Cardiopulmonary Responses of Middle-Aged Men Without Cardiopulmonary Disease to Steady-Rate Positive and Negative Work Performed on a Cycle Ergometer.

Key Words: Cardiopulmonary responses, Concentric contraction, Cycle ergometry, Eccentric contraction, Negative work, Older men, Positive work, Speed.

Although our knowledge of the physiological characteristics of negative work has grown since the work of Abbott and colleagues more than 45 years ago,[1] the literature on negative work is scant compared with the literature on positive work. Research on negative work has primarily focused on muscle mechanics,[2] the relationship of force to power generation and training effects[3-8] energy consumption and efficiency,[1,9-15] electromyographic (EMG) activity,[8,10,12,13,16] pulmonary and cardiovascular responses,[17-19] heat regulation,[15] muscle damage,[12,20-26] and delayed-onset muscle soreness.[27-31] Negative work is less physiologically demanding and less metabolically costly than positive work at the same power output. A dramatic example of the reduced metabolic cost of negative work was reported in 1960 by Hill.[32] He had a small woman resist the forward pedaling of a large athletic man. The woman performing negative work resisted the man's forward pedaling with ease, supporting Hill's conclusion that torque production during eccentric muscle contractions requires less total energy than during positive work. This observation has been confirmed for other forms of negative work, including descending stairs,[33] climbing down a laddermill,[34] and lowering a weight,[35] As a result of the reduced energy cost for a given workload, there is a corresponding reduction in heart rate (HR) and cardiac output during negative work.[4,15,36,37]

During negative work, there is a decrease in minute ventilation (VE) that parallels the decrease in oxygen consumption ([VO.sub.2]) obtained when these variables are compared with positive work across the same work rates.[17-19] Little is known, however, about the effects of negative work on the components of VE, namely, tidal volume (VT) and breathing frequency (fb). Dean and Ross[38] reported rapid shallow breathing in subjects without cardiopulmonary impairments during downhill walking on a treadmill at 3.5 mph with a -7% grade. They further observed that the fb and VT responses were not consistent with the notion that downhill walking is merely a low-intensity form of positive work. Dean and Ross[38] proposed that postural adjustments of the chest wall or restriction of abdominal wall motion, or both, during downhill walking may contribute to the observed rapid shallow breathing response.

In addition to the reduced energy cost of negative work compared with positive work, the energy cost of a given work rate may be affected by the pedaling frequency when a bicycle ergometer is used for exercise. Optimal values for walking speed and pedaling frequency have been reported.[17,33,39-41] Banister and Jackson[40] reported that a low work rate achieved with a high pedaling frequency and low resistance is metabolically equivalent to a much higher work rate achieved with a low pedaling frequency and high resistance. Although optimal pedaling frequencies for different work rates during positive work have been identified, such values for negative work are less clear.

The literature to date on both negative and positive work includes descriptions of energy cost at different work rates, but these descriptions are based primarily on the responses of young subjects without cardiopulmonary impairments.[8,9,13,42-45] Despite the well-known physiological consequences of aging[46-48] and age-related changes in response to conventional exercise during positive work, the responses of older adults to negative work are not known.

The purpose of our study was to compare the responses of middle-aged men (39-65 years of age) to negative and positive work performed on a cycle ergometer at different pedaling frequencies. We studied an older age group because a large proportion of individuals with chronic disabilities who are frequently seen in rehabilitation settings are in their middle to later years. We used a cycle ergometer to minimize postural adjustments that may affect breathing patterns when performing negative work, such as those that occur during downhill walking on a treadmill.[38] We hypothesized that negative work performed on a cycle ergometer by older men without cardiopulmonary impairments is not physiologically equivalent to low-intensity positive work when performed at the same work rate and that the physiological responses to negative work are physiologically distinct compared with physiological responses to positive work.

Method

Research Design

We used a 2 (positive and negative work) by 3 (35-, 55-, and 75-rpm pedaling frequencies) factorial design for repeated measures on both factors to examine differences between positive and negative work for the 3 pedaling frequencies.

Each subject was tested 6 times, once at each of the 3 pedaling frequencies for each of the 2 types of work. Power was held constant at 60 W. This work rate was selected in order to shed light on the physiologic responses to negative work at rates that would be achievable by patients with extremely low functional work capacity (ie, those individuals most likely to benefit from negative work). Subjects were randomly selected to perform positive or negative work on the first of 2 test days, which were scheduled 1 week apart. The order of the pedaling frequencies was also randomly selected. At the beginning of each test, each subject selected a card with the order of pedaling frequencies written on it. The exercise responses that were of particular interest were [VO.sub.2], HR, VE, VT, and fb.

Subjects

Twelve middle-aged men with no history of cardiopulmonary disease, based on a medical history questionnaire, volunteered to participate in the study. The subjects were aged 39 to 65 years ([bar] X = 49.7). No subject had participated in any formal exercise program or training over the past year. A detailed explanation of the research design and purpose of the study was given to each subject. Each subject signed an informed consent form.

Equipment and Measures

The subjects' height and weight were recorded, and their body mass index (BMI) was calculated (ie, weight [in kilograms]/height [in square meters]).

Cycle ergometer. The assembly of the bicycle ergometer is shown in Figure 1. All parts were bolted onto 2 connected rigid wooden frames with wheels. The wheels of the wooden frames were securely braked during testing to maintain the stability of the assembly.

