Training Characteristics and Power Profile of Professional U23 Cyclists throughout a Competitive Season

Abstract

Background:
The purpose of this study was to investigate differences in the power profile derived from training and racing, the training characteristics across a competitive season and the relationships between training and power profile in U23 professional cyclists.

Methods:
Thirty male U23 professional cyclists (age, 20.0 ± 1.0 years; weight, 68.9 ± 6.9 kg; V˙O2max, 73.7 ± 2.5 mL·kg-1·min-1) participated in this study. The cycling season was split into pre-, early-, mid- and late-season periods. Power data 2, 5, 12 min mean maximum power (MMP), critical power (CP) and training characteristics (Hours, Total Work, eTRIMP, Work·h-1, eTRIMP·h-1, Time<VT1, TimeVT1-2 and Time>VT2) were recorded for each period. Power profiles derived exclusively from either training or racing data and training characteristics were compared between periods. The relationships between the changes in training characteristics and changes in the power profile were also investigated.

Results:
The absolute and relative power profiles were higher during racing than training at all periods (p ≤ 0.001-0.020). Training characteristics were significantly different between periods, with the lowest values in pre-season followed by late-season (p ≤ 0.001-0.040). Changes in the power profile between early- and mid-season significantly correlated with the changes in training characteristics (p < 0.05, r = -0.59 to 0.45).

Conclusion:
These findings reveal that a higher power profile was recorded during racing than training. In addition, training characteristics were lowest in pre-season followed by late-season. Changes in training characteristics correlated with changes in the power profile in early- and mid-season, but not in late-season. Practitioners should consider the influence of racing on the derived power profile and adequately balance training programs throughout a competitive season.

Full text

sports
Article
Training Characteristics and Power Profile of
Professional U23 Cyclists throughout a
Competitive Season
Peter Leo 1, *, James Spragg 2, Dieter Simon 3, Justin S. Lawley 1and Iñigo Mujika 4,5
1Department Sport Science, University Innsbruck, 6020 Innsbruck, Austria; justin.lawley@uibk.ac.at
2Spragg Cycle Coaching, Exeter 03833, UK; james@spraggcyclecoaching.com
3
Training and Exercise Sciences, University of Applied Sciences Wiener Neustadt, 2700 Wiener Neustadt, Austria;
dieter.simon@fhwn.ac.at
4Department of Physiology, Faculty of Medicine and Nursing, University of the Basque Country,
48940 Leioa, Spain; inigo.mujika@inigomujika.com
5Exercise Science Laboratory, School of Kinesiology, Faculty of Medicine, Universidad Finis Terrae,
Santiago 8320000, Chile
*Correspondence: peter.leo@student.uibk.ac.at
Received: 11 November 2020; Accepted: 15 December 2020; Published: 17 December 2020


Abstract:
Background: The purpose of this study was to investigate differences in the power profile
derived from training and racing, the training characteristics across a competitive season and the
relationships between training and power profile in U23 professional cyclists.
Methods: Thirty male
U23 professional cyclists (age, 20.0
±
1.0 years; weight, 68.9
±
6.9 kg;
.
V
O
2max
, 73.7
±
2.5 mL
·
kg
−1·
min
−1
)
participated in this study. The cycling season was split into pre-, early-, mid- and late-season
periods. Power data 2, 5, 12 min mean maximum power (MMP), critical power (CP) and training
characteristics (Hours, Total Work, eTRIMP, Work
·
h
−1
, eTRIMP
·
h
−1
, Time
<VT1
, Time
VT1-2
and
Time
>VT2
) were recorded for each period. Power profiles derived exclusively from either training or
racing data and training characteristics were compared between periods. The relationships between
the changes in training characteristics and changes in the power profile were also investigated.
Results: The absolute and relative power profiles were higher during racing than training at all
periods (p
≤
0.001–0.020). Training characteristics were significantly different between periods,
with the lowest values in pre-season followed by late-season (p
≤
0.001–0.040). Changes in the
power profile between early- and mid-season significantly correlated with the changes in training
characteristics (p<0.05, r =
−
0.59 to 0.45). Conclusion: These findings reveal that a higher power
profile was recorded during racing than training. In addition, training characteristics were lowest
in pre-season followed by late-season. Changes in training characteristics correlated with changes
in the power profile in early- and mid-season, but not in late-season. Practitioners should consider
the influence of racing on the derived power profile and adequately balance training programs
throughout a competitive season.
Keywords: cycling; racing; volume; intensity; periodization; performance
1. Introduction
The power profile [
1
–
3
] and training characteristics [
4
–
9
] of professional cyclists have been well
documented in previous research. Depending on the rider type (sprinter, time trialist, allrounder
or climber), professional cyclists require high absolute and relative power outputs over short (1 s to
5 min) to long durations (5 min to >4 h), at various levels of acute and chronic fatigue [
1
–
3
,
10
–
12
].
Differences in race topography (e.g., flat, semi- and high-mountainous stage profiles) [
1
,
13
] and race
Sports 2020,8, 167; doi:10.3390/sports8120167 www.mdpi.com/journal/sports

Sports 2020,8, 167 2 of 12
category [
14
] have been shown to influence the power profile derived from racing in professional
cyclists. Power profiling can be used to predict race performance, especially during mountainous
races, and can distinguish between different rider levels (e.g., U23 vs. professional) and rider
types (e.g., allrounders, domestiques, general classification contenders) [
12
]. However, while it has
been shown that the power profile can be derived from laboratory tests [
2
] and field tests, [
3
] it is
unclear if the power profile can be derived from training data alone. The primary goal of training
is to induce improvements in the power profile by prescribing work bouts based on the athlete’s
strength and weaknesses. Therefore, training should be manipulated accordingly to bring about such
improvements. Training and racing can be quantified through completed hours or distance, and using
external (
power output
, total work) and internal (heart rate, rating of perceived exertion) training load
measures [
5
,
6
,
15
]. Recent research demonstrated that training load can also be quantified using ratios
of volume and intensity [
16
]. Additionally, the organization of training can be described using training
intensity distribution (TID) [
17
]. Training volume and TID have been documented in U23 [
18
] and
professional cyclists [
5
,
8
,
9
,
19
], but the relationship between changes in training characteristics and
changes in the power profile has not yet been investigated.
Therefore, the first aim of this study was to compare the power profile between training and
racing to assess the differences of the power profile derived from training data. The second aim was to
assess the variation in the training characteristics across a competitive season and to investigate the
relationship between changes in training and changes in the power profile of professional U23 cyclists.
2. Materials & Methods
2.1. Participants
Thirty male U23 professional cyclists participated in this study (age, 20.0
±
1.0 years; height, 182.6
±
6.1 cm; weight, 68.9
±
6.9 kg;
.
V
O
2max
, 73.7
±
2.5 mL
·
kg
−1·
min
−1
). All participants were members of a
UCI Continental U23 development team during the cycling season(s) analyzed. Rider type classification
was as follows: allrounders (n=21) and climbers (n=9) [
20
]. Recruitment was based on voluntary
interest. In cases of any prolonged illness, injury (defined as no race days in a given period), or
termination of cycling career, participants were excluded from all analysis. If a cyclist was included in
the analysis over three consecutive years, he was treated as a separate participant for each year due to
varying body mass and performance capacity.
Informed written consent was obtained after each participant was given a verbal and written
explanation of the experimental protocol and fully understood the possible risks involved in taking
part in the study. The study protocol was approved (ID 382019) by the Ethical Review Board at the
University of Innsbruck and followed the principles as set out in the declaration of Helsinki.
2.2. Design
Power data files were collected from the participants during every training and racing session
of a competitive season for 3 consecutive years. Each season was split into 4 periods: pre-season
(November to January), early-season (February to April), mid-season (May to July) and late-season
(August to October).
Anthropometric data were collected in pre-season in conjunction with laboratory measures and
at a randomized point within all other periods. During pre-season, participants performed both a
laboratory-based graded incremental exercise test (GXT) and a critical power (CP) test consisting of 2,
5 and 12 min maximal efforts as per Leo et al. [11].
Throughout the season, 2, 5, and 12 min mean maximum power (MMP) data from each period
were identified from power output files to produce a power profile for each cyclist. CP and work above
CP (W0) estimates were derived from MMP values for training and racing.
Data were then processed in three ways: (1) differences in the absolute and relative power profile
metrics; 2, 5 and 12 min MMP, CP and W
0
values (absolute only) between training and racing for each

