International Journal of Sports Physiology and Performance, 2016, 11, 707  -714

http://dx.doi.org/10.1123/ijspp.2016-0174
© 2016 Human Kinetics, Inc. BRIEF REVIEW

A Systematic Review of Submaximal Cycle Tests to Predict,

Monitor, and Optimize Cycling Performance
Benoit Capostagno, Michael I. Lambert, and Robert P. Lamberts

Finding the optimal balance between high training loads and recovery is a constant challenge for cyclists and their coaches.
Monitoring improvements in performance and levels of fatigue is recommended to correctly adjust training to ensure optimal
adaptation. However, many performance tests require a maximal or exhaustive effort, which reduces their real-world application.
The purpose of this review was to investigate the development and use of submaximal cycling tests that can be used to predict and
monitor cycling performance and training status. Twelve studies met the inclusion criteria, and 3 separate submaximal cycling.
tests were identified from within those 12. Submaximal variables including gross mechanical efficiency, oxygen uptake (VO2),
heart rate, lactate, predicted time to exhaustion (pTE), rating of perceived exertion (RPE), power output, and heart-rate recovery
(HRR) were the components of the 3 tests. pTE, submaximal power output, RPE, and HRR appear to have the most value for
monitoring improvements in performance and indicate a state of fatigue. This literature review shows that several submaximal
cycle tests have been developed over the last decade with the aim to predict, monitor, and optimize cycling performance. To be
able to conduct a submaximal test on a regular basis, the test needs to be short in duration and as noninvasive as possible. In
addition, a test should capture multiple variables and use multivariate analyses to interpret the submaximal outcomes correctly
and alter training prescription if needed.

Keywords: cyclist, monitoring, training status, training prescription, LSCT

Cyclists are often exposed to high training loads and, together to the introduction of new technologies such as valid and reliable
with their coaches, are constantly challenged to find the optimal cycle ergometers and heart-rate monitors, new submaximal tests
balance between training load and recovery.1 This challenge have been developed with the aim to predict, monitor, and fine-tune
is enhanced by the varied response to identical training loads training prescription in cyclists.5,9,12
between athletes.2 Individualized training prescription, based on a Elite cyclists have to sustain high training loads to stimulate
cyclist’s training status and levels of fatigue, is likely to improve adaptation, which may intentionally cause fatigue.13 However, if
adaptation to training.3,4 Therefore, it is important for athletes and this training-induced fatigue is excessive, cyclists could require
coaches to regularly monitor fatigue levels and training status, to an extensive period of reduced training load and recovery,14 which
promote the best response to training. Performance tests such as would negatively affect their training program and race schedule.
the peak-power-output (PPO) test and time trials are able to detect This emphasizes the importance of regular monitoring of the rela-
meaningful changes in training status in cyclists.5,6 However, these tionship between training load and symptoms of fatigue to ensure
tests require a maximal effort, can be disruptive to training and that the cyclists are managed appropriately. Therefore, the aim of
racing programs,1,7 and are therefore only performed 2 or 3 times this brief review was to provide a systematic overview of submax-
a year,7,8 which reduces their application for regular monitoring of imal cycling tests that are currently used to predict and monitor
training programs. cycling performance.
A nonexhaustive, submaximal, cycling-specific test that can
reliably predict cycling performance and/or fatigue would certainly
be an appealing option.9 The use of submaximal cycle tests dates Methods
back to the 1950s, when Åstrand and Ryhming10 developed . the
Åstrand test, which predicts maximal oxygen uptake (VO2max) for Data Sources
a submaximal performance level. Another test that has been around A comprehensive literature review was performed by using 2 bibli-
for a long time (since the 1970s) is the Physical Workload Capacity ographic databases, Medline (PubMed) and Scopus. The following
test,11 which aims
. to determine power output at 170 beats/min and combination of keywords was used for the search: (“cycling” OR
from which VO2max can also be predicted. More recently and due “cyclist” OR “cyclists” OR cycle test) AND (“submaximal” OR
“sub-maximal” OR “sub maximal”) AND (“athletes” OR “cyclists”
OR “trained” OR “well-trained” OR “elite”). These search terms
returned 440 articles, which were reduced to 102 articles after the
The authors are with the Div for Exercise Science and Sports Medicine, removal of duplicates and consideration of the article titles. The
University of Cape Town, Newlands, South Africa. Lamberts is also with the abstracts of the remaining 102 articles were reviewed independently
Div of Orthopaedic Surgery and the Inst of Sports and Exercise Medicine, by 2 of the authors (B.C. and R.L.) to ensure that the inclusion
Stellenbosch University, South Africa. Address author correspondence to criteria were met. Seventy-four articles were excluded from the
Robert Lamberts at [email protected]. study after this step. The remaining 28 articles were downloaded