[Figure 1 ILLUSTRATION OMITTED]

A sturdy chair with arm support was mounted onto the wooden frame on an adjustable linkage to allow optimal leg extension and comfort for each subject. The chair was fixed so that the subject could sit approximately level with the pedals with his legs were extended forward, rather than downward. A chair with arm support was used to minimize the contribution of sitting and the movements associated with balancing and body fixation to the total energy cost of cycling.[13,18] The ball of the subject's foot was positioned over the pedal axis.[10] Foot straps were applied in an effort to ensure that the subject maintained the required foot position throughout the test.

A standard cycle ergometer (Monark model 817 E*) was modified to perform positive and negative work according to the description by Bigland-Ritchie and colleagues.[49] The pedaling frequency was held constant by a constant current feedback control, the output of which was also displayed by a digital output gauge. This control allowed a pedaling frequency ranging from 0 to 120 rpm. The maximum allowable torque that could be registered at the pedals was 70 N for pedaling frequencies ranging from 40 to 120 rpm. At pedaling frequencies less than 40 rpm, the maximum allowable torque decreased in proportion to the decrease in pedaling frequency. The original friction belt braking system of the ergometer was retained for forward pedaling or positive work and for calibration. In this motor-driven cycle ergometer, the pedaling frequency is controlled by the motor and the load is controlled by the subject. A second set of pedaling frequency and torque output gauges were placed in front of the subject to provide direct visual feedback.

Both pedaling frequency and torque output gauges were calibrated prior to testing and recalibrated whenever there was a change in pedaling frequency or work setting. The pedaling frequency output gauge was calibrated by matching the pedaling frequency measured with a stopwatch to the pedaling frequency that was registered on the pedaling frequency output gauge. Subsequently, the torque reading was zeroed at each test speed. Torque calibration was done in the forward direction using the original friction band on the ergometer as reference. The friction band on the ergometer was tightened to a known resistive force, and the torque digital output gauge was calibrated to match the resistive force. Torque calibration for the reverse direction was tested using the original friction band and a known resistive force and was identical to torque calibration for the forward direction. Therefore, for a given pedaling frequency, the torque calibration performed in the forward direction also served to calibrate the reverse direction. These procedures constituted the calibration for each pedaling speed for both positive and negative work (ie, pedaling in the forward and backward directions).

Metabolic measurement cart. A Sensormedics Metabolic Measurement Cart (MMC)([dagger]) was used during exercise testing to perform breath-by-breath gas sampling. Subjects were connected to the MMC by a headpiece assembly and a mouthpiece. A noseclip was used in a attempt to avoid air leakage through the nose. Expired gas was then analyzed at 15-second intervals during testing.

Measures included [VO.sub.2], VE, VT, and fb. Before each test, the MMC was calibrated according to the operator-guided calibration procedures recommended by the manufacturer, using specific calibration gases. Heart rate and rhythm were continuously monitored using a 3-lead electrocardiographic (ECG) monitor,([double dagger]) which was calibrated before each test.

Other measures. For safety reasons, arterial oxygen saturation (Sa[O.sub.2]), blood pressure, and perceived exertion were monitored throughout the test sessions. Arterial oxygen saturation was measured using an oximeter([sections]) with an earlobe sensor. Blood pressure was measured manually using a brachial cuff and stethoscope at 1-minute intervals throughout each test by the same experienced individual. Subject reports of breathing difficulty were recorded every minute. A modified Borg Rating of Perceived Exertion Scale[50] was used to measure breathing difficulty. The scale ranged from 0 ("nothing at all") to 10 ("very, very strong"). Tests were carried out in a temperature-controlled exercise laboratory (21 [degrees] [+ or -] 2 [degrees] C).

General Procedures

Performance of positive work. Subjects were seated for forward cycling or positive work in the same manner as for negative work. The clutch of the motor mechanism was engaged in the forward position. An extra 10-N load was added to the predetermined resistance for each subject to prevent damage to the motor by inadvertently driving it above its set speed. A lap seat belt was fastened to stabilize the subject. The standard distance of the chair from the pedals was determined by the length of the fully extended lower extremity with the pedal positioned vertically and the pedal arm positioned horizontally. The distance of the seat from the pedals was then adjusted for subject comfort prior to the start of testing (no more than 10 degrees of knee flexion was allowed). Because the tension generated by a muscle is, in part, based on its resting length,[51] care was taken to ensure that each test was performed at a comparable seat position for each subject. The subject's feet were strapped into position on the pedals. The subject was encouraged to relax the upper body and trunk while allowing the lower extremities to cycle. After the test pedaling frequency was set by the tester, the subject pedaled forward to assist the cycle ergometer with sufficient effort so that the torque output fell to 10 N. The motor maintained the pedaling frequency set by the tester while the subject assisted the movement of the pedals until the desired torque reading was registered. The subject then maintained the same effort for the duration of the test.

Performance of negative work. The friction band was left slack while the motor drove the pedals in the reverse (backward) direction. At the designated pedaling frequency set by the tester, the subject was directed to resist the movement of the pedals until the desired torque reading was registered. A forced stretch, therefore, was imposed on the same muscles that were used to generate power during conventional cycling using concentric muscle contractions. The rate at which the muscles were "worked upon" was the power being transferred to the subject.[37] The subject was then told to maintain the required torque throughout the test.