Sports 2020,8, 167 3 of 12
period; (2) differences in training characteristics between periods; (3) relationship between the changes
in the power profile metrics and the changes in training characteristics for each period.
2.3. Methodology
2.3.1. Laboratory Testing
Participants were asked to avoid any exhaustive exercise 24 h prior to the test. They were also
encouraged to appear in a fully rested, hydrated and fueled state.
Open circuit spiroergometry with a breath by breath technique (ZAN600, nSpire Health GmbH,
Oberthulba, Germany) was used. For the GXT, volume and flow were calibrated with a 1 L syringe
before each trial. Gas analyzer calibration was completed before each measurement according to the
manufacturer’s recommendations (4.9 Vol% CO
2
, 15.9 Vol% O
2
, 79.2 Vol% N
2
, nSpire Health GmbH,
Oberthulba, Germany). All participants continuously wore a facemask and breathed through a flow
sensor (FlowSensor Type II, nSpire Health GmbH, Oberthulba, Germany).
.
V
O
2max
was defined as
the highest 30 s rolling average achieved before volitional exhaustion. The first ventilatory threshold
(VT1) was defined as the point where the ventilation rate (VE) increased compared to
.
V
O
2
(VE/
.
V
O
2
).
The second ventilatory threshold (VT2) was defined as the onset of hyperventilation during the
GXT [
21
], with an increase in VE compared to the volume of carbon dioxide (
.
V
CO
2
) release, known as
the VE/
.
V
CO
2
ratio. Continuous recordings of heart rate (HR) were performed via short range telemetry
with a 1 Hz sampling rate (V800, Polar Electro Oy, Kempele, Finnland). The HR values corresponding
to VT1 and VT2 were also determined.
GXT was performed in a controlled environment (temperature, 19–22
◦
C) on the participant’s personal
road bike (Alto Prestige, KTM Fahrrad GmbH, Mattighofen, Austria) mounted on an electromagnetically
braked stationary trainer with a 1 Hz sampling rate (Cyclus2, RBM Elektronik-automation GmbH,
Leipzig, Germany). The following GXT protocol was applied: initial workload 150 W, increment
20 W
·
min
−1
. In the case of a last uncompleted stage, maximal power output (P
max
) was calculated as per
Kuipers et al. [22].
Pmax=PL +t
60×20(1)
Equation (1), P
max
=maximum power output, PL =final completed stage in watts, t=associated time
for the uncompleted work stage in seconds.
2.3.2. CP Test Protocol
The CP test was carried out in the field during pre-season within two weeks of the initial laboratory
visit on a standardized uphill climb with an average gradient of 5.5%, on two consecutive days, with an
ambient temperature of 15–20 ◦C.
The CP test consisted of 2, 5 and 12 min maximum efforts in randomized order. The 2 and 5 min
efforts were performed on the same day interspersed by 30 min active recovery in which athletes
were instructed to ride at a rating of perceived exertion (RPE) <2 out of 10. Prior to each effort the
participants were encouraged to produce the highest possible workload and asked to maintain a
cadence between 80 and 100 revolutions per minute (rev·min−1).
The inverse of time model, using a least sum of squares linear regression analysis, was used to
derive CP and W
0
. The intercept of the regression line with the y axis represented CP and the slope W
0
.
The following equation was applied:
P=W0×1
t+CP (2)
Equation (2), P =power output (W), t=duration of field test (s).

Sports 2020,8, 167 4 of 12
2.3.3. Power Output
Power output in the field was recorded using a standardized crank system (SRAM Red, Quarq,
Spearfish, SD, USA) with a 1 Hz sampling rate, monitored on a portable head unit device (Garmin Edge
520, Schaffhausen, Switzerland). A static calibration of the power meter was undertaken prior to
the laboratory visit in the pre-season according to Wooles et al. [
23
]. Participants were instructed to
perform a “zero-offset” before each training or racing session.
2.3.4. Training Characteristics
Total accumulated cycling duration in both training and racing (Hours) was recorded for each period.
External workload was quantified via Total Work (session duration (s) multiplied by the power output with
a 1 Hz sampling rate). Internal workload was quantified using a five-zone model as per Edwards’ training
impulse [
24
] (eTRIMP): zone 1, 50–59% HR
peak
; zone 2, 60–69% HR
peak
; zone 3, 70–79% HR
peak
; zone 4,
80–89% HR
peak
; and zone 5, 90–100% HR
peak
. HR
peak
was defined as the highest HR recorded during early-,
mid- or late-season period or during the GXT in pre-season. eTRIMP was then calculated by multiplying a
zone-specific weighting factor (zone 1 =1, zone 2 =2, zone 3 =3, zone 4 =4, zone 5 =5) by the total time
accumulated in that zone. Workload ratios for external (Total Work) and internal (eTRIMP) workloads were
divided by Hours for each period (Work
·
h
−1
and eTRIMP
·
h
−1
respectively). Total time accumulated at a HR
below that corresponding to VT1 (Time
<VT1
), between VT1 and VT2 (Time
VT1-2
) and above VT2 (Time
>VT2
)
were analyzed for each period. The number of race days was also recorded for each period.
2.3.5. Data Analysis
Absolute and relative training and racing 2, 5 and 12 min MMP values during every period were
identified using a cycling software platform (WKO5 Build 562, TrainingPeaks LLC, Boulder, CO, USA).
Each MMP output was manually checked for data spikes. The MMP values for each period were used
to estimate CP and W0for the relevant period for training and racing by applying the inverse of time
CP model, using a least sum of squares linear regression analysis.
Training characteristics including Hours, Total Work, eTRIMP, Work
·
h
−1
, eTRIMP
·
h
−1
, Time
<VT1
,
TimeVT1-2 and Time>VT2 were computed for each period.
Delta values (
∆
) for the power profile and the training characteristics were derived. For comparisons
between power profile, values between early- and pre-season CP test were used due to there being no
racing in pre-season from which to derive values.
2.3.6. Statistical Analyses
All values are expressed as mean
±
standard deviation and or mean difference (
∆
). Normal distribution
was tested using the Shapiro–Wilk test (p>0.05). Statistical significance was established at p
≤
0.05, two-tailed.
Differences in the power profile including absolute and relative MMP values, CP and W
0
parameter
estimates were compared between training and racing for each period using a one-way repeated
measure analysis of variance (ANOVA). Differences in body mass and training characteristics between
periods were also assessed using a repeated ANOVA. In the case of significance, a post-hoc Holm
correction was applied for both. The repeated measures ANOVA was also applied if the data were not
normally distributed, as was shown appropriate by Norman [25].
Pearson product correlation was used to investigate the relationship in changes between the
power profile and training characteristics.
All statistical analyses were completed using JASP statistics software (version 0.13.1 for Mac OS,
JASP Team, Amsterdam, The Netherlands). All graphs and figures were created using GraphPad
Prism (version 8.0.0 for Mac OS, GraphPad Software, San Diego, CA, USA).
3. Results
Descriptive data of the laboratory GXT and field CP tests are presented in Table 1.