707
 708  Capostagno, Lambert, and Lamberts

and reviewed for all inclusion criteria. The 12 articles that remained Results
after that step were included as part of this systematic review.
The descriptive parameters of the original articles that were included
Study Selection: Inclusion and Exclusion Criteria for this review are presented in Table 1. The participants in the
studies included both male and female athletes, all of whom were
All relevant articles published before 2016 and written in the English trained to elite level. The submaximal protocols measured a variety
language were considered for this review. All searches and reading of physiological and performance
. variables including gross mechan-
of abstracts and full text articles were done independently by 2 of ical efficiency, submaximal VO2 and heart rate, rating of perceived
the authors. After each step (see also Figure 1), they authors would exertion (RPE) and power output at submaximal heart rates, and
discuss their findings to reach a consensus on which articles would lactate responses to constant submaximal workloads.
be excluded or included. Four of the 12 studies that met the inclusion criteria for this
The following inclusion criteria were used to select the articles review investigated either the reliability or the predictive validity of
for the review: the submaximal protocol used, while the other 8 studies investigated
• The research participants were cyclists or triathletes trained to longitudinal data on the changes in these variables over time. Two
elite level (performance level ≥2).15,16 of these longitudinal studies included training interventions, and
1 study used a submaximal test to monitor the effect of 2 surgical
• The mode of exercise during the test was cycling. interventions.
• The test was no longer than 60 minutes. Table 2 provides a brief overview of the study design, main
Downloaded by Bibliotheque De L'Univ Laval on 07/27/17, Volume 11, Article Number 6

• The test was truly submaximal and did not require a maximal parameters measured in the submaximal test, and the main findings
effort or prolonged periods of high-intensity cycling. of the reliability and predictive value of the studies. All 4 studies
included in Table 2 investigated the relationship between submax-
• The test aimed to predict performance and/or monitor changes imal measures of performance and more traditional measures of
in training status. performance.
The 12 articles that met all the inclusion criteria were read Table 3 provides a brief overview of the study design, main
and their reference lists checked for articles the search strategy parameters measured in the submaximal test, and the main findings
may have missed. of the longitudinal studies included in this review. Most of the

Figure 1 — Flowchart showing search strategy.

IJSPP Vol. 11, No. 6, 2016

 Submaximal Monitoring to Optimize Performance  709

Table 1 General Overview of the Studies Included in the Review

.
Relative VO2max,
N mL · min–1 · kg–1
Study (gender) Study type (PPO) Submaximal-test design
Reliability and Predictive Value
Lamberts et al1 15 (M) Reliability of a novel submaximal test 57.0 (382 W) 6 min @ 60%, 6 min @ 80%, 3 min @
90% HRmax + 60 rec. (17 min total)
Lamberts19 82 (M), Validity of predicting PPO and 40-km TT M: 57.5 (385 W); 6 min @ 60%, 6 min @ 80%, 3 min @
20 (F) from a submaximal cycling test F: 50.5 (288 W) 90% HRmax + 60 rec. (17 min total)
Lamberts and 20 (F) Validity of predicting PPO and 40-km TT 50.5 (288 W) 6 min @ 60%, 6 min @ 80%, 3 min @
Davidowitz20 from a submaximal cycling test 90% HRmax + 60 rec. (17 min total)
Sassi et al9 8 (M) Reliability of predicted time to exhaustion 64.9 (375 W) 20-min warm-up + 12 min of submaximal
from submaximal lactate cycling at a constant workload (60–85% of
PPO) (32 min total)
Longitudinal Performance and Fatigue
Monitoring
Downloaded by Bibliotheque De L'Univ Laval on 07/27/17, Volume 11, Article Number 6

Barbeau et al12 7 (M) Longitudinal monitoring of submaximal 73.7 10-min warm-up + 2-min stages @ 150,
and maximal test data 200, 250 W and cadence 50–120 rpm (58
min total)
Capostagno et al4 15 (M) High-intensity interval training guided by 58.4 (389 W), 6 min @ 60%, 6 min @ 80%, 3 min @
data from a submaximal cycling test 53.9 (378 W) 90% HRmax + 60 rec. (17 min total)
Hammes et al13 23 (M) Monitoring the response of submaximal 59.4 (338 W) 6 min @ 60%, 6 min @ 80%, 3 min @
cycling test to intensified training 90% HRmax + 60 rec. (17 min total)
Kunorozva et al25 20 (M) Monitoring the effect of time of day on 51.0 (354 W) 6 min @ 60%, 6 min @ 80%, 3 min @
submaximal cycling test 90% HRmax + 60 rec. (17 min in total)
Lamberts et al6 1 (M) Monitoring changes in submaximal and 71.9 (460 W) 6 min @ 60%; 6 min @ 80%; 3 min @
maximal cycling test 90% HRmax + 60 rec. (17 min total)
Lamberts et al24 1 (M) Monitoring the effect of 2 surgical inter- 71.9 (460 W) 6 min @ 60%; 6 min @ 80%; 3 min @
ventions to improve leg blood flow 90% HRmax + 60 rec. (17 min total)
Otter et al26 20 (F) Monitoring the relationship between sub- 50.3 (301 W) 6 min @ 60%; 6 min @ 80%; 3 min @
maximal cycling test and perceived stress, 90% HRmax + 60 rec. (17 min total)
rec., and cycling performance
Sassi et al9 12 (M) Monitoring the changes in predicted time 74.6 (433 W) 20-min warm-up + 12 min of submaximal
to exhaustion during a competitive season cycling at a constant workload (60–85% of
PPO) (32 min total)
Sassi et al21 8 (M) Monitoring seasonal changes in submaxi- 76.7 20-min warm-up + 12 min of submaximal
mal and maximal cycling tests cycling at a constant workload (60–85% of
PPO) (32 min total)
.
Abbreviations: VO2max, maximal oxygen uptake; PPO, peak power output; 40-km TT, simulated 40-km time trial; HRmax, maximal heart rate; rec, recovery.

