for 16 weeks

Control: no change to their physical activity habits for the length of the trial
Outcomes Letter search RT task (short-term memory)

Word comparison RT task (long-term memory)
VO2 max

Aerobic capacity remained constant for the yoga and control groups between pre- and
post-test (respectively from 18.8 to 18.6 mL/kg/min and from 19.1 to 18.6 mL/kg/
min), whereas the aerobic exercise group showed a significant 11% increase in VO2 max
(from 19.7 to 21.9 mL/kg/min)
Risk of bias

bias)
Participants
Incomplete outcome data (attrition bias) High risk 3/28 were lost from the exercise group, 2/
All outcomes 30 from the yoga group and 1/27 from the
controls. Main outcomes were not analysed
according to the ITT principle

ruled out.
Moul 1995
males, 19 females), the walking, weight training and control group all contained 10
participants
Participants 30 participants aged 65 to 72 years.

Inclusion criteria: nondiseased (no current symptoms or signs suggestive of heart disease)
, non active (defined as < 2 moderate to vigorous aerobic or resistance training sessions
of > 20 minutes per week)
Interventions Walking: 5 sessions per week. Walking 30 minutes at 60% of HRR (as determined by
treadmill testing). Walking duration was increased 2 minutes per week until they reached
40 minutes and HRR were adjusted after 8 weeks of training to 65% of HRR
Weight training: 5 sessions per week of upper and lower body exercises on alternate
days of the week. Abdominal crunches and back extensions were performed in each
session. Weight group employed a daily adjusted progressive resistive exercise programme
(DAPRE) using weights
Controls: 5 sessions per week mild stretching exercises for 30 to 40 minutes. Minimal
challenge to the cardiovascular or muscular systems
Outcomes Immediate Memory (Ross Information Processing Assessment)

Recent Memory (RIPA)
Temporal Orientation (RIPA)
Problem Solving and abstract reasoning (RIPA)
Organization (RIPA)
Auditory Processing (RIPA)
Weight (kg)
Sum of seven skinfolds (mm)
VO2 max (mL/kg/min)
Time on treadmill (min)
Ventilation (mL/min)
RER
Knee extension (lb)
Elbow flexion (lb)
Notes Testing took place at the Human Performance Laboratory and Athletic Training Labo-
ratory, Appalachian State University
Post-test data revealed that the subjects in the walking condition significantly increased
their VO2 max by an average of 16% (from 22.4 to 26.6 mL/kg/min), whereas there
were no significant changes in VO2 max for the other two groups (weight training group
from 21.4 to 20.4 mL/kg/min and controls from 20.9 to 19.3 mL/kg/min)
Risk of bias

bias)
Moul 1995 (Continued)
Participants
All outcomes

ruled out.
Oken 2006
Methods Parallel-group RCT: 2 intervention groups and 1 control group. 135 participants were
randomised in a yoga class (N = 47), exercise group (N = 44) or a wait-list control group
(N = 44)
Follow-up: 6 months.
Participants Participants were healthy adults in the age of 65 to 85 years

Inclusion criteria: other than age not described.
Exclusion criteria: subjects were screened for significant medical problems, had a physical
examination and routine ECG to ensure the safety of the intervention and to exclude
participants with pathology with might impair cognition. Subjects were excluded for the
following reasons: insulin-dependent diabetes, uncontrolled hypertension, evidence of
liver or kidney failure, significant lung disease, alcoholism or other drug abuse, symptoms
or signs of congestive heart failure, symptomatic ischemic heart disease, or significant
valvular disease and significant visual impairment. Subjects also were excluded if they
were actively practicing yoga or had taken a yoga or tai-chi class in the last 6 months or
if they were regularly performing aerobic exercise more than 210 minutes per week.
Interventions Yoga was taught in one class per week along with home practice. The yoga classes were
90 minutes in duration and designed by a certified Iyengar yoga teacher, an Iyengar
trained teacher and a physician. Over all weeks, eighteen poses were taught. Each class
ended with a 10-minute deep relaxation period with the participant lying supine. Daily
home practice was strongly encouraged and participants were encouraged to honour
their individual limits
A certified personal trainer with experience in the geriatric population directed the aerobic
Oken 2006 (Continued)
exercise intervention arm of the trial. The aerobic intervention consisted of 1 class per
week along with home exercise. The aerobic exercise consisted of walking on an outdoor
400-metre track for endurance training. The 1-hour class began with walking 2 laps to
warm up and then progressed to mild leg stretches. Intensity of exercise was determined
by heart rate and modified Borg Rate of Perceived Exertion scale (Borg CR10 Scale)
. Participants wore a heart-rate monitor, and target heart rate was initially estimated
as 70% of maximum based on morning resting heart rate and age. Participants were
instructed to exercise at a level of 6/7 on the Borg scale. Based on perceived exertion, the
heart rate target was adjusted slightly. Participants were strongly encouraged to exercise
daily at least 5 times per week in addition to the weekly class session
Participants in the wait-list control group received no intervention
Outcomes Stroop colour word

Covert orienting of spatial attention
Simple RT (msec)
Choice RT (msec)
10-words learning task (delayed recall)
Letter-number sequencing
SF-36
Stanford sleepiness scale (SSS)
Profile of mood states (POMS)
Multidimensional fatigue inventory (MFI)
Centre for epidemiologic studies depression scale (CESD-10)
State-trait anxiety inventory
One-leg stand (sec)
Chair sit and reach (cm)
Sit and stand (sec)
¼ mile walk (sec)
Notes Testing took place at the Oregon Health and Science University
After 6 months there were no significant differences in time at a ¼ mile walk between
all three groups
Risk of bias
Random sequence generation (selection Low risk “Subjects were randomly assigned to treat-
bias) ment groups in this study with a planned
modified minimization scheme”
Participants
Oken 2006 (Continued)
Blinding (performance bias and detection Low risk Outcome assessors were adequately
bias) blinded.
Outcome assessors
Incomplete outcome data (attrition bias) High risk 9/47 dropped out from yoga, 6/44 from
All outcomes exercise and 2/44 from the wait-list group.
ITT analysis was not performed. Quote:
“No attempt was made to input missing
variables”

ruled out.
Panton 1990
Methods Parallel-group RCT: 2 intervention groups and 1 control group. 57 participants were
randomised in a walk/jog group, a strength group and a control group. Analyses were
performed on 17 participants in the walk/jog group, 20 participants in the strength
group and 12 controls.
Participants were not blinded; it is unclear whether the outcome assessor and the caregiver
were blinded
Participants Participants were retired professionals from the university community of Gainesville, FL
and 70 to 79 years of age.
Inclusion criteria: sedentary non-smokers who had no contraindications to exercise test-
ing or training. Free of any overt evidence of coronary artery disease and other conditions
that would limit their participation in a vigorous exercise programme as tested with a
diagnostic graded exercise test (using a modified Naughton protocol).
Exclusion criteria were not described.
Interventions The walk/jog group participated in three exercise sessions per week for the duration
of the trial. All training sessions were preceded by 5 to 10 minutes of stretching and
warm-up and ended with 5 min of cool-down exercises. Initially, participants started
walking/jogging for 20 minutes at 50% of their maximal heart rate reserve (HRRmax).
The duration was increased by 5 min every 2 weeks until the participants walked for 40
minutes. Training intensity was gradually increased until participants could walk at 60 to
70% of their HRRmax. During the 14th week of training exercise intensity was further
increased by alternating fast walk/moderate walk or fast walk/slow jog intervals. Five
participants increased their training intensity by increasing the slope of the treadmill. By
the 26th week of training, all participants performed at 85% of HRRmax for 35 to 45
min.
Panton 1990 (Continued)
Participants in the strength group participated in 30 min sessions, 3 times a week for
26 weeks. Workouts consisted of one set of 10 variable resistance Nautilus exercises (leg,
arm and torso muscles). During the first 13 weeks, participants used light to moderate
weights and performed 8 to 12 repetitions for each exercise. During the last 13 weeks,
resistance was increased substantially and participants were encouraged to train to voli-
tional muscular fatigue. When participants could complete 12 or more repetitions, the
resistance was increased.
Participants in the control group were asked not to change their lifestyle over the 6
month duration of the trial
Outcomes Total simple Reaction time

Fractionated Reaction time (PreMotorTime and MotorTime)
Speed of Movement measurements
Fat percentage (predicted from body density)
Body density (7 skinfolds)
1RM muscle strength test
VO2 max
Notes Testing took place at the University of Florida College of Medicine

Aerobic capacity significantly improved by 20.4% (from 22.5 to 27.1 mL/kg/min) for
subjects in the walk/jog group; participants in both the strength as well as the control
group showed no significant changes in VO2 max (from 22.5 to 23.3 and 22.2 to 22.0,
respectively).
Risk of bias

bias)
Participants
Blinding (performance bias and detection Low risk Cognitive function was assessed by com-
bias) puter and therefore adequately blinded
Outcome assessors
Incomplete outcome data (attrition bias) High risk 8/57 participants were lost to follow-up;
All outcomes it is unclear from which condition these
participants were lost. Main outcomes were
not analysed according to the ITT principle
Panton 1990 (Continued)

ruled out.
Whitehurst 1991
Methods Parallel-group RCT: 1 intervention group and 1 control group. 14 participants at ran-
domisation (all females): 7 in both the exercise and the control group
Follow-up: 8 weeks.
Participants Females in the age range of 61 to 73 years living in a rural community in North Carolina.
Inclusion criteria: did not participate in aerobic exercise more than one time per week
prior to the trial. Medical clearance from a physician (resting ECG and physical exami-
nation). Free of primary cardiovascular risk factors. Maintained the household
Interventions Exercise: 3 supervised sessions per week for 8 weeks (total of 24 sessions). 5 to 10 minutes
of warming up and cooling down. The participants cycled for 8 to 10 minutes the first
week to provide acclimatization. Thereafter, 3 to 5 minutes was added to subsequent
sessions so that by week 4 all participants were cycling for 35 to 40 minutes at their
target heart rate
Control: did not engage in any form of vigorous physical activity during the course of
the trial
Outcomes Simple reaction time

Choice reaction time
Estimated VO2 max
Notes Testing took place at the Human Performance Laboratory, Florida Atlantic University
The subjects in the exercise group significantly increased their VO2 max values by an
average 16% (from 25.4 to 29.7 mL/kg/min), whereas the subjects in the control group
increased their VO2 max by a (non significant) 2% (from 24.7 to 25.4 mL/kg/min)
Risk of bias

bias)
Participants
Whitehurst 1991 (Continued)
Blinding (performance bias and detection Low risk Cognitive function was assessed by com-
bias) puter (quote: “a standard choice reaction-
Outcome assessors time apparatus was used”) and therefore ad-
equately blinded
All outcomes

ruled out.
Characteristics of excluded studies [ordered by study ID]
Study Reason for exclusion
Alessi 1999 No pre to post-intervention cognitive data. Mean MMSE scores of the participants was below the range
of what is considered ’normal’ cognition (mean MMSE of 13.6 ± 8.5)
Alves 2012 No measure of cardiorespiratory fitness.
Barry 1966 Not a RCT but a CCT.
Blumenthal 1988 Participants were too young to meet the given inclusion criteria of this review
Blumenthal 1991 Data were already published in Blumenthal 1989.
Bream 1996 No published data.
Carles 2007 Participants were too young to meet the given inclusion criteria of this review
Cassilhas 2007 Exercise was not intended to improve aerobic fitness and no fitness parameters present
Colcombe 2006 No pre to post-intervention cognitive data.
Deary 2006 Not a RCT but a longitudinal survey.
Dietrich 2004 Data could not test the cardiovascular fitness hypothesis since cognition was assessed during exercise
(Continued)
Dorner 2007 Participants had cognitive impairment; cognitive impairment was an exclusion criterium for our review
Dustman 1984 Not a RCT but a quasi-randomised study (participants “alternately assigned”)
Emery 1990b Perceived (subjective) measurements of cognition were analysed according to groups but the objective
measures of cognition were analysed according to perceived measures of cognition
Emery 1998 Only had combined intervention groups and no pure aerobic exercise intervention group
Emery 2003 Within participants repeated measures design to evaluate the influence of music and exercise on cognition.
No control group
Erickson 2007 Not a RCT but a cross-sectional study.
Etnier 1997a No exercise intervention.
Etnier 2001 The control group was encouraged to continue exercising; however no formal programme was provided
Fabre 1999 No means and SDs for cognitive data. These results are described in Fabre 2002.
Gates 2011 Selected participants that have early changes in memory without diagnosis and excluded people with
perfect MMSE
Glisky 1997 No published data.
Hassmén 1992 Not RCT but “matched controls”.
Hassmén 1997 Not a RCT, participants matched on cognitive performance in pairs, then randomised
Hawkins 1992 No fitness parameter present.
Hill 1993 Not a RCT but a quasi-randomised study (participants “assigned to intervention group”)
Ijuin 2013 No measure of aerobic fitness and not peer-reviewed.
Jacobson 2007 No fitness parameter present and not published in a peer reviewed journal
Jedrziewski 2007 Not a RCT but a narrative review.
Kerschan 2002 Both groups followed aerobic training intervention.
Kharti 2001 Study participants were depressed older men and women: depression was an exclusion criterion for our
review
Kramer 1999 This article provides no quantitative data on which an analysis can be based. Quantitative data of the
RCT of this research group is provided in Kramer 2001, which is included in our review.
(Continued)
Kramer 2007 Not a RCT but a narrative review.
Larson 2006 Not a RCT but a prospective cohort study.
Lautenschlager 2008 No assessment of fitness parameters.
Leinonen 2007 No pre- to post-intervention data present, only selected baseline results
Littbrand 2006 No pre- to post-intervention cognitive data. Applicability study for the evaluation of attendance and
adverse events of an exercise programme
Liu-Ambrose 2010 No measure of cardiorespiratory fitness.
Masley 2009 RCT included younger, middle-aged and older adults.
McAuley 2006 No pre- to post-intervention cognitive parameters.
McKenzie 2000 No published data.
Molloy 1988 Exercise was not intended to improve aerobic fitness.
Mortimer 2012 No aerobic fitness measure.
Munguía-Izquierdo 2007 Participants were too young to meet the given inclusion criteria of this review. No fitness parameters
present
Netz 2007 Data could not test the cardiovascular fitness hypothesis since cognition was assessed during exercise
O’Dwyer 2007 Not a RCT but description of a study protocol.
Oken 2004 Participants were too young to meet the given inclusion criteria of this review
Okumiya 1996 No fitness parameter present.
Palleschi 1996 Participants were elderly patients with senile dementia of the Alzheimer type: this was an exclusion
criterion for our review
Palmer 1995 No published data.
Perri 1984 Not a true RCT but a clinical trial.
Pierce 1993 Participants were too young to meet the inclusion criteria of this review
Plati 2006 Not a RCT but “matched controls”.
Powell 1975 Narrative review; this article provides no quantitative data
(Continued)
Powers 2007 No published data.
Predovan 2012 Not a RCT, group selection was based on order of recruitment and willingness to participate in an exercise
programme
Prohaska 2007 Not a RCT but a narrative article.
Querry 1998 No published data.
Rikli 1991 Balance, sit and reach flexibility, shoulder flexibility, and grip strength were given as fitness parameters.
We excluded this RCT since neither of the two fitness parameters reflect aerobic fitness
Rosendahl 2006 No pre- to post-intervention cognitive data. Mean MMSE scores of the participants was below the range
of what is considered ’normal’ cognition (mean MMSE of 17.8 ± 5.1)
Russell 1984 No published data.
Sato 2007 No cognition or fitness parameters, or both.
Shatil 2013 No objective measure of aerobic fitness in all groups.
Sibley 2007 Not a RCT and data could not test the cardiovascular fitness hypothesis since cognition was assessed
during exercise
Small 2006 No fitness parameters present.
Smiley-Oyen 2008 Not a RCT but a quasi-randomised study (“Group allocation alternated between CARDIO and FLEX-
TONE”)
Stevenson 1990 Both intervention groups received aerobic training (different levels of intensity)
van Uffelen 2007 Participants had mild cognitive impairment; cognitive impairment was an exclusion criterion for our
review
Verghese 2006 Not a RCT but a prospective cohort study.
Wallman 2004 Participants were too young to meet the inclusion criteria of this review
Wilbur 2005 No objective measures of cognitive parameters (symptom impact inventory)
Williams 1997 No objective measures of fitness, only subjective measures (Perceived General Fitness)
Williamson 2009 No assessment of fitness parameters. What could have been used to assess fitness (400 m walk) was taken
here as part of an assessment of functionality (specifically normal gait speed) and because of how this
measure was implemented (walked at usual pace, allowed to rest, allowed to not complete), it could not
be used for fitness assessment
(Continued)
Winter 2007 Data could not test the cardiovascular fitness hypothesis since cognition was assessed during exercise
Zlomanczuk 2006 No fitness parameters present.
Characteristics of studies awaiting assessment [ordered by study ID]
Chapman 2013
Methods Parallel-group, randomised, controlled trial: 1 intervention group, 1 wait-list control group. 37 participants at ran-
domisation; 18 (13 female) were put into the physical training group, 19 (14 female) were put into the control group
Follow-up: 12 weeks
Participants Participants were 57 to 75 years of age.