Sports 2020,8, 167 5 of 12
Table 1.
Physiological characteristics of professional U23 cyclists from the GXT (graded incremental
exercise test) and CP (critical power) test (mean ±SD).
Variables Absolute Values Relative Values
Pmax 458 ±38 [W] 6.6 ±0.4 [W·kg−1]
.
VO2max 5076 ±424 [mL·min−1] 73.7 ±2.5 [mL·kg−1·min−1]
VT1 256 ±22 [W] 3.8 ±0.3 [W·kg−1]
VT2 367 ±38 [W] 5.3 ±0.5 [W·kg−1]
CP 382 ±33 [W] 5.5 ±0.4 [W·kg−1]
W017.8 ±3.6 [kJ] n/a
GXT—graded incremental exercise test, P
max
—maximum power output;
.
V
O
2max
—maximum oxygen uptake;
VT1—first ventilatory threshold; VT2—second ventilatory threshold; CP—critical power; W
0
—work above
critical power
3.1. Power Profile
Absolute 2, 5 and 12 min MMP were higher in racing compared to training for early- (
∆
=33 W,
p
≤
0.001;
∆
=23 W, p
≤
0.001;
∆
=23 W, p
≤
0.001), mid- (
∆
=23 W, p=0.005;
∆
=26 W, p
≤
0.001;
∆
=28 W,
p=0.020) and late-season (
∆
=31 W, p
≤
0.001;
∆
=27 W, p
≤
0.001;
∆
=27 W, p
≤
0.001). Absolute CP
was also higher in racing compared to training for early- (
∆
=21 W, p
≤
0.001), mid- (
∆
=29 W, p
≤
0.001)
and late-season (
∆
=27 W, p
≤
0.001) (Figure 1). No significant differences were found for W
0
between
racing and training for either early-, mid- or late-season (p≥0.05).
Sports 2020, 8, x FOR PEER REVIEW 5 of 12
Table 1. Physiological characteristics of professional U23 cyclists from the GXT (graded incremental
exercise test) and CP (critical power) test (mean ± SD).
Variables Absolute Values Relative Values
Pmax 458 ± 38 [W] 6.6 ± 0.4 [W⋅kg−1]
VO2max 5076 ± 424 [mL⋅min−1] 73.7 ± 2.5 [mL⋅kg−1⋅min−1]
VT1 256 ± 22 [W] 3.8 ± 0.3 [W⋅kg−1]
VT2 367 ± 38 [W] 5.3 ± 0.5 [W⋅kg−1]
CP 382 ± 33 [W] 5.5 ± 0.4 [W⋅kg−1]
W’ 17.8 ± 3.6 [kJ] n/a
GXT—graded incremental exercise test, Pmax—maximum power output; VO2max—maximum oxygen
uptake; VT1—first ventilatory threshold; VT2—second ventilatory threshold; CP—critical power;
W’—work above critical power
3.1. Power Profile
Absolute 2, 5 and 12 min MMP were higher in racing compared to training for early- (∆ = 33 W,
p ≤ 0.001; ∆ = 23 W, p ≤ 0.001; ∆ = 23 W, p ≤ 0.001), mid- (∆ = 23 W, p = 0.005; ∆ = 26 W, p ≤ 0.001; ∆ = 28
W, p = 0.020) and late-season (∆ = 31 W, p ≤ 0.001; ∆ = 27 W, p ≤ 0.001; ∆ = 27 W, p ≤ 0.001). Absolute
CP was also higher in racing compared to training for early- (∆ = 21 W, p ≤ 0.001), mid- (∆ = 29 W, p ≤
0.001) and late-season (∆ = 27 W, p ≤ 0.001) (Figure 1). No significant differences were found for W’
between racing and training for either early-, mid- or late-season (p ≥ 0.05).
Figure 1. Differences in the absolute power profile between training and racing across periods; (A)—
2-min power output, (B)—5-min power output, (C)—12-min power output, (D)—critical power;
Early-season
Mid-season
Late-season
300
400
500
600
700
2 min Power Output (W)
Training
Racing
  
Early-season
Mid-season
Late-season
250
350
450
550
12 min Power Output (W)
Training
Racing
  
Early-season
Mid-season
Late-season
250
350
450
550
5 min Power Output (W)
Training
Racing
  
Early-season
Mid-season
Late-season
250
300
350
400
450
500
Critical Power (W)
Training
Racing
  
AB
C D
Figure 1.
Differences in the absolute power profile between training and racing across periods;
(
A
)—2-min power output, (
B
)—5-min power output, (
C
)—12-min power output, (
D
)—critical power.

