Exercise intensity during competition time

trials in professional road cycling

SABINO PADILLA, IÑIGO MUJIKA, JAVIER ORBAÑANOS, and FRANCISCO ANGULO
Departamento de Investigación y Desarrollo, Servicios Médicos, Athletic Club de Bilbao, Basque Country, SPAIN;
MEDIPLAN SPORT S.L., Vitoria—Gasteiz, Basque Country, SPAIN; and Departamento de Alto Rendimiento, Instituto
Vasco de Educación Fı́sica (IVEF—SHEE), Vitoria—Gasteiz, Basque Country, SPAIN

ABSTRACT
PADILLA, S., I. MUJIKA, J. ORBAÑANOS, and F. ANGULO. Exercise intensity during competition time trials in professional road
cycling. Med. Sci. Sports Exerc., Vol. 32, No. 4, pp. 850 – 856, 2000. Purpose: To estimate, upon competition heart rate (HR), exercise
intensity during time trials (TT) in professional road cycling. Methods: Eighteen world-class cyclists completed an incremental
laboratory cycling test to assess maximal power output (Wmax), maximal HR (HRmax), onset of blood lactate accumulation (OBLA),
lactate threshold (LT), and a HR-power output relationship. An OBLAZONE (HROBLA ⫾ 3 beats䡠min⫺1) and a LTZONE (HRLT ⫾ 3
beats䡠min⫺1) were described. HR was monitored during 12 prologue (⬍10 km, PTT), 18 short (⬍40 km, STT), 19 long (⬎40 km, LTT),
eight uphill (UTT), and seven team (TTT) time trials. A HR-power output relationship was computed to estimate each cyclist’s power
output during TT racing from competition HR. Competition training impulse (TRIMP) values were estimated from HR and race
duration. Results: %HRmax were 89 ⫾ 3%, 85 ⫾ 5%, 80 ⫾ 5%, 78 ⫾ 3%, and 82 ⫾ 2% in PTT, STT, LTT, UTT, and TTT,
respectively. The amount of TRIMP were, respectively, 21 ⫾ 3, 77 ⫾ 23, 122 ⫾ 27, 129 ⫾ 14, and 146 ⫾ 6. Competition HR values
relative to HROBLA and HRLT were, respectively, 100 ⫾ 3%, 114 ⫾ 8% in PTT, 95 ⫾ 7%, 108 ⫾ 9% in STT, 89 ⫾ 5%, 103 ⫾ 8%
in LTT, 87 ⫾ 2%, 101 ⫾ 5% in UTT, and 91 ⫾ 4%, 105 ⫾ 11% in TTT. Conclusions: %HRmax, TRIMP and time distribution around
HROBLA and HRLT reflected the physiological demands of different TT categories. HROBLA and HRLT were accurate intensity markers
in events lasting, respectively, ⱕ30 (PTT and STT) and ⱖ30 min (LTT, UTT, TTT). Key Words: HEART RATE, OBLA, LACTATE
THRESHOLD

I
n 3-wk professional road cycling stage races, perfor-
mance in the time trials (TT) is of paramount importance
to the final overall standings of the race. This racing
format, in which the cyclist most often races individually
and attempts to achieve the shortest possible time to cover
a fixed distance, has often been simulated under laboratory
conditions (4,5,8,13,19,20,22), usually with the aim of pre-
dicting cycling performance in the field. However, there are
very few reports on the intensity of actual road cycling
competition in general, and the TT in particular (23). This
lack of data is mainly due to the technical difficulties of
determining oxygen uptake (V̇O2) and blood lactate con-
centration ([La]) during competition, which are two of the
main methods used by exercise scientists to quantify exer-
cise intensity (14). In this past decade, the advent of accurate
portable telemetric heart rate (HR) monitors (18) has made
it possible to estimate exercise intensity both during training
and competition by relating individual HR values measured
in the field with those previously obtained in a laboratory
setting (9,10,23).
0195-9131/00/3203-0850/0
MEDICINE & SCIENCE IN SPORTS & EXERCISE®
Copyright © 2000 by the American College of Sports Medicine
Submitted for publication February 1999.
Accepted for publication June 1999.
850
In contrast with most laboratory simulations, TT cycling
competition represents a unique experience during which
the athlete’s power output (i.e., exercise intensity) is freely
chosen. Although several methods based on HR values have
been described to quantify the load undertaken by athletes
during training (1,16), the principle of training specificity
with regard to the intensity at which a cyclist trains cannot
be met unless the intensity and physiological demands of
competition are determined. The aim of the present study
was therefore to estimate, using competition HR data, ex-
ercise intensity during the TT in professional cycling, and to
compare the cyclists’ physiological responses to the differ-
ent competition TT categories and formats.
METHODS
Subjects. After receiving verbal and written explana-
tion of the purpose, procedures, and potential risks of the
present study, 18 international level professional road cy-
clists gave their written informed consent to participate in
this investigation. All subjects were members of the same
cycling team and had finished at least one of the main
professional 3-wk stage races (i.e., Tour de France, Giro
d’Italia, or Vuelta a España). The main physical character-
istics of the participating subjects are presented in Table 1.
