Content uploaded by Ernst Albin Hansen
Author content
All content in this area was uploaded by Ernst Albin Hansen on Jan 23, 2017
Content may be subject to copyright.
1
Effects of cycling training at imposed low cadences - a systematic review
1
2
Ernst A. Hansena and Bent R. Rønnestadb
3
4
Affiliations:
5
a Research Interest Group of Physical Activity and Human Performance, SMI®, Department of
6
Health Science and Technology, Aalborg University, Denmark
7
b Section for Sport Science, Lillehammer University College, Lillehammer, Norway
8
9
Corresponding author:
10
Ernst Albin Hansen, MS, PhD, DSc
11
Associate Professor
12
Research Interest Group of Physical Activity and Human Performance, SMI®, Department of
13
Health Science and Technology, Aalborg University
14
Fredrik Bajers Vej 7D, DK-9220 Aalborg, Denmark
15
Phone: (+45) 50653439. E-mail: eah@hst.aau.dk
16
17
18
As accepted for publication in International Journal of Sports Physiology and Performance, ©Human Kinetics
19
20
doi: 10.1123/ijspp.2016-0574
21
22
Running head: Cycling training at imposed low cadences
23
24
Word count: 4663
25
Abstract word count: 243
26
Number of tables: 1
27
Number of figures: 1
28
29
2
ABSTRACT
30
The present article reviews effects of training at low imposed cadences in cycling. We
31
performed a systematic literature search of MEDLINE and SPORTDiscus up to April 2016 to
32
identify potentially relevant articles. Based on the titles and abstracts of the identified articles,
33
a subset of articles was selected for evaluation. These articles constituted original research
34
articles on adaptation to training at different imposed cadences in cycling. Seven articles were
35
selected for evaluation. With regard to the terminology in the present article, “low cadences”
36
refers to cadences below the freely chosen cadence. Eighty rpm can for example be
37
considered a low cadence if effort is maximal. On the other hand, the cadence has to be lower
38
than 80 rpm (e.g. 40-70 rpm) to be considered low if cycling is performed at low power
39
output. The reason is that the choice of cadence is dependent on power output. In conclusion,
40
there is presently no strong evidence for a benefit of training at low cadences. It can
41
tentatively be recommended to consider including training bouts of cycling at low cadence at
42
moderate to maximal intensity. The reason for the restrained recommendation is the
43
following. Some of the selected studies indicate no clear performance enhancing effect of
44
training at low cadence, or even indicate a superior effect from training at freely chosen
45
cadence. Furthermore, the selected studies are considerably dissimilar with respect to e.g.
46
participant characteristics and to the applied training regimens.
47
48
Keywords: Endurance, Exercise, Pedal rate, Pedaling frequency, Pedalling frequency,
49
50
51
3
INTRODUCTION
52
Cyclists’ exercise consists primarily of outdoor cycling, which is typically performed
53
for 2-30 hours per week, depending on the category of the cyclists.1 Besides, a number of
54
alternative types of exercise are occasionally applied. These types of exercise comprise e.g.
55
indoor ergometer cycling,2 strength training,3 and alternative endurance training activities
56
such as cross country skiing and running.4
57
A special kind of cycling training consists of one or more cycling bouts, each lasting in
58
the range of approximately 5-20 min, and being performed at an imposed low cadence. In the
59
present context, the cadence is considered to be low when it is below the freely chosen
60
cadence (Fig. 1). This kind of training is commonly applied by competitive cyclists and
61
widely denoted “power training”, “heavy low cadence training”, or “functional strength
62
training”. Power output during the cycling bouts is moderate to high. Our experience with
63
competitive cyclists is that the bouts are often performed on an ascent. The low cadence
64
causes pedal force to be high in each pedal thrust. Exactly how high the pedal force in each
65
pedal thrust becomes is a result of the combination of power output and cadence.
66
Based on our experience with competitive cycling environments, it appears evident that
67
cycling training at imposed low cadences is currently widely applied, and has been for a
68
number of years. The question is whether or not this training in fact enhances endurance
69
cycling performance? A number of research studies have been performed to investigate that
70
question during the recent years. However, the scientific literature within the field has not
71
been systematically reviewed previously. Therefore, that is the purpose of the present review
72
article.
73
74
METHODS
75
4
A systematic search for relevant literature using MEDLINE, through pubmed.com, and
76
SPORTDiscus was performed on 5. April, 2016. Keywords used were ((cycling OR cyclists)
77
AND training) AND (cadence OR pedal rate OR pedal frequency OR pedalling OR pedaling).
78
The search identified 222 articles from MEDLINE and 190 articles from SPORTDiscus.
79
Seven of the articles were considered of primary relevance based on the titles and abstracts
80
and selected for further evaluation. These 7 articles were original-research peer-reviewed
81
articles on adaptation to training at different imposed cadences. In addition, the reference lists
82
in the 7 selected articles were studied to identify additional articles of secondary relevance
83
that were eligible for this review. Finally, our own article archives were scrutinised for
84
relevant articles. Because of the modest amount of identified relevant literature, all 7 selected
85
studies were evaluated regardless of the participants’ characteristics and training background.
86
87
RESULTS
88
This chapter contains an evaluation of the selected articles. Key information from the 7
89
articles which were selected based on the systematic literature search is presented in Table 1.
90
With regard to the characteristics of the participants, these ranged across the studies
91
from healthy sedentary males5 to endurance-trained cyclists with an average maximal oxygen
92
uptake (VO2max) of 65 ml kg-1 min-1.6 In the present context, it could generally be expected
93
that less trained individuals would obtain training adaptations faster, and of larger relative
94
magnitude, as compared with more trained individuals. The duration of the training periods
95
ranged from 2 weeks5 to 12 weeks.7, 8 In the present context, 2 weeks could generally be
96
considered relatively short while 4 weeks could be considered sufficient for initial
97
neuromuscular adaptations. Twelve weeks could be considered appropriate for initial
98
musculoskeletal adaptations. One of the selected studies was performed in the main
99
competition season,9 while other studies were performed in the off-season period.7, 8 Not all
100
5
articles contained information on this aspect, which is unfortunate since indications on
101
training status at the beginning of the training intervention including information on training
102
performed during the months prior to the intervention is of great relevance. To further meet
103
this challenge, a control group should obviously be included in training intervention studies,
104
whenever it is possible.
105
With regard to the training regimens, the selected studies showed great variation.
