EURO PEAN

SO CIETY O F
Review CARDIOLOGY ®

European Journal of Preventive

Cardiology

Endurance training and maximal 0(00) 1–11

! The European Society of
Cardiology 2015
oxygen consumption with ageing: Reprints and permissions:
sagepub.co.uk/journalsPermissions.nav
Role of maximal cardiac DOI: 10.1177/2047487315617118
ejpc.sagepub.com
output and oxygen extraction

David Montero1 and Candela Dı́az-Cañestro2

Abstract
Background: The increase in maximal oxygen consumption (VO2max) with endurance training is associated with that of
maximal cardiac output (Qmax), but not oxygen extraction, in young individuals. Whether such a relationship is altered
with ageing remains unclear. Therefore, we sought systematically to review and determine the effect of endurance
training on and the associations among VO2max, Qmax and arteriovenous oxygen difference at maximal exercise
(Ca-vO2max) in healthy aged individuals.
Design and methods: We conducted a systematic search of MEDLINE, Scopus and Web of Science, from their
inceptions until May 2015 for articles assessing the effect of endurance training lasting 3 weeks or longer on VO2max
and Qmax and/or Ca-vO2max in healthy middle-aged and/or older individuals (mean age 40 years). Meta-analyses were
performed to determine the standardised mean difference (SMD) in VO2max, Qmax and Ca-vO2max between post and
pre-training measurements. Subgroup and meta-regression analyses were used to evaluate the associations among SMDs
and potential moderating factors.
Results: Sixteen studies were included after systematic review, comprising a total of 153 primarily untrained healthy
middle-aged and older subjects (mean age 42–71 years). Endurance training programmes ranged from 8 to 52 weeks of
duration. After data pooling, VO2max (SMD 0.89; P < 0.0001) and Qmax (SMD 0.61; P < 0.0001) were increased after
endurance training; no heterogeneity among studies was detected. Ca-vO2max was only increased with endurance
training interventions lasting more than 12 weeks (SMD 0.62; P ¼ 0.001). In meta-regression, the SMD in Qmax was
positively associated with the SMD in VO2max (B ¼ 0.79, P ¼ 0.04). The SMD in Ca-vO2max was not associated with the
SMD in VO2max (B ¼ 0.09, P ¼ 0.84).
Conclusions: The improvement in VO2max following endurance training is a linear function of Qmax, but not Ca-vO2max,
through healthy ageing.

Keywords
Exercise training, maximal oxygen consumption, ageing, meta-analysis
Received 3 August 2015; accepted 21 October 2015

contributing to VO2max via enhancing convective oxygen

Introduction
Maximal oxygen consumption (VO2max) is consistently
increased following endurance training (ET) in healthy
individuals across all ages.1–15 In this regard, ET provokes
multiple phenotypic modifications including increases in
cardiac structure/function,16,17 blood volume,18,19 oxygen
carrying capacity of the blood,20 skeletal muscle capillar-
isation,21,22 mitochondrial content/function23–25 as well as
decreases in peripheral vascular resistance26–28 and more
efficient blood flow distribution,29 all of them potentially
delivery to and/or extraction by active tissue. Whichever
adaptations to ET underlie the increase in VO2max, these
1
Zurich Center for Integrative Human Physiology (ZIHP), University of
Zurich, Switzerland
2
Center for Molecular Cardiology, University of Zurich, Switzerland
Corresponding author:
David Montero, Institute of Physiology, ZIHP, University of Zurich, Office
23 J 64, Winterthurerstrasse 190, 8057 Zurich, Switzerland.
Email: david.montero.barril@gmail.com

Downloaded from cpr.sagepub.com at The University of Hong Kong Libraries on December 4, 2015
2 European Journal of Preventive Cardiology 0(00)
must result in higher maximal cardiac output (Qmax) and/
or arteriovenous oxygen difference at maximal exercise
(Ca-vO2max) –following the Fick principle, respectively,
and primarily reflecting convective oxygen delivery and
extraction. However, whether the increase in VO2max
with ET is associated with that of Qmax and/or Ca-
vO2max remains to be established, despite related research
spanning over the past five decades.30–33
To begin addressing this uncertainty, we adopted a
meta-analysis approach to determine the effects of ET
on and associations between the components of VO2max
in young adults (mean age <40 years), in order to limit
the influence of ageing. We found that Qmax, but not
Ca-vO2max, was increased and linearly correlated with
the increase in VO2max following ET,34 suggesting that
oxygen extraction is not affected by ET in young indi-
viduals. In this sense, oxygen extraction may present
near optimal levels in untrained youth, as it is sup-
ported by the approximate Ca-vO2max observed in
young athletes and sedentary individuals.35,36 Thus,
improvements in oxygen extraction may not occur as
a result of a ‘ceiling effect’. With ageing, in turn, there
may be potential for improvement, in view of the higher
Ca-vO2max in master athletes versus healthy age-
matched untrained individual.35,37,38 Accordingly, it
has been proposed that the increase in VO2max with
ET in aged individuals may be a function of the
improvement in Ca-vO2max.1,7 Such contention is, how-
ever, unclear based on individual studies in that a vari-
able response to ET is apparently observed in VO2max
components in healthy middle-aged and older individ-
uals.2–6,8–14 Presumably, the small sample size as well as
distinct training characteristics, study population and
methodology of individual studies may have com-
pounded the impact of ET on Qmax and Ca-vO2max.
Therefore, the purpose of this study was systematic-
ally to review and meta-analyse the effects of ET on
VO2max, Qmax and Ca-vO2max in healthy middle-aged
and older subjects, as well as to determine the associ-
ations among VO2max, Qmax and Ca-vO2max and poten-
tial moderating factors.
Methods
The review is reported according to the meta-analysis
of observational studies in epidemiology group
guidelines.39
Data sources and searches
Our systematic search included MEDLINE, Scopus and
Web of Science, from their inceptions until May 2015.
We used combinations of the subject headings ‘healthy’,
‘middle’, ‘older’, ‘elderly’, ‘training’, ‘VO2max’, ‘max-
imal’, ‘peak’, ‘oxygen’, ‘aerobic’, ‘cardiac output’,
Downloaded from cpr.sagepub.com at The University of Hong Kong Libraries on December 4, 2015
‘arteriovenous’ and ‘extraction’; the search strategy for
MEDLINE is shown in Supplementary Figure 1. We
also performed hand searching in identified reviews, art-
icles included in meta-analysis, related citations in
MEDLINE and Google.
Article selection
To be included in the analysis, an original research art-
icle had to assess VO2max along with Qmax and/or Ca-
vO2max with dynamic exercise involving a large muscle
mass (e.g. running, cycling), before and after an ET
intervention lasting 3 weeks or more in healthy
middle-aged and/or older subjects (mean age 40
years). We excluded ET interventions of less than 3
weeks’ duration because Qmax may not be significantly
increased by this very short ET period.9,24 Studies
following the above criteria but including additional
interventions deemed likely to influence VO2max, Qmax
or Ca-vO2max were excluded. In addition, VO2max
(and Qmax if applicable) absolute values (volume/time)
had to be available or if normalised by anthropometri-
cal variables (e.g. weight, body surface area) these latter
had to be not significantly altered by the training inter-
vention. In the event of multiple publications pertaining
to the same research, the first published or more com-
prehensive report was included. The inclusion of art-
icles in our analysis was not limited by publication
status or language. Article selection was performed
independently and in duplicate by two investigators
(DM and CD).
Data extraction and quality assessment
The following variables were summarised in a pre-for-
matted spreadsheet: authors, year of publication, char-
acteristics of study participants (n, age, gender, height,
weight, body surface area, body mass index, haemoglo-
bin concentration, haematocrit, fitness status, health
status), cardiovascular/respiratory variables (heart
rate, blood pressure, total peripheral resistance, stroke
volume, Qmax, Ca-vO2max, VO2max), characteristics of
the assessment of VO2max, Qmax and Ca-vO2max (tech-
nique, test protocol) and ET features (type, modality,
frequency, intensity, session length, duration). The Fick
equation was used to calculate Ca-vO2max in studies
reporting only VO2max and Qmax values.4,8 Missing
standard deviations (SD) of Ca-vO2max in the latter
studies were imputed from linear regression analysis
of log (SD of Ca-vO2max) on log (mean Ca-vO2max)
from studies that reported complete data,1–3,5–7,9–14,25
as previously described.40
A systematic appraisal of quality for observational
research (SAQOR)41 previously applied in meta-analysis
of observational studies evaluating cardiovascular
Montero and Dı́az-Cañestro 3