Inclusion criteria: “no prior history of neurological or psychiatric conditions, average IQ range, native English speaker,
and minimum of a high school diploma”
Exclusion criteria: “MR scanning contraindications, cognitive impairment (TICS-M < 28 and MoCA < 26), elevated
depressive symptoms (BDI >14), left-handedness, increased body mass BMI > 40, abnormal electrocardiographic
response, significant hypertensive blood pressure response to exercise, or inability to reach 85% of maximum predicted
heart rate for age... if they reported regular aerobic activity of more than twice a week for 20 min or more.They could
not have regularly exercised for at least 3 months prior to enrolling in the study.”
Interventions Physical Training: “The training regimen consisted of three 60 min sessions of aerobic exercise training per week for
a period of 12 weeks. The participants’ aerobic exercise alternated each session between exercise bike and treadmill.
The exercise bike routine included: 5 min warm up at 43 watts, cycling for 50 min at a speed that increased their
heart rate to 50-75% of their maximum achieved heart rate on VO2 max testing, and a 5 min cool down at 43 watts.
The treadmill workout included: 5 min warmup at 2 miles per h (mph), walking on treadmill for 50 min at a speed
that increased their heart rate to 50-75% of their maximum achieved heart rate on VO2 max testing, and a 5 min
cool down at 2mph.”
Control: Wait-list
Outcomes Wechsler Abbreviated Scale of Intelligence (WASI)

BDI
MoCA
Tics-M
Trails B - Trails A
CVLT-II
WMS-IV immediate/delayed memory
Delis-Kaplan Executive Function System-Color Word Interference subtest (DKEFS- color word)
Backward Digit Span
BMI
Structural MRI
absolute Cerebral Blood Flow (aCBF)
RPE
VO2 max
Chapman 2013 (Continued)
Notes Testing took place at the The University of Texas at Dallas, The University of Texas Southwestern Medical Center,
and The Cooper Institute
VO2 max increased significantly to a greater extent for the Physical Training group than the Control group at the
mid-point (p = .03), however at the endpoint change in VO2 max did not differ between the groups.
Linde 2014
Methods Parallel-group, randomised, controlled trial: 3 intervention groups, 1 wait-list control group. 70 participants were
randomised: 19 (11 female) to a physical intervention group, 18 (9 female) to a cognitive intervention group, 17 (11
female) to a combined physical and cognitive intervention group, and 16 (10 female) to a wait-list control group
Follow-ups: 16-weeks and 3 months after conclusion of intervention
Participants Participants were aged 60-75 from a “medium-sized German city”

Inclusion criteria: “age of 60-75 years”
Exclusion criteria: “dementia, depression, and possible medical conditions (e.g., coronary heart
diseases, hypertension, stroke, pulmonary diseases) that would not allow for participation in a regular exercise program.
”
Interventions “The interventions took place in groups of 8-10 participants and were hosted at the facilities of the Faculty of Sport
Sciences and its campus.”
Physical Activity Intervention: “Participants trained two times per week, each session lasting 60 min, for a period of 16
weeks.” Sessions consisted of 20 mins progressive strength training of each major muscle group and 40 mins aerobic
endurance training - 5 minute warm-up, 30 minute walking or running, 5 minute cool down. “Each individual was
asked to exercise at an intensity of 40-50% heart rate reserve (moderate intensity) during the beginner’s stage; the
intensity of activity was then incrementally increased to 60-70% (moderate to vigorous intensity) by the end of the
developmental stage.”
Cognitive Activity Intervention: “Cognitive training took place once a week for approximately 30 min... The primary
element of the cognitive intervention consisted of the individual editing of worksheets. In addition, some partner
and group exercises were carried out. During the first 5 min, warm-up exercises were performed as a group (e.g.
, training of short-term memory) or homework was discussed. Some small amount of information was then given
relating to one of the following topics: information processing speed, attention, introduction to the memory model,
sensory memory, short-term memory, mnemonics, long-term memory, and memory aids. Following the distribution
of information, the following cognitive abilities were practiced for 25 min: short-term memory, visuospatial skills,
information processing speed, concentration, and logical reasoning. At the end of each session two additional exercises
were provided as homework.”
Combined Physical and Cognitive Activity Intervention: “The combined intervention consisted of the physical plus
cognitive interventions and took place twice a week. The cognitive training program was carried out at the first
training session of the week, before the physical training. The total duration of the first training session each week
therefore was 90 min, while the second session lasted only 60 min (consisting only of physical training).”
Wait-list Control: “An inactive waiting control group was selected to act as a comparison group. Study participants
in the control group were asked to continue their daily routines as before. To increase the motivation to participate
in the study, a 12-week fitness class was offered after the follow-up assessment.”
Outcomes Reasoning subtest of Leistungs-Prüf-System 50+ (LPS 50+)

Spatial relations subtest of LPS 50+
d2: Test of Attention
Trail Making Test Part A
Digit-Symbol Substitution Test (DSST from NAI)
Linde 2014 (Continued)
Word List test (NAI subtest)