Performance of free pedaling. For free pedaling prior to exercise with positive work, the seated subject placed his feet in the appropriate position on the pedals, and then the feet were strapped into place. He was instructed to relax while the motor drove the feet around at the desired pedaling frequency. The subject maintained foot contact with the pedals while passively allowing the pedals to carry the legs around forward. During the test session, the subject free pedaled at the test pedaling frequency during the warm-up and cool-down periods. The procedure was the same for free pedaling prior to exercise with negative work, except that the subject allowed the pedals to carry the legs around in the backward direction.

Practice sessions. All subjects attended at least 2 practice sessions (2 subjects required 3 practices) of positive and negative work. A subject was deemed to have learned the negative work cycling technique when he could maintain a given torque output for 1 minute over a range of torques. Our initial pilot work suggested that negative work cycling was more difficult to learn than positive work cycling. No major difficulty, however, was encountered with the practice and test sessions. Practice sessions were also used to familiarize the subjects with the testing environment, general procedures, and monitoring equipment. Practice sessions lasted an average of 25 minutes and were designed to promote a learning effect while minimizing any training effect.

Test protocol. After completion of the practice sessions, subjects completed the two test conditions (ie, positive and negative work) using the cycle ergometer. Tests were conducted at least 1 week apart.

Subjects were requested not to have a large meal at least 3 hours prior to testing and not to consume any substances that contained stimulants (eg, coffee, soft drinks with caffeine). They were asked to have a restful 24-hour period prior to testing and to wear comfortable attire and shoes when they visited the exercise laboratory.

When a subject arrived at the exercise laboratory on each test day, his height and weight were measured and pulmonary function testing was done, including 3 trials of forced expiratory maneuvers for calculation of forced vital capacity and forced expiratory volume in 1 second. These routine pulmonary function measurements were taken to rule out pulmonary dysfunction.

For each test, the subject relaxed while seated on the testing chair, which was adjusted for leg length and comfort, for at least 5 minutes. The ECG electrodes were attached. The subject was then connected to the MMC by means of the headpiece assembly and the mouthpiece, and a noseclip was applied. The subject inspired room air via a low-resistance, one-way valve. The subject was then again asked to relax in this comfortable position for another 3 minutes while resting metabolic measurements were taken. Baseline physiological measurements were usually taken in the second to third minute of the test when [VO.sub.2] and HR of the subject had stabilized, that is, deviation was less than 2%.

After warming up with free pedaling for 2 minutes, the subject pedaled at the assigned pedaling frequency (the first of the 3 randomized pedaling frequencies) and power output. After a steady-state was reached (usually within 3 minutes) data, were collected for 5 to 7 minutes before the test at the assigned pedaling frequency was terminated. The subject free pedaled during the cool-down period for several minutes. Vital signs were continuously monitored until they were within 15% of the baseline values. The subject then rested for 30 minutes prior to being tested at the next pedaling frequency. This procedure was continued until tests at all 3 pedaling frequencies were completed. This protocol was repeated on the second test day for the other type of work.

Data Analysis

Descriptive statistics for the 5 dependent variables for each of the 6 steady-state tests were calculated. Arterial saturation and blood pressure were recorded as a safety precaution, and the data were not included in the analysis of the present study.

A 2 x 3 (2 types of work and 3 pedaling frequencies) factorial design for repeated measures on both factors was used to analyze the data. An analysis of variance (ANOVA) for repeated measures was used for each of the 5 dependent variables. Multivariate ANOVAs for repeated measures were used to analyze the main effects for pedaling frequency and the interaction when the assumptions for ANOVA for repeated measures were violated, resulting in inflated probability values.[52] Newman-Keuls post hoc tests were used to test the effect of pedaling frequency when a significant omnibus F value was found.

Linear regression was used to determine the relationship between VT and VE, fb and VE, HR and [VO.sub.2], and VE and [VO.sub.2] during both positive and negative work. The regression lines for each of the 4 relationships were tested for homogeneity of the regression coefficients between positive and negative work. An alpha of less than .05 was used as the critical value for the statistical tests.

Results

Subject Characteristics

The BMIs of the subjects averaged 26.1 kg/[m.sup.2] and ranged from 21.8 to 31.5 kg/[m.sup.2]. Pulmonary function test results were within the normal range for each individual, based on the conventional standards of the American Thoracic Society.[53]

All exercise tests were performed without any untoward episodes. The average Sa[O.sub.2] was 97%, and this variable remained stable throughout exercise testing. The ECG and blood pressure measurements were within normal limits during rest, during exercise, and after exercise recovery for all subjects. The Borg subjective rating of breathing difficulty averaged one unit for both positive and negative work during the steady-state portions of the exercise tests. The baseline values preceding positive and negative work represent a composite mean of the 3 baseline periods for each test day.

Oxygen Consumption

Descriptive statistics (means [+ or -] standard deviations) for [VO.sub.2] are shown in Figure 2. During positive work, the [VO.sub.2] values were 1.04 [+ or -] 0.13, 1.12 [+ or -] 0.13, and 1.25 [+ or -] 0.13 L/min for pedaling frequencies of 35, 55, and 75 rpm, respectively. During negative work, the [VO.sub.2] values were 0.64 [+ or -] 0.17, 0.55 [+ or -] 0.14, and 0.6 [+ or -] 0.24 L/min for pedaling frequencies of 35, 55, and 75 rpm, respectively.