Values shown in the table were computed from each cy-
clist’s best laboratory test (i.e., that in which the highest
TABLE 1. Subject characteristics and laboratory measurements (N ⫽ 18). Values
were computed from each cyclist’s best laboratory test (i.e., that in which the high-
est Wmax was attained).
Variables Mean ⴞ SD
Age (yr) 26 ⫾ 3
Height (cm) 180 ⫾ 7
Mass (kg) 68.8 ⫾ 7.0
Wmax (W) 439 ⫾ 45
Wmax䡠BM⫺1 (W䡠kg⫺1) 6.4 ⫾ 0.3
WOBLA (W) 386 ⫾ 52
WLT (W) 334 ⫾ 50
HRmax (beats䡠min⫺1) 194 ⫾ 5
HROBLA (beats䡠min⫺1) 178 ⫾ 6
HRLT (beats䡠min⫺1) 163 ⫾ 8
[La]max (mmol䡠L⫺1) 9.9 ⫾ 2.1
Wmax, maximal power output; BM, body mass; WOBLA, power output at the onset of
blood lactate accumulation; WLT, power output at the individual lactate threshold;
HRmax, maximal heart rate; HROBLA, heart rate at the onset of blood lactate accumu-
lation; HRLT, heart rate at the individual lactate threshold; [La]max, maximal blood lactate
concentration.
maximal power output was attained). All experimental pro-
cedures were approved by the ethics committee of the Uni-
versidad del Paı́s Vasco.
Laboratory protocol. Three to four weeks before the
beginning of the race in which each TT was studied, cyclists
performed an incremental maximal laboratory test on a
mechanically-braked cycle ergometer (Monark 818 E,
Varberg, Sweden) adapted with a racing saddle, drop han-
dlebars, and clip-in pedals. Initial resistance was set at 110
W, and it was increased by 35 W every 4 min, with 1-min
recovery intervals between workloads. Pedal rate was main-
tained constant at 75 rpm throughout the test. Subjects kept
cadence with a metronome. Testing continued until the
subjects were no longer able to maintain the required pedal
rate. Heart rate was recorded every 5 s during the whole test
(Sport Tester PE 3000, Polar Electro Oy, Kempele, Fin-
land). Forty-eight hours prior to each laboratory test, sub-
jects were instructed to refrain from strenuous physical
training and to ingest a standardized food and fluid plan
based on individual body mass. Training and nutritional logs
were kept to assess subjects’ compliance with these require-
ments. No food intake was allowed during the 3 hr prior to
testing.
Maximal power output (Wmax) was determined as the
highest workload a cyclist could maintain for a complete
4-min period. When the last workload was not maintained 4
full min, maximal power output was calculated as follows
(17):
Wmax ⫽ Wf ⫹ 关共 t/ 240 兲 ⫻ 35 兴
in which Wf is the value of the last complete workload (W),
t is the time the last workload was maintained (s), and 35 is
the power output difference between the last two workloads
(W).
Based on individual laboratory data obtained in the last
test prior to each race under investigation, a HR-power
output relationship was computed for each cyclist from a
linear regression equation. Each cyclist’s individual equa-
tion was used to estimate mean power output during actual
TT racing from competition mean HR values. As in previ-
ously reported investigations, the HR-power output relation-
ship was considered to be linear (3,23,27).
PROFESSIONAL TIME TRIAL CYCLING
Immediately after completion of each workload, capillary
blood samples (25 ␮l) were withdrawn from a previously
Range
hyperemized ear lobe (Finalgon, Laboratorios FHER, Bar-
22–32
celona, Spain) for the determination of blood lactate con-
160–190 centration ([La]), using an electroenzymatic technique with
53.0–80.0 an automatic analyzer (YSI威 1500 Sport, Yellow Springs
349–525
5.7–6.8 Instruments, Yellow Springs, OH). Following the recom-
275–478 mendations of the manufacturer, the analyzer was calibrated
202–417
187–204 before each test with standard solutions of known lactate
168–191 concentrations (0, 5, and 15 mmol䡠L⫺1).
146–174
6.9–13.7
The exercise intensity corresponding to the onset of blood
lactate accumulation (OBLA) was identified on the [La]-
power output curve by straight line interpolation between
the two closest points as the power output eliciting a blood
lactate concentration of 4 mmol䡠L⫺1 (24). The lactate
threshold (LT) was identified on each subject’s [La]-power
output curve as the exercise intensity that elicited a 1
mmol䡠L⫺1 increase in [La] above average baseline lactate
values measured when exercising at 40 – 60% of the maxi-
mal aerobic power output (11). Heart rate and power output
values at OBLA (HROBLA and WOBLA, respectively) and LT
(HRLT and WLT, respectively) were determined by straight
line interpolation. Based on these values, two exercise in-
tensity zones were considered: OBLAZONE and LTZONE,
described respectively as HROBLA ⫾ 3 beats䡠min⫺1 and
HRLT ⫾ 3 beats䡠min⫺1.
Competition time trials. For the purpose of the study,
competition TT were classified into five categories accord-
ing to their distance, type of terrain, and racing format.
