METHODS

Subjects

Twenty four college aged volunteers (12 men, 12 women) served as subjects. None had been trained using either the arm or cycle ergometer. Twelve (seven men, five women) were tested on a cycle ergometer, and the remaining 12 (five men, seven women) were tested on an arm ergometer. All subjects were healthy and free from any orthopaedic injury at the time of the investigation. They were informed of the purpose of the experiment and gave their written consent to participate. All experimental procedures were evaluated and approved by the Institutional Review Board for Human Subjects Experimentation at The College of New Jersey. Table 1 presents the physical and physiological characteristics of the subjects.

Experimental design

Subjects within each group completed a peak oxygen uptake (Vo₂peak) test and three submaximal exercise trials on four separate laboratory visits. All subjects were instructed to have eaten a minimum of three hours before each laboratory visit. The Vo₂peak test was always conducted before the three submaximal exercise trials. A minimum of 48 hours separated each visit, and all four visits were completed within a three week period. The three submaximal exercise trials were conducted at the same time of day and were presented in a randomised and counterbalanced order. During each submaximal exercise trial, subjects underwent a 10 minute arm or leg ergometry test during which both metabolic and cardiorespiratory variables were measured once the steady state had been achieved.

Exercise trials

Equipment set up

The upper body exercise (UE) was conducted on a modified Monark cycle ergometer (Monark 864, Monark Exercise AB, Varberg, Sweden). The sprocket of the ergometer was adjusted so that the crank axle was at shoulder level. Selection of this leg ergometer for arm use was to equate the size of flywheel so that the flywheel per se would not confound the results.¹⁴ The lower body exercise (LE) was conducted on a standard Monark cycle ergometer (Monark 828E). For both UE and LE, subjects were positioned so that their respective limbs were extended but not locked when the pedal was at its farthest end from the subjects.

Vo₂peak tests

During the arm and leg Vo₂peak tests, subjects performed a continuous incremental protocol that was similar to those reported previously.¹⁵ The protocol for UE was designed by Sawka et al,¹⁷ and that for LE was developed by Åstrand.¹⁸ The test was preceded by a 10 minute warm up period. The initial power output was 25 W for the arm test and 50 W for the leg test. For both tests, power output was increased by 25 W every two minutes. The pedal rate was set at 50 rev/min for both tests, with the cadence guided by a metronome. All subjects were verbally encouraged to continue exercise until volitional exhaustion. A plateau in Vo₂ was not generally observed in the arm tests. This was also observed in our previous study,¹⁵ the reason being that the muscle mass involved is small and muscular fatigue normally occurs before the cardiorespiratory system reaches its peak. In this case, the termination criterion is simply when the subject fails to maintain a target crank frequency for 15 seconds. However, the plateau as well as RER greater than 1.1 were shown in all of the leg tests. During both arm and leg tests, Vo₂ and heart rate (HR) were recorded every 20 seconds, and Vo₂peak was determined by averaging the two consecutive highest measurements.

Submaximal exercise trials

During each submaximal exercise trial, subjects performed a 10 minute steady state exercise at 40, 60, or 80 rev/min. The target pedal rate for each trial was maintained by using a metronome. The power output was set at 50 W for both types of exercise; this intensity was chosen to ensure that all subjects were able to achieve a steady state especially during the arm exercise. Beginning at the 8th minute of each exercise, Vo₂, carbon dioxide production (Vco₂), expiratory ventilation (Ve), RER, and HR were recorded at 20 second intervals, and the data then averaged over the next two minute period. Measurements of these variables were also taken at rest before the exercise during the first visit to the laboratory.

Measurements

Vo₂, Vco₂, Ve, and RER were determined using a two way T shape breathing valve (2700 series; Hans Rudolph, Inc, Kansas City, Missouri, USA) and an open circuit respiratory metabolic system (Metabolic Measurement Cart 2900; SensorMedics, Inc, Yorba Linda, California, USA). HR was measured using a wireless HR monitor (Pacer; Polar CIC, Inc, Port Washington, New York, USA), which updated values every five seconds. All trials were conducted in the laboratory where the mean barometric pressure and laboratory temperature were 755 (2) mm Hg and 23.2 (1.2)°C respectively.

Calculations

Energy expenditure during exercise was calculated from the steady state Vo₂ adjusted for substrate metabolism using the RER.¹⁶ The rates of carbohydrate and fat oxidation were calculated from Vo₂ and RER based on the assumption that protein breakdown contributes little to energy metabolism during exercise.

Statistical analysis

Statistical analyses were performed separately for each mode of exercise. Within each mode of ergometry, a one way analysis of variance with repeated measures was used to compare all dependent variables including those measured and calculated across three levels of pedal frequency. A significant F ratio was followed by a post hoc analysis using the Sheffé procedure. All statistical analyses were performed using SPSS software (SPSS Inc, Chicago, Illinois, USA).

RESULTS

Upper body exercise

During UE, no differences in Vo₂, HR, Ve, %Vo₂peak, and %HRpeak were observed between 40 and 60 rev/min (table 2). However, all these variables were higher (p<0.05) at 80 rev/min than at either 40 or 60 rev/min. This greater physiological response at the higher pedal rate was not observed for RER. In fact, there was a trend toward a higher RER at 40 and 60 rev/min than at 80 rev/min.

The total energy expenditure was higher (p<0.05) at 80 rev/min than at 40 rev/min, whereas no difference was found between 40 and 60 rev/min and between 60 and 80 rev/min (fig 1). However, both carbohydrate and fat oxidation remained the same across all three pedal rates.