[Figure 2 ILLUSTRATION OMITTED]

The results of the ANOVA showed that [VO.sub.2] was lower during negative work than during positive work (P [is less than] .001), and there was an effect of pedaling frequency on [VO.sub.2] (P = .002). Post hoc tests revealed that [VO.sub.2] during positive work was higher at 75 rpm (1.25 [+ or -] 0.13 L/min) than at 35 rpm (1.04 [+ or -] 0.13 L/min) and 55 rpm (1.12 [+ or -] 0.13 L/min) (P [is less than] .05). During negative work, [VO.sub.2] was higher at 75 rpm (0.67 [+ or -] 0.24 L/min) than at 55 rpm (0.55 [+ or -] 0.14 L/min). There was also an interaction (P [is less than] .01) showing that [VO.sub.2] increased linearly during positive work and that [VO.sub.2] was lowest at 55 rpm (0.55 [+ or -] 0.14 L/min) during negative work.

Heart Rate

Descriptive statistics (means [+ or -] standard deviations) for HR are shown in Figure 3. During positive work, the HR values were 93.0 [+ or -] 9.4, 95.4 [+ or -] 10.5, and 99.2 [+ or -] 12.1 bpm for pedaling frequencies of 35, 55, and 75 rpm, respectively. For negative work, the HR values were 82.5 [+ or -] 11.5, 77.8 [+ or -] 14.0, and 85.1 [+ or -] 15.4 bpm for pedaling frequencies of 35, 55, and 75 rpm, respectively.

[Figure 3 ILLUSTRATION OMITTED]

The results of the ANOVA showed that HR was lower during negative work than during positive work (P [is less than] .001), and there was an effect of pedaling frequency on HR (P = .003). Post hoc tests revealed that HR during positive work was higher at 75 rpm (99.2 [+ or -] 12.1 bpm) than at 55 rpm (95.44 [+ or -] 10.5 bpm) and 35 rpm (93.04 [+ or -] 9.4 bpm) (P [is less than] .05). During negative work, HR was higher at 75 rpm (85.1 [+ or -] 15.4 bpm) than 55 rpm (77.8 [+ or -] 14.0 bpm). There was also an interaction (P [is less than] .05) showing that HR increased linearly during positive work and that HR was lowest at 55 rpm (77.8 [+ or -] 14.0 bpm) during negative work.

Minute Ventilation

Descriptive statistics (means [+ or -] standard deviations) for VE are shown in Figure 4. During positive work, the VE values were 24.6 [+ or -] 4.9, 26.2 [+ or -] 5.7, and 29.4 [+ or -] 6.0 L/min for pedaling frequencies of 35, 55, and 75 rpm, respectively. For negative work, the VE values were 16.0 [+ or -] 4.1, 14.8 [+ or -] 4.2, and 18.6 [+ or -] 7.2 L/min for pedaling frequencies of 35, 55, and 75 rpm, respectively.

[Figure 4 ILLUSTRATION OMITTED]

The results of the ANOVA showed that VE was lower during negative than during positive work (P [is less than] .001), and there was an effect of pedaling frequency on VE (P [is less than] .01). Post hoc tests revealed that VE during both positive and negative work was higher at 75 rpm (29.4 [+ or -] 6.0 L/min and 18.6 [+ or -] 7.2 L/min, respectively) than at 55 rpm (26.2 [+ or -] 5.7 L/min and 14.8 [+ or -] 4.2 L/min, respectively) and 35 rpm (24.6 [+ or -] 4.9 L/min and 16.0 [+ or -] 4.1 L/min, respectively) (P [is less than] .05). There was no interaction.

Tidal Volume

Descriptive statistics (means [+ or -] standard deviations) for VT are shown in Figure 5. During positive work, the VT values were 1.40 [+ or -] 0.31, 1.41 [+ or -] 0.30, and 1.57 [+ or -] 0.35 L/breath for pedaling frequencies of 35, 55, and 75 rpm, respectively. For negative work, the VT values were 0.93 [+ or -] 0.24, 0.93 [+ or -] 0.29, and 1.12 [+ or -] 0.41 L/breath for pedaling frequencies of 35, 55, and 75 rpm, respectively.

[Figure 5 ILLUSTRATION OMITTED]

The results of the ANOVA showed that VT was lower during negative than during positive work (P [is less than] .001), and there was an effect of pedaling frequency on VT (P [is less than] .03). Post hoc tests revealed that VT during both positive and negative work at 75 rpm was higher (1.57 [+ or -] 0.35 L/breath and 1.12 [+ or -] 0.41 L/breath, respectively) than at 55 rpm (1.41 [+ or -] 0.30 L/breath and 0.93 [+ or -] 0.29 L/breath, respectively) and 35 rpm (1.40 [+ or -] 0.31 L/breath and 0.93 [+ or -] 0.24 L/breath, respectively) (P [is less than] .05). There was no interaction.

Breathing Frequency

Descriptive statistics (means [+ or -] standard deviations) for fb are shown in Figure 6. During positive work, the fb values were 18 [+ or -] 4.1, 19 [+ or -] 3.6, and 19 [+ or -] 4.2 breaths/min for pedaling frequencies of 35, 55, and 75 rpm, respectively. For negative work, the fb values were 18 [+ or -] 6.3, 17 [+ or -] 4.7, and 18 [+ or -] 5.8 breaths/min for pedaling frequencies of 35, 55, and 75 rpm, respectively. The results of the ANOVA showed that there were no differences in fb between positive and negative work or among pedaling frequencies.