106
Imposed cadences from 35 crank revolutions per min (rpm)5 to 140 rpm10 were applied. The
107
intensity during the cycling bouts at specified cadences ranged from submaximal at lactate
108
threshold, corresponding to on average approximately 63 W,5 to maximal effort.8, 9 The
109
characteristics of the performed bouts ranged from 4-8 sets of 12 maximal crank revolutions
110
which were completed twice per week8 to a single bout of 60 min which was completed 5
111
times per week.5 The major outcomes of the training in the selected articles were a mixture
112
between performance and indicators of performance. An example of a performance
113
measurement is average power output in a time trial. For comparison, the physiological
114
response of lactate concentration during cycling at a submaximal power output is considered
115
an example of an indicator of performance because of its indirect nature.
116
Interestingly, two of the selected studies used an approach, which is somewhat unusual
117
and more similar to the approach of heavy strength training. Koninckx et al. (2010)8
118
compared 4-8 sets of 12 maximum crank revolutions at a relatively low cadence (80 rpm) and
119
high power output (775-875 W) with traditional heavy strength training (3 sets, 8-15RM).
120
Both groups showed similar improvements in performance and indicators of performance,
121
except for maximal power output at a high cadence, where the strength-training group showed
122
improvement whereas the low cadence group did not (and actually showed impaired pedal
123
stroke efficiency). In another study by Paton et al. (2009), 5 sets of 30-s maximal cycling
124
bouts at either high or low cadence were compared.9 The low cadence group was able to
125
6
perform a higher power output during the training sessions. After 4 weeks of training, this
126
group achieved a “likely beneficial” effect on performance, and on indicators of performance,
127
as compared to the high cadence group. The authors observed a larger upregulation of the
128
anabolic hormone testosterone after low cadence sessions compared with the high cadence
129
sessions. This made them suggest that some of the superior adaptations to the low cadence
130
training may be related to this anabolic hormone. However, the impact of acute upregulation
131
of anabolic hormones on training adaptations is uncertain. Some results indicate an additive
132
effect on training adaptations,11 while other results show no additive effect.12 Indeed in the
133
study by Paton et al. (2009),9 the low cadence group increased their VO2max, while no change
134
occurred in the high cadence group. This is, as the authors’ stated, the most likely explanation
135
for the observed group differences.9
136
Kristoffersen et al. (2014)7 investigated the effects of 12 weeks of two different types of
137
training. Two weekly sessions of 5×6 min intervals at a moderate intensity was performed.
138
One group performed the intervals at low cadence (40 rpm) while another applied a freely
139
chosen cadence. The low cadence group showed no change on performance and indicators of
140
performance. The freely chosen cadence group showed increases in all these variables. In this
141
case, in trained master cyclists (≈47 years old), intervals performed with freely chosen
142
cadence appears to be better than intervals at imposed low cadence. In another study by
143
Ludyga et al. (2016), a comparison was made between training with high and low cadence
144
intervals. Similar increases in indicators of performance were observed in the two groups.10
145
Both groups showed larger improvements than a control group that merely performed basic
146
low-intensity endurance training with a lower volume and no high intensity sessions (while
147
the high and low cadence group included high intensity sessions). The authors concluded that
148
both high and low cadence training provide effective training stimuli, when identical exercise
149
intensities are prescribed.10 Another study by Nimmerichter et al. (2012) which applied a
150
7
similar design as the latter, observed no significant differences in improvements during a
151
graded exercise test between a low (60 rpm) and a high (100 rpm) cadence interval-training
152
group (6×5 min at 300 W) and a control group that did not perform the intervals.13 A fine
153
detail of that particular study that actually increases the ecological validity of the study was
154
that the low cadence intervals were performed uphill and the high cadence intervals were
155
performed in flat terrain. The study revealed larger improvements in time trials performed in
156
the terrain in which the interval-training sessions were performed. Interestingly, only the low
157
cadence group increased time trial performance in both flat and uphill terrain.13 Therefore, the
158
authors concluded that training at imposed low cadences results in a potentially higher
159
training stimulus with a crossover effect to flat time trials. The same research group
160
performed a longitudinal study of elite cyclists in which the training time spent on intervals
161
(lasting 2-20 min and performed at 40-60 rpm) was strongly correlated with the classification
162
of the riders (r=–0.86) and the improvement of 20-min time trial power output during the
163
season (r=0.83).14 In line with Nimmerichter et al.’s (2012)13 confirmation of the training
164
principle of specificity, it has been suggested that in order to improve performance in both
165
uphill and flat terrain, a cyclist should train in the specific terrains like flat and uphill road
166
with high and low cadence.15
167
Somewhat different training adaptations to high and low cadence training was observed
168
by Whitty et al. (2016) in trained cyclists. A low cadence group trained specific interval (4-
169
6×4 min at 70% of maximal aerobic power) sessions of 45-60 min duration at a cadence 20%
170
below the freely chosen cadence. For comparison, another group performed the interval
171
sessions at a cadence 20% above the freely chosen cadence.6 Both groups achieved similar
172
increases in VO2max and maximal aerobic power, but only the high cadence group improved
173
the efficiency at 90 and 110 rpm. Despite the seemingly better adaptations in the high cadence
174
group on efficiency, the low cadence group achieved a greater improvement of average power
175
8
output during a 15-min all-out trial as compared to the high cadence group (16% vs 8%,
176
respectively).
177
Hirano et al. (2015) observed that values of oxygen uptake (VO2) and heart rate were
178
lower during cycling at low cadences compared with higher cadences, despite of identical
179
power output at the two cadences.5 In addition, pedal force was higher at low cadence at the
180
same time as peripheral oxygenation was lower. After 2 weeks of 5 weekly sessions with low
181
cadence cycling, the increases in power output at the lactate threshold was larger than after
182
high cadence sessions in previously untrained males.5 The authors suggested that high pedal
183
forces with concomitant low muscle oxygenation caused by pedalling at low cadence (35
184
rpm) constituted the peripheral stimuli for aerobic improvements.5 Whether the same would
185
be the case for well-trained cyclists remains to be investigated.
186
187
DISCUSSION
188
During prolonged strenuous cycling, or during intensive cycling at high effort, cyclists
189
typically find it difficult to apply anything other than their freely chosen cadence. Still, it is of
190
course possible, with volitional control and exertion, to apply particular cadences that are
191
considerably higher or lower than the freely chosen cadence. However, if for example a low
192
cadence is applied during cycling at high power output, performance will most likely be
193
impaired.16 It follows, that a cyclist rarely would do that in a test or competition situation.