Sports 2020,8, 167 6 of 12
Relative 2, 5 and 12 min MMP were higher in racing compared to training for early- (
∆
=0.4 W
·
kg
−1
,
p
≤
0.001;
∆
=0.3 W
·
kg
−1
,p
≤
0.001;
∆
=0.3 W
·
kg
−1
,p
≤
0.001), mid- (
∆
=0.3 W
·
kg
−1
,p=0.005;
∆
=0.4 W
·
kg
−1
,p
≤
0.001;
∆
=0.4 W
·
kg
−1
,p
≤
0.001) and late-season (
∆
=0.4 W
·
kg
−1
,p
≤
0.001;
∆
=0.5 W
·
kg
−1
,p
≤
0.001;
∆
=0.4 W
·
kg
−1
,p
≤
0.001). Relative CP was also higher in racing compared
to training for early- (
∆
=0.3 W
·
kg
−1
,p
≤
0.001), mid- (
∆
=0.4 W
·
kg
−1
,p
≤
0.001) and late-season
(
∆
=0.4 W
·
kg
−1
,p
≤
0.001) (Figure 2). Body mass was the lowest in late-season compared to pre-
(∆=0.8 kg, p=0.031), early- (∆=1.1 kg, p≤0.001) and mid-season (∆=1.0 kg, p=0.003).
Sports 2020, 8, x FOR PEER REVIEW 6 of 12
Relative 2, 5 and 12 min MMP were higher in racing compared to training for early- (∆ = 0.4
W⋅kg−1, p ≤ 0.001; ∆ = 0.3 W⋅kg−1, p ≤ 0.001; ∆ = 0.3 W⋅kg−1, p ≤ 0.001), mid- (∆ = 0.3 W⋅kg−1, p = 0.005; ∆
= 0.4 W⋅kg−1, p ≤ 0.001; ∆ = 0.4 W⋅kg−1, p ≤ 0.001) and late-season (∆ = 0.4 W⋅kg−1, p ≤ 0.001; ∆ = 0.5
W⋅kg−1, p ≤ 0.001; ∆ = 0.4 W⋅kg−1, p ≤ 0.001). Relative CP was also higher in racing compared to training
for early- (∆ = 0.3 W⋅kg−1, p ≤ 0.001), mid- (∆ = 0.4 W⋅kg−1, p ≤ 0.001) and late-season (∆ = 0.4 W⋅kg−1, p
≤ 0.001) (Figure 2). Body mass was the lowest in late-season compared to pre- (∆ = 0.8 kg, p = 0.031),
early- (∆ = 1.1 kg, p ≤ 0.001) and mid-season (∆ = 1.0 kg, p = 0.003).
Figure 2. Differences in the relative power profile between training and racing across periods; (A)—2
min relative power output, (B)—5 min relative power output, (C)—12 min relative power output,
(D)—relative critical power.
3.2. Training Characteristics
Training characteristics are presented in Table 2.
Table 2. Training characteristics of professional U23 cyclists across periods (mean ± SD).
Variables Pre-Season Early-Season Mid-Season Late-Season
Hours (h) 167 ± 46 202 ± 28 * 219 ± 26 *,# 150 ± 36 *,#,†
Total Work (kJ) 90.507 ± 45.622 132.825 ± 36.738 * 147.983 ± 33.497 * 97.539 ± 35.832 #,†
Work⋅h−1(kJ⋅h−1) 529 ± 182 658 ± 143 * 674 ± 119 * 642 ± 142 *
eTRIMP (A.U.) 31.477 ± 9.543 37.356 ± 6.416 * 39.036 ± 8.007 *,# 25.325 ± 7.960 *,#
eTRIMP⋅h−1(A.U.⋅h−1) 192 ± 33 186 ± 28 * 178 ± 31 168 ± 33 *,#
Early-season
Mid-season
Late-season
5.0
6.0
7.0
8.0
9.0
10.0
2 min Power Output (W⋅kg
–1
)
Training
Racing
  
Early-season
Mid-season
Late-season
4.0
4.5
5.0
5.5
6.0
6.5
7.0
12 min Power Output (W⋅kg
–1
)
Training
Racing
  
Early-season
Mid-season
Late-season
4.0
5.0
6.0
7.0
8.0
5 min Power Output (W⋅kg
–1
)
Training
Racing
  
Early-season
Mid-season
Late-season
3.0
4.0
5.0
6.0
7.0
Critical Power (W⋅kg
–1
)
Training
Racing
  
AB
C D
Figure 2.
Differences in the relative power profile between training and racing across periods;
(
A
)—2 min relative power output, (
B
)—5 min relative power output, (
C
)—12 min relative power
output, (D)—relative critical power.
3.2. Training Characteristics
Training characteristics are presented in Table 2.
Table 2. Training characteristics of professional U23 cyclists across periods (mean ±SD).
Variables Pre-Season Early-Season Mid-Season Late-Season
Hours (h) 167 ±46 202 ±28 * 219 ±26 *,# 150 ±36 *,#,†
Total Work (kJ) 90.507 ±45.622
132.825
±
36.738 *
147.983 ±33.497 * 97.539 ±35.832 #,†
Work·h−1(kJ·h−1)529 ±182 658 ±143 * 674 ±119 * 642 ±142 *
eTRIMP (A.U.) 31.477 ±9.543 37.356 ±6.416 * 39.036 ±8.007 *,# 25.325 ±7.960 *,#
eTRIMP
·
h
−1
(A.U.
·
h
−1
)
192 ±33 186 ±28 * 178 ±31 168 ±33 *,#
Time<VT1 (h) 29.2 ±11.7 39.4 ±19.1 * 41.8 ±16.4 29.2 ±15.8 *,#,†

Sports 2020,8, 167 7 of 12
Table 2. Cont.
Variables Pre-Season Early-Season Mid-Season Late-Season
TimeVT1-2 (h) 104.5 ±36.0 124.7 ±21.0 * 125.4 ±27.0 * 76.2 ±29.4 *,#,†
Time>VT2 (h) 15.3 ±8.0 18.9 ±5.2 * 20.4 ±6.6 * 12.0 ±5.6 *
Race Days n/a13 ±5#20 ±614 ±7#
Hours—training hours, eTRIMP—Edwald’s training impulse; VT—ventilatory threshold, * significantly different
from pre-season; #significantly different from early-season, †significantly different from mid-season (p≤0.05).
Hours were lower in pre-season compared to early- (
∆
=34 h, p
≤
0.001) and mid-season (
∆
=52 h,
p
≤
0.001) but higher than in late-season (
∆
=17 h, p=0.027); Hours were higher in mid-season than
in early-season (
∆
=17 h, p=0.002); and Hours were lower in late-season than in early- (
∆
=51 h,
p≤0.001) and mid-season (∆=69 h, p≤0.001).
Total work was lower in pre-season compared to early- (
∆
=42,319 kJ, p=0.002) and mid-season
(
∆
=57,477, p
≤
0.001) and was lower in late-season than in early- (
∆
=35,286 kJ, p=0.002) and
mid-season (
∆
=50,444, p
≤
0.001). Work
·
h
−1
was lower in pre-season compared to early- (
∆
=129 kJ
·
h
−1
,
p=0.034), mid- (∆=145 kJ·h−1,p=0.034) and late-season (∆=112 kJ·h−1,p=0.015).
eTRIMP was lower in pre- compared to early- (
∆
=5878 arbitrary unit (A.U.), p=0.003) and mid-
(
∆
=7558 A.U., p
≤
0.001) but higher than in late-season (
∆
=6152 A.U., p=0.003), and was lower in
late- than in early- (
∆
=12,030 A.U., p
≤
0.001) and mid-season (
∆
=13,710 A.U., p
≤
0.001). eTRIMP
·
h
−1
was lower in late- compared to pre- (
∆
=23.6 A.U.
·
h
−1
,p=0.007) and early-season (
∆
=17.5 A.U.
·
h
−1
,
p=0.007).
Time
<VT1
was lower in pre- compared to early- (
∆
=10.1 h, p=0.002) and late-season (
∆
=12.6 h,
p
≤
0.001) and lower in late- compared to early- (
∆
=10.1 h, p=0.004) and mid-season (
∆
=12.6 h,
p≤0.001).
Time
VT1-2
was lower in pre- compared to early- (
∆
=20 h, p=0.005) and mid-season (
∆
=21 h,
p=0.011) but higher than in late-season (
∆
=28.3 h, p=0.002); it was lower in late- compared to early-
(∆=48.5 h, p≤0.001) and mid-season (∆=49.2 h, p≤0.001).
Time
>VT2
was lower in pre- compared to early- (
∆
=3.5 h, p=0.040) and mid-season (
∆
=5.1 h,
p=0.017) but higher than in late-season (
∆
=3.2 h, p=0.040); it was lower in late- compared to early-
(∆=6.8 h, p≤0.001) and mid-season (∆=8.4 h, p≤0.001).
The number of race days was higher in mid- compared to early- (
∆
=6, p
≤
0.001) and late-season
(∆=6, p≤0.001).
3.3. Relationship between Changes in Training Characteristics and Changes in Power Profile
The
∆
in 2 and 5 min MMP between early- and pre-season significantly correlated with
∆
Work
(r =
−
0.53, p=0.002; r =
−
0.59, p
≤
0.001, respectively, Figure 3A,B) and
∆
Work
·
h
−1
(r =
−
0.42, p=0.019,
Figure 3C). The
∆
in 12 min MMP and CP between early- and pre-season significantly correlated with
race days (r =−0.44, p=0.014; r =−0.40, p=0.027, Figure 3D,E).