studies monitored changes in performance, except for 3, where no Studies on Reliability and Predictive Value
performance measure was included in the study. In 1 of those studies,
the relationship between the changes in submaximal variables and Sassi et al9 investigated the relationship between blood lactate during
perceived stress and recovery was measured. submaximal constant workloads and endurance cycling performance
(time to exhaustion). They specifically chose blood lactate (La–) because
it is a popular physiological measure of exercise intensity17 and is rel-
Discussion atively easy to measure. The cyclists exercised at a randomly selected
constant workload of 60% to 85% of their PPO while maintaining a
The purpose of this review was to investigate the development and cadence of 94 to 98 rpm. Capillary blood samples were collected from
use of submaximal cycling tests that are currently used to predict the ear lobe at the end of the fifth ([La–]t5) and tenth ([La–]t10) minutes.
and monitor cycling performance. Despite acknowledgment of the The test was stopped after the [La–]t10 sample was collected. A predicted
need for a submaximal test that can be performed regularly, without time to exhaustion (pTE) was then calculated from a regression equa-
disruption of cyclists’ training and racing programs,1,4,9,13 very few tion that included [La–]t5, [La–]t10, and the rate of change of [La–]slope.
of these types of tests have been developed.1,9,12 In addition, only 2 The coefficient of variation (CV) of pTE is 11.7%, which could lead
of the 3 tests identified in this review have had their reliability and to a high error in the estimation of endurance performance. However,
validity investigated.1,9 time-to-exhaustion efforts may have an even higher CV (~25%).18 In

IJSPP Vol. 11, No. 6, 2016

 710  Capostagno, Lambert, and Lamberts

Table 2 Outcome Variables and Main Findings of the Reliability and Predictive Value Studies Included in This
Review
Study Groups Study design Duration Parameters Main findings
Lamberts et al1 1 Reliability of submaximal variables in a sub- 5 wk PO80, PO90; RPE80, PO ICC (.91–1.00); RPE
maximal cycling test RPE90; HRR ICC (.85–.88); HRR ICC
(.99)
Validity of predicting endurance performance 5 wk PO80, PO90; RPE80, PO PPO r = .88–0.94; PO
from submaximal variables in a submaximal RPE90; HRR 40-km TT r = .84–.92
cycling test
Lamberts and Dav- 1 Validity of predicting endurance performance 1 wk PO80, PO90; RPE80, PO PPO r = .88–.94; PO
idowitz20 from submaximal variables in a submaximal RPE90; HRR 40-km TT r = .84–.92
cycling test
Lamberts19 1 Validity of predicting endurance performance 1 wk PO80, PO90; RPE80, PO PPO r = .88–.94; PO
from submaximal variables in a submaximal RPE90; HRR 40-km TT r = .84–.92
cycling test
Sassi et al9 1 Predicting performance from variables in a ~10 d PO; Submaximal [La] @ [La–]t10 PO r = .87
Downloaded by Bibliotheque De L'Univ Laval on 07/27/17, Volume 11, Article Number 6

submaximal cycling test 10 minutes ([La]t10)

Abbreviations: PO, power output; PO80, average PO at 80% of maximal heart rate; PO90, average PO at 90% of maximal heart rate; RPE, rating of perceived exertion; RPE80,
RPE at 80% of maximal heart rate; RPE90, RPE at 90% of maximal heart rate; HRR, heart-rate recovery; [La–], blood lactate concentration; [La–]t10, [La–] at the tenth minute.