194
During training, on the other hand, where adaptation rather than performance is in focus, there
195
might be a point in applying a low or high cadence in certain cycling bouts. Consequently, it
196
is important to acknowledge the difference between governing the cadence during training
197
versus during test or competition.
198
With regard to the terminology in the present article, 80 rpm can for example be
199
considered a low cadence if effort is maximal. On the other hand, the cadence has to be lower
200
9
than 80 rpm (e.g. 40-70 rpm) to be considered low if cycling is performed at low power
201
output (Fig. 1). The reason is that the choice of cadence is dependent on power output. As an
202
example, freely chosen cadence in 10 professional cyclists during uphill road cycling was
203
reported to be on average approximately 78 rpm at 50 W as compared to an average of
204
approximately 88 rpm at 750 W.17 At an even higher power output of on average 1020 W,
205
during sprinting at maximal effort in road cycling, freely chosen cadence was on average 110
206
rpm in 6 professional cyclists.18 The freely chosen cadence during cycling appears to be
207
basically consistent within an individual.19, 20 However, at the same time it varies considerably
208
between individuals. As an example of the latter, the freely chosen cadence in 10 professional
209
cyclists ranged from average values of 62-89 rpm during a 1. category mountain ascent in
210
Tour De France.17 Slope also influences the freely chosen cadence. As an example, the freely
211
chosen cadence in 8 well-trained cyclists was on average 91 and 59 rpm during flat and
212
uphill cycling, respectively, at a power output of on average 280-292 W.21 For a more
213
thorough review of factors external and internal to the cyclist, which affect the freely chosen
214
cadence, the reader is referred to a previously published review article.22
215
Terms like “heavy low cadence training” or “functional strength training” signals that
216
this kind of training has a substantial effect on muscle strength. Further, that any potential
217
performance enhancing effect is related to mechanisms involved in adaptations to heavy
218
strength training. However, the effect on maximal muscle strength of cycling training at
219
imposed low cadence is rarely investigated. In the few cases where it has been investigated,
220
there seems to be no effect on maximal muscle strength.7, 8 The seemingly lack of effect of
221
imposed low cadence training on maximal strength is due to the relatively low force
222
development during low cadence cycling as compared to the maximal force capacity of the
223
leg muscles. In that context, 40-60 repetitions per min for 5-20 min is more similar to
224
endurance training than to heavy strength training. Therefore, potential mechanisms
225
10
underpinning positive effects of heavy strength training on endurance performance, including
226
increases in maximal strength, rate of force development and muscle mass (reviewed
227
previously),23, 24 are somewhat unlikely to explain any potential benefits of training at
228
imposed low cadence. It should be noted that the study by Konincks et al. (2010) indicated
229
that it is possible to obtain almost similar performance enhancing effects from heavy strength
230
training and cycling at imposed low cadence when the protocol for the latter aims at maximal
231
force production for a short duration of time. In addition from a strict scientific point of view,
232
it should be remarked that the two studies by Konincks et al. (2010) and Paton et al. (2009)
233
did not include control groups, which trained with freely chosen cadence. Therefore, for these
234
two studies we cannot conclude that intervals performed at imposed low cadence are more
235
advantageous than the same intervals performed at the same intensity at a freely chosen
236
cadence.
237
The power output is the most important variable to improve with regard to performance,
238
but the study by Whitty et al. (2016) highlights some potential different adaptations by
239
imposed low and high cadences. Further, this indicates that a combination of high and low
240
cadence training might be a good choice. Nimmericher et al. (2012)13 observed no differences
241
between high and low cadence training on improvements of indicators of performance, but
242
they too noticed a slight advantage of low cadence training on overall time trial performance.
243
It might be speculated that there are some slight effects of the low cadence training that are
244
only detectable when the cyclists are becoming quite fatigued as compared to standard
245
indicators of performance being tested in a more unfatigued condition. Cycling at a low
246
cadence increases quadriceps muscle activation25 and recruitment of type II muscle fibres26 as
247
compared with cycling at a higher cadence at the same power output. Therefore, it can be
248
speculated that cycling at imposed low cadences results in increased neuromuscular stimulus
249
and an increased stimulus on type II muscle fibres. It could also be suggested that during
250
11
imposed low cadence bouts at all-out intensity for 15-30 s (like applied in the studies by
251
Koninckx et al. (2010)8 and Paton et al. (2009)9), type IIX fibres can be activated and thus
252
converted to the more efficient type IIA fibres. Indeed, heavy strength training induced a
253
decrease in the proportion of type IIX fibres that was associated with improvement of power
254
output during a 40-min all-out trial in well-trained female cyclists (r=–0.63).27 This potential
255
mechanism can, in theory, have a larger effect on performance than on indicators of
256
performance.
257
By imposing low or high cadences during cycling, a number of acute responses can be
258
influenced. These responses can turn out to be of relevance with respect to for example the
259
loading, perception, function, and performance of the cyclist. As a consequence, the cadence
260
applied during training can affect the training adaptation. The following paragraphs focus on
261
responses in some selected key variables of biomechanical and physiological characteristics.
262
When cycling at a constant power output, pedal force increases with a decrease in
263
cadence.28, 29 In terms of magnitude of force, peak pedal force amounted to approximately 440
264
N during cycling at 61 rpm at a power output of 260 W.19 That force corresponded to
265
approximately 47% of the peak pedal force applied during cycling at maximal effort at the
266
same cadence.19 Similar absolute values of pedal force have been reported by others.29 It is
267
possible that a change from high to low cadences, and thereby from low to high pedal forces,
268
shifts the muscle fibre recruitment pattern in direction of more type II muscle fibre and less
269
type I fibre recruitment. This is supported by a previous report of more glycogen depletion of
270
type II muscle fibres after cycling at 50 rpm as compared to 100 rpm at an intensity of 80% of
271
the VO2max.26
272
Brain activity also varies with cadence. As an example, the cerebral activation during
273
ergometer cycling was investigated by oxygen-15-labelled H2O positron emission
274
tomography (PET) in 7 healthy individuals.30 The study showed that compared to rest, active
275
12
cycling at an ergometer “loading” of 1-12 kg (corresponding to not stated power outputs)
276
significantly activated sites bilaterally in the primary sensory cortex, primary motor cortex,
277
supplementary motor cortex and the anterior part of cerebellum. Comparing passive cycling
278
movements with rest, an almost equal activity was observed. When subtracting activity
279
recordings during passive cycling movements from recordings during active cycling,
280
significant activity was only observed in the leg area of the primary motor cortex and the
281
precuneus, whereas not in the primary sensory cortex. The primary motor cortex activity was
282
positively correlated with the cadence of the active cycling. However, with regard to that
283
particular result, it should be noted that power output was most likely increased with
284
increasing cycling cadence as load rather than power output was maintained constant.