2-km walking test to estimate VO2 max
Notes Testing was conducted at the Faculty of Sport Science at the University of Leipzig
Increase of cardiovascular fitness were not significantly different between the control and intervention groups
DATA AND ANALYSES
Comparison 1. Aerobic exercise versus any active intervention
No. of No. of
Outcome or subgroup title studies participants Statistical method Effect size
1 Cognitive speed 6 389 Std. Mean Difference (IV, Random, 95% CI) 0.12 [-0.08, 0.33]
1.1 Simple reaction time 2 113 Std. Mean Difference (IV, Random, 95% CI) 0.09 [-0.28, 0.46]
1.2 Choice reaction time 0 0 Std. Mean Difference (IV, Random, 95% CI) 0.0 [0.0, 0.0]
1.3 Trailmaking part A 1 49 Std. Mean Difference (IV, Random, 95% CI) -0.36 [-0.96, 0.24]
1.4 Digit symbol substitution 3 227 Std. Mean Difference (IV, Random, 95% CI) 0.24 [-0.03, 0.50]
2 Verbal memory functions 5 292 Std. Mean Difference (IV, Random, 95% CI) 0.08 [-0.38, 0.55]
(immediate)
2.1 16 words immediate recall 0 0 Std. Mean Difference (IV, Random, 95% CI) 0.0 [0.0, 0.0]
2.2 Randt Memory test story 1 65 Std. Mean Difference (IV, Random, 95% CI) 0.34 [-0.15, 0.83]
recall
2.3 Ross Information 1 30 Std. Mean Difference (IV, Random, 95% CI) 0.60 [-0.18, 1.37]
Processing Assessment
immediate memory
2.4 Wechsler Adult 1 24 Std. Mean Difference (IV, Random, 95% CI) -1.41 [-2.36, -0.45]
Intelligence Scales logical
memory immediate recall
2.5 Rey auditory verbal 1 124 Std. Mean Difference (IV, Random, 95% CI) 0.10 [-0.25, 0.45]
learning trial I-V
2.6 Hopkins Verbal Learning 1 49 Std. Mean Difference (IV, Random, 95% CI) 0.34 [-0.27, 0.94]
Test (immediate)
3 Visual memory functions 2 89 Std. Mean Difference (IV, Random, 95% CI) -0.26 [-0.97, 0.44]
(immediate)
3.1 Benton visual retention 1 65 Std. Mean Difference (IV, Random, 95% CI) 0.02 [-0.47, 0.50]
(#error)
3.2 Wechsler Memory Scales 1 24 Std. Mean Difference (IV, Random, 95% CI) -0.73 [-1.61, 0.15]
visual reproduction
4 Working memory 3 238 Std. Mean Difference (IV, Random, 95% CI) 0.10 [-0.16, 0.36]
4.1 Digit span backward 2 189 Std. Mean Difference (IV, Random, 95% CI) 0.16 [-0.13, 0.45]
4.2 2-Back (accuracy, Hits - 1 49 Std. Mean Difference (IV, Random, 95% CI) -0.14 [-0.74, 0.46]
False Alarms)
5 Memory functions (delayed) 3 249 Std. Mean Difference (IV, Random, 95% CI) 0.10 [-0.16, 0.35]
5.1 16 words delayed recall 0 0 Std. Mean Difference (IV, Random, 95% CI) 0.0 [0.0, 0.0]
learning delayed recall trial
5.3 10 words delayed recall 1 76 Std. Mean Difference (IV, Random, 95% CI) -0.10 [-0.55, 0.35]
Test - 12 words (delayed)
6 Executive functions 6 367 Std. Mean Difference (IV, Random, 95% CI) 0.38 [-0.14, 0.90]
6.1 Trailmaking part B 2 113 Std. Mean Difference (IV, Random, 95% CI) 0.27 [-0.11, 0.65]
6.2 Ross Information 1 30 Std. Mean Difference (IV, Random, 95% CI) 2.75 [1.69, 3.82]
Processing Assessment problem
solving and abstract reasoning
mental control
6.4 Task switching paradigm 1 124 Std. Mean Difference (IV, Random, 95% CI) 0.03 [-0.32, 0.38]
(accuracy)
6.5 Verbal fluency 0 0 Std. Mean Difference (IV, Random, 95% CI) 0.0 [0.0, 0.0]
6.6 Letter number sequencing 1 76 Std. Mean Difference (IV, Random, 95% CI) 0.07 [-0.38, 0.52]
7 Perception 3 178 Std. Mean Difference (IV, Random, 95% CI) -0.01 [-0.50, 0.48]
7.1 Face recognition (delayed 1 124 Std. Mean Difference (IV, Random, 95% CI) 0.17 [-0.18, 0.53]
recall)
Processing Assessment auditory
processing
7.3 Wechsler Adult 1 24 Std. Mean Difference (IV, Random, 95% CI) -0.73 [-1.61, 0.15]
Intelligence Scales visual
reproduction
8 Cognitive inhibition 4 314 Std. Mean Difference (IV, Random, 95% CI) -0.06 [-0.28, 0.17]
8.1 Stroop colour word 2 141 Std. Mean Difference (IV, Random, 95% CI) -0.13 [-0.46, 0.20]
(interference)
8.2 Stopping task (accuracy 1 124 Std. Mean Difference (IV, Random, 95% CI) 0.01 [-0.35, 0.36]
choice RT)
8.3 Flanker Task (Incongruent 1 49 Std. Mean Difference (IV, Random, 95% CI) 0.00 [-0.59, 0.60]
RT)
9 Visual attention 3 265 Std. Mean Difference (IV, Random, 95% CI) 0.22 [-0.03, 0.46]
9.1 Digit vigilance 0 0 Std. Mean Difference (IV, Random, 95% CI) 0.0 [0.0, 0.0]
9.2 Tracking (accuracy index) 0 0 Std. Mean Difference (IV, Random, 95% CI) 0.0 [0.0, 0.0]
9.3 2&7 test 1 65 Std. Mean Difference (IV, Random, 95% CI) 0.30 [-0.19, 0.79]
9.4 Visual search (accuracy) 1 124 Std. Mean Difference (IV, Random, 95% CI) 0.25 [-0.10, 0.60]
9.5 Covert orienting of 1 76 Std. Mean Difference (IV, Random, 95% CI) 0.09 [-0.36, 0.54]
visuospatial attention
10 Auditory attention 4 251 Mean Difference (IV, Random, 95% CI) 0.15 [-0.38, 0.69]
10.1 Digit span forward 4 251 Mean Difference (IV, Random, 95% CI) 0.15 [-0.38, 0.69]
11 Motor function 2 189 Std. Mean Difference (IV, Random, 95% CI) 0.08 [-0.20, 0.37]
11.1 Finger tapping 1 65 Std. Mean Difference (IV, Random, 95% CI) 0.19 [-0.30, 0.68]
11.2 Pursuit rotor task 1 124 Std. Mean Difference (IV, Random, 95% CI) 0.02 [-0.33, 0.38]
(tracking error)
12 Drop-out 7 469 Odds Ratio (M-H, Random, 95% CI) 0.96 [0.44, 2.10]
Comparison 2. Aerobic exercise versus no intervention
No. of No. of
1.1 Simple reaction time 2 109 Std. Mean Difference (IV, Random, 95% CI) -0.09 [-0.47, 0.29]
1.2 Choice reaction time 1 14 Std. Mean Difference (IV, Random, 95% CI) -0.53 [-1.60, 0.54]
1.3 Trailmaking part A 1 72 Std. Mean Difference (IV, Random, 95% CI) 0.31 [-0.15, 0.78]
(immediate)
2.1 Randt Memory test story 1 65 Std. Mean Difference (IV, Random, 95% CI) -0.04 [-0.53, 0.45]
recall
immediate memory
2.4 Wechsler Adult 0 0 Std. Mean Difference (IV, Random, 95% CI) 0.0 [0.0, 0.0]
learning trial I-V
2.6 Hopkins Verbal Learning 0 0 Std. Mean Difference (IV, Random, 95% CI) 0.0 [0.0, 0.0]
Test (immediate)
3 Visual memory functions 1 65 Std. Mean Difference (IV, Random, 95% CI) -0.09 [-0.57, 0.40]
(immediate)
3.1 Benton visual retention 1 65 Std. Mean Difference (IV, Random, 95% CI) -0.09 [-0.57, 0.40]
(#error)
3.2 Wechsler Memory Scales 0 0 Std. Mean Difference (IV, Random, 95% CI) 0.0 [0.0, 0.0]
visual reproduction
4 Working memory 2 137 Mean Difference (IV, Random, 95% CI) 0.30 [-0.54, 1.15]
4.1 Digit span backward 2 137 Mean Difference (IV, Random, 95% CI) 0.30 [-0.54, 1.15]
4.2 2-Back (accuracy, Hits - 0 0 Mean Difference (IV, Random, 95% CI) 0.0 [0.0, 0.0]
False Alarms)
5 Memory functions (delayed) 2 152 Std. Mean Difference (IV, Fixed, 95% CI) 0.09 [-0.23, 0.41]
5.1 16 words delayed recall 0 0 Std. Mean Difference (IV, Fixed, 95% CI) 0.0 [0.0, 0.0]
5.2 Rey auditory verbal 1 72 Std. Mean Difference (IV, Fixed, 95% CI) 0.25 [-0.21, 0.72]
5.3 10 words delayed recall 1 80 Std. Mean Difference (IV, Fixed, 95% CI) -0.05 [-0.49, 0.38]
5.4 Hopkins Verbal Learning 0 0 Std. Mean Difference (IV, Fixed, 95% CI) 0.0 [0.0, 0.0]
mental control
6.4 Task switching paradigm 0 0 Std. Mean Difference (IV, Random, 95% CI) 0.0 [0.0, 0.0]
(accuracy)
6.6 Letter number sequencing 1 80 Std. Mean Difference (IV, Random, 95% CI) -0.03 [-0.47, 0.41]
7 Cognitive inhibition 3 217 Std. Mean Difference (IV, Random, 95% CI) 0.20 [-0.06, 0.47]
7.1 Stroop colour word 3 217 Std. Mean Difference (IV, Random, 95% CI) 0.20 [-0.06, 0.47]
(interference)
7.2 Stopping task (accuracy 0 0 Std. Mean Difference (IV, Random, 95% CI) 0.0 [0.0, 0.0]
choice RT)
7.3 Flanker Task (Incongruent 0 0 Std. Mean Difference (IV, Random, 95% CI) 0.0 [0.0, 0.0]
RT)
8.2 Tracking (accuracy index) 1 10 Std. Mean Difference (IV, Random, 95% CI) 0.76 [-0.55, 2.07]
8.4 Visual search (accuracy) 0 0 Std. Mean Difference (IV, Random, 95% CI) 0.0 [0.0, 0.0]
8.5 Covert orienting of 1 80 Std. Mean Difference (IV, Random, 95% CI) -0.02 [-0.45, 0.42]
9 Auditory attention 1 65 Mean Difference (IV, Fixed, 95% CI) 0.16 [-1.01, 1.33]
9.1 Digit span forward 1 65 Mean Difference (IV, Fixed, 95% CI) 0.16 [-1.01, 1.33]
10 Motor function 1 65 Mean Difference (IV, Fixed, 95% CI) 0.10 [-7.87, 8.08]
10.1 Finger tapping 1 65 Mean Difference (IV, Fixed, 95% CI) 0.10 [-7.87, 8.08]
10.2 Pursuit rotor task 0 0 Mean Difference (IV, Fixed, 95% CI) 0.0 [0.0, 0.0]
(tracking error)
11 Drop-out 5 267 Odds Ratio (IV, Random, 95% CI) 1.84 [0.79, 4.29]
Comparison 3. Aerobic exercise versus flexibility/balance programme
No. of No. of
1.3 Trailmaking part A 0 0 Std. Mean Difference (IV, Random, 95% CI) 0.0 [0.0, 0.0]
(immediate)
recall
immediate memory
learning trial I-V
3 Visual memory functions 1 65 Mean Difference (IV, Fixed, 95% CI) 0.05 [-1.65, 1.76]
(immediate)
3.1 Benton visual retention 1 65 Mean Difference (IV, Fixed, 95% CI) 0.05 [-1.65, 1.76]
(#error)
3.2 Wechsler Memory Scales 0 0 Mean Difference (IV, Fixed, 95% CI) 0.0 [0.0, 0.0]
visual reproduction
4 Working memory 2 189 Mean Difference (IV, Random, 95% CI) 0.36 [-0.41, 1.12]
4.1 Digit span backward 2 189 Mean Difference (IV, Random, 95% CI) 0.36 [-0.41, 1.12]
5 Memory functions (delayed) 2 200 Std. Mean Difference (IV, Random, 95% CI) 0.08 [-0.20, 0.36]
5.1 16 words delayed recall 0 0 Std. Mean Difference (IV, Random, 95% CI) 0.0 [0.0, 0.0]
5.3 10 words delayed recall 1 76 Std. Mean Difference (IV, Random, 95% CI) -0.10 [-0.55, 0.35]
mental control
(accuracy)
6.6 Letter number sequencing 1 76 Std. Mean Difference (IV, Random, 95% CI) 0.07 [-0.38, 0.52]
7 Perception 2 144 Std. Mean Difference (IV, Random, 95% CI) 0.22 [-0.11, 0.54]
recall)
processing
reproduction
(interference)
choice RT)
9.5 Covert orienting of 1 76 Std. Mean Difference (IV, Random, 95% CI) 0.09 [-0.36, 0.54]
10 Auditory attention 2 189 Mean Difference (IV, Random, 95% CI) -0.17 [-0.83, 0.49]
10.1 Digit span forward 2 189 Mean Difference (IV, Random, 95% CI) -0.17 [-0.83, 0.49]
(tracking error)
Comparison 4. Aerobic exercise versus strength programme
No. of No. of
1 Verbal memory functions 1 20 Mean Difference (IV, Fixed, 95% CI) 0.30 [-4.17, 4.77]
(immediate)
1.1 Randt Memory test story 0 0 Mean Difference (IV, Fixed, 95% CI) 0.0 [0.0, 0.0]
recall
1.2 16 words immediate recall 0 0 Mean Difference (IV, Fixed, 95% CI) 0.0 [0.0, 0.0]
1.3 Ross Information 1 20 Mean Difference (IV, Fixed, 95% CI) 0.30 [-4.17, 4.77]
immediate memory
1.4 Wechsler Adult 0 0 Mean Difference (IV, Fixed, 95% CI) 0.0 [0.0, 0.0]
1.5 Rey auditory verbal 0 0 Mean Difference (IV, Fixed, 95% CI) 0.0 [0.0, 0.0]
learning trial I-V
2 Executive functions 1 20 Mean Difference (IV, Fixed, 95% CI) -2.30 [-4.49, -0.11]
2.1 Trailmaking part B 0 0 Mean Difference (IV, Fixed, 95% CI) 0.0 [0.0, 0.0]
2.2 Ross Information 1 20 Mean Difference (IV, Fixed, 95% CI) -2.30 [-4.49, -0.11]
2.3 Wechsler Memory Scales 0 0 Mean Difference (IV, Fixed, 95% CI) 0.0 [0.0, 0.0]
mental control
2.4 Word comparison (#error) 0 0 Mean Difference (IV, Fixed, 95% CI) 0.0 [0.0, 0.0]
2.5 Task switching paradigm 0 0 Mean Difference (IV, Fixed, 95% CI) 0.0 [0.0, 0.0]
(accuracy)
2.6 Verbal fluency 0 0 Mean Difference (IV, Fixed, 95% CI) 0.0 [0.0, 0.0]
3 Perception 1 20 Mean Difference (IV, Fixed, 95% CI) -0.5 [-2.93, 1.93]
3.1 Face recognition (delayed 0 0 Mean Difference (IV, Fixed, 95% CI) 0.0 [0.0, 0.0]
recall)
3.2 Ross Information 1 20 Mean Difference (IV, Fixed, 95% CI) -0.5 [-2.93, 1.93]
processing
3.3 Wechsler Adult 0 0 Mean Difference (IV, Fixed, 95% CI) 0.0 [0.0, 0.0]
reproduction
4 Cognitive speed 1 37 Mean Difference (IV, Fixed, 95% CI) -4.0 [-27.93, 19.93]
4.1 Simple reaction time 1 37 Mean Difference (IV, Fixed, 95% CI) -4.0 [-27.93, 19.93]
Comparison 5. Fitness Improved: aerobic exercise versus any active intervention
No. of No. of
1.1 Simple reaction time 1 37 Std. Mean Difference (IV, Random, 95% CI) -0.10 [-0.75, 0.54]
1.3 Trailmaking part A 1 49 Std. Mean Difference (IV, Random, 95% CI) -0.36 [-0.96, 0.24]
(immediate)
recall
immediate memory
2.4 Wechsler Adult 1 24 Std. Mean Difference (IV, Random, 95% CI) -1.41 [-2.36, -0.45]
learning trial I-V
Test (immediate)
3 Visual memory functions 2 89 Mean Difference (IV, Random, 95% CI) -0.59 [-2.04, 0.87]
(immediate)
3.1 Benton visual retention 1 65 Mean Difference (IV, Random, 95% CI) 0.05 [-1.65, 1.76]
(#error)
3.2 Wechsler Memory Scales 1 24 Mean Difference (IV, Random, 95% CI) -1.45 [-3.50, 0.60]
visual reproduction
4.2 2-Back (accuracy, Hits - 1 49 Std. Mean Difference (IV, Random, 95% CI) -0.14 [-0.74, 0.46]
False Alarms)
5 Memory functions (delayed) 2 173 Mean Difference (IV, Random, 95% CI) 0.48 [-0.29, 1.25]
5.1 16 words delayed recall 0 0 Mean Difference (IV, Random, 95% CI) 0.0 [0.0, 0.0]
5.2 Rey auditory verbal 1 124 Mean Difference (IV, Random, 95% CI) 0.5 [-0.44, 1.44]
5.3 10 words delayed recall 0 0 Mean Difference (IV, Random, 95% CI) 0.0 [0.0, 0.0]
5.4 Hopkins Verbal Learning 1 49 Mean Difference (IV, Random, 95% CI) 0.44 [-0.94, 1.82]
mental control
(accuracy)
6.6 Letter number sequencing 0 0 Std. Mean Difference (IV, Random, 95% CI) 0.0 [0.0, 0.0]
7 Perception 3 178 Std. Mean Difference (IV, Random, 95% CI) -0.01 [-0.50, 0.48]
recall)
processing
7.3 Wechsler Adult 1 24 Std. Mean Difference (IV, Random, 95% CI) -0.73 [-1.61, 0.15]
reproduction
(interference)
choice RT)
8.3 Flanker Task (Incongruent 1 49 Std. Mean Difference (IV, Random, 95% CI) 0.00 [-0.59, 0.60]
RT)
9.5 Covert orienting of 0 0 Std. Mean Difference (IV, Random, 95% CI) 0.0 [0.0, 0.0]
10 Auditory attention 3 213 Mean Difference (IV, Random, 95% CI) 0.15 [-0.49, 0.79]
10.1 Digit span forward 3 213 Mean Difference (IV, Random, 95% CI) 0.15 [-0.49, 0.79]
(tracking error)
Comparison 6. Fitness improved: aerobic exercise versus no intervention
No. of No. of
1.2 Choice reaction time 1 14 Std. Mean Difference (IV, Random, 95% CI) -0.53 [-1.60, 0.54]
1.3 Trailmaking part A 1 72 Std. Mean Difference (IV, Random, 95% CI) 0.31 [-0.15, 0.78]
(immediate)
2.1 Randt Memory test story 1 65 Std. Mean Difference (IV, Random, 95% CI) -0.04 [-0.53, 0.45]
recall
immediate memory
learning trial I-V
2.6 Hopkins Verbal Learning 0 0 Std. Mean Difference (IV, Random, 95% CI) 0.0 [0.0, 0.0]
Test (immediate)
3 Visual memory functions 1 65 Mean Difference (IV, Random, 95% CI) -0.28 [-1.87, 1.30]
(immediate)
3.1 Benton visual retention 1 65 Mean Difference (IV, Random, 95% CI) -0.28 [-1.87, 1.30]
(#error)
3.2 Wechsler Memory Scales 0 0 Mean Difference (IV, Random, 95% CI) 0.0 [0.0, 0.0]
visual reproduction
4.2 2-Back (accuracy, Hits - 0 0 Std. Mean Difference (IV, Random, 95% CI) 0.0 [0.0, 0.0]
False Alarms)
5 Memory functions (delayed) 1 72 Mean Difference (IV, Fixed, 95% CI) 0.92 [-0.75, 2.59]
5.1 Rey auditory verbal 1 72 Mean Difference (IV, Fixed, 95% CI) 0.92 [-0.75, 2.59]
mental control
6.4 Task switching paradigm 0 0 Std. Mean Difference (IV, Random, 95% CI) 0.0 [0.0, 0.0]
(accuracy)
6.6 Letter number sequencing 0 0 Std. Mean Difference (IV, Random, 95% CI) 0.0 [0.0, 0.0]
7 Cognitive inhibition 2 137 Std. Mean Difference (IV, Random, 95% CI) 0.29 [-0.04, 0.63]
7.1 Stroop colour word 2 137 Std. Mean Difference (IV, Random, 95% CI) 0.29 [-0.04, 0.63]
(interference)
7.2 Stopping task (accuracy 0 0 Std. Mean Difference (IV, Random, 95% CI) 0.0 [0.0, 0.0]
choice RT)
7.3 Flanker Task (Incongruent 0 0 Std. Mean Difference (IV, Random, 95% CI) 0.0 [0.0, 0.0]
RT)
8.4 Visual search (accuracy) 0 0 Std. Mean Difference (IV, Random, 95% CI) 0.0 [0.0, 0.0]
8.5 Covert orienting of 0 0 Std. Mean Difference (IV, Random, 95% CI) 0.0 [0.0, 0.0]
9 Auditory attention 1 65 Mean Difference (IV, Fixed, 95% CI) 0.16 [-1.01, 1.33]
9.1 Digit span forward 1 65 Mean Difference (IV, Fixed, 95% CI) 0.16 [-1.01, 1.33]
10 Motor function 1 65 Mean Difference (IV, Fixed, 95% CI) 0.10 [-7.87, 8.08]
10.1 Finger tapping 1 65 Mean Difference (IV, Fixed, 95% CI) 0.10 [-7.87, 8.08]
10.2 Pursuit rotor task 0 0 Mean Difference (IV, Fixed, 95% CI) 0.0 [0.0, 0.0]
(tracking error)
11 Drop-out 3 164 Odds Ratio (IV, Random, 95% CI) 1.50 [0.50, 4.50]
Analysis 1.1. Comparison 1 Aerobic exercise versus any active intervention, Outcome 1 Cognitive speed.
Review: Aerobic exercise to improve cognitive function in older people without known cognitive impairment
Comparison: 1 Aerobic exercise versus any active intervention
Outcome: 1 Cognitive speed
Std. Std.
Mean Mean
Study or subgroup Treatment Control Difference Weight Difference
N Mean(SD) N Mean(SD) IV,Random,95% CI IV,Random,95% CI
1 Simple reaction time

Oken 2006 38 -321.9 (79.7) 38 -335.6 (74) 20.1 % 0.18 [ -0.27, 0.63 ]
Panton 1990 17 -274 (28.9) 20 -270 (44.7) 9.7 % -0.10 [ -0.75, 0.54 ]
Subtotal (95% CI) 55 58 29.8 % 0.09 [ -0.28, 0.46 ]

Heterogeneity: Tau2 = 0.0; Chi2 = 0.48, df = 1 (P = 0.49); I2 =0.0%
Test for overall effect: Z = 0.45 (P = 0.65)
2 Choice reaction time
Subtotal (95% CI) 0 0 Not estimable
Heterogeneity: not applicable
Test for overall effect: not applicable
3 Trailmaking part A
Legault 2011 16 -37.44 (11.331) 33 -34.24 (7.166013) 11.3 % -0.36 [ -0.96, 0.24 ]
Subtotal (95% CI) 16 33 11.3 % -0.36 [ -0.96, 0.24 ]