Prologue TT (PTT) were the race opening TT, which are
most often raced over flat terrain and cover distances no
longer than 10 km. Individual TT raced over predominantly
level terrain were classified as short (STT) and long (LTT)
when the distance covered was shorter or longer than 40 km,
respectively. Individual TT in which the altitude change
between the start and the finish line was higher than 500 m
were considered as uphill TT (UTT). Finally, team TT
(TTT) were the TT races in which, instead of racing indi-
vidually as in all previously described competition formats,
all members of the team rode together. Table 2 indicates the
professional races during which this investigation was car-
ried out, their date of initiation, the number of cyclists
studied in each of the races, as well as the distance of each
TT.
In professional road cycling, not all TT are raced at the
fastest possible pace. Indeed, due to team racing strategies and
according to their specific role in competition (21), some cy-
clists ride all out in the competition TT (TTLIMIT), whereas
some other cyclists choose, following the indications of the
team coach, a more conservative approach (TTSTRATEGY). In
fact, the only cyclists racing TT all out are those who have a
real chance of winning the stage, are competing for a top
position in the final overall standings, or are involved in the
fight for the “best team” standings. This represents no more
than 15–20% of the racers. Therefore, possible differences in
the physical and physiological responses to these two ap-
proaches were studied.
Medicine & Science in Sports & Exercise姞 851
TABLE 2. Professional cycling time trial (TT) races during which heart rate was monitored.
Prologue TT Short TT Long TT Uphill TT Team TT
Laboratory Distance Distance Distance Distance Distance
Race Race start date test date N (km) N (km) N (km) N (km) N (km)
Tour de France 1993 July 3, 1993 June 1993 2 6.8
Vuelta a Catalunya 1993 September 9, 1993 August 1993 3 18.9
Vuelta a Castilla 1993 August 14, 1993 July 1993 2 27.0
Tour de France 1994 July 2, 1994 June 1994 2 7.2 3 64.0 1 47.5 1 66.5
Giro d’Italia 1994 May 21, 1994 April 1994 1 7.0 6 44.0 3 35.0
Vuelta a España 1994 April 25, 1994 April 1994 3 9.0 3 39.5 1 53.0
Tour de France 1995 July 1, 1995 June 1995 4 54.0 6 67.0
Giro d’Italia 1995 May 12, 1995 April 1995 2 6.1 5 19.0 5 42.0 4 43.0
Dauphinée-Libéré 1995 June 2, 1995 April 1995 2 6.7 5 36.5
Total observations 12 18 19 8 7
Mean ⫾ SD 7.1 ⫾ 1.0 28.2 ⫾ 9.6 51.4 ⫾ 8.8 41.8 ⫾ 6.3 66.8 ⫾ 0.4
N, number of observations.

In all studied competition TT, the HR of the cyclist was
recorded every 5 s throughout the race using a HR monitor
(Sport Tester PE 3000, Polar Electro Oy, Kempele, Fin-
land). Recorded data were then transferred to a computer
and analyzed with a computer program (Polar Precision
Performance Software, Polar Electro Oy, Kempele, Fin-
land). HR files including more than 5% of the values falling
outside a physiological range of 50 –220 beats䡠min⫺1 were
excluded from the study. Moreover, files included in the
computations were filtered to eliminate values falling out-
side the above-mentioned range using the correction func-
tion of the software (23). Competition HR values were
expressed as a percentage of HR values determined in the
laboratory test (%HRmax, %HRLT, and %HROBLA) using the
following equation:
%HR1,2,3 ⫽ 共 HRT ⫺ HRB兲 / 共 HR1,2,3 ⫺ HRB兲
in which 1, 2, and 3 are HRmax, HROBLA, and HRLT,
respectively, HRT is the average HR during competition,
and HRB is the basal HR (16). The latter was determined as
the average of the HR values recorded immediately after
awakening for the seven consecutive days prior to the lab-
oratory tests with the cyclists still lying in bed.
The distribution of the exercise time during competition
TT was analyzed in relation to the above-mentioned
OBLAZONE and LTZONE exercise intensity indices. In ad-
dition, the training impulse (TRIMP) values, an integrative
marker of the exercise load undertaken during competition,
were estimated from TT competition HR measurements and
race duration, using the following formula, as previously
described by Banister (1):
TRIMP ⫽ A 䡠 B 䡠 C
in which A is competition time (in min), B is [(HRT ⫺
HRB)/(HRmax ⫺ HRB)], and C equals 0.64䡠e1.92B.
Statistical analyses. Descriptive statistics are ex-
pressed as mean ⫾ SD. ANOVA, followed by Fisher’s post
hoc test, was used for all comparisons among different TT
categories, as well as between TTLIMIT and TTSTRATEGY.
Correlations between variables were computed from linear
regression. The level of statistical significance was set at
P ⬍ 0.05.
RESULTS
Laboratory measurements. As shown in Table 1, the
present group of cyclists attained Wmax values of 439 ⫾ 45
W and 6.4 ⫾ 0.3 W䡠kg⫺1 in the incremental laboratory test,
and HRmax was 194 ⫾ 5 beats䡠min⫺1. Submaximal labora-
tory power output and HR values (WOBLA, WLT, HROBLA,
and HRLT) can also be seen in Table 1.