addition,. while more traditional variables of endurance performance, and female cyclists. The strong correlations between actual and
such as VO2max, may have lower CVs associated with their measure- predicted measures of performance support the use of the LSCT
ment (CV < 3%), these variables are relatively stable in professional as a submaximal protocol to monitor performance in both male
cyclists and require a maximal effort to be determined.9 and female cyclists.
The Lamberts and Lambert Submaximal Cycle Test (LSCT) is a
novel submaximal cycling protocol designed to monitor and predict Studies on Longitudinal Performance and Fatigue
cycling performance.1 The LSCT consists of 3 stages of increasing
Monitoring
intensity in which cyclists are asked to ride at predetermined fixed
percentages of their maximal heart rate (HRmax).1,5,19 Stage 1 is 6 Barbeau et al12 monitored 7 competitive male cyclists during
minutes at an intensity of 60% of HRmax, stage 2 is 6 minutes at an a competitive season using a submaximal cycling test. The
intensity of 80% of HRmax, and stage 3 is 3 minutes at an intensity cyclists were tested during the preparation period (November),
of 90% of HRmax.1 RPEs are recorded in the final 30 seconds of each the specific preparation period (February), and the competition
stage, while heart rate, power output, and cadence are continuously period (May and July). The submaximal test consisted of three
recorded throughout the test. On completion of the third stage, the 16-minute stages of increasing workloads of 150, 200, and 250
cyclists are instructed to stop pedaling, sit upright, and not talk while W. Each stage was further divided into eight 2-minute periods
a 60-second heart-rate recovery (HRR) is recorded. at a set cadence. The initial target cadence was 50 rpm, and this
The reliability and validity of the LSCT have been confirmed was increased by 10 rpm every 2 minutes until the . final stage at
by Lamberts et al.1 Fifteen well-trained male cyclists completed 3 120 rpm. The submaximal variables of heart rate, VO2, mechanical
PPO tests and 3 simulated 40-km time trials (TTs) over a 5-week efficiency, and ventilation were measured during the test. Including
period. The LSCT was performed as a standardized warm-up proce- the 10-minute warm-up at 100 W, the total duration of the submax-
dure before both the PPO and the 40-km TT. Interclass correlation imal test is 58 minutes. .
coefficients (ICCs) and 95% confidence intervals were used to The cyclists in this study showed no changes in either VO2max
determine the reliability of the variables of the LSCT. In addition, or ventilatory thresholds during the entire training and competition
the typical error of measurement (TEM) was determined for each period. Heart rate decreased at 200 and 250 W between February
variable. Statistical analysis showed that the LSCT was highly and May, which the authors interpreted as a positive adaptation
. to
reliable, with ICCs ranging from .85 to 1.00 and TEMs ranging training. This was further supported by a decrease in VO2 at 200
from 0.8% to 4.4% during stages 2 and 3 of the LSCT. In addition, W in May and 250 W in July. Mechanical efficiency increased at
there
. were good correlations between the LSCT and PPO (r = .94), 150 W from February to July and at 250 W from February until the
VO2max (r = .91), and 40-km-TT time (r = .92). A similar study design end of. the season. The results of this study support the hypothesis
was completed on 20 trained to well-trained female cyclists,20 the that VO2max and other variables from maximal-effort tests may have
results of which showed strong relationships between LSCT-based
. limited use in monitoring elite cyclists. However, the reliability
predicted values and actual values for PPO (r = .98), VO2max (r = of this test is not known, and this greatly reduces the value of this
.98), and 40-km TT (r = .98). assessment.
Lamberts19 expanded on these results by examining the rela- Sassi et al9 investigated the variations in pTE during a compet-
tionships in a larger cohort of male (n = 82) and female (n = 20) itive season in 12 professional male cyclists. An protocol identical
cyclists. There were similar relationships found between LSCT- . to that of the study described above was used, except the constant
based predicted values and actual values for PPO (r = .98), VO2max workload was 5.5 W/kg. Four tests were used to monitor seasonal
(r = .96), and 40-km TT (r = .98). .In addition, there were no gender changes in pTE: December (basic training period), February (begin-
differences for 40-km TT and VO2max, which indicates that the ning of competitive period), and the best and worst tests performed
same multivariate regression equations can be used for both male during the competitive season (March to October). There were

IJSPP Vol. 11, No. 6, 2016

 Submaximal Monitoring to Optimize Performance  711

Table 3 Changes in Submaximal and Performance Parameters in the Longitudinal Studies Included in the Review
Changes in Changes
submaximal in performance
Study Groups Study design Duration parameters parameter
.
Barbeau 1 Longitudinal monitoring of submaximal- and 1 season MEG ↓ VO2max ↔
et al12 maximal-test data
HR ↓
.
VO2 ↓
Capostagno Structured High-intensity-training intervention 4 wk No change 40-km TT ↔
et al4
Flexible 4 wk PO80 ↑ (tendency) 40-km TT ↑
HRR ↑
Hammes 1 Intensified-training intervention 6d Day 8: PO80 ↑, PO90 ↑ Day 8:
et al13 (d 1–8)
RPE80 ↑, RPE90 ↑ HRmax ↓
HRR ↑ 40-km TT ↓
Downloaded by Bibliotheque De L'Univ Laval on 07/27/17, Volume 11, Article Number 6

