A Biomechanical Analysis of the Effect of Pedal Width on the Thigh Muscle Activity During

A biomechanical analysis of the effect of pedal width on the thigh muscle activity during a cyclist’s pedalling motion

Henrik Sjöstedt1, Tobias Lestrell1, Lars Hanson1, Ingrid Svensson2,

Åsa Segerström3and Gert-Åke Hansson4

1 Ergonomics, Department of Design Sciences, Lund University, Sweden

2 Solid mechanics, Department of Construction Sciences, Lund University, Sweden

3Physiotherapy, Department of Health Sciences, Lund University, Sweden

4 Ergonomics, Occupational and Environmental Medicine, Lund University, Sweden

Abstract

Numerous studies have been done optimising the seat and foot pedal positions. These studies have all focussed on variables in either the vertical or horizontal, inline with the bike frame, directions, but there exists a third dimension that has been ignored; the horizontal direction perpendicular to the bike frame. The aim of the study is to investigate the influence of pedal width (Q-factor) on muscle activity in the thigh. A bicycle was mounted on a bicycle trainer. Ten test subjects with a hip width span ranging from 135 mm to 184 mm were asked to, with a fixed cadence and power out, bike with four different Q-factors ranging from 140-340 mm. Muscle activity was measured on six muscles with surface electromyography. The only muscle activity significantly affected by the relation of hip breadth and pedal position was gluteus medius. The trend was that the muscles tend to work less than pedals move further out from the hip joint until an extreme outward position there the muscle activity increase again. The result hint that, at least on an individual level, muscle activity could well be significantly reduced by tailoring the pedal width to a unique rider. On a general level the results at least give grounds for further, in-depth study of the topic.

1. Introduction

Bicycles are constructed to efficiently produce forward motion through a mechanical input by the rider and through the years, many developments have been made to refine the basic geometry. Numerous studies have been done on the subject of optimising the seat and foot pedal positions, for example Gonzalez & Hull [1] and Mimmi et al. [2], to produce the highest output for a given input. These studies have all focussed on variables in either the vertical or horizontal, inline with the bike frame, directions, but there exists a third dimension that has been ignored; the horizontal direction perpendicular to the bike frame. Studying the distance in between the pedals unlocks this third dimension. The general trend with bicycle manufacturers is to minimise the lateral distance between the outside of the crank arms, also known as the Q-factor, in order to minimise lateral flex in the bottom bracket and crank arm area. This philosophy reaches its zenith on track racing bicycles, where the Q-factor can be as low as 128 mm [3], on road racing bikes it’s around 135-145, and on mountain bikes it’s generally even higher, typically over 165mm to make space for wider tires and a third chain ring. In recent years the bottom bracket and crank area have undergone major design developments within the racing community, away from the old square-tapered axle mould, and their general stiffness has increased greatly. The largest factors in this development are the move towards oversized hollow axles combined with moving the bearing casings outside the frame [4], but also through hollow crank arms made from carbon fibre. This new material’s low mass and rigidity makes it possible to experiment with the Q-factor without significantly compromising stability. Still the general trend is to keep the Q-factor as low as possible because it's thought to produce the most natural position, mimicking that of inline gaits. It is believed that the muscles would naturally be optimized for this, since walking or running is what humans have done throughout evolution [5].

The hypothesis states that eliminating the horizontal force component of pedalling a bicycle, caused by angling the legs inward to accommodate narrow Q-factors, will reduce the overall muscle activity in the thigh, and the aim of the study is to investigate this hypothesis by conducting tests on subjects to measure the influence of pedal width on muscle activity in the buttock and thigh.

2. Method

2.1 Equipment

The bicycle used during the project was a road racing model, with a frame size of 57cm, a size that could accommodate all the test subjects, by adjusting the saddle height. The bicycle featured clipless pedals [6], which require a special cycling shoe with a stiff sole and a snap binding to affix the shoe to the pedal. The binding ensures the foot stays in the optimal position on the pedal whilst allowing a slight rotational float. Clipless pedals of the SPD (Shimano Pedalling Dynamics) type were fitted to the bike.

A trainer rig is mounted on the rear of the bicycle, and lifts the rear wheel so that it has no contact with the ground and the bicycle won’t move during pedalling. The rear wheel spins a flywheel that can be regulated to give resistance to its motion, simulating the rolling resistance experienced on a real bike. Using a simple computer, power output generated at the rear wheel, the virtual speed of the bicycle and other properties can be displayed to the rider. The computer can also be set to give a fixed power output, an objective achieved by varying the resistance of the flywheel according to the rotational speed of the rear wheel. The bicycle was mounted on the bicycle trainer and calibrated. The trip computer was configured to ensure a constant output from the rider, 200W for the men and 150 W for the women. It was decided through testing that these values were high enough for the muscles to give discernable values, but low enough that most subjects were not fatigued during the procedure

Axles were produced that were especially fitted so that they could be inserted between the pedals and the crank arms. It was of critical importance to the data that the pedals were pushed wider apart than was practical for the riders, so it was decided to produce four sets of axles 25mm, 50mm, 75mm and 100mm long each. The 100mm axles extend the q-factor by 200mm, to 340mm, which was assumed to be significantly wider than the optimum width for most subjects that would be tested.