[Figure 6 ILLUSTRATION OMITTED]

Linear Regression Analysis

The relationship between HR and [VO.sub.2] during positive work is described by the equation HR = 62.1 + (29.8) [VO.sub.2], and the Pearson product-moment correlation coefficient (r) was .42 (P [is less than] .05). The HR and [VO.sub.2] relationship during negative work is described by the equation HR = 52.0+ (47.8) [VO.sub.2], and the Pearson r was .66 (P [is less than] .05). The slopes and intercepts of the regression lines between positive and negative work for the relationship of HR and [VO.sub.2] did not differ (P [is greater than] .05).

The relationship between VE and [VO.sub.2] during positive work is described by the equation 4E [+ or -] 8.86+ (31.4) [VO.sub.2], and the Pearson was .81 (P [is less than] .05). The V and V relationship during negative work is described by the equation VE = 1.24 + (24.4) [VO.sub.2], and the Pearson r was .85 (P [is less than] .05). The slopes of the regression lines between positive and negative work for the VE and [VO.sub.2] relationship did not differ, and the intercept was smaller during positive work (P [is less than] .05).

The relationship between VT and VE during positive work is described by the equation VT = 0.89 + (0.022) VE, and the Pearson r was .39 (P [is less than] .05). The VT and VE relationship during negative work is described by the equation VT = 0.51 + (0.030) VE, and the Pearson r was .49 (P [is less than] .05). The slopes and intercepts of the regression lines between positive and negative work for the VT and VE relationship did not differ.

The relationship between fb and VE during positive work is described by the equation fb = 9.62 + (0.34) VE, and the Pearson r was .51 (P [is less than] .05). The relationship between fb and VE during negative work is described by the equation fb = 11.08 + (0.393) VE, and the Pearson r was .39 (P [is less than] .05). The slopes and intercepts of the regression lines between positive and negative work for the fb and VE relationship did not differ.

Discussion

The 12 subjects who participated in this study represented a cross-section of middle-aged men without cardiopulmonary impairments, based on pulmonary function tests, BMI (although these values tended to be high; normal range = 20-25), ECGs, and the absence of resting and exercise-induced arterial desaturation. As predicted for these subjects, the constant work rate of only 60 W for both positive and negative work was generally associated with reports of minimal exertion.

Effects of Steady-Rate Cycling on Oxygen Consumption

Oxygen consumption during negative work was about 55% that for positive work at the constant work output of 60 W. There was more variation in [VO.sub.2] among subjects during negative work than during positive work. The lower [VO.sub.2] associated with negative work is well-established in the literature.[1,3,4,8-10,13,17,18] At the correspondingly low work rate of 60 W, these studies showed that the energy cost measured by [VO.sub.2] during negative work ranged from 45% to 65% of the energy cost measured by [VO.sub.2] during positive work.[1,10,13,17,18] The coefficients of variation (CVs) (standard variation divided by the mean) reported in the literature for [VO.sub.2] during negative work range from 1.4 to 4.4 times those reported for [VO.sub.2] during positive work.[10,13,37,43] In our study, the average CV for [VO.sub.2] during negative work was about 2.5 times that for [VO.sub.2] during positive work. The lower [VO.sub.2] is explained by the increased elastic energy generated during negative work compared with positive work.[1,3,4,8-10,13,17,18] The higher CV during negative work, however, has not been explained previously. It could reflect the novelty involved in this activity compared with more conventional forward pedaling during positive work.

Banister and Jackson[40] observed large variations in [VO.sub.2] when pedaling frequency varied during positive work on a cycle ergometer at a constant power output. A subsequent study by Gaesser and Brooks[54] further illustrated this point. As observed in our study, [VO.sub.2] increased with pedaling frequency during positive work, with [VO.sub.2] being greatest at the highest pedaling frequency of 75 rpm. Less is known about the effect of pedaling frequency on [VO.sub.2] during negative work performed at a constant work rate. Researchers[8,17] have reported a gentle rise in [VO.sub.2] with increasing work rate during negative work. Furthermore, Knuttgen et al[17,18] reported that lower [VO.sub.2], HR, and VE occurred at 60 rpm during negative work when compared with 20 and 100 rpm at a low work rate similar to that used in our study. This finding, however, was not explained. Our work needs to be extended to study in detail the [VO.sub.2] during negative work at intermediate pedaling frequencies such as 55 rpm. Even though [VO.sub.2] at 55 rpm was not different from [VO.sub.2] at 35 rpm, we believe further study is needed to determine whether a pedaling frequency of 55 rpm is more efficient than lower and higher pedaling frequencies using our model.

When VE and HR were plotted against [VO.sub.2] during both positive and negative work, linear relationships were found (P [is less than] .05). These results also showed that, at a work rate of 60 W, VE and HR during both positive and negative work increased correspondingly with [VO.sub.2]. The intercepts of the VE and [VO.sub.2] regression lines differed between positive and negative work (P [is less than] .05). Thomson[19] reported comparable VE when comparing positive and negative work using a cycle ergometer at the same [VO.sub.2]. Other researchers,[17,18,38] however, have reported that both VE and HR were higher during negative work than during positive work when compared at the same [VO.sub.2]. The VE and [VO.sub.2] relationship that we reported is similar to previously reported relationships.[17,18,38] Studying a wider range of exercise intensities would have enabled us to examine these relationships more thoroughly.