285
Imagination of cycling, compared to rest, activated bilaterally sites in the supplementary
286
motor cortex. Christensen et al. (2000)30 suggested that higher motor centres, including the
287
primary and supplementary motor cortices as well as the cerebellum, take an active part in the
288
generation and control of rhythmic motor tasks such as cycling. Still, it was also noted by the
289
authors that perhaps the main part of the cerebral activity observed during the passive cycling
290
was in fact caused by the sensory feedback evoked by the moving limbs and that a similar
291
mechanism also explains a very significant part of the cerebral activity during active cycling.
292
As another example, functional magnetic resonance imaging (fMRI) was applied to record
293
human brain activity during ergometer cycling at 30 rpm, 60 rpm, a variable cadence, and
294
during passive pedalling at 30 rpm.31 Ten healthy adults participated and exercised at, again, a
295
not stated power output. After identifying regions of interest, the intensity and volume of
296
brain activity in each region was calculated and compared across conditions. The results
297
showed that the primary sensory and motor cortices, supplementary motor area and
298
cerebellum were active during cycling. The intensity of activity in these areas increased with
299
increasing cadence and complexity. The cerebellum was the only brain region that showed
300
13
significantly lower activity during passive as compared to active cycling. The authors
301
concluded that primary sensory and motor cortices, supplementary motor area and cerebellum
302
have a role in modifying continuous, bilateral and multi-joint lower extremity movements,
303
and further, that much of this brain activity may be driven by sensory signals from the moving
304
limbs. A recent study highlights the importance of including high cadence cycling in the
305
training.32 That study found enhanced neural efficiency, in the form of reduced cortical
306
activity, after training at imposed high cadence while not after low cadence training.32 This
307
adaptation might allow a reservation of cortical resources for for extending the cycling or
308
increasing the power output and therefore enhancing the capacity for allocating resources in a
309
sports-specific task.32
310
Electromyographic activity is yet another variable that varies with cadence. As an
311
example, surface linear envelope electromyographic activity normalized to electromyographic
312
activity during isometric maximal voluntary contraction was investigated in 11 healthy
313
individuals during ergometer cycling at 0-240 W at 40-100 rpm.33 The study showed that an
314
increase in cadence increased the electromyographic activity over the gluteus maximus,
315
gluteus medius, vastus medialis, medial hamstring, gastrocnemius medialis, and soleus
316
muscles. As another example, root-mean-square electromyographic activity was investigated
317
from 7 leg muscles in 8 active male individuals during ergometer cycling at 100-400 W at 50-
318
120 rpm.34 The study showed that a second-order polynomial equation fitted the average root-
319
means-square electromyographic activity data of all muscles vs. cadence (r2 ranging from
320
0.870-0.996) for each power output. The cadence with the lowest amplitude of
321
electromyographic activity (denoted the optimal cadence by the authors) for a given power
322
output increased with increases in power output from on average 57 rpm at 100 W to on
323
average 99 rpm at 400 W. Thus, cycling at low cadences results in relatively high values of
324
muscle activation.
325
14
VO2, energy turnover, and efficiency all vary with cadence. Values of VO2 can be
326
converted to estimates of rate of energy turnover, alternatively termed metabolic power
327
output, by taking into account respiratory exchange ratio. Further, when mechanical power
328
output is known from cycle ergometer recordings, gross efficiency can be calculated.35 In
329
addition, alternative forms of efficiency can be calculated. These can, for example, account
330
for estimates of “internal power”, generated by muscles to overcome energy changes of
331
moving body segments, and/or account for the resting VO2.36, 37 When VO2 is depicted as a
332
function of cadence, the relationship is approximately U-shaped.38 For comparison, the
333
relationship is inverted-U-shaped when gross efficiency is depicted as a function of cadence
334
39. The cadence with the lowest VO2 or highest gross efficiency can be termed the
335
energetically optimal cadence and it occurs at approximately 50-80 rpm. Thus, cycling at
336
lower cadences results in relatively high values of VO2 and energy turnover as well as low
337
values of efficiency. In more details, the energetically optimal cadence increases with
338
increasing exercise intensity.40 The exact reasons for VO2 to be relatively high at low and
339
high cadences are unknown. However, it is possible that a large type II muscle fibre
340
recruitment at low cadences26 accounts for the high VO2 in that condition since type II muscle
341
fibres are less efficient than type I muscle fibres.19, 35 Besides, it is possible that high internal
342
power accounts for the high VO2 during cycling at relatively high cadences.37
343
Perceived exertion also varies with cadence and constitutes the final variable reviewed
344
here. Subjective rating of perceived exertion (RPE) during exercise can be determined by
345
having individuals indicating their rating on a 15-point (6-20) scale.41 As an example, overall
346
RPE was investigated in 6 healthy male individuals during ergometer cycling at a range of
347
cadences from 40-100 rpm at two power outputs corresponding to 70% and 100% of
348
VO2max.42 After fitting of parabolic curves to the group average scores, the RPE data revealed
349
U-shaped relationships. Minimum RPE-values occurred at 65 and 73 rpm at the low and high
350
15
power output, respectively. As another example, overall RPE was investigated in 20 healthy
351
males during treadmill cycling at a range of cadences from 61-115 rpm at two power outputs
352
corresponding to 40% and 70% of the power output at which VO2max was attained.19 After
353
fitting of second-degree polynomials to the group average scores, the RPE-data revealed
354
parabolic relationships with minimum RPE values occurring at 63 and 72 rpm at the low and
355
high power output, respectively. Thus, cycling at cadences below 60 rpm results in relatively
356
high RPE-values.