Competition speed. A comparison among the differ-
ent TT categories (Table 3) showed that competition speed
was significantly faster in TTT than in all other categories.
PTT were raced significantly faster than STT and UTT, but
not than LTT. UTT were also slower than STT and LTT.

TABLE 3. Physical and physiological responses to the different competition time trial (TT) categories.
Variables Prologue TT (N ⴝ 12) Short TT (N ⴝ 18) Long TT (N ⴝ 19) Uphill TT (N ⴝ 8) Team TT (N ⴝ 7)
Distance (km) 7.3 ⫾ 1.1 28.0 ⫾ 8.6* 49.2 ⫾ 8.0*† 40.6 ⫾ 4.8*†§ 67.0 ⫾ 0.5*†§‡
Time (s) 594 ⫾ 99 2,324 ⫾ 677* 3,975 ⫾ 713*† 4,495 ⫾ 493*† 4,521 ⫾ 150*†
Speed (km䡠h⫺1) 46.3 ⫾ 2.8 43.1 ⫾ 3.0* 44.7 ⫾ 2.0 32.5 ⫾ 2.0*†§ 53.4 ⫾ 1.8*†§‡
HR (beats䡠min⫺1) 177 ⫾ 5 172 ⫾ 9* 162 ⫾ 6*† 158 ⫾ 7*† 165 ⫾ 5*†‡
%HRmax 89 ⫾ 3 85 ⫾ 5* 80 ⫾ 5*† 78 ⫾ 3*† 82 ⫾ 2*‡
%HROBLA 100 ⫾ 3 95 ⫾ 7* 89 ⫾ 5*† 87 ⫾ 2*† 92 ⫾ 4*
%HRLT 114 ⫾ 8 108 ⫾ 9* 103 ⫾ 8* 101 ⫾ 5*† 105 ⫾ 11*
TRIMP 21 ⫾ 3 77 ⫾ 23* 122 ⫾ 27*† 129 ⫾ 14*† 146 ⫾ 6*†§‡
W (W) 380 ⫾ 62 362 ⫾ 59 347 ⫾ 46 342 ⫾ 32 353 ⫾ 42
%Wmax 89 ⫾ 6 84 ⫾ 7* 79 ⫾ 5*† 77 ⫾ 5*† 80 ⫾ 5*
Values are means ⫾ SD.
N, number of observations; HR, heart rate; HRmax, maximal heart rate; HROBLA, heart rate at the onset of blood lactate accumulation; HRLT, heart rate at the individual lactate
threshold; TRIMP, training impulse (1); W, power output; Wmax, maximal power output.
* Significantly different from Prologue TT.
† Significantly different from Short TT.
§ Significantly different from Long TT.
‡ Significantly different from Uphill TT.

852 Official Journal of the American College of Sports Medicine http://www.msse.org

 Figure 2—Mean time (expressed as a percentage of the total racing
time) spent at heart rates (HR) above, equal, and below the exercise
intensity corresponding to the HR value at the onset of blood lactate
accumulation ⴞ 3 beats䡠minⴚ1 (OBLAZONE) during each competition
time trial category. PTT, prologue time trial; STT, short time trial;
LTT, long time trial; UTT, uphill time trial; TTT, team time trial.
Standard deviations and statistical significance symbols have been
omitted for the clarity of the figure.

and 7 ⫾ 6% and 21 ⫾ 11% during TTT). Percentage cycling

time above OBLAZONE was longer during STT compared
with LTT and UTT. Moreover, relative time above and at
OBLAZONE was longer in TTT than in LTT and UTT. When
expressed in absolute values, the longest time above
OBLAZONE was spent in STT (412 ⫾ 612 s); this time was
significantly longer than in LTT (32 ⫾ 83 s) and UTT (0 s)
but was not significantly longer than in PTT (234 ⫾ 159 s)
and TTT (310 ⫾ 255 s). However, time at OBLAZONE in
Figure 1—Individual examples of raw heart rate profiles recorded
TTT (937 ⫾ 469 s) was the longest compared with PTT
during a short time trial (A), a long time trial (B), and a team time trial (229 ⫾ 129 s), STT (568 ⫾ 451 s), LTT (589 ⫾ 858 s), and
(C). HROBLA, heart rate at the onset of blood lactate accumulation; UTT (97 ⫾ 138 s), although only the difference with the
HRLT, heart rate at the individual lactate threshold.
latter reached statistical significance. As depicted in Fig. 3,
92 ⫾ 7% of the racing time was spent above LTZONE in
Competition heart rate. PTT elicited significantly PTT, which was higher than in STT (71 ⫾ 32%) and
higher mean HR, %HRmax, and estimated %Wmax than all significantly higher than in LTT (51 ⫾ 38%), UTT (47 ⫾
other TT categories. These three variables were also differ-
ent between STT on the one hand and LTT and UTT on the
other hand. Mean HR was also statistically different be-
tween STT and TTT and between UTT and TTT. The latter
group difference was detected in the %Wmax values as well
(Table 3). Individual examples of raw HR profiles recorded
during different competition TT categories can be seen in
Figure 1.
Metabolic thresholds. When mean competition HR
was expressed relative to HROBLA and HRLT, PTT values
were significantly higher than STT, LTT, UTT, and TTT.