Effects of Steady-Rate Cycling on Heart Rate

In this study, the HR response followed the same trends as [VO.sub.2] during positive and negative work. Because [VO.sub.2] and HR are highly correlated,[55,56] this similarity could be expected. Although not statistically significantly lower, such a trend in HR measured at 55 rpm during negative work may reflect the lower [VO.sub.2]. From our observation, a cadence of 55 rpm, which approximates an intermediate speed that an individual would self-select, may be associated with less co-contraction, which may reduce peripheral vascular resistance, thus lowering the HR. Whether there is a true difference needs to be established in a study with a larger sample size.

Effects of Steady-Rate Cycling on Minute Ventilation and Its Components

Minute ventilation during both positive and negative work was not different at 35 and 55 rpm, but an increase in VE was observed at 75 rpm. Moreover, there was no interaction in this study reflecting a negligible difference in trend between positive and negative work. A lower VE at 55 rpm was also reported by Knuttgen et al.[17,18] This phenomenon was likely related to a lower metabolic demand at 55 rpm.

Changes in VE reflect changes in its components, VT and fb. In the steady-state component of this study, this fb was statistically the same for all 3 pedaling frequencies in both positive and negative work. Variation in VT, therefore, largely explained the change in VE. When VT and fb were plotted against VE during positive and negative work, however, positive linear relationships between VT and VE and between fb and VE were found (P [is less than] .05). The potential importance of these findings needs further clarification, given that the ranges for all 3 variables (ie, VE, VT, and fb) were relatively small. To illustrate, the mean difference for VE between positive and negative work was 15 L/min and the mean difference for fb was only 2 breaths/min. In addition, the mean increase in VT during exercise ranged from 18% to 32% of mean force vital capacity. Thus, the change in VT could explain the increase in VE in this ventilatory range. Furthermore, the lowest mean VE, observed at 55 rpm during negative work, could be explained by the low fb (Fig. 4). The ventilatory responses observed in our study during low-intensity exercise resemble those reported in the literature.[57-59] In theory, for a given VE, there is an optimal combination of VT and fb that minimizes the work of the respiratory muscles.[57-61] Ventilatory responses have been described as having 2 stages (range 1 and range 2), which enable the respiratory system to adapt efficiently to increasing demands imposed by exertion. Range 1 is characterized by a small change in fb with a linear increase in VT to account for the increase in VE observed during low levels of exercise (ie, associated with a VT of less than 50% of vital capacity).[57-59] In addition, range 1 is characterized by shortening of the expiratory duration (Te), with no change in inspiratory duration (Ti).[61] Range 2 is characterized by a relatively constant VT of about 50% of vital capacity, whereas fb accounts for most of the increase in VE associated with moderate to high intensities of exercise.[57,59] Both Ti and Te are shortened in range 2.[61]

In our study, the mean VT was below 50% of the mean forced vital capacity during both steady-state positive and negative work. In addition, the changes in fb and VT followed closely those described above for range 1. The strong positive linear relationship between VT and VE in this study, regardless of the type of work, is consistent with the characteristics of range 1 described by Hey et al.[58] The results of our study, therefore, showed that at a low work rate of 60 W, the VT and VE relationship during positive work is qualitatively similar to that during negative work for the 3 pedaling frequencies.

The negligible change in fb observed in our study over the 3 pedaling frequencies for each type of work probably reflects the relatively low intensity of the exercise; therefore, we believe that it represents the flat plateau stage of range 1. The negligible change in fb may reflect entrainment or coordination of fb to exercise rhythm.[58,62-64] Entrainment of fb to exercise rhythm is characterized by adopting an fb that is a multiple (subharmonic frequency) of the exercise rhythm.[64] The fb values in our study were similar for the 3 pedaling frequencies for each type of work, resulting in an overall mean fb of 18 breaths/min. The mean fb of 18 breaths/ min was approximately the second, third, and fourth subharmonics of the pedaling frequencies 35, 55, and 75 rpm chosen in our study. The subjects' fb may have been entrained onto the subharmonic of the exercise rhythm. Examination of the individual data indicates that only 2 subjects entrained their fb to 18 breaths/min, which is a multiple of their exercise rhythm. Furthermore, the VT and VE relationship had a positive intercept, which does not support any contribution of entrainment of fb to exercise rhythm[59] In summary, the relationship of both fb and VT to VE during both positive and negative work can be described by the range 1 ventilatory response described in the literature. To verify this conclusion, Ti and Te will need to be measured in future studies.