Sports 2020,8, 167 8 of 12
Sports 2020, 8, x FOR PEER REVIEW 8 of 12
Figure 3. Relationship between the change in the power profile and training characteristics for the ∆
early- vs. pre-season. MMP—mean maximum power, CP—critical power; (A)—Work and 2 min
MMP, (B)—Work⋅h−1 and 5 min MMP, (C)—Work and 5 min MMP, (D)—Race Days and 12 min MMP,
(E)—Race Days and CP.
The ∆ in 2 and 12 min MMP between mid- and early-season significantly correlated with the
∆Time<VT1 (r = 0.41, p = 0.022; r = 0.42, p = 0.020—Figure 4A,B). The ∆ in CP between mid- and early-
season significantly correlated with ∆Time>VT2 (r = 0.45, p = 0.012, Figure 4C).
Figure 4. Relationship between the change in power profile and training characteristics for ∆ mid- vs.
early-season. MMP—mean maximum power, CP—critical power, Time<VT1—time below the first
ventilatory threshold, Time>VT2—time above the second ventilatory threshold; (A)—Time<VT1 and 2
min MMP, (B)—Time<VT1 and 12 min MMP, (C)—Time>VT2 and CP.
4. Discussion
This study aimed to investigate differences in the power profile between training and racing,
changes in training characteristics between different seasonal periods and correlations between these
variables. The absolute and relative power profile was higher in racing compared to training for
early-, mid- and late-season. Training characteristics were significantly different between periods and
changes in these training characteristics significantly influenced the changes seen in the power
profile.
In previous work, Leo et al. [11] found no differences in the absolute power profile of U23 cyclists
across a competitive season. Changes in the relative power profile were primarily due to varying
body mass. Those changes in relative MMP and CP values are accompanied by a reduction in body
mass the longer the season lasts due to accumulated race days and training volume. However, the
authors did not differentiate between training and racing efforts. In the present study, a higher
absolute (4.6–8.5%) and relative (4.2–8.4%) power profile was recorded during racing than training
–250 –200 –150 –100 –50
0
–300,000
–200,000
–100,000
100,000
200,000
Early-season Delta in 2min MMP (W)
Early-season Delta in Work (kJ)
r
=
p
=0.002
–0.536
0 50 100
–100 –50
0
200
400
600
–200
Early-season Delta in 5min MMP (W)
Early-season Delta inWork⋅h–1
p
=0.009
r
= –0.470
0 50 100
–100 –50
0
–300,000
–200,000
–100,000
100,000
200,000
Early-season Delta in 5min MMP (W)
Early-season Delta in Work (kJ)
p
=<0.001
r
= –0.598
ABC
02040
–60 –40 –20
0
5
10
15
20
25
Early-season Delta in 12min MMP (W)
Early-season Race Days
p
=0.014
r
= –0.445
02040
–60 –40 –20
0
5
10
15
20
25
Early-season Delta in CP (W)
Early-season Race Days
p
=0.027
r
= –0.404
D E
0100
–200 –100
0
20
40
–40
–20
Mid-season Delta in 2minMMP (W)
Mid-season Delta in Time
<VT1
(h)
p
=0.022
r
= 0.417
0 50 100
–50
0
20
40
–40
–20
Mid-season Delta in 12min MMP (W)
Mid-season Delta in Time
<VT1
(h)
p
=0.020
r
= 0.424
0 204060
–60 –40 –20
0
5
10
15
–15
–10
–5
Mid-season Delta in CP (W)
Mid-season Delta in Time
>VT2
(h)
p
=0.012
r
= 0.452
4 A 4 B 4 C
Figure 3.
Relationship between the change in the power profile and training characteristics for the
∆
early- vs. pre-season. MMP—mean maximum power, CP—critical power; (
A
)—Work and 2 min
MMP, (
B
)—Work
·
h
−1
and 5 min MMP, (
C
)—Work and 5 min MMP, (
D
)—Race Days and 12 min MMP,
(E)—Race Days and CP.
The
∆
in 2 and 12 min MMP between mid- and early-season significantly correlated with the
∆
Time
<VT1
(r =0.41, p=0.022; r =0.42, p=0.020—Figure 4A,B). The
∆
in CP between mid- and
early-season significantly correlated with ∆Time>VT2 (r =0.45, p=0.012, Figure 4C).
Sports 2020, 8, x FOR PEER REVIEW 8 of 12
Figure 3. Relationship between the change in the power profile and training characteristics for the ∆
early- vs. pre-season. MMP—mean maximum power, CP—critical power; (A)—Work and 2 min
MMP, (B)—Work⋅h−1 and 5 min MMP, (C)—Work and 5 min MMP, (D)—Race Days and 12 min MMP,
(E)—Race Days and CP.
The ∆ in 2 and 12 min MMP between mid- and early-season significantly correlated with the
∆Time<VT1 (r = 0.41, p = 0.022; r = 0.42, p = 0.020—Figure 4A,B). The ∆ in CP between mid- and early-
season significantly correlated with ∆Time>VT2 (r = 0.45, p = 0.012, Figure 4C).
Figure 4. Relationship between the change in power profile and training characteristics for ∆ mid- vs.
early-season. MMP—mean maximum power, CP—critical power, Time<VT1—time below the first
ventilatory threshold, Time>VT2—time above the second ventilatory threshold; (A)—Time<VT1 and 2
min MMP, (B)—Time<VT1 and 12 min MMP, (C)—Time>VT2 and CP.
4. Discussion
This study aimed to investigate differences in the power profile between training and racing,
changes in training characteristics between different seasonal periods and correlations between these
variables. The absolute and relative power profile was higher in racing compared to training for
early-, mid- and late-season. Training characteristics were significantly different between periods and
changes in these training characteristics significantly influenced the changes seen in the power
profile.
In previous work, Leo et al. [11] found no differences in the absolute power profile of U23 cyclists
across a competitive season. Changes in the relative power profile were primarily due to varying
body mass. Those changes in relative MMP and CP values are accompanied by a reduction in body
mass the longer the season lasts due to accumulated race days and training volume. However, the
authors did not differentiate between training and racing efforts. In the present study, a higher
absolute (4.6–8.5%) and relative (4.2–8.4%) power profile was recorded during racing than training
–250 –200 –150 –100 –50
0
–300,000
–200,000
–100,000
100,000
200,000
Early-season Delta in 2min MMP (W)
Early-season Delta in Work (kJ)
r
=
.002
p
=
-0.536
0 50 100
–100 –50
0
200
400
600
–200
Early-season Delta in 5min MMP (W)
Early-season Delta inWork⋅h–1
.009
p
=
r
= -0.470
0 50 100
–100 –50
0
–300,000
–200,000
–100,000
100,000
200,000
Early-season Delta in 5min MMP (W)
Early-season Delta in Work(kJ)
<.001
p
=
r
= -0.598
ABC
02040
–60 –40 –20
0
5
10
15
20
25
Early-season Delta in 12min MMP (W)
Early-season Race Days
.014
p
=
r
= -0.445
02040
–60 –40 –20
0
5
10
15
20
25
Early-season Delta in CP (W)
Early-season Race Days
.027
p
=
r
= -0.404
D E
0100
–200 –100
0
20
40
–40
–20
Mid-season Delta in 2min MMP (W)
Mid-season Delta in Time
<VT1
(h)
p
=0.022
r
= 0.417
0 50 100
–50
0
20
40
–40
–20
Mid-season Delta in 12min MMP (W)
Mid-season Delta in Time
<VT1
(h)
p
=0.020
r
= 0.424
0 204060
–60 –40 –20
0
5
10
15
–15
–10
–5
Mid-season Delta in CP (W)
Mid-season Delta in Time
>VT2
(h)
p
=0.012
r
= 0.452
ABC
Figure 4.
Relationship between the change in power profile and training characteristics for
∆
mid- vs.
early-season. MMP—mean maximum power, CP—critical power, Time
<VT1
—time below the first
ventilatory threshold, Time
>VT2
—time above the second ventilatory threshold; (
A
)—Time
<VT1
and
2 min MMP, (B)—Time<VT1 and 12 min MMP, (C)—Time>VT2 and CP.
4. Discussion
This study aimed to investigate differences in the power profile between training and racing,
changes in training characteristics between different seasonal periods and correlations between these
variables. The absolute and relative power profile was higher in racing compared to training for early-,
mid- and late-season. Training characteristics were significantly different between periods and changes
in these training characteristics significantly influenced the changes seen in the power profile.
In previous work, Leo et al. [
11
] found no differences in the absolute power profile of U23 cyclists
across a competitive season. Changes in the relative power profile were primarily due to varying body
mass. Those changes in relative MMP and CP values are accompanied by a reduction in body mass the
longer the season lasts due to accumulated race days and training volume. However, the authors did
not differentiate between training and racing efforts. In the present study, a higher absolute (4.6–8.5%)
and relative (4.2–8.4%) power profile was recorded during racing than training for all periods of the
season. These findings suggest that power outputs recorded in training alone are not reflective of a true