4 Digit symbol substitution
Blumenthal 1989 31 52.4839 (8.6) 34 48.65 (9.9985) 16.9 % 0.40 [ -0.09, 0.90 ]
Emery 1990a 14 35.8 (12.6) 24 32.9 (11.3) 9.3 % 0.24 [ -0.42, 0.90 ]
Kramer 2001 58 96.5 (3.88) 66 95.7 (6.34) 32.7 % 0.15 [ -0.20, 0.50 ]
Subtotal (95% CI) 103 124 58.9 % 0.24 [ -0.03, 0.50 ]

Total (95% CI) 174 215 100.0 % 0.12 [ -0.08, 0.33 ]
Test for subgroup differences: Chi2 = 3.24, df = 2 (P = 0.20), I2 =38%
-4 -2 0 2 4
Favours control Favours aerobic
Analysis 1.2. Comparison 1 Aerobic exercise versus any active intervention, Outcome 2 Verbal memory
functions (immediate).
Outcome: 2 Verbal memory functions (immediate)
Std. Std.
Mean Mean
1 16 words immediate recall

2 Randt Memory test story recall
Blumenthal 1989 11.06129 (3.866701) 31 34 9.8 (3.450362) 23.1 % 0.34 [ -0.15, 0.83 ]
Subtotal (95% CI) 31 34 23.1 % 0.34 [ -0.15, 0.83 ]

3 Ross Information Processing Assessment immediate memory
Moul 1995 10 23.5 (3.48) 20 19.6 (7.323196) 16.6 % 0.60 [ -0.18, 1.37 ]
Subtotal (95% CI) 10 20 16.6 % 0.60 [ -0.18, 1.37 ]

4 Wechsler Adult Intelligence Scales logical memory immediate recall
Fabre 2002 8 7.4 (1) 16 9.1 (1.236932) 13.4 % -1.41 [ -2.36, -0.45 ]
Subtotal (95% CI) 8 16 13.4 % -1.41 [ -2.36, -0.45 ]

5 Rey auditory verbal learning trial I-V
Kramer 2001 58 9.95 (1.8) 66 9.75 (2.1) 26.4 % 0.10 [ -0.25, 0.45 ]
Subtotal (95% CI) 58 66 26.4 % 0.10 [ -0.25, 0.45 ]

6 Hopkins Verbal Learning Test (immediate)
Legault 2011 16 6.38 (1.784) 33 5.82 (1.595437) 20.4 % 0.34 [ -0.27, 0.94 ]
Subtotal (95% CI) 16 33 20.4 % 0.34 [ -0.27, 0.94 ]

Total (95% CI) 123 169 100.0 % 0.08 [ -0.38, 0.55 ]
Heterogeneity: Tau2 = 0.18; Chi2 = 12.53, df = 4 (P = 0.01); I2 =68%
-4 -2 0 2 4
Analysis 1.3. Comparison 1 Aerobic exercise versus any active intervention, Outcome 3 Visual memory
Outcome: 3 Visual memory functions (immediate)
Std. Std.
Mean Mean
1 Benton visual retention (#error)

Blumenthal 1989 -5.545161 (3.348185) 31 34 -5.6 (3.653081) 62.6 % 0.02 [ -0.47, 0.50 ]
Subtotal (95% CI) 31 34 62.6 % 0.02 [ -0.47, 0.50 ]

2 Wechsler Memory Scales visual reproduction
Fabre 2002 8 10.2 (2.8) 16 11.65 (1.334166) 37.4 % -0.73 [ -1.61, 0.15 ]
Subtotal (95% CI) 8 16 37.4 % -0.73 [ -1.61, 0.15 ]

Total (95% CI) 39 50 100.0 % -0.26 [ -0.97, 0.44 ]
-4 -2 0 2 4
Analysis 1.4. Comparison 1 Aerobic exercise versus any active intervention, Outcome 4 Working memory.
Outcome: 4 Working memory
Std. Std.
Mean Mean
1 Digit span backward

Blumenthal 1989 7.851613 (2.852162) 31 34 6.9 (2.280351) 27.7 % 0.37 [ -0.12, 0.86 ]
Kramer 2001 58 7.2 (1.83) 66 7.1 (2.11) 53.6 % 0.05 [ -0.30, 0.40 ]
Subtotal (95% CI) 89 100 81.3 % 0.16 [ -0.13, 0.45 ]

2 2-Back (accuracy, Hits - False Alarms)
Legault 2011 16 0.5723 (0.18167) 33 0.6 (0.231733) 18.7 % -0.14 [ -0.74, 0.46 ]
Subtotal (95% CI) 16 33 18.7 % -0.14 [ -0.74, 0.46 ]

Total (95% CI) 105 133 100.0 % 0.10 [ -0.16, 0.36 ]
Test for subgroup differences: Chi2 = 0.77, df = 1 (P = 0.38), I2 =0.0%
-4 -2 0 2 4
Analysis 1.5. Comparison 1 Aerobic exercise versus any active intervention, Outcome 5 Memory functions
(delayed).
Outcome: 5 Memory functions (delayed)
Std. Std.
Mean Mean
1 16 words delayed recall

2 Rey auditory verbal learning delayed recall trial
Kramer 2001 58 11.4 (2.7) 66 10.9 (2.6) 50.8 % 0.19 [ -0.17, 0.54 ]
Subtotal (95% CI) 58 66 50.8 % 0.19 [ -0.17, 0.54 ]

Oken 2006 38 7 (2) 38 7.2 (1.9) 31.4 % -0.10 [ -0.55, 0.35 ]
Subtotal (95% CI) 38 38 31.4 % -0.10 [ -0.55, 0.35 ]

4 Hopkins Verbal Learning Test - 12 words (delayed)
Legault 2011 16 9.38 (2.187) 33 8.94 (2.539838) 17.8 % 0.18 [ -0.42, 0.78 ]
Subtotal (95% CI) 16 33 17.8 % 0.18 [ -0.42, 0.78 ]

Total (95% CI) 112 137 100.0 % 0.10 [ -0.16, 0.35 ]
-4 -2 0 2 4
Analysis 1.6. Comparison 1 Aerobic exercise versus any active intervention, Outcome 6 Executive functions.
Outcome: 6 Executive functions
Std. Std.
Mean Mean
1 Trailmaking part B
Blumenthal 1989 -79.487097 (27.626574) 31 34 -90.9 (34.001912) 18.5 % 0.36 [ -0.13, 0.85 ]
Legault 2011 15 -81.53 (49.11) 33 -86.18 (28.530309) 17.0 % 0.13 [ -0.48, 0.74 ]
Subtotal (95% CI) 46 67 35.5 % 0.27 [ -0.11, 0.65 ]

2 Ross Information Processing Assessment problem solving and abstract reasoning
Moul 1995 10 27.3 (3.48) 20 18.46 (2.940153) 11.4 % 2.75 [ 1.69, 3.82 ]
Subtotal (95% CI) 10 20 11.4 % 2.75 [ 1.69, 3.82 ]

Test for overall effect: Z = 5.06 (P < 0.00001)
3 Wechsler Memory Scales mental control
Fabre 2002 8 5.9 (0.9) 16 6.2 (0.961769) 13.8 % -0.31 [ -1.16, 0.55 ]
Subtotal (95% CI) 8 16 13.8 % -0.31 [ -1.16, 0.55 ]

4 Task switching paradigm (accuracy)
Kramer 2001 58 95.8 (6.32) 66 95.6 (7.39) 20.2 % 0.03 [ -0.32, 0.38 ]
Subtotal (95% CI) 58 66 20.2 % 0.03 [ -0.32, 0.38 ]

5 Verbal fluency
6 Letter number sequencing
Oken 2006 38 10.7 (2.8) 38 10.5 (2.6) 19.0 % 0.07 [ -0.38, 0.52 ]
Subtotal (95% CI) 38 38 19.0 % 0.07 [ -0.38, 0.52 ]

Total (95% CI) 160 207 100.0 % 0.38 [ -0.14, 0.90 ]
-4 -2 0 2 4
Analysis 1.7. Comparison 1 Aerobic exercise versus any active intervention, Outcome 7 Perception.
Outcome: 7 Perception
Std. Std.
Mean Mean
1 Face recognition (delayed recall)

Kramer 2001 58 75.1 (18.28) 66 71.4 (23.56) 52.0 % 0.17 [ -0.18, 0.53 ]
Subtotal (95% CI) 58 66 52.0 % 0.17 [ -0.18, 0.53 ]

2 Ross Information Processing Assessment auditory processing
Moul 1995 10 27.7 (3.8) 20 27.1 (2.124265) 26.3 % 0.21 [ -0.55, 0.97 ]
Subtotal (95% CI) 10 20 26.3 % 0.21 [ -0.55, 0.97 ]

3 Wechsler Adult Intelligence Scales visual reproduction
Fabre 2002 8 10.2 (2.8) 16 11.65 (1.334166) 21.7 % -0.73 [ -1.61, 0.15 ]
Subtotal (95% CI) 8 16 21.7 % -0.73 [ -1.61, 0.15 ]

Total (95% CI) 76 102 100.0 % -0.01 [ -0.50, 0.48 ]
-4 -2 0 2 4
Analysis 1.8. Comparison 1 Aerobic exercise versus any active intervention, Outcome 8 Cognitive inhibition.
Outcome: 8 Cognitive inhibition
Std. Std.
Mean Mean
1 Stroop colour word (interference)

Blumenthal 1989 -1.670968 (6.298905) 31 34 -1.25 (6.140033) 21.1 % -0.07 [ -0.55, 0.42 ]
Oken 2006 38 -10.8 (4.3) 38 -10 (4.6) 24.6 % -0.18 [ -0.63, 0.27 ]
Subtotal (95% CI) 69 72 45.8 % -0.13 [ -0.46, 0.20 ]

2 Stopping task (accuracy choice RT)
Kramer 2001 58 93.1 (132.51) 66 92.3 (150.3) 40.2 % 0.01 [ -0.35, 0.36 ]
Subtotal (95% CI) 58 66 40.2 % 0.01 [ -0.35, 0.36 ]

3 Flanker Task (Incongruent RT)
Legault 2011 -538.8438 (67.86966) 16 -539.08 (80.523323) 33 14.0 % 0.00 [ -0.59, 0.60 ]
Subtotal (95% CI) 16 33 14.0 % 0.00 [ -0.59, 0.60 ]

Total (95% CI) 143 171 100.0 % -0.06 [ -0.28, 0.17 ]
-4 -2 0 2 4
Analysis 1.9. Comparison 1 Aerobic exercise versus any active intervention, Outcome 9 Visual attention.
Outcome: 9 Visual attention
Std. Std.
Mean Mean
1 Digit vigilance
2 Tracking (accuracy index)
3 2%7 test
Blumenthal 1989 -6.790323 (5.064251) 31 34 -9.1 (9.257699) 24.4 % 0.30 [ -0.19, 0.79 ]
Subtotal (95% CI) 31 34 24.4 % 0.30 [ -0.19, 0.79 ]

4 Visual search (accuracy)
Kramer 2001 58 97.75 (3.34) 66 96.95 (3.02) 46.7 % 0.25 [ -0.10, 0.60 ]
Subtotal (95% CI) 58 66 46.7 % 0.25 [ -0.10, 0.60 ]

5 Covert orienting of visuospatial attention
Oken 2006 38 -37.1 (34.1) 38 -40.3 (36.8) 28.9 % 0.09 [ -0.36, 0.54 ]
Subtotal (95% CI) 38 38 28.9 % 0.09 [ -0.36, 0.54 ]

Total (95% CI) 127 138 100.0 % 0.22 [ -0.03, 0.46 ]
-4 -2 0 2 4
Analysis 1.10. Comparison 1 Aerobic exercise versus any active intervention, Outcome 10 Auditory
attention.
Outcome: 10 Auditory attention
Mean Mean
1 Digit span forward

Blumenthal 1989 8.929032 (2.570791) 31 34 8.6 (2.039608) 18.1 % 0.33 [ -0.81, 1.46 ]
Emery 1990a 14 11.5 (4.3) 24 11.4 (4.2) 3.4 % 0.10 [ -2.71, 2.91 ]
Fabre 2002 8 6.1 (0.7) 16 5.55 (0.806226) 42.7 % 0.55 [ -0.08, 1.18 ]
Kramer 2001 58 8 (1.98) 66 8.4 (2.11) 35.8 % -0.40 [ -1.12, 0.32 ]
Total (95% CI) 111 140 100.0 % 0.15 [ -0.38, 0.69 ]

Test for subgroup differences: Not applicable
-4 -2 0 2 4
Analysis 1.11. Comparison 1 Aerobic exercise versus any active intervention, Outcome 11 Motor function.
Outcome: 11 Motor function
Std. Std.
Mean Mean
1 Finger tapping
Blumenthal 1989 123.616129 (14.348363) 31 34 120.35 (18.774051) 34.3 % 0.19 [ -0.30, 0.68 ]
Subtotal (95% CI) 31 34 34.3 % 0.19 [ -0.30, 0.68 ]

2 Pursuit rotor task (tracking error)
Kramer 2001 58 -29.65 (6.25) 66 -29.8 (6.56) 65.7 % 0.02 [ -0.33, 0.38 ]
Subtotal (95% CI) 58 66 65.7 % 0.02 [ -0.33, 0.38 ]

Total (95% CI) 89 100 100.0 % 0.08 [ -0.20, 0.37 ]
-4 -2 0 2 4
Analysis 1.12. Comparison 1 Aerobic exercise versus any active intervention, Outcome 12 Drop-out.
Outcome: 12 Drop-out
Study or subgroup Treatment Control Odds Ratio Weight Odds Ratio

M- M-
H,Random,95% H,Random,95%
n/N n/N CI CI
Blumenthal 1989 2/33 2/33 12.1 % 1.00 [ 0.13, 7.55 ]
Emery 1990a 1/15 4/15 9.6 % 0.20 [ 0.02, 2.02 ]
Fabre 2002 0/8 0/16 Not estimable
Kramer 2001 25/83 25/91 44.6 % 1.14 [ 0.59, 2.20 ]
Legault 2011 3/18 0/36 6.1 % 16.48 [ 0.80, 338.51 ]
Moul 1995 0/10 0/20 Not estimable
Oken 2006 6/44 9/47 27.6 % 0.67 [ 0.22, 2.06 ]
Total (95% CI) 211 258 100.0 % 0.96 [ 0.44, 2.10 ]

Total events: 37 (Treatment), 40 (Control)
0.01 0.1 1 10 100

Favours aerobic Favours control
Analysis 2.1. Comparison 2 Aerobic exercise versus no intervention, Outcome 1 Cognitive speed.
Comparison: 2 Aerobic exercise versus no intervention
Std. Std.
Mean Mean

Oken 2006 38 -321.9 (79.7) 42 -311.4 (78.5) 29.1 % -0.13 [ -0.57, 0.31 ]
Panton 1990 17 -274 (28.9) 12 -275 (52) 12.8 % 0.02 [ -0.71, 0.76 ]
Subtotal (95% CI) 55 54 41.9 % -0.09 [ -0.47, 0.29 ]

Whitehurst 1991 7 -0.41 (0.03) 7 -0.39 (0.04) 6.5 % -0.53 [ -1.60, 0.54 ]
Subtotal (95% CI) 7 7 6.5 % -0.53 [ -1.60, 0.54 ]