Records identified from MEDLINE,

scopus and web of science
(n=101 after duplicate removal)

Excluded based on abstract/title screening (n=67)

• Cross-sectional studies (n=18)
• Non-healthy subjects (n=15)
• Mean age <40 years (n=9)
• Irrelevant (n=9)
• Reviews (n=8)
• Acute exercise (n=4)
• Strength training (n=2)
• Non-training intervention (n=2)

Articles selected for full-text

retrieval (n=34)

Additional articles identified through hand

searching (n=5)

Excluded after full-text screening (n=25)

• No report of Qmax nor Ca-vO2max (n=19)
• No report of VO2max (n=3)
• Non-healthy subjects (n=1)
• Strength training (n=1)
• Data not available (n=1)

Articles included in meta-

analysis (n=14)

Figure 1. Flow diagram of the process of article selection.

Ca-vO2max: arteriovenous oxygen difference at maximal exercise; Qmax: maximal cardiac output; VO2max: maximal oxygen
consumption.

function42,43 was performed to provide an assessment of
study quality. The SAQOR was adjusted to assess: (a)
the study sample; (b) quality of VO2max assessment; (c)
quality of Qmax assessment; (d) quality of Ca-vO2max
assessment; (e) confounding variables and (f) data.
Overall, the SAQOR was scored out of 16, quality
deemed better with a greater score. Data extraction
and quality assessment were performed independently
and in duplicate by two investigators (DM and CD).
Data synthesis and analysis
The meta-analysis and related analyses were performed
using Review Manager software (RevMan 5.3; Cochrane
Collaboration, Oxford, UK) and Comprehensive
Meta-analysis software (version 2; Biostat, Englewood,
New Jersey, USA). The primary outcomes were the stan-
dardised mean difference (SMD) in VO2max, Qmax and
Ca-vO2max between post and pre-ET measurements.
The SMD summary statistic allowed us to standardise
values obtained using different methods into a uniform
scale to complete the meta-analysis.44 Each SMD was
weighted by the inverse variance and they were pooled
with a random effects model.44,45 Heterogeneity among
studies was assessed using the chi-squared test for hetero-
geneity and I2 statistics. Potential moderating factors
influencing the SMD in VO2max, SMD in Qmax and
SMD in Ca-vO2max (gender, weight, training characteris-
tics, methodological quality) were evaluated by subgroup
analysis comparing studies grouped by dichotomous or
continuous variables. Median values of continuous vari-
ables were used as cut-off values for grouping studies.

Downloaded from cpr.sagepub.com at The University of Hong Kong Libraries on December 4, 2015
4 European Journal of Preventive Cardiology 0(00)

Table 1. Main characteristics of studies included in the meta-analysis.

Fitness status

Gender Age Weight VO2max Maximal Qmax Ca-vO2max

Study n (%,) (years) (kg) Description (ml/kg/min) exercise test assessment assessment

Wang et al., 201414 13 0 64  3 76  6 Moderately T 48  4 Cycle ergometer single-breath C2H2 fick equation
Fujimoto et al., 20103 9 33 71  3 76  9 UT 23  3 Treadmill C2H2 rebreathing fick equation
Murias et al. (1), 201010 6 100 69  7 72  6 UT 24  2 Cycle ergometer open-circuit C2H2 fick equation
Murias et al. (2), 20109 8 0 68  7 82  8 UT 28  7 Cycle ergometer open-circuit C2H2 fick equation
Morris et al. (A), 20028 10 0 63  3 81  7 N/A 27  4 Cycle ergometer C2H2 rebreathing fick equation
Morris et al. (B), 20028 10 0 65  3 83  10 N/A 28  3 Cycle ergometer C2H2 rebreathing fick equation
Marshall et al., 20016 9 11 58  7 80  7 N/A 33  5 Treadmill C2H2 rebreathing fick equation
McGuire et al., 20017 5 0 50–51 100  37 T/UT 36 treadmill C2H2 rebreathing fick equation
Beere et al., 19991 10 0 66  4 85 N/A 18  7 Cycle ergometer thermodilution catheterization
Hijazi et al., 19985 8 63 52  7 N/A N/A N/A Cycle ergometer N/A catheterization
Spina et al., 199813 10 0 65  3 84  12 UT 29  4 Treadmill/cycle C2H2 rebreathing fick equation
ergometer
Spina et al. (1A), 199311 15 0 63  12 83  12 N/A 28 Treadmill C2H2 rebreathing fick equation
Spina et al. (1B), 199311 16 100 64  12 65  12 N/A 21 Treadmill C2H2 rebreathing fick equation
Spina et al. (2), 19931,2 7 100 63  13 67  9 UT 18  4 Cycle ergometer radioactive fick equation
indicator-dilution
Ehsani et al., 19912 9 0 64  3 80  6 UT 30  4 Cycle ergometer radioactive fick equation
indicator-dilution
Haennel et al., 19894 8 0 42  12 83  24 N/A 33  8 Cycle ergometer radioactive fick equation
indicator-dilution

Data are n, % of women, mean  SD, mean or range. Two articles presented separate study groups that were distinguished by A and B.8,11 Ca-vO2diff:
arteriovenous oxygen difference at maximal exercise; C2H2: acetylene; N/A: data not available; Qmax: maximal cardiac output; T: trained; UT: untrained;
VO2max: maximal oxygen consumption.