357
358
Limitations
359
Based on the chapters above, it appears challenging to firmly point at certain
360
recommendable commonalities in the training regimens. The reasons are for example that
361
there is a considerable variation between studies in participant characteristics, as well as in
362
study design and applied procedures. Other limitations with regard to interpretation of the
363
outcomes of the selected studies are that some articles are missing key information on for
364
example the time of the year at which the study was conducted, the training performed prior
365
to the intervention, as well as on the training performed in addition to the specific
366
intervention. Besides, some of the studies lack a pure control group, which merely should
367
have continued the usual training.
368
369
Suggestions for future research
370
More research is needed to enhance our knowledge on adaptations to cycling training at
371
different imposed cadences. Especially, research is needed that in particular systematically
372
investigates the influence of the following variables:
373
- training background and fitness status of the studied participants.
374
- season in which the training is performed (pre-, in-, or post-competition).
375
16
- duration of the training period.
376
- characteristics of the applied training regimens including for example the frequency of
377
training, the duration and number of bouts, the intensity of training, and the applied
378
cadences.
379
- training and/or races performed in addition to the specific imposed cadence
380
intervention.
381
In addition, it appears relevant to include a control group that simply continues the usual
382
training (as done by Nimmerichter et al., 201213 and Ludyga et al., 201610) or performs
383
similar intervals (at the same intensity, but at a freely chosen cadence) as the different
384
high/low cadence groups.7
385
Such new research will be useful with regard to providing additional valuable
386
information on how different groups of individuals can benefit from cycling training at
387
different imposed cadences.
388
389
Practical applications
390
With respect to practical applications, it is first and foremost worth noting that training
391
at a large range of cadences, from very low to very high, probably will be beneficial for
392
performance in competitive cycling. A reason is that work demands in competitive cycling are
393
multifaceted and include pedalling at both low and high cadences. Furthermore, from a
394
practical point of view, it appears to make sense for a single individual cyclist to perform a
395
particular kind of cycling training if that training is assessed to improve the cyclist’s
396
performance. Though, to be able to provide evidence-based training advice, which can be
397
recommended for cyclists in general, it is necessary to evaluate randomized controlled studies
398
involving groups of cyclists. That was the focus of the present review.
399
400
17
Conclusions
401
To conclude, new knowledge was obtained from the present systematic review. Seven
402
original articles were considered relevant and selected for further evaluation. The 7 articles
403
contained information on adaptations in endurance performance (and in indicators of
404
endurance performance) to cycling training at different imposed cadences, including low
405
cadence. For example, competitive male cyclists who trained 30-s bouts at maximal effort at
406
low cadence (60-70 rpm) twice per week for 4 weeks obtained increased peak power in an
407
incremental cycling test as well as increased power output at set submaximal physiological
408
responses.9 In addition, healthy sedentary males who trained 60-min bouts at low cadence (35
409
rpm) 5 times per week for 2 weeks improved their power output at the lactate threshold.5
410
Another study observed no differences in training adaptations between groups of cyclists that
411
performed high and low cadence training,10 while yet another actually favoured freely chosen
412
cadence training over low cadence training in cyclists.7 There is presently no strong evidence
413
for a benefit of training at low cadences. The overall interpretation is that it tentatively can be
414
recommended to consider including training bouts of cycling at low cadence at moderate to
415
maximal intensity.6, 8, 9, 13 The reason for the restrained recommendation is the following.
416
Some of the selected studies indicate no clear performance enhancing effect of training at low
417
cadence or even indicate a superior effect from training at freely chosen cadence.
418
Furthermore, the selected studies are considerably dissimilar with respect to for example
419
participant characteristics as well as to the applied training regimens.
420
421
18
References
422
423
1. Jeukendrup AE, Martin J. Improving cycling performance: how should we spend our
424
time and money. Sports Med. 2001; 31(7):559-569.
425
2. Faria EW, Parker DL, Faria IE. The science of cycling: physiology and training - part
426
1. Sports Med. 2005; 35(4):285-312.
427
3. Faria IE. Applied physiology of cycling. Sports Med. 1984; 1(3):187-204.
428
4. Rønnestad BR, Hansen EA, Raastad T. Effect of heavy strength training on thigh
429
muscle cross-sectional area, performance determinants, and performance in well-
430
trained cyclists. Eur. J. Appl. Physiol. 2010; 108(5):965-975.
431
5. Hirano M, Shindo M, Mishima S, et al. Effects of 2 weeks of low-intensity cycle
432
training with different pedaling rates on the work rate at lactate threshold. Eur. J.
433
Appl. Physiol. 2015; 115(5):1005-1013.
434
6. Whitty AG, Murphy AJ, Coutts AJ, Watsford ML. The effect of low vs high cadence
435
interval training on the freely chosen cadence and performance in endurance trained
436
cyclists. Appl. Physiol. Nutr. Metab. 2016; 41(6):666-673.
437
7. Kristoffersen M, Gundersen H, Leirdal S, Iversen VV. Low cadence interval training
438
at moderate intensity does not improve cycling performance in highly trained veteran
439
cyclists. Front. Physiol. 2014; 5:34:1-7.
440
8. Koninckx E, Van Leemputte M, Hespel P. Effect of isokinetic cycling versus weight
441
training on maximal power output and endurance performance in cycling. Eur. J.
442
Appl. Physiol. 2010; 109(4):699-708.
443
9. Paton CD, Hopkins WG, Cook C. Effects of Low- vs. High-Cadence Interval Training
444
on Cycling Performance. J. Strength Cond. Res. 2009; 23(6):1758-1763.
445
19
10. Ludyga S, Gronwald T, Hottenrott K. Effects of high vs. low cadence training on
446
cyclists' brain cortical activity during exercise. J. Sci. Med. Sport. 2016; 19(4):342-
447
347.
448
11. Rønnestad BR, Nygaard H, Raastad T. Physiological elevation of endogenous
449
hormones results in superior strength training adaptation. Eur. J. Appl. Physiol. 2011;
450
111(9):2249-2259.
451
12. West DW, Burd NA, Tang JE, et al. Elevations in ostensibly anabolic hormones with
452
resistance exercise enhance neither training-induced muscle hypertrophy nor strength
453
of the elbow flexors. J. Appl. Physiol. 2010; 108(1):60-67.
454
13. Nimmerichter A, Eston R, Bachl N, Williams C. Effects of low and high cadence
455
interval training on power output in flat and uphill cycling time-trials. Eur. J. Appl.
456
Physiol. 2012; 112(1):69-78.