Surface electromyography (EMG) was used to record muscle activity. The muscles selected for study are all located above the knee, it having been observed that muscles below the knee were unaffected by the varying thigh incidence angles measured in the test. The muscles selected were: Gluteus Maximus, Gluteus Medius, Vastus Lateralis, Vastus Medialis, Semitendinosus and Biceps Femoris. Other studies that have focussed on bicycle ergonomics targeted these selected muscles along with others discarded for this project by the above criterion [1] [2]. The muscle activities were sampled in loggers with a sampling frequency of 1024 Hz.

2.2 Subjects

Ten test subjects, two female and eight male were acquired through the process of advertising with flyers around the university campus. This led to a fairly mixed group of individuals, comprising both sexes and amateur enthusiast cyclists as well as subjects who used bikes mainly for transport. The bike use frequency range from once a month to everyday and the distance range from under 10 km a week to above 300 km a week. The average hip width of the test subjects was 164 mm (s.d.=13 mm). The hip width covered a span of 49 mm ranging from 135 mm to 184 mm.

2.3 Procedure

The following procedure was used:

• The subjects were briefed on the purpose of the study and the procedure that would be used to perform the tests.
• The subject’s body measurements were taken. Hip width was measured between the two anterior superior iliac spines, after it was decided that this distance is related to the span between the femoral heads. The inside leg measurement was also taken.
• The electrode sites on the muscles were prepared with sandpaper and alcohol to remove dead skin cells and fats from the skin surface. The SEMG electrodes were then affixed to positions suggested by SENIAM [7] on the selected muscles. Reference values of muscular rest and maximum voluntarily contraction was recorded for all six muscles.
• The saddle position was adjusted according to a quick method used by enthusiast cyclists [8], the height of the saddle, measured from the centre of the crank to the top of the saddle surface along the supporting tube of the frame, being put at 0.885*(inside leg measurement). The horizontal position was then set so that the surface of the knee cap was vertically in line with the centre of the pedal, with the pedal at it’s most forward position, 90 degrees from top dead centre.
• The cyclist was given time to warm up on the bicycle. The cyclist was told to choose a gear that gave him/her a comfortable cadence than producing the desired effect (150 W for woman and 200 W for man). The selected gear was kept during the remaining tests.
• The loggers were switched on.
• The tests were performed using a randomly generated sequence for different pedal widths. The different widths were achieved by inserting varying sizes of spacer axle between the pedal and crank arm. The SEMG values were recorded for two minutes at each of the five widths, the two minutes beginning once the subject had attained a speed at which the power output was comfortable to maintain. The first tested widths was repeated at the end of the test to check that the subjects not were fatigued and were still producing similar activity at both the start and end of the proceedings.
• The subjects were allowed to cool down. The loggers were switched off and the EMG equipment was removed and subject was asked to answer a questionnaire about pedal width. When completed the questionnaire subjects who desired took a shower and they were afterwards thanked and rewarded for participating.

2.4 Statistical Analysis

After the testing was complete the data on the memory cards was analysed using a software developed for detail analysis of long term EMG recording [9]. The results were presented as 50th and 90th percentiles– i.e. a value, relative to maximum voluntary muscle contraction, below which the muscle activity was recorded for 50% or 90% of the two minute period for each width.The EMG data was further normalised to the highest value in the 90th percentile range, and the pedal width was then measured against the hip width of the subject for each set of results. The normalising against the 90th percentile values was done in order to eliminate extreme values which could otherwise wrongly influence the results.

The results were then tested to display significant trends (α=0.05). This procedure involved gathering together results from similar hip width – pedal width relationships into groups, or clusters and ANOVA-tests were performed to identify the statistical properties and differences of the muscle activity values within each cluster. The tests also calculated the significance values for each cluster, showing if there was a meaningful correlation between the muscle activity and the widths.

3. Results

Table 1 shows that muscles activity in Gluteus Medius was significantly affected by the relation of hip breadth and pedal position. The trend is that the muscles tend to work less when pedals move further out from the hip joint until an extreme outward position where the muscle activity increase again.

Another interesting result of this experiment is the apparent differences in muscle activity levels depending on the pedal widths in individual test subjects. Eight out of ten subjects showed slight drops in activity from their initial narrowest position. These results are not significant, but do give indication of a slight trend that would seem to suggest that there is at least a difference in muscle activity between the different widths, which in turn could mean that the pedal width could be uniquely tailored to individual cyclists.

Six subjects expressed in questionnaire that they felt discomfort on one or more pedal set ups. All six subjects felt uncomfortable when biking with the highest Q-factor. The number of subjects feeling discomfort decreased with decreased Q-factor. One subject felt discomfort when biking without pedal extenders. Most natural feel subject when biking without any extenders (4 subjects) and when biking on 25 mm extenders (4 subjects).