Meta-regression analyses were performed to evaluate the
associations among the SMD in VO2max, SMD in Qmax,
SMD in Ca-vO2max and potential moderating factors. In
all meta-regression models, studies were weighted by the
inverse variance of the dependent variable. Potential
moderating factors were entered as independent variables
in regression models with the SMD in VO2max, SMD in
Qmax or SMD in Ca-vO2max as the dependent variable.
Publication and/or other biases were evaluated by the
Begg and Mazumdar’s rank correlation test and
Egger’s regression test.46 A P value less than 0.05 was
considered statistically significant.
Results
Study selection and characteristics
The flow diagram of the process of article selection is
shown in Figure 1, which resulted in the inclusion of 14
articles. Two of the articles presented separate study
groups,8,11 each of which was evaluated as an individ-
ual study. Table 1 illustrates the main characteristics of
the resulting 16 studies, comprising a total of 153
healthy middle-aged and older subjects (mean age ran-
ging from 42 to 71 years). Two studies included trained
subjects,7,14 six studies only involved untrained sub-
jects,2,3,9,10,12,13 while eight studies did not report on
fitness status but presented poor to very poor mean
group VO2max baseline values1,4–6,8,11 conforming to
established guidelines.47 ET programmes consisted
of endurance continuous training (ECT) and/or
endurance interval training (EIT), respectively, of vari-
able intensity performed through walking, running,
cycling and/or rowing exercise, ranging from 1.4 to
4.0 hours per week and from 8 to 52 weeks’ duration
(Table 2).
The mean quality of the studies was moderate
to high. The average score was 10.1  1.7 out of a pos-
sible 16 points (Supplementary Table 1). As for the
evaluation of potential biases, the funnel plot
(Supplementary Figure 2), Begg and Mazumdar’s
rank correlation test and the Egger’s regression test
did not suggest the presence of publication bias and/
or other biases for the SMD in VO2max in the studies
included in the meta-analysis (P ¼ 0.19, P ¼ 0.40,
respectively). Likewise, there was no evidence of publi-
cation and/or other biases with regard to the SMD in
Qmax (Supplementary Figure 3) and the SMD in Ca-
vO2max (Supplementary Figure 4) in the studies
included in the meta-analysis.

Downloaded from cpr.sagepub.com at The University of Hong Kong Libraries on December 4, 2015
Montero and Dı́az-Cañestro 5

Table 2. Endurance training characteristics of studies included in the meta-analysis.

Intensity Hours/ Duration Load

Study Type, modality Intensity description status (0–5) wk (wk) (AU)a

Wang et al., 201414 Cycling, EIT 4 min 90–95% HRmax, 5.0 1.40 8.0 56.0
3 min 70% HRmax
Fujimoto et al., 20103 Walking/running, 75–90% HRmax 3.0 2.38 52.1 372.3
ECT/EIT
Murias et al. (1), 201010 Cycling, ECT/EIT 70% VO2max (10 wk); 4.0 2.06 12.0 98.9
1 min 90–100% Wattmax,
1 min rest (2 wk)
Murias et al. (2), 20109 Cycling, ECT/EIT 70% VO2max (10 wk); 4.0 2.06 12.0 98.9
1 min 90–100% Wattmax,
1 min rest (2 wk)
Morris et al. (A), 20028 Cycling, ECT 87% HRmax 3.5 1.45 10.0 50.8
Morris et al. (B), 20028 Cycling, ECT 66% HRmax 1.0 1.45 10.0 14.5
Marshall et al., 20016 Cycling, ECT 75–80% VO2max 3.5 1.67 8.0 46.7
McGuire et al., 20017 Walking/running/cycling, 75% HRmax 3.0 2.86 19.6 168.2
ECT
Beere et al., 19991 (Arm–leg) cycling, ECT 75–90% HRmax 3.5 1.50 12.9 67.5
Hijazi et al., 19985 Cycling, ECT 65% VO2max 3.0 2.50 8.0 60.0
Spina et al., 199813 Walking/running/ 85% HRmaxb 3.5 4.00 36.0 504.0
cycling/rowing,
ECT/EIT
Spina et al. (1A), 199311 Walking/running/ 81% HRmax 3.0 2.93 45.0 396.0
cycling/rowing,
ECT/EIT
Spina et al. (1B), 199311 Walking/running/ 77% HRmax 3.0 2.91 45.0 393.5
cycling/rowing,
ECT/EIT
Spina et al. (2), 199312 Walking/running/ 81% HRmax 3.0 3.85 45.0 519.8
cycling/rowing,
ECT/EIT
Ehsani et al., 19912 Walking/running/ 60–80% VO2max, 93% 3.5 4.00 50.6 708.0
cycling, ECT/EIT VO2max (short bouts)
Haennel et al., 19894 Cycling, ECT 80% HRR 3.5 1.35 9.0 42.5
Weighted mean – – 3.3 2.39 25.5 236.3
Data are mean or range. Two articles presented separate study groups that were distinguished by A and B.8,11
a
Calculated by multiplying intensity status, hours/week and duration (weeks) of training. bDuring last 3 months of the exercise intervention. AU:
arbitrary units; ECT: endurance continuous training; EIT: endurance interval training; HRmax: maximal heart rate; HRR: heart rate reserve; VO2max:
maximal oxygen consumption; Wattmax: maximal power output; wk: week.

Effect of ET on VO2max Effect of ET on Qmax

VO2max was determined in all studies during cycle erg-
ometer or treadmill incremental exercise (Table 1).
After data pooling, the meta-analysis revealed an
increase in VO2max after ET (SMD 0.89, 95% confi-
dence interval (CI) 0.65, 1.12; P < 0.0001) (Figure 2).
There was no heterogeneity between studies (I2 ¼ 0%,
P ¼ 0.98). In subgroup analyses, none of the assessed
potential moderating factors (gender, training charac-
teristics, methodological quality) significantly influ-
enced the SMD in VO2max (Table 3).
Qmax was evaluated via acetylene (C2H2) rebreathing
(n ¼ 8), radioactive indicator dilution (n ¼ 3), open-circuit
C2H2 (n ¼ 2), single-breath C2H2 (n ¼ 1) and thermodilu-
tion (n ¼ 1) techniques during upright (n ¼ 13) or supine
(n ¼ 2) exercise (Table 1). After data pooling, Qmax was
enhanced following ET (SMD 0.61, 95% CI 0.37, 0.85;
P < 0.0001) (Figure 2). Heterogeneity between studies was
not detected (I2 ¼ 0%, P ¼ 0.84). As for the intrinsic
determinants of Qmax, maximal stroke volume (SVmax)
was increased (SMD 0.52, 95% CI 0.28, 0.77, I2 ¼ 0%;

Downloaded from cpr.sagepub.com at The University of Hong Kong Libraries on December 4, 2015
6 European Journal of Preventive Cardiology 0(00)