457
14. Nimmerichter A, Eston RG, Bachl N, Williams C. Longitudinal monitoring of power
458
output and heart rate profiles in elite cyclists. J. Sports Sci. 2011; 29(8):831-840.
459
15. Bertucci W, Grappe F, Girard A, Betik A, Rouillon JD. Effects on the crank torque
460
profile when changing pedalling cadence in level ground and uphill road cycling. J.
461
Biomech. 2005; 38(5):1003-1010.
462
16. Zoladz JA, Rademaker ACHJ, Sargeant AJ. Human muscle power generating
463
capability during cycling at different pedalling rates. Exp. Physiol. 2000; 85(1):117-
464
124.
465
17. Vogt S, Roecker K, Schumacher YO, et al. Cadence-power-relationship during
466
decisive mountain ascents at the Tour de France. Int. J. Sports Med. 2008; 29(3):244-
467
250.
468
18. Menaspà P, Quod M, Martin DT, Peiffer JJ, Abbiss CR. Physical Demands of
469
Sprinting in Professional Road Cycling. Int. J. Sports Med. 2015; 36(13):1058-1062.
470
20
19. Hansen EA, Andersen JL, Nielsen JS, Sjøgaard G. Muscle fibre type, efficiency, and
471
mechanical optima affect freely chosen pedal rate during cycling. Acta Physiol. Scand.
472
2002; 176:185-194.
473
20. Hansen EA, Ohnstad AE. Evidence for freely chosen pedalling rate during
474
submaximal cycling to be a robust innate voluntary motor rhythm. Exp. Brain Res.
475
2008; 186(3):365-373.
476
21. Millet GP, Tronche C, Fuster N, Candau R. Level ground and uphill cycling efficiency
477
in seated and standing cycling. Med. Sci. Sports Exerc. 2002; 34(10):1645-1652.
478
22. Hansen EA, Smith G. Factors affecting cadence choice during submaximal cycling
479
and cadence influence on performance. Int. J. Sports Physiol. Perform. 2009; 4(1):3-
480
17.
481
23. Mujika I, Rønnestad BR, Martin DT. Effects of Increased Muscle Strength and
482
Muscle Mass on Endurance-Cycling Performance. Int. J. Sports Physiol. Perform.
483
2016; 11(3):283-289.
484
24. Aagaard P, Andersen JL. Effects of strength training on endurance capacity in top-
485
level endurance athletes. Scand. J. Med. Sci. Sports. 2010; 20 Suppl 2:39-47.
486
25. Lucia A, San Juan AF, Montilla M, et al. In professional road cyclists, low pedaling
487
cadences are less efficient. Med. Sci. Sports Exerc. 2004; 36(6):1048-1054.
488
26. Ahlquist LE, Bassett Jr DR, Sufit R, Nagle FJ, Thomas DP. The effect of pedaling
489
frequency on glycogen depletion rates in type I and II quadriceps muscle fibers during
490
submaximal cycling exercise. Eur. J. Appl. Physiol. 1992; 65(4):360-364.
491
27. Vikmoen O, Ellefsen S, Trøen O, et al. Strength training improves cycling
492
performance, fractional utilization of VO2max and cycling economy in female
493
cyclists. Scand. J. Med. Sci. Sports. 2016; 26(4):384-396.
494
21
28. Patterson RP, Moreno MI. Bicycle pedalling forces as a function of pedalling rate and
495
power output. Med. Sci. Sports Exerc. 1990; 22(4):512-516.
496
29. Candotti CT, Ribeiro J, Soares DP, De Oliveira AR, Loss JF, Guimarães ACS.
497
Effective force and economy of triathletes and cyclists. Sports Biomech. 2007;
498
6(1):31-43.
499
30. Christensen LOD, Johannsen P, Sinkjaer T, Petersen N, Pyndt HS, Nielsen JB.
500
Cerebral activation during bicycle movements in man. Exp. Brain Res. 2000;
501
135(1):66-72.
502
31. Mehta JP, Verber MD, Wieser JA, Schmit BD, Schindler-Ivens SM. The Effect of
503
Movement Rate and Complexity on Functional Magnetic Resonance Signal Change
504
During Pedaling. Motor Control. 2012; 16(2):158-175.
505
32. Ludyga S, Hottenrott K, Gronwald T. Four weeks of high cadence training alter brain
506
cortical activity in cyclists. J. Sports Sci. 2016; doi: 10.1080/02640414.2016.1198045.
507
33. Ericson MO, Nisell R, Arborelius UP, Ekholm J. Muscular activity during ergometer
508
cycling. Scand. J. Rehab. Med. 1985; 17(2):53-61.
509
34. Macintosh BR, Neptune RR, Horton JF. Cadence, power, and muscle activation in
510
cycle ergometry. Med. Sci. Sports Exerc. 2000; 32(7):1281-1287.
511
35. Coyle EF, Sidossis LS, Horowitz JF, Beltz JD. Cycling efficiency is related to the
512
percentage of type I muscle fibers. Med. Sci. Sports Exerc. 1992; 24(7):782-788.
513
36. Winter DA. Biomechanics and motor control of human movement, 2 ed, New York,
514
John Wiley & Sons, Inc.; 1990.
515
37. Hansen EA, Jørgensen LV, Sjøgaard G. A physiological counterpoint to mechanistic
516
estimates of "internal power" during cycling at different pedal rates. Eur. J. Appl.
517
Physiol. 2004; 91(4):435-442.
518
22
38. Coast JR, Welch HG. Linear increase in optimal pedal rate with increased power
519
output in cycle ergometry. Eur. J. Appl. Physiol. 1985; 53(4):339-342.
520
39. Böning D, Gönen Y, Maassen N. Relationship between work load, pedal frequency,
521
and physical fitness. Int. J. Sports Med. 1984; 5(2):92-97.
522
40. Nielsen JS, Hansen EA, Sjøgaard G. Pedalling rate affects endurance performance
523
during high-intensity cycling. Eur. J. Appl. Physiol. 2004; 92(1-2):114-120.
524
41. Borg G. Perceived exertion as an indicator of somatic stress. Scand. J. Rehab. Med.
525
1970; 2(2):92-98.
526
42. Löllgen H, Graham T, Sjogaard G. Muscle metabolites, force, and perceived exertion
527
bicycling at varying pedal rates. Med. Sci. Sports Exerc. 1980; 12(5):345-351.