Langlois 2012 36 -44.08 (13.37) 36 -49.49 (20.21) 26.9 % 0.31 [ -0.15, 0.78 ]
Subtotal (95% CI) 36 36 26.9 % 0.31 [ -0.15, 0.78 ]

Blumenthal 1989 52.483871 (8.968893) 31 34 48.16 (10.18025) 24.7 % 0.44 [ -0.05, 0.94 ]
Subtotal (95% CI) 31 34 24.7 % 0.44 [ -0.05, 0.94 ]

Total (95% CI) 129 131 100.0 % 0.12 [ -0.16, 0.41 ]
-4 -2 0 2 4
Analysis 2.2. Comparison 2 Aerobic exercise versus no intervention, Outcome 2 Verbal memory functions
(immediate).
Std. Std.
Mean Mean

Blumenthal 1989 11.06129 (3.866701) 31 34 11.21 (3.845777) 47.5 % -0.04 [ -0.53, 0.45 ]
Subtotal (95% CI) 31 34 47.5 % -0.04 [ -0.53, 0.45 ]

Langlois 2012 36 10.92 (3.11) 36 10.19 (3.78) 52.5 % 0.21 [ -0.25, 0.67 ]
Subtotal (95% CI) 36 36 52.5 % 0.21 [ -0.25, 0.67 ]

Total (95% CI) 67 70 100.0 % 0.09 [ -0.24, 0.43 ]
-4 -2 0 2 4
Analysis 2.3. Comparison 2 Aerobic exercise versus no intervention, Outcome 3 Visual memory functions
(immediate).
Std. Std.
Mean Mean

Blumenthal 1989 -5.545161 (3.348185) 31 34 -5.26 (3.15352) 100.0 % -0.09 [ -0.57, 0.40 ]
Subtotal (95% CI) 31 34 100.0 % -0.09 [ -0.57, 0.40 ]

Total (95% CI) 31 34 100.0 % -0.09 [ -0.57, 0.40 ]
-4 -2 0 2 4
Analysis 2.4. Comparison 2 Aerobic exercise versus no intervention, Outcome 4 Working memory.
Mean Mean

Blumenthal 1989 31 7.8 (2.852162) 34 7.34 (2.211334) 45.4 % 0.46 [ -0.78, 1.71 ]
Langlois 2012 36 6.64 (2.47) 36 6.47 (2.46) 54.6 % 0.17 [ -0.97, 1.31 ]
Subtotal (95% CI) 67 70 100.0 % 0.30 [ -0.54, 1.15 ]

Total (95% CI) 67 70 100.0 % 0.30 [ -0.54, 1.15 ]
-4 -2 0 2 4
Analysis 2.5. Comparison 2 Aerobic exercise versus no intervention, Outcome 5 Memory functions
(delayed).
Std. Std.
Mean Mean
N Mean(SD) N Mean(SD) IV,Fixed,95% CI IV,Fixed,95% CI

Langlois 2012 36 10.56 (3.21) 36 9.64 (3.99) 47.2 % 0.25 [ -0.21, 0.72 ]
Subtotal (95% CI) 36 36 47.2 % 0.25 [ -0.21, 0.72 ]

Oken 2006 38 7 (2) 42 7.1 (1.6) 52.8 % -0.05 [ -0.49, 0.38 ]
Subtotal (95% CI) 38 42 52.8 % -0.05 [ -0.49, 0.38 ]

Total (95% CI) 74 78 100.0 % 0.09 [ -0.23, 0.41 ]
Heterogeneity: Chi2 = 0.88, df = 1 (P = 0.35); I2 =0.0%
-4 -2 0 2 4
Analysis 2.6. Comparison 2 Aerobic exercise versus no intervention, Outcome 6 Executive functions.
Std. Std.
Mean Mean
Blumenthal 1989 -79.487097 (27.626574) 31 34 -81.48 (36.321516) 31.5 % 0.06 [ -0.43, 0.55 ]
Langlois 2012 36 -104.95 (37.85) 36 -136.2 (72.86) 32.8 % 0.53 [ 0.06, 1.00 ]
Subtotal (95% CI) 67 70 64.3 % 0.30 [ -0.16, 0.76 ]

5 Verbal fluency
Oken 2006 38 10.7 (2.8) 42 10.8 (4) 35.7 % -0.03 [ -0.47, 0.41 ]
Subtotal (95% CI) 38 42 35.7 % -0.03 [ -0.47, 0.41 ]

Total (95% CI) 105 112 100.0 % 0.18 [ -0.16, 0.53 ]
-4 -2 0 2 4
Analysis 2.7. Comparison 2 Aerobic exercise versus no intervention, Outcome 7 Cognitive inhibition.
Std. Std.
Mean Mean

Blumenthal 1989 -1.670968 (6.298905) 31 34 -4.53 (6.657397) 29.5 % 0.44 [ -0.06, 0.93 ]
Langlois 2012 36 -141.67 (46.5) 36 -149.38 (42.95) 33.4 % 0.17 [ -0.29, 0.63 ]
Oken 2006 38 -10.8 (4.3) 42 -11 (3.7) 37.1 % 0.05 [ -0.39, 0.49 ]
Subtotal (95% CI) 105 112 100.0 % 0.20 [ -0.06, 0.47 ]

Total (95% CI) 105 112 100.0 % 0.20 [ -0.06, 0.47 ]
-4 -2 0 2 4
Analysis 2.8. Comparison 2 Aerobic exercise versus no intervention, Outcome 8 Visual attention.
Std. Std.
Mean Mean
1 Digit vigilance
Bakken 2001 5 10.1 (23.4) 5 -16.2 (37.3) 5.8 % 0.76 [ -0.55, 2.07 ]
Subtotal (95% CI) 5 5 5.8 % 0.76 [ -0.55, 2.07 ]

3 2%7 test
Blumenthal 1989 6.790323 (5.064251) 31 34 6.57 (4.97189) 42.2 % 0.04 [ -0.44, 0.53 ]
Subtotal (95% CI) 31 34 42.2 % 0.04 [ -0.44, 0.53 ]

Oken 2006 38 -37.1 (34.1) 42 -36.6 (30) 52.0 % -0.02 [ -0.45, 0.42 ]
Subtotal (95% CI) 38 42 52.0 % -0.02 [ -0.45, 0.42 ]

Total (95% CI) 74 81 100.0 % 0.05 [ -0.26, 0.37 ]
-4 -2 0 2 4
Analysis 2.9. Comparison 2 Aerobic exercise versus no intervention, Outcome 9 Auditory attention.
Mean Mean

Blumenthal 1989 8.929032 (2.570791) 31 34 8.77 (2.196588) 100.0 % 0.16 [ -1.01, 1.33 ]
Total (95% CI) 31 34 100.0 % 0.16 [ -1.01, 1.33 ]

-4 -2 0 2 4
Analysis 2.10. Comparison 2 Aerobic exercise versus no intervention, Outcome 10 Motor function.
Mean Mean
1 Finger tapping
Blumenthal 1989 123.616129 (14.348363) 31 34 123.51 (18.367362) 100.0 % 0.10 [ -7.87, 8.08 ]
Subtotal (95% CI) 31 34 100.0 % 0.10 [ -7.87, 8.08 ]

Total (95% CI) 31 34 100.0 % 0.10 [ -7.87, 8.08 ]
-10 -5 0 5 10
Analysis 2.11. Comparison 2 Aerobic exercise versus no intervention, Outcome 11 Drop-out.

n/N n/N IV,Random,95% CI IV,Random,95% CI
Bakken 2001 3/8 2/7 15.1 % 1.50 [ 0.17, 13.23 ]
Blumenthal 1989 2/33 2/34 17.5 % 1.03 [ 0.14, 7.79 ]
Langlois 2012 7/43 4/40 41.5 % 1.75 [ 0.47, 6.50 ]
Oken 2006 6/44 2/44 25.9 % 3.32 [ 0.63, 17.43 ]
Whitehurst 1991 0/7 0/7 Not estimable
Total (95% CI) 135 132 100.0 % 1.84 [ 0.79, 4.29 ]

0.01 0.1 1 10 100

Analysis 3.1. Comparison 3 Aerobic exercise versus flexibility/balance programme, Outcome 1 Cognitive
speed.
Comparison: 3 Aerobic exercise versus flexibility/balance programme
Std. Std.
Mean Mean

Oken 2006 38 -321.9 (79.7) 38 -335.6 (74) 28.9 % 0.18 [ -0.27, 0.63 ]
Subtotal (95% CI) 38 38 28.9 % 0.18 [ -0.27, 0.63 ]

Blumenthal 1989 31 52.4839 (8.6) 34 48.65 (9.9985) 24.2 % 0.40 [ -0.09, 0.90 ]
Kramer 2001 58 96.6 (3.88) 66 95.7 (6.34) 46.9 % 0.17 [ -0.19, 0.52 ]
Subtotal (95% CI) 89 100 71.1 % 0.25 [ -0.04, 0.54 ]

Total (95% CI) 127 138 100.0 % 0.23 [ -0.01, 0.47 ]
-4 -2 0 2 4
Favours control Favours treatment
Analysis 3.2. Comparison 3 Aerobic exercise versus flexibility/balance programme, Outcome 2 Verbal
Std. Std.
Mean Mean

Blumenthal 1989 11.06129 (3.866701) 31 34 9.8 (3.450362) 36.8 % 0.34 [ -0.15, 0.83 ]
Subtotal (95% CI) 31 34 36.8 % 0.34 [ -0.15, 0.83 ]

Moul 1995 10 23.5 (3.48) 10 16 (8.22) 16.2 % 1.14 [ 0.18, 2.10 ]
Subtotal (95% CI) 10 10 16.2 % 1.14 [ 0.18, 2.10 ]

Kramer 2001 58 9.95 (1.8) 66 9.75 (2.1) 46.9 % 0.10 [ -0.25, 0.45 ]
Subtotal (95% CI) 58 66 46.9 % 0.10 [ -0.25, 0.45 ]

Total (95% CI) 99 110 100.0 % 0.36 [ -0.09, 0.80 ]
-4 -2 0 2 4
Analysis 3.3. Comparison 3 Aerobic exercise versus flexibility/balance programme, Outcome 3 Visual
Mean Mean

Blumenthal 1989 -5.545161 (3.348185) 31 34 -5.6 (3.653081) 100.0 % 0.05 [ -1.65, 1.76 ]
Subtotal (95% CI) 31 34 100.0 % 0.05 [ -1.65, 1.76 ]

Total (95% CI) 31 34 100.0 % 0.05 [ -1.65, 1.76 ]
-4 -2 0 2 4
Analysis 3.4. Comparison 3 Aerobic exercise versus flexibility/balance programme, Outcome 4 Working
memory.
Mean Mean

Blumenthal 1989 7.851613 (2.852162) 31 34 6.9 (2.280351) 30.0 % 0.95 [ -0.31, 2.21 ]
Kramer 2001 58 7.2 (1.83) 66 7.1 (2.11) 70.0 % 0.10 [ -0.59, 0.79 ]
Total (95% CI) 89 100 100.0 % 0.36 [ -0.41, 1.12 ]

-4 -2 0 2 4
Analysis 3.5. Comparison 3 Aerobic exercise versus flexibility/balance programme, Outcome 5 Memory
Std. Std.
Mean Mean

Kramer 2001 58 11.4 (2.7) 66 10.9 (2.6) 61.8 % 0.19 [ -0.17, 0.54 ]
Subtotal (95% CI) 58 66 61.8 % 0.19 [ -0.17, 0.54 ]

Oken 2006 38 7 (2) 38 7.2 (1.9) 38.2 % -0.10 [ -0.55, 0.35 ]
Subtotal (95% CI) 38 38 38.2 % -0.10 [ -0.55, 0.35 ]

Total (95% CI) 96 104 100.0 % 0.08 [ -0.20, 0.36 ]
-4 -2 0 2 4
Analysis 3.6. Comparison 3 Aerobic exercise versus flexibility/balance programme, Outcome 6 Executive
functions.
Std. Std.
Mean Mean
Blumenthal 1989 -79.487097 (27.626574) 31 34 -90.9 (34.001912) 25.6 % 0.36 [ -0.13, 0.85 ]
Subtotal (95% CI) 31 34 25.6 % 0.36 [ -0.13, 0.85 ]

Moul 1995 10 27.3 (3.48) 10 23 (4.11) 9.5 % 1.08 [ 0.13, 2.03 ]
Subtotal (95% CI) 10 10 9.5 % 1.08 [ 0.13, 2.03 ]

Kramer 2001 58 95.8 (6.32) 66 95.6 (7.39) 36.4 % 0.03 [ -0.32, 0.38 ]
Subtotal (95% CI) 58 66 36.4 % 0.03 [ -0.32, 0.38 ]

5 Verbal fluency
Oken 2006 38 10.7 (2.8) 38 10.5 (2.6) 28.4 % 0.07 [ -0.38, 0.52 ]
Subtotal (95% CI) 38 38 28.4 % 0.07 [ -0.38, 0.52 ]

Total (95% CI) 137 148 100.0 % 0.23 [ -0.09, 0.55 ]
-4 -2 0 2 4
Analysis 3.7. Comparison 3 Aerobic exercise versus flexibility/balance programme, Outcome 7 Perception.
Std. Std.
Mean Mean

Kramer 2001 58 75.1 (18.28) 66 71.4 (23.56) 86.4 % 0.17 [ -0.18, 0.53 ]
Subtotal (95% CI) 58 66 86.4 % 0.17 [ -0.18, 0.53 ]

Moul 1995 10 27.7 (3.8) 10 26 (2.85) 13.6 % 0.48 [ -0.41, 1.38 ]
Subtotal (95% CI) 10 10 13.6 % 0.48 [ -0.41, 1.38 ]

Total (95% CI) 68 76 100.0 % 0.22 [ -0.11, 0.54 ]
-4 -2 0 2 4
Analysis 3.8. Comparison 3 Aerobic exercise versus flexibility/balance programme, Outcome 8 Cognitive
inhibition.
Std. Std.
Mean Mean

Blumenthal 1989 -1.670968 (6.298905) 31 34 -1.25 (6.140033) 24.6 % -0.07 [ -0.55, 0.42 ]
Oken 2006 38 -10.8 (4.3) 38 -10 (4.6) 28.7 % -0.18 [ -0.63, 0.27 ]
Subtotal (95% CI) 69 72 53.2 % -0.13 [ -0.46, 0.20 ]

Kramer 2001 58 93.1 (132.51) 66 92.3 (150.3) 46.8 % 0.01 [ -0.35, 0.36 ]
Subtotal (95% CI) 58 66 46.8 % 0.01 [ -0.35, 0.36 ]

Total (95% CI) 127 138 100.0 % -0.06 [ -0.31, 0.18 ]
-4 -2 0 2 4
Analysis 3.9. Comparison 3 Aerobic exercise versus flexibility/balance programme, Outcome 9 Visual
attention.
Std. Std.
Mean Mean
1 Digit vigilance
3 2%7 test
Blumenthal 1989 -6.790323 (5.064251) 31 34 -9.1 (9.257699) 24.4 % 0.30 [ -0.19, 0.79 ]
Subtotal (95% CI) 31 34 24.4 % 0.30 [ -0.19, 0.79 ]

Kramer 2001 58 97.75 (3.34) 66 96.95 (3.02) 46.7 % 0.25 [ -0.10, 0.60 ]
Subtotal (95% CI) 58 66 46.7 % 0.25 [ -0.10, 0.60 ]

Oken 2006 38 -37.1 (34.1) 38 -40.3 (36.8) 28.9 % 0.09 [ -0.36, 0.54 ]
Subtotal (95% CI) 38 38 28.9 % 0.09 [ -0.36, 0.54 ]

Total (95% CI) 127 138 100.0 % 0.22 [ -0.03, 0.46 ]
-4 -2 0 2 4
Analysis 3.10. Comparison 3 Aerobic exercise versus flexibility/balance programme, Outcome 10 Auditory
attention.
Mean Mean

Blumenthal 1989 8.929032 (2.570791) 31 34 8.6 (2.039608) 31.1 % 0.33 [ -0.81, 1.46 ]
Kramer 2001 58 8 (1.98) 66 8.4 (2.11) 68.9 % -0.40 [ -1.12, 0.32 ]
Total (95% CI) 89 100 100.0 % -0.17 [ -0.83, 0.49 ]

-4 -2 0 2 4
Analysis 3.11. Comparison 3 Aerobic exercise versus flexibility/balance programme, Outcome 11 Motor
function.
Std. Std.
Mean Mean
1 Finger tapping
Blumenthal 1989 123.616129 (14.348363) 31 34 120.35 (18.774051) 34.3 % 0.19 [ -0.30, 0.68 ]
Subtotal (95% CI) 31 34 34.3 % 0.19 [ -0.30, 0.68 ]

Kramer 2001 58 -29.65 (6.25) 66 -29.8 (6.56) 65.7 % 0.02 [ -0.33, 0.38 ]
Subtotal (95% CI) 58 66 65.7 % 0.02 [ -0.33, 0.38 ]

Total (95% CI) 89 100 100.0 % 0.08 [ -0.20, 0.37 ]
-4 -2 0 2 4
Analysis 3.12. Comparison 3 Aerobic exercise versus flexibility/balance programme, Outcome 12 Drop-out.