Post training Pre training Std. Mean difference Std. Mean difference
Study or subgroup Mean SD Total Mean SD Total Weight IV, random, 95% CI IV, random, 95% CI
1.1.1 Effect of endurance training on maximal oxygen consumption
Wang et al, 2014 3.9 0.24 13 3.69 0.3 13 8.9% 0.75 [–0.05, 1.55]
Fujimoto et al, 2010 1.97 0.31 9 1.73 0.3 9 6.1% 0.75 [–0.22, 1.71]
Murias et al 1, 2010 2.01 0.29 6 1.73 0.25 6 3.8% 0.95 [–0.27, 2.18]
Murias et al 2, 2010 2.95 0.48 8 2.29 0.49 8 4.7% 1.29 [0.18, 2.39]
Morris et al A, 2002 2.48 0.25 10 2.21 0.32 10 6.6% 0.90 [–0.03, 1.83]
Morris et al B, 2002 2.57 0.22 10 2.3 0.22 10 6.1% 1.18 [0.21, 2.14]
Marshall et al, 2001 2.85 0.54 9 2.64 0.51 9 6.5% 0.38 [–0.55, 1.32]
McGuire et al, 2001 3.26 0.64 5 2.96 0.65 5 3.6% 0.42 [–0.84,1.68]
Beere et al, 1999 1.84 0.53 10 1.57 0.54 10 7.2% 0.48 [–0.41, 1.38]
Hijazi et al, 1998 1.23 0.25 8 1 0.25 8 5.3% 0.87 [–0.17, 1.91]
Spina et al, 1998 2.87 0.35 10 2.39 0.35 10 5.9% 1.31 [0.33, 2.30]
Spina et al 1 A, 1993 2.8 0.39 15 2.35 0.39 15 9.4% 1.12 [0.34, 1.90]
Spina et al 1 B, 1993 1.66 0.4 16 1.36 0.4 16 11.1% 0.73 [0.01, 1.45]
Spina et al 2, 1993 1.27 0.16 7 1.08 0.16 7 4.3% 1.11 [–0.04, 2.27]
Ehsani et al, 1991 2.05 0.26 9 1.85 0.2 9 6.0% 0.82 [–0.15, 1.79]
Haennel et al, 1989 3.22 0.28 8 2.82 0.25 8 4.5% 1.42 [0.29, 2.56]
Subtotal (95% CI) 153 153 100.0% 0.89 [0.65, 1.12]
Heterogeneity: Tau2 = 0.00; Chi2 = 5.77, df = 15 (P = 0.98); I2 = 0%
Test for overall effect: Z = 7.26 (P < 0.00001)

1.1.2 Effect of endurance training on maximal cardiac output

Wang et al, 2014 24 2.5 13 21.8 2.7 13 9.0% 0.82 [0.01, 1.62]
Fujimoto et al, 2010 19.23 4.72 9 17.46 4.11 9 6.7% 0.38 [–0.55, 1.32]
Murias et al 1, 2010 16.7 1.6 6 15.8 1.6 6 4.4% 0.52 [–0.64, 1.68]
Murias et al 2, 2010 20.3 1.8 8 16.8 3 8 4.7% 1.34 [0.22, 2.45]
Morris et al A, 2002 14.68 1.01 10 13.29 1.26 10 6.3% 1.17 [0.20, 2.13]
Morris et al B, 2002 15.27 2.06 10 14.08 1.61 10 7.2% 0.62 [–0.29, 1.52]
Marshall et al, 2001 18.92 3.33 9 17.14 3.78 9 6.6% 0.48 [–0.46,1.42]
McGuire et al, 2001 21.68 5.44 5 21.36 5.05 5 3.8% 0.06 [–1.18, 1.30]
Beere et al, 1999 12.8 2.4 10 12.3 3 10 7.6% 0.18 [–0.70, 1.05]
Spina et al, 1998 18.3 2.45 6 15.4 1.96 8 3.6% 1.21 [–0.07, 2.48]
Spina et al 1 A, 1993 19 3.87 15 17 3.87 15 11.0% 0.50 [–0.23, 1.23]
Spina et al 1 B, 1993 11.5 1.6 16 11.2 1.2 16 12.1% 0.21 [–0.49, 0.90]
Spina et al 2, 1993 14.5 2.65 7 13.5 2.12 7 5.2% 0.39 [–0.67, 1.45]
Ehsani et al, 1991 18.5 4.1 10 15.7 2.6 19 7.0% 0.78 [–0.14, 1.70]
Haennel et al, 1989 20.6 1.41 8 18.6 1.7 8 4.9% 1.21 [0.12, 2.30]
Subtotal (95% CI) 142 142 100.0% 0.61 [0.37, 0.85]
Heterogeneity: Tau2 = 0.00; Chi2 = 8.88, df = 14 (P = 0.84); I2 = 0%
Test for overall effect: Z = 4.92 (P < 0.00001)

1.1.3 Effect of endurance training on arteriovenous oxygen difference at maximal exercise

Wang et al, 2014 16.3 2.3 13 17.3 2.7 13 8.3% –0.39 [–1.16, 0.39]
Fujimoto et al, 2010 10.4 1 9 10.1 1.4 9 6.5% 0.23 [–0.69, 1.16]
Murias et al 1, 2010 12 1.7 6 11 1.8 6 4.6% 0.53 [–0.63, 1.69]
Murias et al 2, 2010 14.7 2.1 8 13.5 2.2 8 5.7% 0.53 [–0.47, 1.53]
Morris et al A, 2002 16.89 1.76 10 16.63 1.77 10 7.0% 0.14 [–0.74, 1.02]
Morris et al B, 2002 16.83 1.77 10 16.34 1.78 10 7.0% 0.26 [–0.62, 1.15]
Marshall et al, 2001 15.05 0.63 9 15.52 0.93 9 6.3% –0.56 [–1.51, 0.38]
McGuire et al, 2001 15.24 1.98 5 13.8 1.97 5 3.8% 0.66 [–0.64, 1.95]
Beere et al, 1999 14.3 2.3 10 12.5 1.7 10 6.5% 0.85 [–0.07, 1.78]
Hijazi et al, 1998 16.2 2.55 8 14.1 1.98 8 5.4% 0.87 [0.17, 1.91]
Spina et al, 1998 14.4 1.47 6 14.7 1.71 6 4.7% –0.17 [–1.31, 0.96]
Spina et al 1 A, 1993 14.8 1.16 15 13.8 0.77 15 8.5% 0.99 [0.22, 1.75]
Spina et al 1 B, 1993 14.4 1.6 16 12.2 1.6 16 8.3% 1.34 [0.56, 2.12]
Spina et al 2, 1993 9 2.12 7 8 1.59 7 5.2% 0.50 [–0.57,1.57]
Ehsani et al, 1991 12 2.7 9 12 2.5 9 6.5% 0.00 [–0.92, 0.92]
Haennel et al, 1989 15.63 1.81 8 15.16 1.83 8 5.9% 0.24 [–0.74, 1.23]
Subtotal (95% CI) 149 149 100.0% 0.38 [0.11, 0.66]
Heterogeneity: Tau2 = 0.08; Chi2 = 2015, df = 15 (P = 0.17); I2 = 26%
Test for overall effect: Z = 2.75 (P = 0.006)

–4 –2 0 2 4
Decrease Increase

Figure 2. Forest plots of the SMD in VO2max, Qmax and Ca-vO2max.