Reduced-training-load intervention 3d Day 11: PO80 ↓ Day 11: PPO ↑

(d 8–11)
RPE80 ↓, RPE90 ↓ HRmax ↑
40-km TT ↑
Kunorozva 1 Monitoring the effect of time of day on submax- 2 wk RPE80 ↑, RPE90 ↑
et al25 imal-cycling-test data
Lamberts 1 Monitoring changes in submaximal- and maxi- 10 wk PO80 ↑, PO90 ↑ PPO ↑
et al6 mal-cycling-test data
HRR ↑ 40-km TT ↑
Otter 1 Monitoring the relationship between submax- 12 mo PO80 ↑ Sport-specific recovery ↑
et al26 imal-cycling-test data and perceived stress,
PO90 ↑ General stress ↓
recovery, and cycling performance
HRR ↑ Sport-specific stress ↓
Sassi et al9 1 Monitoring the changes in predicted time to 10 mo [La–] ↓ pTE ↑
exhaustion during a competitive season
.
Sassi et al21 1 Monitoring seasonal changes in submaximal- 10 mo [La–] ↓ VO2max ↑
and maximal-cycling-test data
pTE ↑ PPO ↑
PO at RCP ↑
.
Lamberts 1 Longitudinal monitoring of submaximal and 11 mo PO80↓, PO90 ↓a PPO ↓,VO2max ↓
et al24 performance variables
RPE80, RPE90 ↔a HRmax ↔
HRR↓a 40-km TT ↓
.
Abbreviations: MEG, gross mechanical efficiency; HR, heart rate; TT, time trial; VO2max, maximal oxygen uptake; PO80, average power output at 80% of maximal HR
(HRmax); PO90, average power output at 90% of HRmax; RPE80, rating of perceived exertion at 80% of HRmax; RPE90, rating of perceived exertion at 90% of HRmax; HRR,
HR recovery; PPO, peak power output; pTE, predicted time to exhaustion; RCP, respiratory-compensation point.
a Not published in paper, but actual data were analyzed.

significant differences in pTE between the December test and 2 over time, .with TCOMP being significantly higher than TREST and
tests from the competitive period (February and best). In addition, TPRECOMP. VO2max, PPO, and power output at the respiratory-com-
there were differences between the best and worst tests during the pensation point were all significantly higher at TCOMP than at TREST,
competitive period, which could indicate relatively large fluctuations suggesting that a positive training adaptation had occurred. These
in performance during a competitive season. The variations the pTE results support the use of pTE as a submaximal measurement of
ranged from 103% (between the worst competition-period test and endurance performance.
the basic-training-period test) to 224% (between the basic period Lamberts et al6 monitored a world-class cyclocross cyclist
and the best tests during the competitive season). during a 10-week preseason period that included some sudden
Sassi
. et al 21 also investigated the seasonal changes in increases in training load. A PPO and 40-km TT were performed
pTE, VO2max, and respiratory-compensation point in a group of 8 at the beginning and end of the 10-week monitoring period. The
national-level cyclists. Three testing sessions took place throughout LSCT was performed as a standardized warm-up procedure before
the season: TREST (beginning of December), TPRECOMP (3 months later each of the performance tests and on a weekly basis during the
after the precompetitive training), and TCOMP (a further 3 months monitoring period. The cyclist’s training status improved during
later during the competitive period). Determination of pTE was the 10-week monitoring period, as demonstrated by increases in
performed as described previously,9 but the target workload was PPO and 40-km-TT performance. The power output during the
70% to 80% of PPO or 4.5 to 5.5 W/kg. pTE significantly increased third stage (90% of HRmax) of the LSCT increased in a magnitude

IJSPP Vol. 11, No. 6, 2016

 712  Capostagno, Lambert, and Lamberts

similar to that of the increases in PPO and average power output criteria were based on the TEM of these variables of the LSCT as
during the 40-km TT. This supports the previous findings from this previously determined.5
group, which showed that average power output during the third The cyclists in the flexible-training group showed a more
stage (90% of HRmax) of the LSCT is strongly correlated to both uniform response to the HIT training than the cyclists in the
PPO and 40-km-TT performance.1 This indicates that power output structured-training group, with the majority of the cyclists in the
during the third stage of the LSCT tracks changes in performance flexible-training group improving their 40-km-TT performance by
in an elite cyclist. Changes in HRR were also shown to have strong more than the TEM. The results of this study support the use of a
correlations with changes in power output during the second (r = submaximal cycling test to monitor fatigue levels and titrate training
.91) and third (r = .98) stages. to promote adaptation to training.3 However, not all the cyclists in
Large increases in training load were detected by some variables the flexible-training group responded favorably to the HIT train-
of the LSCT.6 Sudden increases in the cyclist’s training load resulted ing, which suggests that other variables in the LSCT may provide
in a sharp increase in power output during the first minute of the further information on fatigue levels in well-trained cyclists. Future
third stage. This indicates that the cyclist had to exercise at a much research should consider the role of submaximal power output and
higher workload to reach the target heart rate. This was confirmed its relationship with RPE when prescribing training sessions.
by an increase in RPE during the second and third stages. Note The LSCT may serve as a submaximal performance measure
that these changes were also accompanied by a faster HRR at the and monitoring tool to determine the effect of certain interventions.
end of the test. When viewed in isolation, a faster HRR could be Kunorozva et al25 used the LSCT to monitor the effect of time of
Downloaded by Bibliotheque De L'Univ Laval on 07/27/17, Volume 11, Article Number 6