Changes in Minute Ventilation and Its Components From BaseLine to Steady-Rate Exercise

With negative work, the increase in VE from baseline to steady-rate exercise was low compared with positive work (Fig. 4). This finding reflects the characteristics of negative work (eg, metabolically less demanding compared with positive work). This change in VE during negative work was associated with relatively small changes in VT (ie, 0.85 L/breath at baseline to 0.93 L/breath at 35 and 55 rpm to 1.12 L/breath at 75 rpm) (Fig. 5). In contrast, for positive work, the increase in VT were more marked (ie, 0.81 L/breath at baseline to 1.4 L/breath at 35 and 55 rpm to 1.57 L/breath at 75 rpm). Figure 6 shows that for negative and positive work, fb increased statistically to the same extent from baseline to steady-rate exercise. Thus, despite the difference in VE between positive and negative work during steady-rate exercise, the ventilatory response from baseline to steady-rate exercise at a power output of 60 W showed a comparable quantitative increase in tb for positive and negative work. The difference in VE for positive and negative work from baseline to steady-rate exercise, therefore, can be explained by a differential increase in VT. For negative work, the increase in VE from baseline to steady-rate exercise was largely effected by an increase in fb with a relatively small increase in VT. In contrast, for positive work, the increase in VE from baseline to steady-rate exercise was effected by increases in both VT and fb.

Although a comparable quantitative increase in fb was observed for positive and negative work, this increase appears to be disproportionately high in negative work, given its relatively low intensity, compared with positive work. In addition, this observation is not consistent with the typical range 1 ventilatory response that was evident in positive work. The predominant increase in fb during negative work at the start of exercise has been reported by Dean and Ross.[38] The mechanism for this ventilatory response to negative work is unclear. Dean and Ross? who observed rapid shallow breathing during downhill walking, acknowledged that this response may be related to factors other than negative work (eg, pastural stabilization). In our study, we believe that we eliminated the role of pastural stabilization because subjects were in a sitting position and were performing negative work on a cycle ergometer.

Exercise Responses of Older Subjects to Steady-Rate Cycling

We found individual variation in response to negative work, especially at the highest pedaling frequency of 75 rpm. Two of the oldest subjects, for example, showed a relatively high [VO.sub.2] (ie, 1.0 and 1.2 L/min) during negative work at 75 rpm. When a box plot, as described. by Wilkinson,[52] was used to map the distribution of [VO.sub.2] at 75 rpm during negative work, the subject with a [VO.sub.2] of 1.2 L/min was an outlier. Tests were repeated for these 2 subjects, and the results were comparable to the results of each of these subjects' initial tests. Two factors could explain this finding. First, one subject appeared to be tense throughout the test sessions, particularly during negative work at 75 rpm. The subject needed to frequently be reminded to relax his upper body. Tension could increase the [VO.sub.2] due to isometric work. The greater [VO.sub.2] observed for the other subject, however, was apparently not associated with an increase in upper-body stabilization. An alternate explanation is that negative work for this subject involved relatively more concentration and coordination to perform than for the other subjects. Shock and Norris[65] reported that neuromuscular coordination generally decreases with age. Other authors also have reported a corresponding decrease in muscle torque at high velocities of muscle contraction[66,67] and in the number of motor units.[66]

Vandervoort et al[48] reported that older subjects do not perform activities requiring rapid muscle contraction because of biomechanical factors. In our study, it is possible that other muscle groups not normally involved in performing negative work were active. This additional work could explain the increase in [VO.sub.2]. Another hypothesis is that age-related changes of the series elastic component reduce the muscle's ability to use elastic energy generated during negative work. Aging has been reported to induce degenerative changes in the fibroelastic tissue,46 which may also affect the series elastic component of muscle. Thus, older people may lose the normal integrity of the fibroelastic connective tissue, making it less able to perform negative work. This hypothesis was not examined in our study.

Clinical Implications

Patients with severe exercise limitations who are unable to benefit from conventional positive work may be able to derive additional benefit from negative work because it can be performed with less active muscle contraction and thus less energy cost and ventilatory demand than the same workload in positive work.[68] The results of our study have confirmed that negative work is not physiologically comparable to low-intensity positive work and that negative work can be tolerated by older individuals. Whether negative work is better tolerated by patients who are prone to rapid, shallow breathing (eg, patients with restrictive lung disease) compared with patients who require prolonged inspiratory and expiratory respiratory phases (eg, patients with chronic air flow limitation) remains to be elucidated.

Conclusions

We studied the responses of older subjects without cardiopulmonary impairments to negative and positive work performed on a cycle ergometer at 3 cadences. We studied these subjects because we believe that an understanding of the normal responses of a cohort of subjects who most resemble patients seen clinically with low functional work capacity and for whom negative work may have some benefit is warranted. A cycle ergometer was used in an attempt to minimize postural stabilization during exercise, and 3 cadences were selected because of the well-documented differences in their energy cost when force is held constant. A low work rate was selected so that we could better understand the energetics of low-intensity work (60 W) that may constitute a maximal workload for many patients with severe disability.

The data confirm and extend previous findings that the ventilatory responses to negative work compared with positive work, especially ventilatory demands, are decreased commensurate with oxygen demand at a given work rate. The effect of pedaling frequency on ventilation during negative work led us to conclude that VE was determined primarily by a change in VT because fb was not different among the 3 pedaling frequencies between the 2 types of work. Considering the effect of negative work versus positive work, changes in VE from baseline in our study appear to be influenced primarily by a disproportionate increase in fb in relation to VT. This finding sheds new light on the ventilatory responses described by range 1 for low-intensity exercise. This finding has implications for patients who require long time constants (ie, prolonged inspiratory and expiratory phases). Thus, further investigation into the unique effects of the work type versus pedaling frequency on ventilation is warranted. In addition, further study is needed to elucidate the mechanism underlying these responses.