Squares represent the SMD for each study. Diamonds represents the pooled SMD across studies.

Downloaded from cpr.sagepub.com at The University of Hong Kong Libraries on December 4, 2015
Montero and Dı́az-Cañestro 7

Table 3. Subgroup analyses of the effect of endurance training on VO2max, Qmax and a-vO2diff at maximal exercise.
VO2max Qmax Ca-vO2max

SMD SMD SMD

Subgroup (95% CI) P value I2 Pdiff (95% CI) P value I2 Pdiff (95 % CI) P I2 Pdiff

Gender
Women (n ¼ 3) 0.86 (0.31, 1.41) 0.002 0 0.75 0.31 (–0.21, 0.83) 0.24 0 0.16 0.92 (0.34, 1.50) 0.002 7 0.07
Men (n ¼ 10) 0.96 (0.66, 1.26) <0.001 0 0.74 (0.44, 1.04) <0.001 0 0.31 (0.01, 0.60) 0.04 2
Training characteristics
Type
Cycling (n ¼ 9) 0.87 (0.54, 1.19) <0.001 0 0.86 0.75 (0.41, 1.09) <0.001 0 0.23 0.22 (0.11, 0.56) 0.18 10 0.20
Including 0.91 (0.56, 1.26) <0.001 0 0.46 (0.11, 0.80) 0.009 0 0.58 (0.15, 1.01) 0.008 30
weight-bearing
(n ¼ 7)
Modality
ECT (n ¼ 7) 0.79 (0.41, 1.17) <0.001 0 0.54 0.62 (0.21, 1.02) 0.003 0 0.96 0.32 (0.06, 0.69) 0.10 2 0.71
Including EIT (n ¼ 9) 0.95 (0.64, 1.25) <0.001 0 0.60 (0.30, 0.91) <0.001 0 0.42 (0.02, 0.83) 0.04 42
Intensity status
>3.5 (n ¼ 3) 0.94 (0.37, 1.51) 0.001 0 0.84 0.88 (0.31, 1.45) 0.002 0 0.30 0.12 (0.52, 0.77) 0.71 26 0.37
3.5 (n ¼ 13) 0.87 (0.61, 1.14) <0.001 0 0.55 (0.28, 0.81) <0.001 0 0.45 (0.14, 0.75) 0.004 25
Duration
>12.4 weeks (n ¼ 8) 0.85 (0.53, 1.18) <0.001 0 0.77 0.42 (0.10, 0.74) 0.01 0 0.08 0.62 (0.25, 1.00) 0.001 20 0.06
12.4 weeks (n ¼ 8) 0.92 (0.57, 1.28) <0.001 0 0.85 (0.49, 1.22) <0.001 0 0.14 (0.20, 0.47) 0.42 0
Load
>98.9 AU (n ¼ 7) 0.91 (0.56, 1.26) <0.001 0 0.86 0.46 (0.11, 0.80) 0.009 0 0.23 0.58 (0.15, 1.01) 0.008 30 0.20
98.9 AU (n ¼ 9) 0.87 (0.54, 1.19) <0.001 0 0.75 (0.41, 1.09) <0.001 0 0.22 (0.11, 0.56) 0.18 10
Maximal exercise test
Cycle ergometer 0.97 (0.67, 1.27) <0.001 0 0.37 0.78 (0.47, 1.10) <0.001 0 0.08 0.26 (0.02, 0.55) 0.07 0 0.43
(n ¼ 10)
Treadmill (n ¼ 5) 0.74 (0.35, 1.14) <0.001 0 0.35 (–0.01, 0.71) 0.05 0 0.56 (0.11, 1.24) 0.10 63
Methodological quality
>10 points (n ¼ 7) 0.96 (0.63, 1.29) <0.001 0 0.52 0.63 (0.30, 0.96) <0.001 0 0.84 0.41 (0.09, 0.91) 0.10 55 0.80
10 points (n ¼ 9) 0.80 (0.46, 1.15) <0.001 0 0.58 (0.22, 0.93) 0.001 0 0.34 (0.01, 0.67) <0.05 0

Median values of continuous variables were used as cut-off values for grouping studies. AU: arbitrary units; Ca-vO2max: arteriovenous oxygen difference
at maximal exercise; CI: confidence interval; ECT: endurance continuous training; EIT: endurance interval training; I2: index of heterogeneity; n: number
of studies; Qmax: maximal cardiac output; SMD: standardised mean difference between post and pre-endurance training measurements; VO2max:
maximal oxygen consumption.

P < 0.0001), whereas maximum heart rate (HRmax) was
not altered (mean difference 0.52 beats/min, 95% CI –
2.17, 3.21, I2 ¼ 17%; P ¼ 0.71) after ET. In subgroup ana-
lyses (Table 3), Qmax was not enhanced after ET in studies
only comprising women (n ¼ 3) (SMD 0.31, 95% CI –
0.21, 0.83; P ¼ 0.24). Moreover, the increase in Qmax
after ET in studies that assessed maximal exercise on a
treadmill (n ¼ 5) did not reach statistical significance
(SMD 0.35, 95% CI –0.01, 0.71; P ¼ 0.05).
Effect of ET on Ca-vO2max
The Fick equation was used to calculate Ca-vO2max in
the majority of the studies (Table 1). Two studies dir-
ectly determined Ca-vO2max by means of invasive tech-
niques.1,5 After data pooling, Ca-vO2max was increased
after ET (SMD 0.38, 95% CI 0.11, 0.66; P ¼ 0.006)
(Figure 2). There was no significant heterogeneity
between studies (I2 ¼ 26 %, P ¼ 0.17). In subgroup ana-
lyses, Ca-vO2max was not increased following ET in
studies (a) consisting of only cycling exercise training
(n ¼ 9) (SMD 0.22, 95% CI –0.11, 0.56; P ¼ 0.18) or
ECT (n ¼ 7) (SMD 0.32, 95% CI –0.06, 0.69;
P ¼ 0.10), (b) with an intensity status above the
median value (n ¼ 3) (SMD 0.12, 95% CI –0.52, 0.77;
P ¼ 0.71), (c) with a duration or load of training below
the median value (n ¼ 8 and n ¼ 9, respectively) (SMD –
0.14, 95% CI –0.20, 0.47; P ¼ 0.42; SMD 0.22, 95% CI
–0.11, 0.56; P ¼ 0.18, respectively) and (d) with a meth-
odological quality score above the median value (n ¼ 7)
(SMD 0.41, 95% CI –0.09, 0.91; P ¼ 0.10).
Meta-regression analyses
A positive association was found between the SMD in
Qmax and the SMD in VO2max (B ¼ 0.79, 95% CI 0.02,

Downloaded from cpr.sagepub.com at The University of Hong Kong Libraries on December 4, 2015
8 European Journal of Preventive Cardiology 0(00)