Sports 2020,8, 167 9 of 12
maximal power profile in U23 professional cyclists. Interval training sessions in cycling are commonly
prescribed by power output, heart rate or perception of effort, and are rarely prescribed as maximal
effort [
26
,
27
]. Recent research has shown that the accumulated time at or above 90% HR
max
during
interval training is sufficient to elicit cardiovascular adaptations such as increased cardiac output
and stroke volume. Therefore efforts do not need to be maximal in nature to induce adaptations [
27
].
Leo et al. [
11
] also reported that efforts in racing might determine the power profile, as they found good
agreement between CP derived from formal testing and field-derived MMP values. This confirmed
previous findings by Karsten et al. [
28
]. In contrast, Pinot and Grappe concluded that MMP values
derived from racing do not reflect a cyclist’s true maximum power profile [
3
,
29
]. A possible explanation
for this is that efforts in racing may be influenced by race scenarios and team tactics [
11
], glycogen
depletion [
30
] or model fitting [
31
]. Race topography and category, as well as short term fatigue in
single day racing or accumulated fatigue in multi-stage racing, have also been shown to influence
the power profile [
12
,
14
,
32
,
33
]. It has thus been recommended to verify field derived MMP values
with a minimum of two maximum effort field tests per season (i.e. CP tests) for baseline comparisons.
In summary, MMP and CP values derived from racing efforts in combination with a prior CP field test
may offer the best way to longitudinally monitor the power profile in elite cyclists.
Training characteristics were significantly different between periods. Indeed, both higher volume
and intensity were seen in the early- and mid-season compared to pre- and late-season. From pre- to
early-season, volume characteristics including Hours, Total Work and eTRIMP increased by 18.6 to
46.7%, while intensity metrics Work
·
h
−1
(+24.3%) and eTRIMP
·
h
−1
(
−
3.1%) diverged. In mid-season,
volume characteristics including Hours, Total Work and eTRIMP were higher by 4.5 to 11.4% compared
to early-season. Intensity metrics including Work
·
h
−1
(+7.9%) and eTRIMP
·
h
−1
(
−
4.4%) were again
divergent. From mid- to late-season, both volume (
−
4.7 to
−
35.1%) and intensity (
−
5.6 to
−
58.5%) metrics
were clearly declining. The number of race days was 53.8% higher during mid- compared to early-season,
while there were 30% fewer race days in late- compared to mid-season. A possible explanation for the
conflicting findings between volume and intensity metrics is that the participants’ heart rate responses
were lowered either due to the accumulated training load and subsequent accumulated fatigue or
improved cardiovascular adaptations [
34
,
35
]. The decline in training characteristics could be triggered
by the residual fatigue accumulated during the entire season through training and racing [
16
,
35
],
precluding athletes from maintaining or increasing training volume and or intensity. In contrast,
the power profile was maintained throughout a competitive season, with some improvements in
the relative power profile due to reduced body mass [
11
]. This may indicate that excessive fatigue
negatively influences training characteristics rather than the power profile, and therefore performance,
in the short term [36].
Research has shown that a polarized training intensity distribution may be the preferred training
strategy in elite endurance athletes [
37
]. Stöggl and Sperlich reported a training intensity distribution of
68
±
12% Time
<VT1
, 6
±
8% Time
VT1-2
and 26
±
7% Time
>VT2
in endurance athletes following a polarized
training approach [
17
]. However, in the present study, training intensity distribution did not follow a
polarized approach. The average distribution across the entire season was 17.4–19.4% in Time
<VT1
,
50.8–62.5% in Time
VT1–2
and 8.0–9.3% in Time
>VT2
and could thus be classified as a threshold training
intensity distribution. We hypothesized that the high percentage of Time
VT1–2
may have been due to
the high number of race days, as in races athletes cannot control the power requirement; however,
post hoc testing showed that there was no correlation between race days and Time
VT1-2
for any period,
nor did athletes with lower CP record more Time
VT1-2
. Therefore, the high percentage of total training
Time
VT1-2
may be due to this distribution being the coaches’ desired training distribution, or it was
due to poor intensity zone discipline with athletes executing low intensity sessions too hard [
38
,
39
].
Another possibility is that the training zones used by coaches were not anchored, or were inaccurately
anchored, to physiological thresholds.
The relationship between changes in the power profile and training characteristics from pre- to
early-season revealed that if the riders increased training load or race days too much, a decrease in