Table 1. Muscle activity and significance level for the six muscles. Muscle activity is split in cluster for Gluteus Medius where increased cluster level corresponds to hip joint further out from pedal.

Muscle activity / Significance
level between cluster
Cluster / N / Mean / Std. Deviation / Minimum / Maximum
Vastus Medialis 50 / Total / 50 / 0.248 / 0.176 / 0.130 / 0.763 / 0.847
Vastus Medialis 90 / Total / 50 / 0.942 / 0.058 / 0.745 / 1.000 / 0.083
Vastus Lateralis 50 / Total / 50 / 0.204 / 0.045 / 0.147 / 0.306 / 0.343
Vastus Lateralis 90 / Total / 50 / 0.944 / 0.069 / 0.721 / 1.000 / 0.781
Biceps Femoris 50 / Total / 50 / 0.322 / 0.085 / 0.177 / 0.501 / 0.987
Biceps Femoris 90 / Total / 50 / 0.905 / 0.096 / 0.620 / 1.000 / 0.738
Semitendinosus 50 / Total / 50 / 0.237 / 0.113 / 0.097 / 0.534 / 0.694
Semitendinosus 90 / Total / 50 / 0.884 / 0.111 / 0.456 / 1.000 / 0.075
Gluteus Maximus 50 / Total / 45 / 0.205 / 0.051 / 0.121 / 0.324 / 0.800
Gluteus Maximus 90 / Total / 45 / 0.928 / 0.060 / 0.776 / 1.000 / 0.189
Gluteus Medius 50 / Total / 45 / 0.241 / 0.099 / 0.068 / 0.434 / 0.070
Gluteus Medius 90 / 1 / 9 / 0.988 / 0.029 / 0.913 / 1.000
2 / 13 / 0.720 / 0.236 / 0.261 / 1.000
3 / 15 / 0.622 / 0.286 / 0.162 / 1.000
4 / 7 / 0.605 / 0.297 / 0.301 / 1.000
5 / 1 / 0.979 / - / 0.979 / 0.979
Total / 45 / 0.729 / 0.273 / 0.162 / 1.000 / 0.008

4. Discussion

Since the selection of test subjects was done through advertising around the university, subjects comprised solely of volunteers. This way of selecting subjects may not reflect the general public as a whole, but given the time scale of the project it was found adequate and resulted in a fair spread of body measurements. Of the ten test subjects, three active male tri-athletes were included and two women, though given more time, more women would have been tested so that a comparison between the activity values could be made between men and women and wider range of hip width. Both women tested displayed noticeably different Gluteus Medius values from the men, whilst being similar to each other, and more testing could have confirmed if this was merely a fluke or a genuine trend.

The body measurements were subject to potential errors, because of difficulties finding the exact points of reference on numerous bodies. It is doubtful the standard relationship used to correlate iliac spine width to hip width holds for all subjects.

From the results, we can conclude that there is no significant correlation between the hip width and pedal width, when taking all six muscles into consideration, as can be seen by the lack of universal significance in the ANOVA-test results. However, when looking at individual muscles there are significant tendencies towards reciprocity in the 90th percentile Gluteus medius muscles, and a potential trend in the 50th percentile Gluteus medius, 90th percentile Semitendinosus and 90th percentile Vastus Medialis, as can be witnessed by their significance values of around 0.05 or below. The tables show that the mean value in the Gluteus Medius decreased by as much as 40% from the highest value cluster to the lowest.

The average hip joint width amongst the test subjects is 164mm which would, in theory, require a Q-factor of around 8 cm (the centre of pressure on the foot is at approx. 4cm from the outer edge of the crank arm) to produce the vertical line between the hip joint and pedals proposed in the hypothesis suggesting that it may not be the hip joint width that dictates the optimum pedal width. In that case it is worth continuing a study to determine exactly which measurement it is that does.

Observing the SEMG results from each individual show no significant traits, rather hints at a trend putting the lowest activity value at 25-50mm extension, in the amateur riders as well as the every day riders. If anything can be deduced from the questionnaire, it is that the sensations the test subjects felt during testing did not necessarily have anything to do with how much their muscles were working.

It must be mentioned that there can be negative aspects to increasing the Q-factor, such as potential loss off lateral stiffness, hampered cornering ability whilst pedalling because of the decreased ground clearance and last but not least, increased wind resistance due to the legs protruding ever more into undisturbed air. Of those factors the stiffness can be easily dealt with by use of exotic materials in racing bikes, but the other issues need further study and development to fully understand their implications

The basic hypothesis is proven to be in need of some slight adjustment, for instance that it is not feasible to have the knee and ankle joints in a vertical line from the hip joints. The results from the subject tests suggest the line should be drawn along the femur instead, making the centre of the trochanter in line with the knee and ankle.