(a) 2.00
1.80
1.60
1.40
SMD in Qmax 1.20
1.00
0.80
0.60
0.40
0.20
0.00
0.28 0.40 0.53 0.65 0.78 0.90 1.02 1.15 1.27 1.40 1.52
(b) 2.00
1.74
1.48
SMD in Ca-vO2max

1.22
0.96
0.70
0.44
0.18
–0.08
–0.34
–0.60
0.28 0.40 0.53 0.65 0.78 0.90 1.02 1.15 1.27 1.40 1.52
SMD in VO2max

Figure 3. Meta-regression plots of the SMD in VO2max according to the SMD in Qmax (B ¼ 0.79, 95% CI 0.02, 1.56; P ¼ 0.04) (a) and
the SMD in Ca-vO2max (B ¼ 0.09, 95% CI –0.84, 1.03; P ¼ 0.84) (b). The size of each circle is proportional to the study’s weight.
Ca-vO2max: arteriovenous oxygen difference at maximal exercise; Qmax: maximal cardiac output; SMD: standardised mean difference
between post and pre-endurance training measurements; VO2max: maximal oxygen consumption.

1.56; P ¼ 0.04) (Figure 3(a)). Likewise, the SMD in
SVmax, but not the mean difference in HRmax, was posi-
tively associated with the SMD in Qmax (B ¼ 0.67, 95%
CI 0.16, 1.19; P ¼ 0.01) and the SMD in VO2max
(B ¼ 0.81, 95% CI 0.03, 1.58; P ¼ 0.04).
In contrast, the SMD in Ca-vO2max was not asso-
ciated with the SMD in VO2max (B ¼ 0.09, 95% CI –
0.84, 1.03; P ¼ 0.84) (Figure 3(b)). None of the assessed
potential moderating factors were significantly asso-
ciated with the SMD in VO2max, the SMD in Qmax or
the SMD in Ca-vO2max.
Discussion
In this systematic review and meta-analysis, we pooled
and analysed data from 16 studies assessing the effects
of ET interventions lasting 8 to 52 weeks on VO2max
and Qmax and/or Ca-vO2max in a total of 153 healthy
middle-aged and older subjects. The main finding of
this meta-analysis is that despite both Qmax and Ca-
vO2max being increased with ET, only Qmax is linearly
associated with the increase in VO2max.
Over the inconsistent findings reported by previous
studies,1–14 this meta-analysis demonstrates increases in
the VO2max components after ET in aged subjects, with
null or low heterogeneity among studies. This indicates
the presence of statistical limitations (e.g. type II errors)
in prior research, which are common to scientific fields
based on small sample size studies48 – and furthermore,
corroborates the adequacy of our meta-analytic
approach. In addition, we found that (a) a linear func-
tion related the ET-induced increases in VO2max and
Qmax but not Ca-vO2max, and (b) the increase in Qmax
entirely resulted from that of SVmax because HRmax was
not altered with ET, all concurring with pooled evi-
dence in young individuals.34 Potential mechanisms
underlying the increase in SVmax induced by ET in
aged individuals comprise adaptations in left ventricu-
lar structure/function,17 blood volume19 as well as
decreased peripheral vascular resistance,28 plausibly
due to enhanced muscle capillarisation,22 vascular
endothelial and/or smooth muscle dilator function.49,50
Taken together, the above findings suggest that the
effect of ET on VO2max is proportional to the increase