Sports 2020,8, 167 10 of 12
the power profile occurred, as Total Work, Work
·
h
−1
and race days negatively correlated with the
power profile. This was evidenced by the previously discussed divergence between external and
internal intensity metrics where Work
·
h
−1
increased by +24.3% whereas eTRIMP
·
h
−1
fell by 3.1%.
From pre- to early-season there was a larger increase in training Time
VT1-VT2
compared with Time
<VT1
.
Previous work has shown that VT1 may represent a threshold intensity in relation to the level of fatigue
induced in the autonomic nervous system [40].
Therefore, a practical recommendation may be that as racing is introduced in early-season,
total work should not be further increased; to achieve this, a reduction in the intensity of the overall
volume may be beneficial. This approach would also induce a shift towards a polarized training
intensity distribution. This is evidenced in the relationship between the change in the power profile and
training characteristics from early- to mid-season, where training Time
<VT1
and Time
>VT2
positively
correlated with the changes in the power profile. Essentially, a shift towards a polarized training
intensity distribution positively influenced the power profile, which has been shown to have a positive
impact on race performance [
12
]. Interestingly, no relationship between the change in the power profile
and training characteristics was found for mid- to late-season. As training load clearly decreased
during late-season, it may not provide any predictive ability to monitor the power profile. It may be
that in late-season, riders enter a maintenance phase whereby the minimum effective training dose [
36
]
is used to maintain the power profile [11].
The authors are aware that the current study is not without limitations. Power output data could
be influenced by stochastic and non-stochastic pacing patterns due to training or racing in mass start
events as well as team strategies or race tactics [
41
]. Using fixed duration for MMP values could
over- or underestimate the maximum capable power output, as it could be part of a longer effort.
MMP values could be recorded right at the beginning, middle or end of a period, but as shown in
previous research [
11
,
42
], the power profile during a competitive season remains relatively consistent.
Regarding training documentation, only cycling-specific workouts were quantified, as they represented
the majority of training, However the cyclists could have still completed additional cross training
activities (XC-skiing, ski touring, running, strength and conditioning) during pre-season which could
not be assessed in this study due to different training devices and training log documentation.
In conclusion, the current study found a higher absolute and relative power profile during racing
compared to training across a competitive season. Training characteristics in volume and intensity
increased from pre- to early- until mid-season, while in late-season a reduction in training volume
could be seen. Changes in training characteristics are predictive for changes in the power profile for pre-
until mid-season. Interestingly, although training characteristics, i.e. volume, decline in late-season,
the riders could maintain their power profile without seeing a reduction compared to previous periods.
Author Contributions:
Conceptualization, P.L., J.S., D.S., J.S.L. and I.M.; methodology, P.L. and J.S.; software, P.L.;
validation, P.L., J.S., D.S., J.S.L. and I.M.; formal analysis, P.L. and J.S.; investigation, P.L.; validation, P.L., J.S., D.S.
and I.M.; resources, J.S.L.; data curation, P.L. and J.S.; writing—original draft preparation, P.L.; writing—review and
editing, J.S., D.S., J.S.L. and I.M.; visualization, P.L.; supervision, J.S.L. and I.M.; project administration, I.M.
All authors have read and agreed to the published version of the manuscript.
Funding: No funding was provided for this project.
Acknowledgments:
We would like to thank the participants for dedicating their time and effort towards this
study. We would also like to thank Tirol KTM Cycling Team and Medalp Group for their support in performing
this project.
Conflicts of Interest: The authors report no conflict of interest.
References
1.
Sanders, D.; Heijboer, M. Physical demands and power profile of different stage types within a cycling grand
tour. Eur. J. Sport Sci. 2019,19, 736–744. [CrossRef] [PubMed]
2.
Quod, M.J.; Martin, D.T.; Martin, J.C.; Laursen, P.B. The power profile predicts road cycling MMP. Int. J.
Sports Med. 2010,31, 397–401. [CrossRef] [PubMed]

Sports 2020,8, 167 11 of 12
3.
Pinot, J.; Grappe, F. The record power profile to assess performance in elite cyclists. Int. J. Sports Med.
2011
,
32, 839–844. [CrossRef] [PubMed]
4.
van Erp, T.; Hoozemans, M.; Foster, C.; de Koning, J.J. The Influence of Exercise Intensity on the Association
between Kilojoules Spent and Various Training Loads in Professional Cycling. Int. J. Sports Physiol. Perform.
2019,14, 1395–1400. [CrossRef] [PubMed]
5.
Sanders, D.; van Erp, T.; de Koning, J.J. Intensity and load characteristics of professional road cycling:
Differences between men’s and women’s races. Int. J. Sports Physiol. Perform. 2019,14, 296–302. [CrossRef]
6.
Passfield, L.; Hopker, J.G.; Jobson, S.; Friel, D.; Zabala, M. Knowledge is power: Issues of measuring training
and performance in cycling. J. Sports Sci. 2017,35, 1426–1434. [CrossRef]
7.
van Erp, T.; Foster, C.; de Koning, J.J. Relationship between various training-load measures in elite cyclists
during training, road races, and time trials. Int. J. Sports Physiol. Perform. 2019,14, 493–500. [CrossRef]
8.
van Erp, T.; Sanders, D.; de Koning, J.J. Training characteristics of male and female professional road cyclists:
A 4-year retrospective analysis. Int. J. Sports Physiol. Perform. 2019,15, 534–540. [CrossRef]
9.
Metcalfe, A.J.; Menaspa, P.; Villerius, V.; Quod, M.; Peiffer, J.J.; Govus, A.D.; Abbiss, C.R. Within-season
distribution of external training and racing workload in professional male road cyclists. Int. J. Sports
Physiol. Perform. 2017,12, S2142–S2146. [CrossRef]
10.
Wahl, P.; Schütt, S.; Volmary, P. Power Profiling als leistungsdiagnostisches Toolim Radsport–Identifizierung
leistungsrelevanter physiologischer Zubringergrößen. In BISp Jahrbuch Forschungsförderung 2016/17;
SPORTVERLAG Strauß: Hellenthal, Germany, 2018; pp. 63–69.
11.
Leo, P.; Spragg, J.; Mujika, I.; Menz, V.; Lawley, J.S. Power profiling in U23 professional cyclists during a
competitive season. Int. J. Sports Physiol. Perform. 2020, in press.
12.
Leo, P.; Spragg, J.; Mujika, I.; Giorgi, A.; Lorang, D.; Simon, D.; Lawley, J.S. Power profiling, workload
characteristics and race performance of U23 and professional cyclists during the multistage race Tour of the
Alps. Int. J. Sports Physiol. Perform. 2020, in press.
13.
Padilla, S.; Mujika, I.; Orbananos, J.; Santisteban, J.; Angulo, F.; Jose Goiriena, J. Exercise intensity and
load during mass-start stage races in professional road cycling. Med. Sci. Sports Exerc.
2001
,33, 796–802.
[CrossRef] [PubMed]
14.
van Erp, T.; Sanders, D. Demands of Professional Cycling Races: Influence of Race Category and Result.
Eur J. Sport Sci. 2020, ahead of print. [CrossRef] [PubMed]
15.
Saw, A.; Halson, S.; Mujika, I. Monitoring athletes during training camps: Observations and translatable
strategies from elite road cyclists and swimmers. Sports 2018,6, 63. [CrossRef] [PubMed]
16.
Sanders, D.; Heijboer, M.; Hesselink, M.K.C.; Myers, T.; Akubat, I. Analysing a cycling grand tour: Can we
monitor fatigue with intensity or load ratios? J. Sports Sci. 2018,36, 1385–1391. [CrossRef]
17.
Stoggl, T.L.; Sperlich, B. The training intensity distribution among well-trained and elite endurance athletes.
Front. Physiol. 2015,6, 295. [CrossRef]
18.
Zapico, A.; Calderon, F.; Benito, P.; Gonzalez, C.; Paris, A.; Pigozzp, F.; Salvo, V.D. Evolution of physiological
and haematological parameters with training load in elite male road cyclists: A longitudinal study. J. Sports
Med. Phys. 2007,47, 191–196.
19.
Seiler, S. What is best practice for training intensity and duration distribution in endurance athletes?
Int. J. Sports Physiol. Perform. 2010,5, 276–291. [CrossRef]
20.
Giorgi, A.; Vicini, M.; Pollastri, L.; Lombardi, E.; Magni, E.; Andreazzoli, A.; Orsini, M.; Bonifazi, M.;
Lukaski, H.; Gatterer, H. Bioimpedance patterns and bioelectrical impedance vector analysis (BIVA) of road
cyclists. J. Sports Sci. 2018,36, 2608–2613. [CrossRef]
21.
Wasserman, K.; Hansen, J.E.; Sue, D.Y.; Stringer, W.W.; Whipp, B.J. Principles of exercise testing and
interpretation: Including pathophysiology and clinical applications. Med. Sci. Sports Exerc. 2005,37, 1249.
22.
Kuipers, H.; Verstappen, F.T.; Keizer, H.A.; Geurten, P.; van Kranenburg, G. Variability of aerobic performance
in the laboratory and its physiologic correlates. Int. J. Sports Med. 1985,6, 197–201. [CrossRef] [PubMed]
23.
Wooles, A.; Robinson, A.; Keen, P. A static method for obtaining a calibration factor for SRM bicycle power
cranks. Sports Eng. 2005,8, 137–144. [CrossRef]
24. Edwards, S. The Heart Rate Monitor Book; Polar Electro Oy: New York, NY, USA, 1993.
25.
Norman, G. Likert scales, levels of measurement and the “laws” of statistics.
Adv. Health Sci. Educ. Theory Pract.
2010,15, 625–632. [CrossRef]