M- M-
n/N n/N CI CI
Blumenthal 1989 2/33 2/33 7.3 % 1.00 [ 0.13, 7.55 ]
Kramer 2001 25/83 25/91 69.2 % 1.14 [ 0.59, 2.20 ]
Oken 2006 6/44 9/47 23.5 % 0.67 [ 0.22, 2.06 ]
Total (95% CI) 170 181 100.0 % 0.99 [ 0.58, 1.72 ]

0.01 0.1 1 10 100

Favours treatment Favours control
Analysis 4.1. Comparison 4 Aerobic exercise versus strength programme, Outcome 1 Verbal memory
Comparison: 4 Aerobic exercise versus strength programme
Mean Mean

Moul 1995 10 23.5 (3.48) 10 23.2 (6.32) 100.0 % 0.30 [ -4.17, 4.77 ]
Subtotal (95% CI) 10 10 100.0 % 0.30 [ -4.17, 4.77 ]

Total (95% CI) 10 10 100.0 % 0.30 [ -4.17, 4.77 ]
-10 -5 0 5 10
Analysis 4.2. Comparison 4 Aerobic exercise versus strength programme, Outcome 2 Executive functions.
Mean Mean
Moul 1995 10 27.3 (3.48) 10 29.6 (0.63) 100.0 % -2.30 [ -4.49, -0.11 ]
Subtotal (95% CI) 10 10 100.0 % -2.30 [ -4.49, -0.11 ]

4 Word comparison (#error)
6 Verbal fluency
Total (95% CI) 10 10 100.0 % -2.30 [ -4.49, -0.11 ]
-10 -5 0 5 10
Analysis 4.3. Comparison 4 Aerobic exercise versus strength programme, Outcome 3 Perception.
Mean Mean

Moul 1995 10 27.7 (3.8) 10 28.2 (0.95) 100.0 % -0.50 [ -2.93, 1.93 ]
Subtotal (95% CI) 10 10 100.0 % -0.50 [ -2.93, 1.93 ]

Total (95% CI) 10 10 100.0 % -0.50 [ -2.93, 1.93 ]
-10 -5 0 5 10
Analysis 4.4. Comparison 4 Aerobic exercise versus strength programme, Outcome 4 Cognitive speed.
Mean Mean

Panton 1990 17 -274 (28.9) 20 -270 (44.7) 100.0 % -4.00 [ -27.93, 19.93 ]
Total (95% CI) 17 20 100.0 % -4.00 [ -27.93, 19.93 ]

-100 -50 0 50 100

Analysis 5.1. Comparison 5 Fitness Improved: aerobic exercise versus any active intervention, Outcome 1
Cognitive speed.
Comparison: 5 Fitness Improved: aerobic exercise versus any active intervention
Std. Std.
Mean Mean

Panton 1990 17 -274 (28.9) 20 -270 (44.7) 16.8 % -0.10 [ -0.75, 0.54 ]
Subtotal (95% CI) 17 20 16.8 % -0.10 [ -0.75, 0.54 ]

Legault 2011 16 -37.44 (11.331) 33 -34.24 (7.166013) 18.9 % -0.36 [ -0.96, 0.24 ]
Subtotal (95% CI) 16 33 18.9 % -0.36 [ -0.96, 0.24 ]

Blumenthal 1989 31 52.4839 (8.6) 34 48.65 (9.9985) 25.5 % 0.40 [ -0.09, 0.90 ]
Kramer 2001 58 96.5 (3.88) 66 95.7 (6.34) 38.8 % 0.15 [ -0.20, 0.50 ]
Subtotal (95% CI) 89 100 64.3 % 0.24 [ -0.05, 0.52 ]

Total (95% CI) 122 153 100.0 % 0.08 [ -0.22, 0.37 ]
-4 -2 0 2 4
Verbal memory functions (immediate).
Std. Std.
Mean Mean

Blumenthal 1989 11.06129 (3.866701) 31 34 9.8 (3.450362) 23.1 % 0.34 [ -0.15, 0.83 ]
Subtotal (95% CI) 31 34 23.1 % 0.34 [ -0.15, 0.83 ]

Moul 1995 10 23.5 (3.5) 20 19.6 (7.323196) 16.6 % 0.60 [ -0.18, 1.37 ]
Subtotal (95% CI) 10 20 16.6 % 0.60 [ -0.18, 1.37 ]

Fabre 2002 8 7.4 (1) 16 9.1 (1.236932) 13.4 % -1.41 [ -2.36, -0.45 ]
Subtotal (95% CI) 8 16 13.4 % -1.41 [ -2.36, -0.45 ]

Kramer 2001 58 9.95 (1.8) 66 9.75 (2.1) 26.4 % 0.10 [ -0.25, 0.45 ]
Subtotal (95% CI) 58 66 26.4 % 0.10 [ -0.25, 0.45 ]

Legault 2011 16 6.38 (1.784) 33 5.82 (1.595437) 20.4 % 0.34 [ -0.27, 0.94 ]
Subtotal (95% CI) 16 33 20.4 % 0.34 [ -0.27, 0.94 ]

Total (95% CI) 123 169 100.0 % 0.08 [ -0.38, 0.55 ]
-4 -2 0 2 4
Visual memory functions (immediate).
Mean Mean

Blumenthal 1989 -5.545161 (3.348185) 31 34 -5.6 (3.653081) 57.4 % 0.05 [ -1.65, 1.76 ]
Subtotal (95% CI) 31 34 57.4 % 0.05 [ -1.65, 1.76 ]

Fabre 2002 8 10.2 (2.8) 16 11.65 (1.334166) 42.6 % -1.45 [ -3.50, 0.60 ]
Subtotal (95% CI) 8 16 42.6 % -1.45 [ -3.50, 0.60 ]

Total (95% CI) 39 50 100.0 % -0.59 [ -2.04, 0.87 ]
-4 -2 0 2 4
Working memory.
Std. Std.
Mean Mean

Blumenthal 1989 7.851613 (2.852162) 31 34 6.9 (2.280351) 27.7 % 0.37 [ -0.12, 0.86 ]
Kramer 2001 58 7.2 (1.83) 66 7.1 (2.11) 53.6 % 0.05 [ -0.30, 0.40 ]
Subtotal (95% CI) 89 100 81.3 % 0.16 [ -0.13, 0.45 ]

Legault 2011 16 0.5723 (0.18167) 33 0.6 (0.231733) 18.7 % -0.14 [ -0.74, 0.46 ]
Subtotal (95% CI) 16 33 18.7 % -0.14 [ -0.74, 0.46 ]

Total (95% CI) 105 133 100.0 % 0.10 [ -0.16, 0.36 ]
-4 -2 0 2 4
Memory functions (delayed).
Mean Mean

Kramer 2001 58 11.4 (2.7) 66 10.9 (2.6) 68.4 % 0.50 [ -0.44, 1.44 ]
Subtotal (95% CI) 58 66 68.4 % 0.50 [ -0.44, 1.44 ]

Legault 2011 16 9.38 (2.187) 33 8.94 (2.539838) 31.6 % 0.44 [ -0.94, 1.82 ]
Subtotal (95% CI) 16 33 31.6 % 0.44 [ -0.94, 1.82 ]

Total (95% CI) 74 99 100.0 % 0.48 [ -0.29, 1.25 ]
-4 -2 0 2 4
Std. Std.
Mean Mean
Blumenthal 1989 -79.487097 (27.626574) 31 34 -90.9 (34.001912) 22.2 % 0.36 [ -0.13, 0.85 ]
Legault 2011 15 -81.53 (49.11) 33 -86.18 (28.530309) 20.9 % 0.13 [ -0.48, 0.74 ]
Subtotal (95% CI) 46 67 43.1 % 0.27 [ -0.11, 0.65 ]

Moul 1995 10 27.3 (3.48) 20 18.46 (2.940153) 15.4 % 2.75 [ 1.69, 3.82 ]
Subtotal (95% CI) 10 20 15.4 % 2.75 [ 1.69, 3.82 ]

Test for overall effect: Z = 5.06 (P < 0.00001)
Fabre 2002 8 5.9 (0.9) 16 6.2 (0.961769) 17.9 % -0.31 [ -1.16, 0.55 ]
Subtotal (95% CI) 8 16 17.9 % -0.31 [ -1.16, 0.55 ]

Kramer 2001 58 95.8 (6.32) 66 95.6 (7.39) 23.6 % 0.03 [ -0.32, 0.38 ]
Subtotal (95% CI) 58 66 23.6 % 0.03 [ -0.32, 0.38 ]

5 Verbal fluency
Total (95% CI) 122 169 100.0 % 0.48 [ -0.18, 1.15 ]
-4 -2 0 2 4
Perception.
Std. Std.
Mean Mean

Kramer 2001 58 75.1 (18.28) 66 71.4 (23.56) 52.0 % 0.17 [ -0.18, 0.53 ]
Subtotal (95% CI) 58 66 52.0 % 0.17 [ -0.18, 0.53 ]

Moul 1995 10 27.7 (3.8) 20 27.1 (2.124265) 26.3 % 0.21 [ -0.55, 0.97 ]
Subtotal (95% CI) 10 20 26.3 % 0.21 [ -0.55, 0.97 ]

Fabre 2002 8 10.2 (2.8) 16 11.65 (1.334166) 21.7 % -0.73 [ -1.61, 0.15 ]
Subtotal (95% CI) 8 16 21.7 % -0.73 [ -1.61, 0.15 ]

Total (95% CI) 76 102 100.0 % -0.01 [ -0.50, 0.48 ]
-4 -2 0 2 4
Std. Std.
Mean Mean

Blumenthal 1989 -1.670968 (6.298905) 31 34 -1.25 (6.140033) 28.0 % -0.07 [ -0.55, 0.42 ]
Subtotal (95% CI) 31 34 28.0 % -0.07 [ -0.55, 0.42 ]

Kramer 2001 58 93.1 (132.51) 66 92.3 (150.3) 53.4 % 0.01 [ -0.35, 0.36 ]
Subtotal (95% CI) 58 66 53.4 % 0.01 [ -0.35, 0.36 ]

Legault 2011 -538.8438 (67.86966) 16 -539.08 (80.523323) 33 18.6 % 0.00 [ -0.59, 0.60 ]
Subtotal (95% CI) 16 33 18.6 % 0.00 [ -0.59, 0.60 ]

Total (95% CI) 105 133 100.0 % -0.02 [ -0.27, 0.24 ]
-4 -2 0 2 4
Visual attention.
Std. Std.
Mean Mean
1 Digit vigilance
3 2%7 test
Blumenthal 1989 -6.790323 (5.064251) 31 34 -9.1 (9.257699) 34.3 % 0.30 [ -0.19, 0.79 ]
Subtotal (95% CI) 31 34 34.3 % 0.30 [ -0.19, 0.79 ]

Kramer 2001 58 97.75 (3.34) 66 96.95 (3.02) 65.7 % 0.25 [ -0.10, 0.60 ]
Subtotal (95% CI) 58 66 65.7 % 0.25 [ -0.10, 0.60 ]

Total (95% CI) 89 100 100.0 % 0.27 [ -0.02, 0.56 ]
-4 -2 0 2 4
Analysis 5.10. Comparison 5 Fitness Improved: aerobic exercise versus any active intervention, Outcome
10 Auditory attention.
Mean Mean

Blumenthal 1989 8.929032 (2.570791) 31 34 8.6 (2.039608) 21.7 % 0.33 [ -0.81, 1.46 ]
Fabre 2002 8 6.1 (0.7) 16 5.55 (0.806226) 41.5 % 0.55 [ -0.08, 1.18 ]
Kramer 2001 58 8 (1.98) 66 8.4 (2.11) 36.8 % -0.40 [ -1.12, 0.32 ]
Total (95% CI) 97 116 100.0 % 0.15 [ -0.49, 0.79 ]

-4 -2 0 2 4
11 Motor function.
Std. Std.
Mean Mean
1 Finger tapping
Blumenthal 1989 123.616129 (14.348363) 31 34 120.35 (18.774051) 34.3 % 0.19 [ -0.30, 0.68 ]
Subtotal (95% CI) 31 34 34.3 % 0.19 [ -0.30, 0.68 ]

Kramer 2001 58 -29.65 (6.25) 66 -29.8 (6.56) 65.7 % 0.02 [ -0.33, 0.38 ]
Subtotal (95% CI) 58 66 65.7 % 0.02 [ -0.33, 0.38 ]

Total (95% CI) 89 100 100.0 % 0.08 [ -0.20, 0.37 ]
-4 -2 0 2 4
12 Drop-out.