Downloaded from cpr.sagepub.com at The University of Hong Kong Libraries on December 4, 2015
Montero and Dı́az-Cañestro
in SVmax throughout adulthood and healthy ageing,
regardless of the pre-training level of oxygen extraction,
which is reduced in aged individuals.35,37,38 Of note,
such a relationship might differ according to
gender,10–12 albeit this remains uncertain on the basis
of scarce data in women (three studies, n ¼ 29) in the
present meta-analysis.
The observation that oxygen extraction is increased
with ET contrasts with our previous meta-analysis in
young individuals.34 Both meta-analyses, however,
coincide in that Ca-vO2max is only improved with
longer (>12 weeks) ET interventions.34 Importantly,
the average duration of ET in studies in aged subjects
was seemingly higher compared with studies in young
subjects (25.5 vs. 8.6 weeks), thus, this may explain, at
least in part, the aforementioned discrepancy.
Moreover, the fact that Ca-vO2max is improved with
ET after, but not prior to, a given period of time in
young and aged individuals is intrinsically noteworthy.
This highlights the existence of prevalent temporal
boundaries for the effect of ET on the determinants
of oxygen extraction in healthy individuals, which
may explain, to some degree, the dissociation between
Ca-vO2max and VO2max. As for the mechanisms, the
increase in Ca-vO2max with ET in aged individuals
could be solely attributed to the improvement in
blood flow distribution to exercising limbs, given that
leg Ca-vO2max may not be affected by ET.1 However,
further invasive studies in aged subjects are required to
discard any contribution of peripheral adaptations
exclusively enhancing microvascular oxygen delivery
and/or diffusion from capillaries in to muscle to the
improvement in VO2max, as previously debated.51
Nonetheless, this meta-analysis buttressed the notion
that VO2max is primarily governed by the convective
oxygen delivery to the exercising limbs.20,52,53
Although the ET programmes of the included stu-
dies presented substantial diversity, the effects asso-
ciated with distinct training features were certainly
homogeneous, except for those regarding oxygen
extraction. Other than duration, the type, modality,
intensity and load of ET influenced Ca-vO2max in sub-
group analyses (Table 3). With respect to type and
modality, the increase in Ca-vO2max reached statistical
significance in studies including, but not limited to,
weight-bearing exercise (walking, running) or EIT but
not in those comprising only cycling or ECT. While
speculative, the effect of ET on Ca-vO2max may be
facilitated by the interaction of multiple adaptations
potentially improving blood flow distribution such as
enhanced skeletal muscle capillarisation, vascular con-
ductance and sympathetic activity, among others,
induced by different types and modalities of exercise
in aged individuals.54–56 Moreover, Ca-vO2max was
increased with ET interventions characterised by an
Downloaded from cpr.sagepub.com at The University of Hong Kong Libraries on December 4, 2015
9
intensity status below the median value. In this line, it
has been demonstrated that isoenergetic ET of lower
versus higher intensity is more effective in enhancing
skeletal muscle capillary density, vascular conductance
and thereby Ca-vO2max.56 In turn, Ca-vO2max was posi-
tively influenced by the load of ET (calculated by multi-
plying intensity status and total hours of training). As
Ca-vO2max was not positively associated with intensity
status, it is implied that the increase in oxygen extrac-
tion in studies with a higher load of training was pri-
marily driven by their prolonged duration, as discussed
above. Furthermore, it should be noted that Ca-vO2max
was only significantly increased in studies with moder-
ate to low methodological quality, which is also
observed in young individuals.34 Ultimately, consider-
ing the nature and relatively small number of studies
comprising the subgroup analyses here applied, these
should be considered exploratory and not as proof of
causality.57
Limitations
There are some limitations and strengths in this meta-
analysis. First, the findings were derived from studies
mostly comprising men, thus our conclusions should be
primarily limited to male individualss. Second, the Ca-
vO2max was calculated by means of the Fick equation
(Ca-vO2max ¼ VO2max/Qmax) in the majority of the stu-
dies, which could have reduced the statistical power to
detect the association between Ca-vO2max and VO2max,
because of the combined error in Qmax and VO2max
measurements.58,59 However, such a source of con-
founding may have been attenuated by the use of the
SMD (between post and pre-ET) as the summary stat-
istic, which is suggested by the null heterogeneity in
VO2max and Qmax among studies. Finally, the mean
methodological quality of the included studies was
determined as moderate to high, and the presence of
publication and/or other biases was discarded.
Conclusions
The current meta-analysis demonstrates that the
improvement in VO2max with ET is a linear function
of Qmax, but not Ca-vO2max, in primarily untrained
healthy aged individuals, a finding that is similarly pre-
sent in young individuals. This suggests that any
increase in systemic oxygen delivery with ET contrib-
utes, in a proportional manner, to the enhancement of
VO2max, independent of healthy ageing and its related
decrease in oxygen extraction capacity. Due to the
small number of women in the studies included in the
meta-analysis, further research is needed to elucidate
the determinant(s) of VO2max improvement with ET
in women.
10
Conflict of interest
The authors declared no potential conflicts of interest with
respect to the research, authorship, and/or publication of this
article.
Funding
The authors received no financial support for the research,
authorship, and/or publication of this article.
References
1. Beere PA, Russell SD, Morey MC, et al. Aerobic exercise
training can reverse age-related peripheral circulatory
changes in healthy older men. Circulation 1999; 100:
1085–1094.
2. Ehsani AA, Ogawa T, Miller TR, et al. Exercise training
improves left ventricular systolic function in older men.
Circulation 1991; 83: 96–103.
3. Fujimoto N, Prasad A, Hastings JL, et al. Cardiovascular
effects of 1 year of progressive and vigorous exercise
training in previously sedentary individuals older than
65 years of age. Circulation 2010; 122: 1797–1805.
4. Haennel R, Teo KK, Quinney A, et al. Effects of hydrau-
lic circuit training on cardiovascular function. Med Sci
Sports Exerc 1989; 21: 605–612.
5. Hijazi OM, Ramanathan M, Estrera AS, et al. Fixed
maximal stroke index in patients after pneumonectomy.
Am J Respir Crit Care Med 1998; 157: 1623–1629.
6. Marshall P, Al-Timman J, Riley R, et al. Randomized
controlled trial of home-based exercise training to evalu-
ate cardiac functional gains. Clin Sci 2001; 101: 477–483.
7. McGuire DK, Levine BD, Williamson JW, et al. A 30-
year follow-up of the Dallas bedrest and training study:
II. Effect of age on cardiovascular adaptation to exercise
training. Circulation 2001; 104: 1358–1366.
8. Morris N, Gass G, Thompson M, et al. Rate and ampli-
tude of adaptation to intermittent and continuous exer-
cise in older men. Med Sci Sports Exerc 2002; 34:
471–477.
9. Murias JM, Kowalchuk JM and Paterson DH. Time
course and mechanisms of adaptations in cardiorespira-
tory fitness with endurance training in older and young
men. J Appl Physiol (1985) 2010; 108: 621–627.
10. Murias JM, Kowalchuk JM and Paterson DH.
Mechanisms for increases in VO2max with endurance
training in older and young women. Med Sci Sports
Exerc 2010; 42: 1891–1898.
11. Spina RJ, Ogawa T, Kohrt WM, et al. Differences in
cardiovascular adaptations to endurance exercise training
between older men and women. J Appl Physiol (1985)
1993; 75: 849–855.
12. Spina RJ, Ogawa T, Miller TR, et al. Effect of exercise
training on left ventricular performance in older women
free of cardiopulmonary disease. Am J Cardiol 1993; 71:
99–104.
13. Spina RJ, Turner MJ and Ehsani AA. Beta-adrenergic-
mediated improvement in left ventricular function by
exercise training in older men. Am J Physiol-Heart
Circulat Physiol 1998; 274: H397–H404.
Downloaded from cpr.sagepub.com at The University of Hong Kong Libraries on December 4, 2015