Sports 2020,8, 167 12 of 12
26.
Laursen, P.; Buchheit, M. Science and Application of High-Intensity Interval Training: Solutions to the Programming
Puzzle; Human Kinetics: Champaign, IL, USA, 2019.
27.
Seiler, S.; Sylta, Ø. How does interval-training prescription affect physiological and perceptual responses?
Int. J. Sports Physiol. Perform. 2017,12, S2–S80. [CrossRef] [PubMed]
28.
Karsten, B.; Jobson, S.A.; Hopker, J.; Jimenez, A.; Beedie, C. High agreement between laboratory and field
estimates of critical power in cycling. Int. J. Sports Med. 2014,35, 298–303. [CrossRef] [PubMed]
29.
Pinot, J.; Grappe, F. A six-year monitoring case study of a top-10 cycling grand tour finisher. J. Sports Sci.
2015,33, 907–914. [CrossRef] [PubMed]
30.
Miura, A.; Sato, H.; Sato, H.; Whipp, B.J.; Fukuba, Y. The effect of glycogen depletion on the curvature
constant parameter of the power-duration curve for cycle ergometry. Ergonomics
2000
,43, 133–141. [CrossRef]
31.
Hill, D.; Smith, J. A method to ensure the accuracy of estimates of anaerobic capacity derived using the
critical power concept. J. Sports Med. Phys. Fit. 1994,34, 23–37.
32.
van Erp, T.; Hoozemans, M.; Foster, C.; de Koning, J.J. Case report: Load, intensity, and performance
characteristics in multiple grand tours. Med. Sci. Sports Exerc. 2020,52, 868–875. [CrossRef]
33.
Rodr
í
guez-Marroyo, J.A.; Villa, J.G.; Pern
í
a, R.; Foster, C. Decrement in professional cyclists’ performance
after a grand tour. Int. J. Sports Physiol. Perform. 2017,12, 1348–1355. [CrossRef]
34.
Plews, D.; Laursen, P.; Kilding, A.; Buchheit, M. Evaluating Training Adaptation with Heart-Rate Measures:
A Methodological Comparison. Int. J. Sports Physiol. Perform. 2013,8, 688–691. [CrossRef] [PubMed]
35.
Achten, J.; Jeukendrup, A.E. Heart rate monitoring: Applications and limitations. Sports Med.
2003
,33,
517–538. [CrossRef] [PubMed]
36.
Hickson, R.C.; Rosenkoetter, M.A. Reduced training frequencies and maintenance of increased aerobic power.
Med. Sci. Sports Exerc. 1981,13, 13–16. [CrossRef] [PubMed]
37.
Rosenblat, M.A.; Perrotta, A.S.; Vicenzino, B. Polarized vs. Threshold Training Intensity Distribution on
Endurance Sport Performance: A Systematic Review and Meta-Analysis of Randomized Controlled Trials.
J. Strength Cond. Res. 2019,33, 3491–3500. [CrossRef]
38.
Scantlebury, S.; Till, K.; Sawczuk, T.; Weakley, J.; Jones, B. Understanding the relationship between coach and
athlete perceptions of training intensity in youth sport. J. Strength Cond. Res.
2018
,32, 3239–3245. [CrossRef]
39.
Brink, M.S.; Frencken, W.G.; Jordet, G.; Lemmink, K.A. Coaches’ and players’ perceptions of training dose:
Not a perfect match. Int. J. Sports Physiol. Perform. 2014,9, 497–502. [CrossRef]
40.
Seiler, S.; Haugen, O.; Kuffel, E. Autonomic Recovery after Exercise in Trained Athletes.
Med. Sci. Sports Exerc.
2007,39, 1366–1373. [CrossRef]
41.
Ouvrard, T.; Groslambert, A.; Ravier, G.; Grospretre, S.; Gimenez, P.; Grappe, F. Mechanisms of performance
improvements due to a leading teammate during uphill cycling. Int. J. Sports Physiol. Perform.
2018
,13,
1215–1222. [CrossRef]
42.
Spiering, B.; Mujika, I.; Sharp, S.; Foulis, A. Maintaining physical performance: The minimum dose of
exercise needed to preserve endurance and strength over time. J. Strength Cond. Res. 2020, in press.
Publisher’s Note:
MDPI stays neutral with regard to jurisdictional claims in published maps and institutional
affiliations.
©
2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).