interpreted as a positive response to training; however, numerous day on cycling performance. The cyclists performed the LSCT on
studies have shown that a faster HRR can also be associated with 5 separate occasions (6 AM, 10 AM, 2 PM, 6 PM, and 10 PM),
fatigue.22,23 These findings highlight the importance of including separated by 24 to 48 hours, in a randomized order. No differences
multiple variables when monitoring athletes.19,22 in submaximal power output and HRR were found between the
The cyclist who took part in the aforementioned study6 con- different times of the day. However, RPE scores were significantly
tinued to be monitored beyond the 10-week precompetition period. higher at 6 PM and 10 PM than at 6 AM and 10 AM. These findings
However, just 2 weeks into the competition period, he began to suggest that mainly subjective RPEs are influenced by time of the
experience painful and “powerless” legs.24 These symptoms became day. However, we would recommend that a submaximal cycle test
so severe that he was unable to train and compete at the level he had such as the LSCT be conducted at the same time of day to minimize
previously. The cyclist was ultimately diagnosed with kinking of the effect of circadian rhythms.
both the left and the right external iliac arteries, which required 2 Recently, Hammes et al13 investigated the effect of train-
separate surgical interventions separated by 3 months. Eight months ing-induced fatigue and recovery on the variables of the LSCT.
after the second surgery and 11 months after the second PPO and Twenty-three male cyclists and triathletes were exposed to a 6-day
40-km-TT tests from the previous study,6 the cyclist performed a intensive-training period, followed by a 65-hour recovery period
third PPO and 40-km-TT test. Once again, the LSCT was performed during which no training took place. The participants completed
as a standardized warm-up before. the PPO and 40-km TT. Actual an LSCT, 40-km TT, and an incremental test on the days preceding
and LSCT-predicted PPO, VO2max, and 40-km-TT performance the start of the training period, after the training period, and after
of the 3 sets of tests were compared with those of 82 apparently the recovery period. There was a marked decrease in performance
healthy male cyclists. . There were no differences between actual and (40-km TT and incremental test) after the intensified training period.
predicted PPO and VO2max before or after the corrective surgery. In The decreases in performance were tracked by changes in several
contrast, there were substantial differences between the actual and variables of the LSCT. Submaximal power output during stages 2
LSCT-predicted 40-km-TT times in the 2 tests performed before the and 3 increased immediately after the intensive-training period and
2 surgical interventions, while postsurgery the actual and predicted this was accompanied by an increase in RPE and HRR. Increased
40-km-TT time were similar. While the training status of the cyclist submaximal power output and HRR are generally associated with an
had certainly decreased after the surgical interventions, the results of improved training status.1,19,20 However, a case study by Lamberts et
this study indicate that the reduced blood flow caused by the kinking al6 and a paper by Hammes et al,13 show that increased submaximal
of the iliac arteries greatly impaired the cyclist’s endurance capacity. power output and faster HRR with increased RPE levels reflect a state
Although not specifically designed for this purpose, discrepancies of overreaching rather than an improved training status. This finding
between actual and LSCT-predicted cycling performance could be highlights that a multifactorial approach is needed when monitoring
an early indicator to support further investigation. performance and that you cannot rely on changes in just 1 parameter.
The application of the LSCT to prescribe training was inves- Otter et al26 monitored the relationship between perceived
tigated by Capostagno et al,4 where the timing of high-intensity stress and recovery and cycling performance in 20 competitive
interval-training (HIT) sessions was adjusted based on the results of female athletes. The athletes completed a RESTQ-Sport question-
the LSCT performed that day. Fifteen male cyclists were divided into naire and an LSCT 8 times during the observation period, with
2 groups: The structured-training group followed a predetermined measurements separated by 6 weeks. High levels of sport-specific
training program, and the flexible-training group had the timing of recovery and self-efficacy were associated with an increased
their HIT sessions adjusted based on their responses to the LSCT. average power output in stage 2 of the LSCT. In contrast, high
Multiple variables of the LSCT were considered before a decision general stress was associated with a decreased average power
on whether to complete the session was made: an increased RPE output during stage 3 of the LSCT. An improved balance of
during the second stage compared with the previous test, failure to recovery stress was associated with an increased average power
reach the target heart rate within the first minute of the third stage, output during stage 3. Improved recovery stress and decreased
and a change in HRR >2 beats/min compared with the previous test. emotional and sport-specific stresses were associated with an
These variables were selected because they are available almost increased HRR. These results highlight the association between
immediately on completion of the LSCT. The cyclists’ session the objective measures of the LSCT and subjective measures of
was postponed if 2 or more of these criteria were met. These 3 stress and recovery (RESTQ-Sport).