528
43. Lamberts RP, Swart J, Noakes TD, Lambert MI. A novel submaximal cycle test to
529
monitor fatigue and predict cycling performance. Br. J. Sports Med. 2011; 45(10):797-
530
804.
531
44. Bertucci WM, Betik AC, Duc S, Grappe F. Gross efficiency and cycling economy are
532
higher in the field as compared with on an Axiom stationary ergometer. J. Appl.
533
Biomech. 2012; 28(6):636-644.
534
45. Emauele U, Horn T, Denoth J. The Relationship Between Freely Chosen Cadence and
535
Optimal Cadence in Cycling. Int. J. Sports Physiol. Perform. 2012; 7(4):375-381.
536
46. Leirdal S, Ettema G. The relationship between cadence, pedalling technique and gross
537
efficiency in cycling. Eur. J. Appl. Physiol. 2011; 111(12):2885-2893.
538
47. Bini RR, Diefenthaeler F. Kinetics and kinematics analysis of incremental cycling to
539
exhaustion. Sports Biomech. 2010; 9(4):223-235.
540
48. Leirdal S, Ettema G. Pedaling technique and energy cost in cycling. Med. Sci. Sports
541
Exerc. 2011; 43(4):701-705.
542
23
49. Hug F, Turpin NA, Guével A, Dorel S. Is interindividual variability of EMG patterns
543
in trained cyclists related to different muscle synergies? J. Appl. Physiol. 2010;
544
108(6):1727-1736.
545
50. Paton CD. Effects of Shoe Cleat Position on Physiology and Performance of
546
Competitive Cyclists. Int. J. Sports Physiol. Perform. 2009; 4(4):517-523.
547
548
549
24
Figure captions
550
Fig. 1
551
Cadence is presented as a function of power output. Data constitute a representative selection
552
of average published values of freely chosen cadences for cyclists during cycling at a range of
553
power outputs on ergometer, treadmill, or road. Data are taken from a number of previously
554
published articles.18, 43-50 The figure illustrates that the freely chosen cadence increases with
555
increasing power output. This means that the term “low cadences” in the present article
556
represents cadences below a value, which is not set but rather increases with power output as
557
the freely chosen cadence.
558
559
25
560
Table 1
561
Key information from the 7 articles, which were selected based on the systematic literature search.
562
563
Article
Participants
Duration of
training period
etc.
Training regimens
Major outcomes
Paton
et al.
(2009)9
Competitive male
cyclists with a
minimum of three
years competitive
experience (n=18)
participated.
Average values of
age, body mass,
VO2max, and
maximal
incremental power
output were 25.9
years, 81.2 kg, 4.46 l
min-1, and 388 W,
respectively.
The training
intervention
lasted 4 weeks.
The study was
performed in
the main
competitive
phase of the
year, during
which all
cyclists were
competing in
endurance (>60
min) road or
mountain bike
races at least
once per week.
The cyclists had
not participated
in any gym-
based strength
training in the 3
months before
the study.
The cyclists were divided (matched pairs) into 2
groups. One group performed training at low
cadence. Another group performed training at high
cadence. Both groups continued their usual
competition program, but replaced part of their
usual training with 30-min laboratory training
sessions performed on an ergometer. The sessions
were performed twice per week and consisted of 3
sets of maximal effort single-leg jumps alternating
with 3 sets of maximal intensity cycling efforts.
The jump part of the training required subjects to
perform 20 explosive step-ups off of a 40-cm box.
The jump efforts were completed for the right and
then left legs consecutively over a 2-min period.
The cycling part required the cyclist to complete
5×30-s maximal intensity cycling efforts at 500-
520 W at a cadence of either 60-70 rpm or 110-120
rpm with 30-s recovery between repetitions. A
transition period of 2 min separated each cycle and
jump set. During the study period, the cyclists spent
approx. 10-15 h wk-1 of training and competing.
The low cadence group, as
compared to the high cadence group,
displayed on average 3.6 percentage
points larger increase in peak power
in an incremental test. The better
training response in the low cadence
group was termed “likely
beneficial”. In line with that, the low
cadence group displayed on average
7 percentage points larger increase
in power output at 4 mM blood
lactate, which was termed “very
likely beneficial”. Further, VO2max
improved on average 3.3 percentage
points more, which was termed
“likely beneficial”. Finally, exercise
economy at 50% of the maximal
incremental peak power output
improved on average 5.1 percentage
points more, which was termed
“likely beneficial”.
Kon-
inckx
et al.
(2010)8
Trained male
cyclists (n=20) with
an average of 6
years of experience
and 7,111 covered
km per year during
those years. Average
values of age,
height, body mass,
VO2max, and
maximal
incremental power
output were 27
years, 1.82 m, 74.0
kg, 60 ml kg-1 min-1,
and 305 W,
respectively.
The training
intervention
lasted 12 weeks.
The study was
performed in
the off-season
following a 3-
week period rest
period. The
cyclists on
average trained
endurance for 4-
5 h week-1 in
addition to the
specific training
intervention.
The cyclists had
no prior
experience with
the specific
training.
The cyclists were divided (matched pairs) into 2
groups. One group performed conventional strength
training (3 sets of parallel half-squat and leg press
exercises at 15RM-8RM) for leg extensor muscles.
Another group performed maximal effort isokinetic
cycling in 4-8 bouts of each 12 crank revolutions at
775-875 W at 80 rpm. Recovery between bouts was
3 min. Isokinetic cycling was performed on a
custom-built ergometer. Both groups performed the
specific training twice per week.
The two different training groups
increased average power output by a
similar magnitude (5-8%) in a 30-
min endurance performance test. In
line with that, lactate threshold
power output and peak power output
obtained in an incremental test were
increased to similar extents for both
groups. The isokinetic (low
cadence) group did not increase
maximal power output during
isokinetic cycling at 120 rpm. That
was in contrast to the strength
training group.
Nimme-
richter
et al.
(2012)13
Trained male
cyclists (n=18) with
a training history of
at least 5 years
participated.
Average values of
age, height, body
mass, VO2max, and
maximal
incremental power
output were 31
years, 1.79 m, 72.6
kg, and 58.4 ml kg-1
min-1, and 392 W,
respectively.
The training
intervention
lasted 4 weeks.
The cyclists on
average trained
for 12 h week-1
during 12 weeks
prior to the
study.