%HROBLA values were also significantly higher in STT than
in LTT and UTT, whereas %HRLT was higher in STT than
in UTT (Table 3).
Figures 2 and 3 show the distribution of racing time
during the different TT categories relative to OBLAZONE Figure 3—Mean time (expressed as a percentage of the total racing
time) spent at heart rates (HR) above, equal, and below the exercise
and LTZONE, respectively. During PTT, percentage time intensity corresponding to the HR value at the individual lactate
spent above (42 ⫾ 28%) and at (40 ⫾ 22%) OBLAZONE threshold ⴞ 3 beats䡠minⴚ1 (LTZONE) during each competition time
(Fig. 2) was significantly longer than during all other TT trial category. PTT, prologue time trial; STT, short time trial; LTT,
long time trial; UTT, uphill time trial; TTT, team time trial. Standard
categories (20 ⫾ 29% and 25 ⫾ 18% during STT, 1 ⫾ 2% deviations and statistical significance symbols have been omitted for
and 14 ⫾ 19% during LTT, 0% and 2 ⫾ 3% during UTT, the clarity of the figure.

PROFESSIONAL TIME TRIAL CYCLING Medicine & Science in Sports & Exercise姞 853
TABLE 4. Comparison between time trials (TT) performed all out (TTLIMIT) and TT performed using a more conservative approach due to team strategies (TTSTRATEGY).
Prologue TT Short TT Long TT Uphill TT
TTLIMIT TTSTRATEGY TTLIMIT TTSTRATEGY TTLIMIT TTSTRATEGY TTLIMIT TTSTRATEGY
Variables (N ⴝ 4) (N ⴝ 8) (N ⴝ 8) (N ⴝ 10) (N ⴝ 4) (N ⴝ 15) (N ⴝ 2) (N ⴝ 6)
Speed (km䡠h⫺1) 47.6 ⫾ 2.3 45.7 ⫾ 3.0 45.3 ⫾ 2.6 41.3 ⫾ 2.0* 46.5 ⫾ 2.4 44.2 ⫾ 1.7* 32.8 ⫾ 3.4 32.4 ⫾ 1.8
HR (beats䡠min⫺1) 181 ⫾ 4 175 ⫾ 4* 176 ⫾ 7 168 ⫾ 8* 167 ⫾ 7 160 ⫾ 5* 166 ⫾ 11 156 ⫾ 4*
%HRmax 90 ⫾ 3 88 ⫾ 3 88 ⫾ 3 82 ⫾ 4 84 ⫾ 3 79 ⫾ 4 81 ⫾ 1 76 ⫾ 2
W (W) 372 ⫾ 88 384 ⫾ 51 374 ⫾ 77 352 ⫾ 41 359 ⫾ 79 344 ⫾ 36 376 ⫾ 5 331 ⫾ 28
%Wmax 89 ⫾ 7 89 ⫾ 6 87 ⫾ 6 81 ⫾ 6* 82 ⫾ 4 78 ⫾ 4 83 ⫾ 1 75 ⫾ 3*
Time ⬎ OBLAZONE (%) 59 ⫾ 23 33 ⫾ 28 38 ⫾ 34 6 ⫾ 13* 3.5 ⫾ 4.0 0.2 ⫾ 0.4* 0 0
Time ⬎ OBLAZONE (s) 344 ⫾ 140 178 ⫾ 144 782 ⫾ 744 115 ⫾ 239* 130 ⫾ 150 6 ⫾ 23* 0 0
Values are means ⫾ SD.
N, number of observations; HR, heart rate; W, power output; Time ⬎ OBLAZONE (%), percentage of total time spent above an exercise intensity corresponding to the HR
value at the onset of blood lactate accumulation ⫾ 3 beats䡠min⫺1; Time ⬎ OBLAZONE (s), amount of seconds spent above an exercise intensity corresponding to the HR
value at the onset of blood lactate accumulation ⫾ 3 beats䡠min⫺1.
* Significantly different from TTLIMIT.

25%), and TTT (57 ⫾ 35%). In absolute values, TTT
required the longest cycling times above LTZONE (2539 ⫾
1550 s), followed by UTT (2140 ⫾ 1221 s), LTT (2126 ⫾
1674 s), STT (1527 ⫾ 742 s), and PTT (528 ⫾ 92 s); the
difference between TTT and PTT reached the level of sta-
tistical significance.
Racing strategies. Differences between TTLIMIT and
TTSTRATEGY are presented in Table 4. Average speed was
consistently faster in the former in all TT categories, al-
though significance was only reached in STT and LTT.
Average HR in TTLIMIT was significantly higher in all TT
categories, %HRmax was higher in STT, LTT, and TTT; and
%Wmax was also higher in STT and UTT. Moreover, time
spent above OBLAZONE in PTT, STT, and LTT was longer
in TTLIMIT than in TTSTRATEGY, both in percentage and
absolute values; the differences were statistically significant
in the latter two.