M- M-
n/N n/N CI CI
Blumenthal 1989 2/33 2/33 9.2 % 1.00 [ 0.13, 7.55 ]
Fabre 2002 0/8 0/16 Not estimable
Kramer 2001 25/83 25/91 86.8 % 1.14 [ 0.59, 2.20 ]
Legault 2011 3/18 0/18 4.1 % 8.35 [ 0.40, 174.50 ]
Total (95% CI) 152 178 100.0 % 1.22 [ 0.66, 2.25 ]

0.01 0.1 1 10 100

Analysis 6.1. Comparison 6 Fitness improved: aerobic exercise versus no intervention, Outcome 1
Cognitive speed.
Comparison: 6 Fitness improved: aerobic exercise versus no intervention
Std. Std.
Mean Mean

Panton 1990 17 -274 (28.9) 12 -275 (52) 16.2 % 0.02 [ -0.71, 0.76 ]
Subtotal (95% CI) 17 12 16.2 % 0.02 [ -0.71, 0.76 ]

Whitehurst 1991 7 -0.41 (0.03) 7 -0.39 (0.04) 7.7 % -0.53 [ -1.60, 0.54 ]
Subtotal (95% CI) 7 7 7.7 % -0.53 [ -1.60, 0.54 ]

Langlois 2012 36 -44.08 (13.37) 36 -49.49 (20.21) 40.2 % 0.31 [ -0.15, 0.78 ]
Subtotal (95% CI) 36 36 40.2 % 0.31 [ -0.15, 0.78 ]

Blumenthal 1989 52.483871 (8.968893) 31 34 48.16 (10.18025) 35.9 % 0.44 [ -0.05, 0.94 ]
Subtotal (95% CI) 31 34 35.9 % 0.44 [ -0.05, 0.94 ]

Total (95% CI) 91 89 100.0 % 0.25 [ -0.05, 0.55 ]
-4 -2 0 2 4
Analysis 6.2. Comparison 6 Fitness improved: aerobic exercise versus no intervention, Outcome 2 Verbal
Std. Std.
Mean Mean

Blumenthal 1989 11.06129 (3.866701) 31 34 11.21 (3.845777) 47.5 % -0.04 [ -0.53, 0.45 ]
Subtotal (95% CI) 31 34 47.5 % -0.04 [ -0.53, 0.45 ]

Langlois 2012 36 10.92 (3.11) 36 10.19 (3.78) 52.5 % 0.21 [ -0.25, 0.67 ]
Subtotal (95% CI) 36 36 52.5 % 0.21 [ -0.25, 0.67 ]

Total (95% CI) 67 70 100.0 % 0.09 [ -0.24, 0.43 ]
-4 -2 0 2 4
Analysis 6.3. Comparison 6 Fitness improved: aerobic exercise versus no intervention, Outcome 3 Visual
Mean Mean

Blumenthal 1989 -5.545161 (3.348185) 31 34 -5.26 (3.15352) 100.0 % -0.28 [ -1.87, 1.30 ]
Subtotal (95% CI) 31 34 100.0 % -0.28 [ -1.87, 1.30 ]

Total (95% CI) 31 34 100.0 % -0.28 [ -1.87, 1.30 ]
-4 -2 0 2 4
Analysis 6.4. Comparison 6 Fitness improved: aerobic exercise versus no intervention, Outcome 4 Working
memory.
Std. Std.
Mean Mean

Blumenthal 1989 31 7.8 (2.852162) 34 7.34 (2.211334) 47.3 % 0.18 [ -0.31, 0.67 ]
Langlois 2012 36 6.64 (2.47) 36 6.47 (2.46) 52.7 % 0.07 [ -0.39, 0.53 ]
Subtotal (95% CI) 67 70 100.0 % 0.12 [ -0.21, 0.46 ]

Total (95% CI) 67 70 100.0 % 0.12 [ -0.21, 0.46 ]
-4 -2 0 2 4
Analysis 6.5. Comparison 6 Fitness improved: aerobic exercise versus no intervention, Outcome 5 Memory
Mean Mean

Langlois 2012 36 10.56 (3.21) 36 9.64 (3.99) 100.0 % 0.92 [ -0.75, 2.59 ]
Total (95% CI) 36 36 100.0 % 0.92 [ -0.75, 2.59 ]

-4 -2 0 2 4
Std. Std.
Mean Mean
Blumenthal 1989 -79.487097 (27.626574) 31 34 -81.48 (36.321516) 49.1 % 0.06 [ -0.43, 0.55 ]
Langlois 2012 36 -104.95 (37.85) 36 -136.2 (72.86) 50.9 % 0.53 [ 0.06, 1.00 ]
Subtotal (95% CI) 67 70 100.0 % 0.30 [ -0.16, 0.76 ]

5 Verbal fluency
Total (95% CI) 67 70 100.0 % 0.30 [ -0.16, 0.76 ]
-4 -2 0 2 4
Std. Std.
Mean Mean

Blumenthal 1989 -1.670968 (6.298905) 31 34 -4.53 (6.657397) 46.9 % 0.44 [ -0.06, 0.93 ]
Langlois 2012 36 -141.67 (46.5) 36 -149.38 (42.95) 53.1 % 0.17 [ -0.29, 0.63 ]
Subtotal (95% CI) 67 70 100.0 % 0.29 [ -0.04, 0.63 ]

Total (95% CI) 67 70 100.0 % 0.29 [ -0.04, 0.63 ]
-4 -2 0 2 4
Analysis 6.8. Comparison 6 Fitness improved: aerobic exercise versus no intervention, Outcome 8 Visual
attention.
Std. Std.
Mean Mean
1 Digit vigilance
3 2%7 test
Blumenthal 1989 6.790323 (5.064251) 31 34 6.57 (4.97189) 100.0 % 0.04 [ -0.44, 0.53 ]
Subtotal (95% CI) 31 34 100.0 % 0.04 [ -0.44, 0.53 ]

Total (95% CI) 31 34 100.0 % 0.04 [ -0.44, 0.53 ]
-4 -2 0 2 4
Analysis 6.9. Comparison 6 Fitness improved: aerobic exercise versus no intervention, Outcome 9 Auditory
attention.
Mean Mean

Blumenthal 1989 8.929032 (2.570791) 31 34 8.77 (2.196588) 100.0 % 0.16 [ -1.01, 1.33 ]
Total (95% CI) 31 34 100.0 % 0.16 [ -1.01, 1.33 ]

-4 -2 0 2 4
Analysis 6.10. Comparison 6 Fitness improved: aerobic exercise versus no intervention, Outcome 10 Motor
function.
Mean Mean
1 Finger tapping
Blumenthal 1989 123.616129 (14.348363) 31 34 123.51 (18.367362) 100.0 % 0.10 [ -7.87, 8.08 ]
Subtotal (95% CI) 31 34 100.0 % 0.10 [ -7.87, 8.08 ]

Total (95% CI) 31 34 100.0 % 0.10 [ -7.87, 8.08 ]
-10 -5 0 5 10
Analysis 6.11. Comparison 6 Fitness improved: aerobic exercise versus no intervention, Outcome 11 Drop-
out.

n/N n/N IV,Random,95% CI IV,Random,95% CI
Blumenthal 1989 2/33 2/34 29.7 % 1.03 [ 0.14, 7.79 ]
Langlois 2012 7/43 4/40 70.3 % 1.75 [ 0.47, 6.50 ]
Whitehurst 1991 0/7 0/7 Not estimable
Total (95% CI) 83 81 100.0 % 1.50 [ 0.50, 4.50 ]

0.01 0.1 1 10 100

ADDITIONAL TABLES
Table 1. Grouping of cognitive tests and studies over cognitive functions
Cognitive function Cognitive tests Trial
Cognitive speed Simple RT Panton 1990, Oken 2006
Choice RT Hassmén 1997, Whitehurst 1991
Trailmaking part A Emery 1998, Legault 2011, Langlois 2012
Digit symbol substitution Blumenthal 1989, Kramer 2001, Emery

1990a
Verbal memory functions (immediate) Randt memory test story recall Blumenthal 1989
16 words immediate recall Hassmén 1997
Ross Information Processing Assessment Moul 1995

Table 1. Grouping of cognitive tests and studies over cognitive functions (Continued)
Wechsler Adult Intelligence Scales logical Fabre 2002

Rey auditory verbal learning test trail I-V Kramer 2001, Langlois 2012
Hopkins Verbal Learning Test Legault 2011
Visual memory functions (immediate) Benton visual retention Blumenthal 1989
Wechsler Memory Scales visual reproduction Fabre 2002

immediate recall
Working memory Digit span backward Blumenthal 1989, Kramer 2001, Langlois
2012
2-Back Legault 2011
Memory function (delayed) 16 words delayed recall Hassmén 1997
Rey auditory verbal learning test delayed re- Kramer 2001, Langlois 2012
call trail
10 words delayed recall Oken 2006
Hopkins Verbal Learning Test - 12 words Legault 2011
Executive functions Trailmaking part B Blumenthal 1989, Legault 2011, Langlois

2012
Ross Information Processing Assessment Moul 1995

problem solving and abstract reasoning
Wechsler Memory Scales mental control Fabre 2002
Task switching paradigm Kramer 2001
Verbal fluency Emery 1990a
Letter number sequencing Oken 2006
Perception Face recognition Hassmén 1997, Kramer 2001
Ross Information Processing Assessment au- Moul 1995

ditory processing
Wechsler Adult Intelligence Scales visual re- Fabre 2002

production
Table 1. Grouping of cognitive tests and studies over cognitive functions (Continued)
Cognitive inhibition Stroop colour word test Blumenthal 1989, Oken 2006, Langlois
2012, Predovan 2012
Stopping task Kramer 2001
Flanker Task Legault 2011
Visual attention Digit vigilance Emery 1990a
Tracking Bakken 2001
2&7 test Blumenthal 1989
Visual search Kramer 2001
Covert orienting of visuospatial attention Oken 2006
Auditory attention Digit span forward Blumenthal 1989, Emery 1990a, Fabre
2002, Hassmén 1997, Kramer 2001
Motor function Finger tapping Bakken 2001, Blumenthal 1989, Emery

1998
Pursuit rotor task Kramer 2001
Table 2. Types of interventions in each trial
Trial Aerobic ex- Strength Flexibility/ Social Cognitive Education Miscellaneous No intervention
ercise balance
Bakken x - - - - - - x
2001
Blumenthal x - x - - - - x
1989
Emery x - - x - - - x
1990a
Fabre 2002 x - - x x - - -
Kramer x - x - - - - -
2001
Langlois x - - - - - - x
2012
Table 2. Types of interventions in each trial (Continued)
Legault x - - - x x - -
2011
Madden x - x - - - - x
1989
Moul 1995 x x x - - - - -
Oken 2006 x - x - - - - x
Panton x x - - - - - x
1990
Whitehurst x - - - - - - x
1991
Table 3. Methodological quality of included trials (CLEAR NPT score)
Study ID Number
1/2 3 4 5 6 / 6.1.1 / 6. 7 / 7.1.1 / 7. 8 / 8.1.1 9 10 Total

1.2 1.2
Bakken 3/3 1 3 1 2/3/2 2/3/2 1/0 1 2 28

2001
Blumen- 3/3 1 3 1 2/2/2 2/2/2 4/3 1 3 34

thal
1989
Emery 3/3 1 3 1 2/2/2 2/2/2 4/3 1 2 33

1990a
Fabre 3/3 1 3 2 2/1/1 2/1/1 4/3 1 1 29

2002
Kramer 3/3 1 3 1 2/3/1 2/3/1 4/3 1 2 33

2001
Langlois 3/3 1 3 3 2/3/2 2/3/2 4/3 1 2 37

2012
Legault 3/3 1 3 1 2/3/1 2/3/1 4/3 1 1 31

2011
Madden 3/3 1 3 1 2/3/2 2/3/2 4/3 1 2 34

1989
Table 3. Methodological quality of included trials (CLEAR NPT score) (Continued)
Moul 3/3 1 3 3 2/3/1 2/3/1 4/3 1 1 34

1995
Oken 1/1 1 3 1 2 / 2 /1 2/2/1 1/3 1 2 24

2006
Panton 3/3 1 3 2 2/3/3 2/3/3 4/3 1 2 38

1990
White- 3/3 1 3 2 2/3/3 2/3/3 4/3 1 3 39

hurst
1991
See Table 4 for CLEAR NPT items.
Table 4. CLEAR NPT checklist items
Number Checklist item
1 Was the generation of allocation sequences adequate?
2 Was the treatment allocation concealed?
3 Were the details of the intervention administered to each group made available?a
4 Were care providers’ experience or skillb in each arm appropriate?c
5 Was participant (i.e. patients) adherence assessed quantitatively?d
6 Were participants adequately blinded?
6.1.1 If participants were not adequately blinded, were all other treatments and care (cointerventions) the same in each ran-
domised group?
6.1.2 If participants were not adequately blinded, were withdrawals and lost to follow-up the same in each randomised group?
7 Were care providers or persons caring for the participants adequately blinded?
7.1.1 If care providers were not adequately blinded, were all other treatments and care (cointerventions) the same in each
randomised group?
7.1.2 If care providers were not adequately blinded, were withdrawals and losses to follow-up the same in each randomised
group?
8 Were outcome assessors adequately blinded to assess the primary outcomes?
8.1.1 If outcome assessors were not adequately blinded, were specific methods used to avoid ascertainment bias?e
Table 4. CLEAR NPT checklist items (Continued)
9 Was the follow-up schedule the same in each group?f
10 Were the main outcomes analysed according to the ITT principle?
a The answer should be “Yes” if these data are either described in the report or made available for each arm (reference to
preliminary report, online addendum, etc.)
b Care provider experience or skill will be assessed only for therapist-dependent interventions (where the success of the
intervention is directly linked to the providers’ technical skill. For other treatment this item is not relevant and should be
answered “Unclear”
c Appropriate experience or skill should be determined according to published data, preliminary studies, guidelines, run-
in period, or a group of experts and should be specified in the protocol for each study arm before the beginning of the
survey
d Treatment adherence will be assessed only for the treatments necessitating iterative interventions (physiotherapy that
supposes several sessions, in contrast to a one-shot treatment such as surgery). For one-shot treatments, this item is not
relevant and should be answered “Unclear”
e The answer is “0” if the answer to 8 is “Yes”. The answer should be “Yes” if the main outcome is objective or hard, or if
outcomes were assessed by a blinded or at least an independent endpoint review committee, or if outcomes were assessed
by an independent outcome assessor trained to perform the measurements in a standardised manner, or if the outcome
assessor was blinded to the study purpose and hypothesis
f This item is not relevant if follow-up is part of the question. For example, this item is not relevant for a trial assessing
frequent versus less frequent follow-up for cancer recurrence. In these situations, this item should be answered “Unclear”
For items 6, 7 and 8 a score of 1 was given for a “Yes”, a score of 2 for “No, because blinding is not feasible”, a score of 3 for “No,
although blinding is feasible” and a score of 4 for “Unclear”. The other items of the checklist (1 to 5, 6.1.1, 6.1.2, 7.1.1, 7.1.2, 8.1.
1, 9 and 10) were given a score of 1 for “Yes”, 2 for “No” and 3 for “Unclear”
APPENDICES
Appendix 1. Search strategies: August 2013
Source Search strategy Hits retrieved
1. ALOIS (www.medicine.ox.ac.uk/alois) - Keyword search: “physical activity” OR ex- 8

all dates to August 2013 ercise
2. MEDLINE In-process and other non- 1. exercise.mp. or Exercise/ 650

indexed citations and MEDLINE 1946 to 2. exercis*.mp.
August 2013 (Ovid SP) 3. motor activit*.mp. [mp=title, abstract,
original title, name of substance word, sub-
ject heading word, keyword heading word,
protocol supplementary concept, rare dis-
ease supplementary concept, unique iden-
tifier]
4. leisure activit*.mp.
5. physical fitness.mp.
6. physical endurance.mp. or Physical En-
durance/
7. exercise tolerance.mp. or Exercise Toler-
ance/
8. aerobic.mp.
9. physical activity.mp.
10. Motor Activity/
11. physical capacity.mp.
12. physical performance training.mp.
13. cognit*.mp.
14. Mental Processes/ or mental process*.
mp.
15. maximal VO2.mp.
16. METS.mp.
17. Watts.mp.
18. treadmill speed.mp.
19. inclination.mp.
20. Adult/ or adult*.mp.
21. middle aged.mp. or Middle Aged/
22. aged.mp. or Aged/
23. elderly.mp.
24. old*.mp.
25. geriatric.mp. or Geriatrics/
26. 1 or 2 or 3 or 4 or 5 or 6 or 7 or 8 or 9
or 10 or 11 or 12
27. 13 or 14 or 15 or 16 or 17 or 18 or 19
28. 20 or 21 or 22 or 23 or 24 or 25
29. 26 and 27 and 28
30. randomised controlled trial.pt.
31. controlled clinical trial.pt.
(Continued)
32. randomized.ab.
33. placebo.ab.
34. drug therapy.fs.
35. randomly.ab.
36. trial.ab.
37. groups.ab.
38. or/30-37
39. 29 and 38
40. (2012* or 2013*).ed.
41. 39 and 40
3. EMBASE 1. exercise.mp. or Exercise/ 750