European Journal of Preventive Cardiology 0(00)
14. Wang E, Næss MS, Hoff J, et al. Exercise-training-
induced changes in metabolic capacity with age: The
role of central cardiovascular plasticity. Age 2014; 36:
665–676.
15. Huang G, Wang R, Chen P, et al. Dose–response rela-
tionship of cardiorespiratory fitness adaptation to con-
trolled endurance training in sedentary older adults. Eur
J Prev Cardiol 2015. Epub ahead of print 21 April 2015.
DOI: 10.1177/2047487315582322.
16. Baggish AL, Wang F, Weiner RB, et al. Training-specific
changes in cardiac structure and function: A prospective
and longitudinal assessment of competitive athletes.
J Appl Physiol (1985) 2008; 104: 1121–1128.
17. Seals DR, Hagberg JM, Spina RJ, et al. Enhanced left
ventricular performance in endurance trained older men.
Circulation 1994; 89: 198–205.
18. Sawka MN, Convertino VA, Eichner ER, et al. Blood
volume: Importance and adaptations to exercise training,
environmental stresses, and trauma/sickness. Med Sci
Sports Exerc 2000; 32: 332–348.
19. Hagberg JM, Goldberg AP, Lakatta L, et al. Expanded
blood volumes contribute to the increased cardiovascular
performance of endurance-trained older men. J Appl
Physiol (1985) 1998; 85: 484–489.
20. Bonne TC, Doucende G, Fluck D, et al. Phlebotomy
eliminates the maximal cardiac output response to six
weeks of exercise training. Am J Physiol Regul Integr
Comp Physiol 2014; 306: R752–760.
21. Murias JM, Kowalchuk JM, Ritchie D, et al.
Adaptations in capillarization and citrate synthase activ-
ity in response to endurance training in older and young
men. J Gerontol A Biol Sci Med Sci 2011; 66: 957–964.
22. Grimby G and Saltin B. The ageing muscle. Clin Physiol
1983; 3: 209–218.
23. Hoppeler H, Howald H, Conley K, et al. Endurance
training in humans: Aerobic capacity and structure of
skeletal muscle. J Appl Physiol (1985) 1985; 59: 320–327.
24. Jacobs RA, Fluck D, Bonne TC, et al. Improvements in
exercise performance with high-intensity interval training
coincide with an increase in skeletal muscle mitochondrial
content and function. J Appl Physiol (1985) 2013; 115:
785–793.
25. Jacobs RA and Lundby C. Mitochondria express
enhanced quality as well as quantity in association with
aerobic fitness across recreationally active individuals up
to elite athletes. J Appl Physiol (1985) 2013; 114:
344–350.
26. Klausen K, Secher NH, Clausen JP, et al. Central
and regional circulatory adaptations to one-leg training.
J Appl Physiol Respir Environ Exerc Physiol 1982; 52:
976–983.
27. Weng TP, Huang SC, Chuang YF, et al. Effects of inter-
val and continuous exercise training on cd4 lymphocyte
apoptotic and autophagic responses to hypoxic stress in
sedentary men. PLoS One 2013; 8: e80248.
28. Makrides L, Heigenhauser GJ and Jones NL. High-intensity
endurance training in 20- to 30- and 60- to 70-yr-old healthy
men. J Appl Physiol (1985) 1990; 69: 1792–1798.
29. Kalliokoski KK, Oikonen V, Takala TO, et al. Enhanced
oxygen extraction and reduced flow heterogeneity in
Montero and Dı́az-Cañestro
exercising muscle in endurance-trained men. Am J
Physiol Endocrinol Metab 2001; 280: E1015–1021.
30. Ekblom B, Astrand PO, Saltin B, et al. Effect of training
on circulatory response to exercise. J Appl Physiol 1968;
24: 518–528.
31. Saltin B, Blomqvist G, Mitchell JH, et al. Response to
exercise after bed rest and after training. Circulation 1968;
38: VII1–78.
32. Saltin B and Calbet JA. Point: In health and in a
normoxic environment, VO2max is limited primarily by
cardiac output and locomotor muscle blood flow.
J Appl Physiol (1985) 2006; 100: 744–745.
33. Arbab-Zadeh A, Perhonen M, Howden E, et al. Cardiac
remodeling in response to 1 year of intensive endurance
training. Circulation 2014; 130: 2152–2161.
34. Montero D, Diaz-Canestro C and Lundby C. Endurance
training and VO2max: Role of maximal cardiac output
and oxygen extraction. Med Sci Sports Exerc 2015; 47:
2024–2033.
35. Ogawa T, Spina RJ, Martin III WH, et al. Effects of
aging, sex, and physical training on cardiovascular
responses to exercise. Circulation 1992; 86: 494–503.
36. Roltsch MH, Brown MD, Hand BD, et al. No associ-
ation between ace i/d polymorphism and cardiovascular
hemodynamics during exercise in young women. Int J
Sports Med 2005; 26: 638–644.
37. Carrick-Ranson G, Hastings JL, Bhella PS, et al. The
effect of lifelong exercise dose on cardiovascular function
during exercise. J Appl Physiol 2014; 116: 736–745.
38. Fleg JL, Schulman SP, O’Connor FC, et al.
Cardiovascular responses to exhaustive upright cycle
exercise in highly trained older men. J Appl Physiol
1994; 77: 1500–1506.
39. Stroup DF, Berlin JA, Morton SC, et al. Meta-analysis of
observational studies in epidemiology: A proposal for
reporting. Meta-analysis of observational studies in
epidemiology (MOOSE) group. JAMA 2000; 283:
2008–2012.
40. Marinho VC, Higgins JP, Sheiham A, et al. Fluoride
toothpastes for preventing dental caries in children
and adolescents. Cochrane Database Syst Rev 2003:
CD002278.
41. Ross LE, Grigoriadis S, Mamisashvili L, et al. Quality
assessment of observational studies in psychiatry: An
example from perinatal psychiatric research. Int J
Methods Psychiatr Res 2011; 20: 224–234.
42. Montero D, Padilla J, Diaz-Canestro C, et al. Flow-
mediated dilation in athletes: Influence of aging. Med
Sci Sports Exerc 2014; 46: 2148–2158.
43. Sprung VS, Atkinson G, Cuthbertson DJ, et al.
Endothelial function measured using flow-mediated dila-
tion in polycystic ovary syndrome: A meta-analysis of the
observational studies. Clin Endocrinol (Oxford) 2013; 78:
438–446.
Downloaded from cpr.sagepub.com at The University of Hong Kong Libraries on December 4, 2015
11
44. Higgins JPT and Green S (eds). Cochrane handbook
for systematic reviews of interventions version 5.1.0
[updated March 2011]. The Cochrane Collaboration,
2011. www.Cochrane-handbook.Org (2011, accessed 25
August 2013).
45. DerSimonian R and Laird N. Meta-analysis in clinical
trials. Control Clin Trials 1986; 7: 177–188.
46. Egger M, Davey Smith G, Schneider M, et al. Bias in
meta-analysis detected by a simple, graphical test. BMJ
1997; 315: 629–634.
47. ACSM. ACSM’s guidelines for exercise testing and pre-
scription. Baltimore, Maryland, USA: Lippincott Raven,
2009.
48. Ioannidis JP. Why most published research findings are
false. PLoS Med 2005; 2: e124.
49. Montero D, Walther G, Diaz-Canestro C, et al.
Microvascular dilator function in athletes: A systematic
review and meta-analysis. Med Sci Sports Exerc 2014; 47:
1485–1494.
50. Montero D. The association of cardiorespiratory fitness
with endothelial or smooth muscle vasodilator function.
Eur J Prev Cardiol 2015; 22: 1200–1211.
51. Lundby C and Montero D. Diffusion limitation of O2
from microvessels in to muscle does not contribute to
the limitation of VO2max. J Physiol 2015; 22: 1200–1211.
52. Gleser MA. Effects of hypoxia and physical training on
hemodynamic adjustments to one-legged exercise. J Appl
Physiol 1973; 34: 655–659.
53. Calbet JA, Radegran G, Boushel R, et al. On the mech-
anisms that limit oxygen uptake during exercise in acute
and chronic hypoxia: Role of muscle mass. J Physiol
2009; 587: 477–490.
54. Daussin FN, Ponsot E, Dufour SP, et al. Improvement of
VO2max by cardiac output and oxygen extraction adapta-
tion during intermittent versus continuous endurance
training. Eur J Appl Physiol 2007; 101: 377–383.
55. Ng AV, Callister R, Johnson DG, et al. Endurance exer-
cise training is associated with elevated basal sympathetic
nerve activity in healthy older humans. J Appl Physiol
(1985) 1994; 77: 1366–1374.
56. Daussin FN, Zoll J, Dufour SP, et al. Effect of interval
versus continuous training on cardiorespiratory and
mitochondrial functions: Relationship to aerobic per-
formance improvements in sedentary subjects. Am J
Physiol Regul Integr Comp Physiol 2008; 295: R264–272.
57. Baker WL, White CM, Cappelleri JC, et al.
Understanding heterogeneity in meta-analysis: The role
of meta-regression. Int J Clin Pract 2009; 63: 1426–1434.
58. Ross RM. ATS/ACCP statement on cardiopulmonary
exercise testing. Am J Respir Crit Care Med 2003; 167:
1451; author reply 1451.
59. Warburton DE, Haykowsky MJ, Quinney HA, et al.
Reliability and validity of measures of cardiac output
during incremental to maximal aerobic exercise. Part I:
Conventional techniques. Sports Med 1999; 27: 23–41.