Training impulse. Table 3 shows the amount of
TRIMP corresponding to each TT category. TTT (146 ⫾ 6)
showed significantly higher values than PTT (21 ⫾ 3), STT
(77 ⫾ 23), and LTT (122 ⫾ 27). TTT values were also
somewhat higher than UTT (129 ⫾ 14), although this dif-
ference was not statistically significant. UTT values were
significantly higher than PTT and STT values, and LTT had
significantly higher values than PTT and STT.
Regression analyses. Significant inverse relation-
ships were observed between the duration of the TT and the
average competition HR (r ⫽ ⫺0.67), and between the
duration of the TT and %HRmax values (r ⫽ ⫺0.64). Com-
petition speed, on the other hand, did not significantly
correlate with exercise intensity-related variables such as
mean HR (r ⫽ 0.34), %HRmax (r ⫽ 0.39), percent time
above OBLAZONE (r ⫽ 0.31), and TRIMP (r ⫽ ⫺0.03).
DISCUSSION
This investigation was undertaken to study the exercise
intensity of road cycling TT performed during professional
stage races. To achieve this goal, and in an attempt to attain
a fair reflection of the actual physiological demands of
professional road cycling, the study was carried out on a
group of world-class cyclists (see the subjects’ physiologi-
cal characteristics in Table 1, and the outstanding average
race speeds in Tables 3 and 4) during some of the most
854 Official Journal of the American College of Sports Medicine
important stage races in the world (Table 2). Considering
%HRmax as a valuable index of exercise intensity
(2,9,10,14,15,25), the present results indicated that profes-
sional road cycling TT were performed at quite high exer-
cise intensity levels, as all of the studied events were raced
above 76%HRmax. This average value is much higher than
the 50% and the 60%HRmax at which level ground stages
and mountain stages, respectively, are usually raced (Padilla
et al., unpublished observations).
Several factors could have an influence on exercise in-
tensity during competition TT, such as duration, whether
they are raced individually or in a group, team strategies,
and race profile. In the present study, there was an inverse
relationship between exercise intensity estimated using av-
erage HR and %HRmax values and the duration of the TT.
Thus, PTT were the events raced at the highest relative
intensity (88 –90%HRmax). This value was close to those
reported by Foster et al. (7) for skating events of a similar
duration (between 84 and 92%HRmax). It is worth mention-
ing that one of the PTT described as TTLIMIT, lasting 492 s,
raced at an average speed of 49.7 km䡠h⫺1 and resulting in an
event victory, elicited in the average 94%HRmax. STT were
raced at an average value of 85%HRmax, which falls within
the range of maximal steady-state values observed by Sny-
der et al. (25) for 30-min exercise tasks (83%, 86%, and
88%HRmax at 10, 20, and 30 min of exercise, respectively).
Palmer et al. (22), on the other hand, reported an average HR
value of 90.3 ⫾ 2.4%HRmax during a 20-km TT simulation
in the laboratory. Moreover, the same authors reported av-
erage values of 94%HRmax for a 16.6-km field TT (23). The
discrepancy between these and the present STT results
could be explained by the longer duration of the latter
(16 –20 km vs 28 km) and by the influence of team strategies
on TT intensity. Indeed, TTLIMIT values measured during
STT reached 88%HRmax, much closer to those values pre-
viously reported (22). More to the point, 27-km and 36.5-km
STT were respectively raced at 93%HRmax (Fig. 1A) and
90%HRmax by individual cyclists in this study, resulting in
a stage victory and a top-three position.
LTT events, which had a duration of about 60 min, were
raced at exercise intensity values that were in keeping with
those previously published for 40 km. Hopkins and McK-
enzie (13) reported an estimated power output of 359 W,
whereas Hoogeveen and Schep (12) reported a power output
http://www.msse.org
of 347 W, estimated using the laboratory HR-power output
relationship. Using the same method, an average power
output of 347 W was estimated in the present investigation
for LTT (average distance 54 km). Coyle et al. (5) reported,
for 1-h laboratory simulations, power outputs of 346 and
311 W in elite and national level cyclists, respectively.
Lindsay et al. (19) observed an average intensity of 326 W
(90%HRmax) during a 40-km TT simulated in the laboratory.
Because of the variability in exercise conditions (laboratory
setting, field test, and amateur or professional competition)
and type of resistance that the athlete must overcome (fixed
resistance in most laboratory settings vs variable resistance
in the field), it is difficult to compare the exercise intensity
values and to justify the differences between the present and
previously published data.
Although, according to race distance and duration, TTT
were similar to LTT, the fact of racing in a group instead of
individually makes this TT format a unique competition
event. Average intensity during TTT was higher than during
LTT and UTT, despite a similar duration. As well, the
average speed attained in TTT was significantly higher than
in all individual TT categories (Table 3). This was undoubt-
edly due to the recovery periods allowed by the drafting
strategies used by all cycling teams during TTT, as shown
by the fluctuations in individual HR values during this type
of event (Fig. 1C). The graph shows repeated HR peaks and
falls, which are indicative of increased and reduced exercise
intensity periods. These periods, however, did not imply
changes in competition speed because they corresponded to
periods during which a cyclist was, respectively, leading the
group or drafting behind his teammates (23). HR values
during TTT were therefore much more irregular than during
individual STT (Fig. 1A) or LTT (Fig. 1B).