Lower body exercise

During LE, Vo₂, HR, Ve, %Vo₂peak, and %HRpeak were higher (p<0.05) at 80 rev/min than at either 40 or 60 rev/min. These changes were similar to those observed during UE, except that RER at 40 rev/min remained marginally the lowest among the three pedal rates (table 2).

The total energy expenditure was also higher (p<0.05) at 80 rev/min than at 40 rev/min, whereas no difference was found between 40 and 60 rev/min and between 60 and 80 rev/min (fig 2). Unlike that observed during UE, carbohydrate oxidation was also higher (p<0.05) at 80 rev/min than at 40 rev/min, while no differences in fat oxidation were observed among the three pedal rates.

DISCUSSION

Pedal frequency as a mechanical constituent of power output reflects the velocity at which movement is taking place. Previous studies have shown the effect of this speed component on physiological processes while power output is held constant. However, many of these studies have either focused on movement economy or athletic performance. With the advance of technology, there has been increased use of exercise apparatus equipped with a servo mechanism which enables the subject to choose a speed without changing workload in both recreational and clinical settings. Although this feature provides exercisers with additional options of conducting a workout, it raises questions about how speed of movement can affect energy metabolism and what would be the optimum speed for energy expenditure. We therefore examined whether and how contraction frequency affects energy expenditure as well as carbohydrate and fat oxidation during both UE and LE. This was accomplished by having subjects pedal on either an arm or a leg ergometer at a constant power output but various frequencies while their metabolic responses were measured using indirect calorimetry.

We found a greater energy expenditure at 80 rev/min than at 40 rev/min during both UE and LE despite an unchanging power output and a concomitant reduction in brake resistance. This speed related increase in total energy expended during both UE and LE can be ascribed to a corresponding increase in muscular demand elicited by an increased rate of muscle contraction. It has been suggested that, as the speed of contraction increases, an increase in internal muscular work is required to overcome the augmented intramuscular friction.² It has also been speculated that, as pedal frequency increases, the recruitment of additional motor units increases despite an unchanging workload.^4,10 This may be supported by the study of MacIntosh et al,¹⁹ who found increased electromyographic activity in the gastrocnemius muscle as pedal frequency increased from 40 to 80 rev/min during cycling at a constant power output. Power output is the product of pedal frequency and brake resistance. From the present findings, it appears that the velocity at which muscle contracts is a more important determinant of energy demand of a given activity, and this relation between velocity and energy expenditure remains similar across different exercise modalities or relative exercise intensities.

Whereas greater energy expenditure was seen at 80 rev/min during both UE and LE, the energy distribution between carbohydrate and fat oxidation differed between the two modes of exercise. During UE, as the total energy expenditure was from carbohydrate and fat oxidation, both these subcomponents remained the same across the three pedal rates. However, during LE, the increase in energy expenditure was accompanied by a corresponding increase in carbohydrate oxidation, whereas fat oxidation remained essentially unchanged. This increase in the amount of energy derived from carbohydrate during LE may simply be the consequence of an increase in Vo₂ or exercise intensity as a result of an increase in pedal frequency. This finding is consistent with the observation by Hoelting et al²⁰ of a reduction in leg blood flow during knee extension at 80 contractions per minute compared with 40 contractions per minute. This reduced circulation in the exercising limbs may force the muscle to depend more on its own glycogen to yield energy. The authors of the previous studies, which used much higher power outputs, argued that most of the additional motor units recruited at higher pedal frequencies probably consist primarily of fast twitch muscle fibres, which are more carbohydrate dependent.^4,10 This, however, does not explain why more energy was derived from carbohydrate, as the intensity used during LE was only about 40% Vo₂peak and did not reach the threshold for mobilising much of the fast twitch muscle fibre of the exercising muscle.²¹

Failure to observe a similar increase in carbohydrate oxidation at 80 rev/min during UE may be due to the higher RERs observed at 40 and 60 rev/min. In other words, when the rate of carbohydrate utilisation was calculated, the lower level of energy expenditure achieved at these lower pedal rates may have been offset by the higher percentage of energy derived from carbohydrate, as determined from the RER. The lower RERs found at 80 rev/min were somewhat unexpected as it was our hypothesis that this speed would be associated with an increase in recruitment of fast twitch muscle fibres and thus utilisation of carbohydrate. The reason why RER responded differently during UE is not readily apparent. The relative intensity achieved during UE corresponded to nearly 70% of mode specific Vo₂peak, and this percentage was much higher than that during LE. As higher RERs at lower pedal rates were also observed during cycle exercise by Hagan et al,¹⁰ who used a much higher workload, it is tempting to speculate that such reversed RER responses during UE may be because the relative intensity experienced was much greater during UE than LE. A lower pedal rate is associated with a greater brake resistance, which dictates the level of muscular tension developed. It is possible that, when exercise is performed at a high intensity, an increase in muscular tension stimulates carbohydrate utilisation to a similar or even greater extent than an increase in pedal frequency. This contention is underscored by Ahlquist et al,¹¹ who showed greater glycogen utilisation of type II muscle fibres after cycle exercise of 85% Vo₂max at 50 than at 100 rev/min.

In conclusion, we found that pedalling at a greater frequency helped to maximise energy expenditure during either UE or LE despite an unchanging power output and a concomitant reduction in brake resistance. This increase in energy expenditure was brought about by a corresponding increase in carbohydrate oxidation during LE when the intensity was rather low. The fact that no differences in either carbohydrate or fat oxidation were observed for the three pedal rates during UE when the intensity was higher suggests that these speed induced changes in carbohydrate utilisation seen during LE can be altered by exercise modality or relative exercise intensity.