IJSPP Vol. 11, No. 6, 2016

 Submaximal Monitoring to Optimize Performance  713

In conclusion, submaximal cycle tests are able to predict and Based on the conducted systematic review, we believe that
monitor changes in training status and cycling performance with rea- submaximal cycle tests play an important practical role in predict-
sonable accuracy in trained to elite cyclists. pTE from submaximal ing, monitoring, and fine-tuning training prescription in trained to
La–, submaximal power output, RPE, and HRR appear to have the elite cyclists.
most value for monitoring improvements in performance and fatigue
levels. Multiple variables captured in the LSCT and multivariable
analyses, such as described by Lamberts et al19 and Otter et al,26 References
are likely the best and most sensitive way to predict, monitor, and
optimize training prescription in cyclists. Although a few initial 1. Lamberts RP, Swart J, Noakes TD, Lambert MI. A novel sub-
studies have started using submaximal cycle tests to monitor6,13 and maximal cycle test to monitor fatigue and predict cycling perfor-
prescribe4 training, more research is needed in this area. mance. Br J Sports Med. 2011;45:797–804. PubMed doi:10.1136/
bjsm.2009.061325
2. Bouchard C, Rankinen T. Individual differences in response to regular
Practical Application
physical activity. Med Sci Sports Exerc. 2001;33:S446–S451. PubMed
To increase compliance, a submaximal test designed to accurately doi:10.1097/00005768-200106001-00013
predict and monitor cycling performance should be simple and 3. O’Connor PJ. Monitoring and titrating symptoms—a science-based
practical to administer. If a test has a long duration, requires costly approach to using your brain to optimise marathon running perfor-
Downloaded by Bibliotheque De L'Univ Laval on 07/27/17, Volume 11, Article Number 6

equipment, or requires the assistance of others to do the test, it is mance. Sports Med. 2007;37:408–411. PubMed
less likely to be performed on a regular basis. 4. Capostagno B, Lambert MI, Lamberts RP. Standardized versus cus-
The duration of the different submaximal tests described in this tomized high-intensity training: effects on cycling performance. Int
systematic review ranged from 17 to 58 minutes.1,12,19,21 A test 58 J Sports Physiol Perform. 2014;9:292–301. PubMed doi:10.1123/
minutes in duration may be too long to be administered on a regular IJSPP.2012-0389
basis, and this reduces the practical application of the assessment. 5. Lamberts RP, Swart J, Woolrich RW, Noakes TD, Lambert MI. Meas-
The frequency of submaximal testing ranged from as little as urement error associated with performance testing in well-trained
4 times during a competitive cycling season12 to multiple times per cyclists: application to the precision of monitoring changes in training
week.1,4,5 The more frequently a test can be completed, without status. Int Sportmed J. 2009;10:33–44.
interfering with training or racing, the more data it can provide. The 6. Lamberts RP, Rietjens GJ, Tijdink HH, Noakes TD, Lambert MI.
noninvasive nature of the LSCT, which does not require capillary Measuring submaximal performance parameters to monitor fatigue
blood samples to determine lactate concentrations9,21 or costly and predict cycling performance: a case study of a world-class cyc-
oxygen analyzers to monitor respiratory gases for the determina- lo-cross cyclist. Eur J Appl Physiol. 2010;108:183–190. PubMed
tion of cycling efficiency,12 makes it a practical test that can be doi:10.1007/s00421-009-1291-3
performed more frequently. In addition, the LSCT can be used as 7. Coutts AJ, Slattery KM, Wallace LK. Practical tests for monitoring
a standardized warm-up protocol before a performance test1,5 and/ performance, fatigue and recovery in triathletes. J Sci Med Sport.
or a training session.1,4 2007;10:372–381. PubMed doi:10.1016/j.jsams.2007.02.007
The correct interpretation of submaximal-test data is a key 8. Lucía A, Hoyos J, Perez M, Chicharro JL. Heart rate and performance
factor to accurately monitor cycling performance. Three important parameters in elite cyclists: a longitudinal study. Med Sci Sports Exerc.
factors to consider are to use the actual submaximal variables rather 2000;32:1777–1782. PubMed doi:10.1097/00005768-200010000-00018
than predicted performance, use multiple submaximal variables and 9. Sassi A, Marcora SM, Rampinini E, Mognoni P, Impellizzeri FM.
if possible multivariate analyses to interpret your submaximal meas- Prediction of time to exhaustion from blood lactate response during
urements, and interpret your findings within an individual rather submaximal exercise in competitive cyclists. Eur J Appl Physiol.
than between athletes. Assessing individual changes in submaximal 2006;97:174–180. PubMed doi:10.1007/s00421-006-0157-1
parameters is important, as some outcome variables are influenced 10. Åstrand PO, Rhyming I. A nomogram for calculation of aerobic
by individual perceptions (eg, RPE) and/or are partly influenced by capacity (physical fitness) from pulse rate during sub-maximal work.
genetic factors (eg, HRR).27 J Appl Physiol. 1954;7:218–221. PubMed
It is critical to know the error of measurement and the smallest 11. Haber P, Schlick W, Schmid P, Mulac K. Estimation of the performance
worthwhile change of any test used to monitor performance. This spectrum of healthy adolescents by using the PWC 170 (Physical Work
will ensure the accurate interpretation of the data and will allow Capacity). Acta Med Austriaca. 1976;3:164–166. PubMed
a coach to make the correct decisions with regard to adapting the 12. Barbeau P, Serresse O, Boulay MR. Using maximal and sub-
training load or not. The use of multiple variables when monitoring maximal aerobic variables to monitor elite cyclists during a
cycling performance and fatigue provides more insight into an ath- season. Med Sci Sports Exerc. 1993;25:1062–1069. PubMed
lete’s training status than the information from a single variable. An doi:10.1249/00005768-199309000-00015
initial case report by Lamberts et al6 and 2 subsequent studies13,22 13. Hammes D, Skorski S, Schwindling S, et al. Can the Lamberts and
have shown that acute fatigue and/or functional overreaching is Lambert Submaximal Cycle Test (LSCT) indicate fatigue and recovery
associated with higher submaximal power outputs at fixed heart- in trained cyclists? Int J Sports Physiol Perform. 2016;11:328–336.
rate intensities (or lower submaximal heart rates at fixed power PubMed doi:10.1123/ijspp.2015-0119
intensities), increased RPE levels, and a faster HRR. When consid- 14. Meeusen R, Duclos M, Foster C, et al. Prevention, diagnosis, and
ering only “point a” and “point c,” one might incorrectly interpret treatment of the overtraining syndrome: joint consensus statement of
that the athlete is responding well to a training stimulus, while if the European College of Sport Science and the American College of
“point b” is also taken into consideration, it becomes clear that the Sports Medicine. Med Sci Sports Exerc. 2013;45:186–205. PubMed
athlete is experiencing a state of acute fatigue and/or functional doi:10.1249/MSS.0b013e318279a10a
overreaching.1,6,14 This example highlights the importance of a 15. De Pauw K, Roelands B, Cheung SS, de Geus B, Rietjens G, Meeusen
multivariate-analysis approach. R. Guidelines to classify subject groups in sport-science research. Int