UTT was the competition category eliciting the lowest
HR response, even in TTLIMIT events. Even though these
HR values as well as estimated power output values were
not significantly different from LTT, the average speed was
remarkably slower in UTT. This discrepancy between speed
and markers of physiological demands was mainly due to
two factors: i) UTT were in the average 13% longer in
duration than LTT; ii) during LTT, which are raced on a
predominantly level terrain, a high percentage of the power
exerted by the cyclist is converted into linear displacement,
quantified as speed. Meanwhile, in UTT, part of this power
is used to perform gravitational work, which depends on the
cyclist’s body mass and is not quantified as linear speed
(6,21,26).
Thus, the average speed attained during competition TT
does not seem to be a fair marker to estimate exercise
intensity (15) because speed of locomotion in cycling de-
pends on multiple factors, such as the type of terrain, envi-
ronmental conditions, or physiological and anthropometri-
cal characteristics of the cyclists (6,21,26). Indeed, the
fastest average competition speed in the present study was
attained in TTT, but the average exercise intensity, as indi-
cated by %HRmax and estimated power output, was lower
than in the next fastest events (i.e., PTT and STT). In
addition, quite low correlations were found between speed
PROFESSIONAL TIME TRIAL CYCLING
on the one hand, and such markers of the physiological
demands of exercise as %HRmax, time spent above
OBLAZONE, or the amount of TRIMP.
As can be inferred from the discussion above, it appears
to be very difficult to estimate TT competition intensity on
average race speed values, or even on a single physiological
marker of exercise intensity such as %HRmax. Exercise
physiologists often attempt to describe how “hard” an ex-
ercise task is using such physiological markers, but “hard-
ness,” in the case of professional road cycling TT, should
integrate not only physiological markers of competition
intensity, but other variables such as the duration of com-
petition as well. Other important but less easily quantifiable
variables include environmental conditions and race profile.
The TRIMP (1) could be considered as an integrative
marker of exercise load. According to this unit, TTT were
the “hardest” TT category of all (Table 3), despite the
drafting strategies that allow intermittent periods of re-
covery. The absolute amount of time spent at and above
OBLAZONE, which has been considered as a hard exer-
cising zone (9,10), was indeed longest in TTT (20 min),
followed by STT (17 min), LTT (10 min), PTT (7.5 min),
and UTT (1.5 min). As a general rule, the organizers of
professional road cycling races choose undulating race
profiles for STT and level roads for LTT in an attempt to
balance hardness by means of the type of terrain in the
former and by duration in the latter.
The percentage racing time spent at the different meta-
bolic zones described in relation to OBLAZONE (Fig. 2) and
LTZONE (Fig. 3), and the average HR values in relation to
HROBLA and HRLT (Table 3), indicate that for TT shorter
than 30 km (duration below 30 – 40 min), such as PTT and
STT, HROBLA could be a fair metabolic index to establish
appropriate training and competition intensities. HRLT, on
the other hand, would be a useful tool to indicate suitable
intensity levels for TT lasting ⱖ60 min, such as LTT and
UTT of the present investigation. However, the fact that
PTT and STT were, respectively, raced at average values of
100% and 95%HROBLA, and UTT, LTT, and TTT at 101%,
103%, and 105% of HRLT, does not imply that the different
TT categories were raced at steady states at the correspond-
ing reference intensities, as shown by the detailed analyses
of the time distribution in relation to each intensity zone,
both in relative and absolute values.
In summary, exercise intensity was not fairly reflected by
the average race speed in competitive road cycling TT.
Markers such as %HRmax, TRIMP, or the time distribution
in relation to the different metabolic zones described using
HROBLA and HRLT seemed to be more accurate indicators of
the physiological demands of this cycling competition for-
mat. Other variables, including duration, whether racing is
individual or in a group, team strategies, and race profile can
also influence the physiological demands of the race. In TT
events lasting ⱕ30 min, HROBLA could be a valuable met-
abolic index to determine appropriate training and compe-
tition pace. For TT lasting ⱖ60 min, suitable intensity levels
would be fairly indicated by HRLT.
Medicine & Science in Sports & Exercise姞 855
 This investigation was supported by a research grant from IBER-
DROLA.
Address for correspondence: Iñigo Mujika, Ph.D., MEDIPLAN
REFERENCES
1. BANISTER, E. W. Modeling elite athletic performance. In: Physio-
logical Testing of Elite Athletes, H. J. Green, J. D. McDougal, and
H. Wenger (Eds.). Champaign, IL: Human Kinetics, 1991, pp.
403– 424.
2. BOULAY, M. R. Physiological monitoring of elite cyclists. Sports
Med. 20:1–11, 1995.
3. CLIFFORD, P. S., J. R. COAST, D. P. SWAIN, M. C. MILLIKEN, P. R.
STRICKER, and J. STRAY-GUNDERSEN. Prediction of oxygen con-
sumption from heart rate during cycling. In: Medical and Scientific
Aspects of Cycling, E. R. Burke and N. M. Newsom (Eds.). Cham-
paign, IL: Human Kinetics, 1988, pp. 93–98.
4. COYLE, E. F., A. R. COGGAN, M. K. HOPPER, and T. J. WALTERS.
Determinants of endurance in well-trained cyclists. J. Appl.