The cyclists were randomised into 3 groups. One
group performed uphill interval training at a low
cadence (60 rpm). Another group performed level-
ground interval training at a high cadence (100
rpm). The training was performed in two interval-
training sessions per week. Intervals were
performed as 6×5 min at 300 W, with 5 min
recovery periods at 90-150 W. A third group acted
as controls and continued their steady training,
without performing any intervals. Training was
performed on the road. During the study period, the
cyclists spent approx. 7-16 h wk-1 of training.
The low cadence group increased
power output during both an uphill
time trial (4.4±5.3%) and a flat time
trial (1.5±4.5%). For comparison,
the changes were -1.3±3.6 and
2.6±6.0%, respectively, for the high
cadence group. For the third group,
the changes were 4.0±4.6% and -
3.5±5.4%, respectively. Time trials
were performed on the road and
lasted approx. 20 min. The authors
stated that these findings suggest
that higher forces during the low-
cadence intervals are potentially
beneficial for performance
enhancement.
26
Article
Participants
Duration of
training period
etc.
Training regimens
Major outcomes
Kristof-
fersen
et al.
(2014)7
Well-trained male
veteran cyclists
(n=22) participated.
Average values of
age, body mass,
VO2max, and
maximal
incremental power
output were 47
years, 78 kg, 57.9
ml kg-1 min-1, and
402 W, respectively.
The training
intervention
lasted 12 weeks.
The study was
performed post
season.
The cyclists were randomly assigned into 2 groups.
One group performed training at low cadence (40
rpm). Another group performed training at freely
chosen cadence. The low cadence group performed
interval training as group sessions on spinning
bikes, twice a week, in addition to their usual
training. Intervals were performed as 5×6 min at
73-82% of the maximal heart rate, with 3 min
active recovery periods at low intensity (60-72% of
maximal heart rate) and freely chosen cadence. In
total, the low cadence group added 60 min week-1
of training to the usual training. The freely chosen
cadence group added 90 min of cycling at moderate
intensity (73-82% of maximal heart rate) and freely
chosen cadence to their usual training. In total, the
cyclists in the low cadence training group on
average completed 91h of training during the 12
weeks, while the cyclists in the freely chosen
cadence group completed on average 88 h of
training.
The low cadence group did not
improve VO2max, maximal
incremental power output, average
power output in a 30-min cycling
trial, and leg strength. For
comparison, the freely chosen
cadence group seemed to adapt in a
beneficial way with respect to
physiological adaptations (e.g.
increased VO2max) and performance
(e.g. increased average power output
in a 30 min cycling trial). Tests were
performed at freely chosen cadence.
Hirano
et al.
(2015)5
Healthy, sedentary
males (n=16)
participated.
Average values of
age, height, body
mass, VO2max, and
maximal
incremental power
output were 23.4
years, 1.71 m, 64.3
kg, 46.1 ml kg-1 min-
1, and 258 W,
respectively.
The training
intervention
lasted 2 weeks.
The participants were randomly assigned into 2
groups. One group performed training at low
cadence (35 rpm). Another group performed
training at high cadence (75 rpm). Sixty-min
training sessions were performed 5 times per week.
Training was conducted on an ergometer at a power
output at lactate threshold (corresponding to on
average approximately 63 W).
The low cadence group displayed
increased power output during
cycling at 50 rpm at the lactate
threshold after the training period.
The same did not occur for the high
cadence group.
Ludyga
et al.
(2016)10
Male and female
cyclists (n=36)
participated.
Average values of
age, height, body
mass, VO2max, and
maximal
incremental power
output were: 27
years, 1.77 m, 71.8
kg, 52.6 ml kg-1 min-
1, and 310 W,
respectively.
The training
intervention
lasted 4 weeks.
The cyclists had
trained at least 4
h week-1 of
cycling during
the 6 months
before the
study.
The cyclists were randomly assigned into 3 groups.
All groups performed 4 h week-1 of cycling
training. The average intensity of the training was
70-80% of the heart rate at the lactate threshold.
One group performed training at low cadence (60
rpm). Another group performed training at high
cadence (120-140 rpm). A third group acted as
controls and performed basic endurance training.
The training of the low and high cadence groups
consisted of 4 weekly 60-min sessions. Two of the
sessions were completed at constant load, while
two other sessions contained 6-8 intervals of 3-min
bouts at high intensity at 60 rpm or 120-140 rpm.
Recovery between bouts consisted of 3 min of
pedalling. Training was performed on an
ergometer.
In contrast to the control group, the
low and high cadence groups
attained similar improvements of
VO2max and power at the lactate
threshold. Besides, there was a
reduction of alpha-, beta-, and
overall-power spectral density in the
high cadence group, which was
more pronounced at high cadences.
Improvements of variables
associated with endurance
performance were correlated with
reductions of EEG spectral power at
90 and 120 rpm.
Whitty
et al.
(2016)6
Endurance-trained
male cyclists (n=16)
participated.
Average values of
age, body mass,
VO2max, and
maximal
incremental power
output were 31.4
years, 74.6 kg, and
4.85 l min-1, and 364
W, respectively.
The training
intervention
lasted 6 weeks.
The cyclists were randomly assigned into 2 groups.
One group performed training at low cadence (20%
below their freely chosen cadence, corresponding
to 60-81 rpm). Another group performed training at
high cadence (20% above their freely chosen
cadence, corresponding to 96-121 rpm). Three
interval sessions were completed each week. Each
session lasted 45-60 min and included 4-6×4 min
bouts at 70% of the maximal incremental power
output corresponding to on average approximately
255 W. Recovery between bouts consisted of 2 min
of pedalling at 100 W. The training described
above replaced a part of the usual training and was
performed on an ergometer. The cyclists trained
>300 km wk-1.
The low cadence group, as
compared to the high cadence group,
displayed a larger increase (on
average 16% vs. 8%) of the average
power output in a 15-min time trial
performed at freely chosen cadence.
The high cadence group increased
the freely chosen cadence from on
average 92 to 101 rpm during
submaximal cycling at 60% of the
maximal incremental power output.
For comparison, no change occurred
in the low cadence group. Both
groups increased the VO2max and
maximal incremental power output
during the training period. The high
cadence group displayed an
increased gross efficiency at 90 and
110 rpm.
Table 1 (Continued)
564
27
565
566
567
Fig. 1
568
569
0
20
40
60
80
100
120
140
0 200 400 600 800 1000 1200
Cadence (rpm)
Power output (W)
Freely chosen
cadence
Low cadence