1974 to 2013 week 27 (Ovid SP) 2. exercis*.mp.
3. motor activit*.mp. [mp=title, abstract,
subject headings, heading word, drug trade
name, original title, device manufacturer,
drug manufacturer, device trade name, key-
word]
durance/
ance/
8. aerobic.mp.
10. Motor Activity/
13. cognit*.mp.
mp.
15. maximal VO2.mp.
16. METS.mp.
17. Watts.mp.
19. inclination.mp.
23. elderly.mp.
24. old*.mp.
or 10 or 11 or 12
27. 13 or 14 or 15 or 16 or 17 or 18 or 19
28. 20 or 21 or 22 or 23 or 24 or 25
29. 26 and 27 and 28
30. “randomi?ed controlled trial”.mp.
(Continued)
[mp=title, abstract, subject headings, head-

ing word, drug trade name, original title,
device manufacturer, drug manufacturer,
device trade name, keyword]
31. “controlled clinical trial”.mp. [mp=
title, abstract, subject headings, heading
word, drug trade name, original title, de-
vice manufacturer, drug manufacturer, de-
vice trade name, keyword]
32. random*.mp.
33. randomised controlled trial/
34. clinical trial.mp.
35. or/30-34
36. 29 and 35
37. (2012* or 2013*).em.
38. 36 and 37
4. PSYCINFO 1. exercise.mp. or Exercise/ 92

1806 to August week 5 2011 (Ovid SP) 2. exercis*.mp.
3. motor activit*.mp. [mp=title, abstract,
heading word, table of contents, key con-
cepts, original title, tests & measures]
durance/
ance/
8. aerobic.mp.
10. Motor Activity/
13. cognit*.mp.
mp.
15. maximal VO2.mp.
16. METS.mp.
17. Watts.mp.
19. inclination.mp.
23. elderly.mp.
24. old*.mp.
or 10 or 11 or 12
(Continued)
27. 13 or 14 or 15 or 16 or 17 or 18 or 19
28. 20 or 21 or 22 or 23 or 24 or 25
29. 26 and 27 and 28
30. “randomi?ed controlled trial”.mp.
[mp=title, abstract, heading word, table of
contents, key concepts, original title, tests
& measures]
31. “controlled clinical trial”.mp. [mp=ti-
tle, abstract, heading word, table of con-
tents, key concepts, original title, tests &
measures]
32. random*.mp.
33. randomised controlled trial/
34. clinical trial.mp.
35. 30 or 31 or 32 or 33 or 34
36. 29 and 35
37. (2012* or 2013*).up.
38. 36 and 37
5. CINAHL (EBSCOhost) to August 2013 S1 TX exercis* 213

S2 TX “physical activit*”
S3 TX cycling
S4 TX swim*
S5 TX gym*
S6 TX walk* OR treadmill
S7 TX danc*
S8 TX yoga*
S9 TX “tai chi”
S10 (MH “Exercise+”)
S11 (MH “Clinical Trials”)
S12 TX trial
S13 TX RCT OR CCT
S14 TX placebo*
S15 TX “double-blind*” OR “single-
blind*”
S16 TX groups OR “control group”
S17 S1 OR S2 OR S3 OR S4 OR S5 OR
S6 OR S7 OR S8 OR S9 OR S10
S18 S11 OR S12 OR S13 OR S14 OR S15
OR S16
S19 TX elderly
S20 (MH “Aged”)
S21 TX geriatric
S22 S19 OR S20 OR S21
S23 TX cognition
S24 cognition
S25 (MH “Cognition”)
S26 TX cognitive
S27 AB brain OR mental OR memory OR
(Continued)
“executive function*”
S28 S23 OR S24 OR S25 OR S26 OR S27
S29 S17 AND S18 AND S22 AND S28
S30 EM 2012
S31 EM 2013
S32 S30 OR S31
S33 S29 AND S32
6. Web of Science (1945 to August 2013) Topic=(“physical activity” OR “physical ex- 869
(ISI Web of Knowledge) ercise” OR cycling OR yoga OR swim*
OR danc* OR aerobic*) AND Topic=
(cogni* OR elderly OR memory OR geri-
atric) AND Topic=(randomly OR trial OR
RCT)
Timespan=2012-2013. Databases=
SCI-EXPANDED, SSCI, A&HCI, CPCI-
S, CPCI-SSH, BKCI-S, BKCI-SSH, CCR-
EXPANDED, IC
7. LILACS (BIREME) All dates to August “exercício físico” OR “physical exercise” 165
2013 OR aerobic$ OR aeróbico OR aerobio OR
yoga OR “physical activit$” OR “actividad
física” OR “atividade física” [Words] and
randomised OR randomized OR trial OR
randomly OR groups [Words] and elderly
OR idoso OR anciano [Words]
8. CENTRAL (the Cochrane Library; Issue #1 “cognit* impair*” 191

2 of 4, 2013) #2 MeSH descriptor: [Cognition Disor-
ders] explode all trees
#3 MCI
#4 ACMI
#5 ARCD
#6 SMC
#7 CIND
#8 BSF
#9 AAMI
#10 LCD
#11 QD or “questionable dementia”
#12 AACD
#13 MNCD
#14 MCD
#15 “N-MCI” or “A-MCI” or “M-MCI”
#16 (cognit* or memory or cerebr* or men-
tal*) near/3 (declin* or impair* or los* or
deteriorat* or degenerat* or complain* or
disturb* or disorder*)
#17 “preclinical AD”
#18 “pre-clinical AD”
(Continued)
#19 “preclinical alzheimer*” or “pre-clini-

cal alzheimer*”
#20 aMCI or MCIa
#21 “CDR 0.5” or “clinical dementia rat-
ing scale 0.5”
#22 “GDS 3” or “stage 3 GDS”
#23 “global deterioration scale” and “stage
3”
#24 “Benign senescent forgetfulness”
#25 “mild neurocognit* disorder*”
#26 (prodrom* near/2 dement*)
#27 episodic* near/2 memory
#28 “preclinical dementia” or “pre-clinical
dementia”
#29 episodic near/2 memory
#30 “pre-clinical dementia” or “preclinical
dementia”
#31 #1 or #2 or #3 or #4 or #5 or #6 or #
7 or #8 or #9 or #10 or #11 or #12 or #13
or #14 or #15 or #16 or #17 or #18 or #
19 or #20 or #21 or #22 or #23 or #24 or
#25 or #26 or #27 or #28 or #29 or #30
#32 “Physical therap*”
#33 “physical activit*”
#34 fitness
#35 exercis*
#36 aerobic
#37 “physical* fit*”
#38 “physical capacity”
#39 “physical training”
#40 Cycling
#41 swim*
#42 gym*
#43 danc*
#44 yoga
#45 “tai chi”
#46 walk*
#47 flexibility
#48 motor*
#49 “leisure activit*”
#50 “physical endurance”
#51 MeSH descriptor: [Exercise Therapy]
explode all trees
#52 #32 or #33 or #34 or #35 or #36 or #
37 or #38 or #39 or #40 or #41 or #42 or
#43 or #44 or #45 or #46 or #47 or #48 or
#49 or #50 or #51 in Trials
#53 #31 and #52 from 2009 to 2011, in
(Continued)
Trials
#54 #52 and (brain or MMSE or cogni-
tion or cognitive or memory) from 2012 to
2013, in Trials
9. ClinicalTrials.gov ( Interventional Studies | cognition OR cog- 273

www.clinicaltrials.gov) All dates to August nitive OR memory OR MMSE OR brain
2013 | “Physical therapy” OR “physical activ-
ity” OR “physical exercise” OR cycling OR
yoga OR swim OR swimming OR dance
OR aerobic | Adult, Senior | received from
01/01/2012 to 08/03/2013
10. ICTRP Search Portal (http:/ Interventional Studies | cognition OR cog- 187
/apps.who.int/trialsearch) [includes: Aus- nitive | “Physical therapy” OR “physical ac-
tralian New Zealand Clinical Trials Reg- tivity” OR “physical exercise” OR cycling
istry; ClinicalTrilas.gov; ISRCTN; Chinese OR yoga OR swim OR swimming OR
Clinical Trial Registry; Clinical Trials Reg- dance OR aerobic | Adult, Senior | received
istry - India; Clinical Research Informa- from 01/01/2012 to 08/03/2013
tion Service - Republic of Korea; German
Clinical Trials Register; Iranian Registry
of Clinical Trials; Japan Primary Registries
Network; Pan African Clinical Trial Reg-
istry; Sri Lanka Clinical Trials Registry; The
Netherlands National Trial Register] All
dates to August 2013
TOTAL before removal of duplicates and first assessment 3398
TOTAL after removal of duplicates and first assessment 352
WHAT’S NEW
Last assessed as up-to-date: 24 August 2013.
Date Event Description
14 April 2015 New search has been performed We performed a literature search update in November
2014. We have put the search results into the Studies
awaiting classification section of this review. We will fully
incorporate them into the next review update
14 April 2015 New citation required but conclusions have not changed We performed a literature search update in November
2014. We have put the search results into the Studies
awaiting classification section of this review. We will fully
(Continued)
incorporate them into the next review update. The con-

clusions are unchanged
HISTORY
Protocol first published: Issue 3, 2005
Review first published: Issue 2, 2008
Date Event Description
24 August 2013 New search has been performed A new update search was performed for this review on
24 August 2013
17 December 2008 New citation required but conclusions have not The update rendered one study (Oken 2006) which
changed met the inclusion criteria. The results of the review
have slightly changed
17 December 2008 Amended Incorporation of the risk of bias tables for all included
studies
15 July 2008 New search has been performed A new update search was performed for this review on
15 July 2008
10 April 2008 New search has been performed The delayed memory functions data have been cor-
rected
10 April 2008 New citation required and conclusions have changed Errors in the data entry for the outcome delayed mem-
ory function have been corrected. The effect of physi-
cal exercise on this outcome are not statistically signif-
icant
CONTRIBUTIONS OF AUTHORS
JY and MA: drafted reviews, obtained copies of trial reports, selected trials for inclusion and exclusion, extracted and entered data, and
interpreted data analyses.
NT: screened trials for inclusion and exclusion, extracted data and interpreted data analyses.
JR: interpreted data analyses.
Consumer Editor: Judith Hoppesteyn-Armstrong
DECLARATIONS OF INTEREST
Jeremy Young - none known
Maaike Angevaren - none known
Jennifer Rusted - none known
Naji Tabet - none known
SOURCES OF SUPPORT
Internal sources
• No sources of support supplied
External sources
• NIHR, UK.
This update was supported by the National Institute for Health Research, via Cochrane Infrastructure funding to the Cochrane
Dementia and Cognitive Improvement group. The views and opinions expressed therein are those of the authors and do not
necessarily reflect those of the Systematic Reviews Programme, NIHR, NHS or the Department of Health
INDEX TERMS
Medical Subject Headings (MeSH)

Cognition [∗ physiology]; Cognition Disorders [prevention & control]; Exercise [∗ physiology]; Memory [physiology]; Oxygen Con-
sumption [physiology]; Physical Fitness [∗ physiology]; Randomized Controlled Trials as Topic
MeSH check words

Aged; Humans; Middle Aged

Young Et Al-2015-The Cochrane Library

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Young Et Al-2015-The Cochrane Library

Uploaded by

Copyright:

Available Formats

Cochrane Database of Systematic Reviews

Aerobic exercise to improve cognitive function in older people

Young J, Angevaren M, Rusted J, Tabet N

Young J, Angevaren M, Rusted J, Tabet N.

Aerobic exercise to improve cognitive function in older people

Jeremy Young1 , Maaike Angevaren2 , Jennifer Rusted1 , Naji Tabet3

Editorial group: Cochrane Dementia and Cognitive Improvement Group.

PLAIN LANGUAGE SUMMARY

Data collection and analysis

The primary outcome measurement was cognitive function, tested

Assessment of risk of bias in included studies Data synthesis

Aerobic exercise versus any active intervention

There was no difference in dropout rates between the aerobic

There was no difference in dropout rates between the aerobic

Aerobic exercise versus strength training intervention

Chapman 2013 {published data only} Churchill 2002

Characteristics of included studies [ordered by study ID]

Methods Parallel-group RCT: 1 intervention group and 1 control group.

Outcomes AI (Accuracy Index by finger movement)

Notes Testing took place at the University of Minnesota.

Bias Authors’ judgement Support for judgement

Random sequence generation (selection Unclear risk No information provided.

Allocation concealment (selection bias) Unclear risk No information provided.

Selective reporting (reporting bias) Unclear risk Insufficient information.

Other bias Low risk However, we cannot rule out contamina-

Participants 101 participants aged 60 to 83 years

Outcomes Tapping (dominant/non-dominant)

of the Randt Memory test could not be traced by the authors.

Notes Testing took place at the Duke University Medical Center.

Bias Authors’ judgement Support for judgement

Random sequence generation (selection Unclear risk No information provided.

Allocation concealment (selection bias) Unclear risk No information provided.

Other bias Low risk Although contamination bias could not be

Participants 48 participants aged 61 to 86 years from a metropolitan inner-city community.

Outcomes Digit Symbol substitution

Notes Testing took place at the Duke University Medical Center.

Bias Authors’ judgement Support for judgement

Random sequence generation (selection Unclear risk No information provided.

Allocation concealment (selection bias) Unclear risk No information provided.

Selective reporting (reporting bias) Unclear risk Insufficient information provided.

Other bias Low risk Although contamination bias could not be

Participants 32 participants in the age range of 60 to 76 years “from clubs”.

Outcomes Memory quotient (= total score of al WAIS subtests)

Notes Testing took place at the University of Montpellier.

Bias Authors’ judgement Support for judgement

Random sequence generation (selection Unclear risk No information provided.

Allocation concealment (selection bias) Unclear risk No information provided.

Selective reporting (reporting bias) Unclear risk Insufficient information provided.

Other bias Low risk Although contamination bias could not be

Participants 124 participants aged 60 to 75 years of age.

Outcomes Visual search task

Time on treadmill (min)

Notes Testing took place at the University of Illinois at Urbana-Champaign

Bias Authors’ judgement Support for judgement

Random sequence generation (selection Unclear risk No information provided.

Allocation concealment (selection bias) Unclear risk No information provided.

Selective reporting (reporting bias) Unclear risk Insufficient information provided.

Other bias Low risk Although contamination bias could not be

Participants 72 participants aged 61 to 89.

Notes Testing took place at the Université du Québec à Montréal.

Bias Authors’ judgement Support for judgement

Random sequence generation (selection Unclear risk No information provided.

Allocation concealment (selection bias) Unclear risk No information provided.

Selective reporting (reporting bias) Unclear risk Insufficient information provided.