Physiol. 64:2622–2630, 1988.
5. COYLE, E. F., M. E. FELTNER, S. A. KAUTZ, et al. Physiological and
biomechanical factors associated with elite endurance cycling
performance. Med. Sci. Sports Exerc. 23:93–107, 1991.
6. DI PRAMPERO, P. E., G. CORTILI, P. MOGNONI, and F. SAIBENE.
Equation of motion of a cyclist. J. Appl. Physiol. 47:201–206,
1979.
7. FOSTER, C., M. P. CROWE, D. HOLUM, et al. The bloodless lactate
profile. Med. Sci. Sports Exerc. 27:927–933, 1995.
8. FOSTER, C., M. A. GREEN, A. C. SNYDER, and N. N. THOMPSON.
Physiological responses during simulated competition. Med. Sci.
Sports Exerc. 25:877– 882, 1993.
9. GILMAN, M. B. The use of heart rates to monitor the intensity of
endurance training. Sports Med. 21:73–79, 1996.
10. GILMAN, M. B., and C. L. WELLS. The use of heart rates to monitor
exercise intensity in relation to metabolic variables. Int. J. Sports
Med. 14:339 –344, 1993.
11. HAGBERG, J. M., and E. C. COYLE. Physiological determinants of
endurance performance as studied in competitive racewalkers.
Med. Sci. Sports Exerc. 15:287–289, 1983.
12. HOOGEVEEN, A. R., and G. SCHEP. The plasma lactate response to
exercise and endurance performance: relationships in elite triath-
letes. Int. J. Sports Med. 18:526 –530, 1997.
13. HOPKINS, S. R., and D. C. MCKENZIE. The laboratory assessment of
endurance performance in cyclists. Can. J. Appl. Physiol. 19:266 –
274, 1994.
14. HOPKINS, W. G. Quantification of training in competitive sports:
methods and applications. Sports Med. 12:161–183, 1991.
856 Official Journal of the American College of Sports Medicine
SPORT S.L., Obdulio López de Uralde 4, bajo, 01008
Vitoria-Gasteiz, Basque Country, Spain. E-mail: imujika@grn.es.
15. JEUKENDRUP, A., and A. VAN DIEMEN. Heart rate monitoring during
training and competition in cyclists. J. Sports Sci. 16:S91–S99,
1998.
16. KARVONEN, J., and T. VUORIMAA. Heart rate and exercise intensity
during sports activities: practical application. Sports Med. 5:303–
312, 1988.
17. KUIPERS, H., F. T. J. VERSTAPPEN, H. A. KEIZER, and P. GUERTEN.
Variability of aerobic performance in the laboratory and its phys-
iological correlates. Int. J. Sports Med. 6:197–201, 1985.
18. LEGER, L., and M. THIVIERGE. Heart rate monitors: validity, stabil-
ity, and functionality. Physician Sportsmed. 16:143–151, 1988.
19. LINDSAY, F. H., J. A. HAWLEY, K. H. MYBURGH, H. H. SCHOMER,
T. D. NOAKES, and S. C. DENNIS. Improved athletic performance in
highly trained cyclists after interval training. Med. Sci. Sports
Exerc. 28:1427–1434, 1996.
20. LOFTIN, M., and B. WARREN. Comparison of a simulated 16.1 km
time trial, V̇O2max and related factors in cyclists with different
ventilatory thresholds. Int. J. Sports Med. 15:498 –503, 1994.
21. PADILLA, S., I. MUJIKA, G. CUESTA, and J. J. GOIRIENA. Level ground
and uphill cycling ability in professional road cycling. Med. Sci.
Sports Exerc. 31:878 – 885, 1999.
22. PALMER, G. S., S. C. DENNIS, T. D. NOAKES, and J. A. HAWLEY.
Assessment of the reproducibility of performance testing on an
air-braked cycle ergometer. Int. J. Sports Med. 17:293–298, 1996.
23. PALMER, G. S., J. A. HAWLEY, S. C. DENNIS, and T. D. NOAKES.
Heart rate responses during a 4-d cycle stage race. Med. Sci. Sports
Exerc. 26:1278 –1283, 1994.
24. SJÖDIN, B., and I. JACOBS. Onset of blood lactate accumulation and
marathon running performance. Int. J. Sports Med. 2:23–26, 1981.
25. SNYDER, A. C., T. WOULFE, R. WELSH, and C. FOSTER. A simplified
approach to estimating the maximal lactate steady state. Int.
J. Sports Med. 14:27–31, 1994.
26. SWAIN, D. P., J. R. COAST, P. S. CLIFFORD, M. C. MILLIKEN, and J.
STRAY-GUNDERSEN. Influence of body size on oxygen consumption
during bicycling. J. Appl. Physiol. 62:668 – 672, 1987.
27. VAN HANDEL, P. J., C. BALDWIN, J. PUHL, A. KATZ, S. DANTINE, and
P. W. BRADLEY. Measurement and interpretation of physiological
parameters associated with cycling performance. In: Medical and
Scientific Aspects of Cycling, E. R. Burke and N. M. Newsom
(Eds.). Champaign, IL: Human Kinetics, 1988, pp. 47–72.
http://www.msse.org