IJSPP Vol. 11, No. 6, 2016

 714  Capostagno, Lambert, and Lamberts

J Sports Physiol Perform. 2013;8(2):111–122. PubMed http://dx.doi. ultra-marathon. J Strength Cond Res. 2015;29:3343–3352. PubMed
org/10.1123/ijspp.8.2.111 doi:10.1519/JSC.0000000000001004
16. Decroix L, De Pauw K, Foster C, Meeusen R. Guidelines to classify 23. Aubry A, Hausswirth C, Louis J, Coutts AJ, Buchheit M, Le Meur
female subject groups in sport-science research. Int J Sports Physiol Y. The development of functional overreaching is associated with
Perform. 2016;11:204–213. PubMed doi:10.1123/ijspp.2015-0153 a faster heart rate recovery in endurance athletes. PLoS One.
17. Weltman A, Seip RL, Snead D, et al. Exercise training at and above 2015;10:e0139754. PubMed doi:10.1371/journal.pone.0139754
the lactate threshold in previously untrained women. Int J Sports Med. 24. Lamberts RP, Mann TN, Rietjens GJ, Tijdink HH. Impairment of
1992;13:257–263. PubMed doi:10.1055/s-2007-1021263 40-km time-trial performance but not peak power output with external
18. Jeukendrup A, Saris WH, Brouns F, Kester AD. A new validated iliac kinking: a case study in a world-class cyclist. Int J Sports Physiol
endurance performance test. Med Sci Sports Exerc. 1996;28:266–270. Perform. 2014;9:720–722. PubMed doi:10.1123/IJSPP.2013-0040b
PubMed doi:10.1097/00005768-199602000-00017 25. Kunorozva L, Roden LC, Rae DE. Perception of effort in morn-
19. Lamberts RP. Predicting cycling performance in trained to elite male ing-type cyclists is lower when exercising in the morning. J Sports
and female cyclists. Int J Sports Physiol Perform. 2014;9(4):610–614. Sci. 2014;32:917–925. PubMed doi:10.1080/02640414.2013.873139
PubMed doi:10.1123/IJSPP.2013-0040a 26. Otter RT, Brink MS, van der Does HT, Lemmink KA. Monitoring
20. Lamberts RP, Davidowitz KJ. Allometric scaling and predicting perceived stress and recovery in relation to cycling performance in
cycling performance in (well-) trained female cyclists. Int J Sports female athletes. Int J Sports Med. 2016;37:12–18. PubMed
Med. 2014;35:217–222. PubMed 27. Hautala AJ, Rankinen T, Kiviniemi AM, et al. Heart rate recovery
Downloaded by Bibliotheque De L'Univ Laval on 07/27/17, Volume 11, Article Number 6

21. Sassi A, Impellizzeri FM, Morelli A, Menaspa P, Rampinini E. Sea- after maximal exercise is associated with acetylcholine receptor M2
sonal changes in aerobic fitness indices in elite cyclists. Appl Physiol (CHRM2) gene polymorphism. Am J Physiol Heart Circ Physiol.
Nutr Metab. 2008;33:735–742. PubMed doi:10.1139/H08-046 2006;291:H459–H466. PubMed doi:10.1152/ajpheart.01193.2005
22. Mann TN, Platt C, Lamberts RP, Lambert MI. Faster heart rate
recovery with increased RPE: paradoxical responses after an 87 km

IJSPP Vol. 11, No. 6, 2016