(*) Sammons-Preston Canada, 3219 Yonge St, Ste 326, Toronto, Ontario, Canada M4N 2L3.

([dagger]) Summit Technologies Inc, 840-71 H Ave SW, Ste 900, Calgary, Alberta, Canada T2P 3G2.

([double dagger]) Hewlett-Packard Co, 3495 Deer Creek Rd, Palo Alto, CA 94304.

([sections]) Datex-Ohmeda (Canada) Inc, 1093 Meyerside Dr, Unit 2, Missisauga, Ontario, Canada, L5T 1J6.

References

[1] Abbott BC, Bigland B, Ritchie JM. The physiological cost of negative work. J Physiol (Lond). 1952;117:380-390.

[2] Kautz SA, Hull ML, Neptune RR. A comparison of muscular mechanical energy expenditure and internal work of cycling. J Biomech. 1994;27:1459-1467.

[3] Abbott BC, Bigland B. The effects of force and speed changes on the rate of oxygen consumption during negative work exercise. J Physiol (Lond). 1953;120:319-325.

[4] Asmussen E. Positive and negative work. Acta Physiol Scand. 1953;28: 364-382.

[5] Colliander EB, Tesch PA. Effects of eccentric and concentric muscle actions in resistance training. Acta Physiol Scand. 1990;140:31-39.

[6] Eloranta V, Komi PV. Function of quadriceps femoris muscle under maximum concentric and eccentric contractions. Electromyogr Clin Neurophysiol. 1980;20:159-174.

[7] Hather BM, Tesch PA, Buchanan P, Dudley GA. Influence of eccentric actions on skeletal muscle adaptations to resistance training: Acta Physiol Scand. 1991;143:177-185.

[8] Komi PV, Kaneko M, Aura O. EMG activity of the leg extensor muscles with special reference to mechanical efficiency in concentric and eccentric exercise. Int J Sports Med. 1987;8(suppl):22-29.

[9] Aura O, Komi PV. Mechanical efficiency of pure positive and pure negative work with special reference to the work intensity. IntJ Sports Med. 1986;7:44-49.

[10] Bigland-Ritchie B, Woods J. Integrated electromyogram and oxygen uptake during positive and negative work. J Physiol (Lond). 1976;260: 267-277.

[11] De Looze MP, Toussaint HM, Commissaris DA, et al. Relationships between energy expenditure and positive and negative mechanical work in repetitive lifting and lowering. J Appl Physiol. 1994:77:420-426.

[12] Clarkson PM, Tremblay I. Exercise-induced muscle damage, repair, and adaptation to humans. J Appl Physiol. 1988;65:1-6.

[13] Hesser CM, Linnarsson D, Bjurstedt H. Cardiorespiratory and metabolic responses to positive, negative, and minimum-load dynamic leg exercise. Respir Physiol. 1977;30:51-67.

[14] Klausen K, Knuttgen HG. Effect of training on oxygen consumption in negative muscular work. Acta Physiol Scand. 1971;83:319-323.

[15] Nielsen B, Nielsen SL, Petersen FB. Thermoregulation during positive and negative work at different environmental temperatures. Acta Physiol Scand. 1972;85:249-257.

[16] Moritani T, Muramatsu S, Muro M. Activity of motor units during concentric and eccentric contractions. Am J Phys Med. 1988;66: 338 -350.

[17] Knuttgen HG, Petersen FB, Klausen K. Exercise with concentric and eccentric muscle contractions. Acta Paediatr Scand Suppl. 1971;217: 42-46.

[18] Knuttgen HG, Petersen FB, Klausen Ii. Oxygen uptake and heart rate responses to exercise performed with concentric and eccentric muscle contraction. Med Sci Sports. 1971;3:1-5.

[19] Thomson DA. Cardiac output during positive and negative work. Scand J Clin Lab Invest. 1971;27:193-200.

[20] Armstrong RB. Mechanisms of exercise-induced delayed onset muscle soreness: a brief review. Med Sci Sports Exerc. 1985;16:529-538.

[21] Asmussen E. Observations on experimental muscle soreness. Acta Rheumatol Scand. 1956;1:109-116.

[22] Croisier JL, Camus G, Deby-Dupont G, et al. Myocellular enzyme leakage, polymorphonuclear neutrophil activation, and delayed onset muscle soreness induced by isokinetic eccentric exercise. Arch Physiol Biochem. 1996;104:322-329.

[23] Friden J, Sjostrom M, Ekblom B. A morphological study of delayed muscle soreness. Experientia. 1981;37:506-507.

[24] Newham DJ. The consequences of eccentric contractions and their relationship to delayed onset muscle pain. Eur J Appl Physiol. 1988;57: 353-359.

[25] Newham DJ,Jones DA, Ghosh G, Aurora P. Muscle fatigue and pain after eccentric contractions at long and short length. Clin Sci. 1988;74: 553-557.

[26] Rodenburg JB, Bar PR, De Boer RW. Relations between muscle soreness and biochemical and functional outcomes of eccentric exercise. J Appl Physiol. 1993;74:2976-2983.

[27] Gleeson M, Blannin AK, Zhu B, et al. Cardiorespiratory, hormonal, and haematological responses to submaximal cycling performed 2 days after eccentric or concentric exercise bouts. J Sports Sci. 1995;13:471-479.

[28] Jakeman P, Maxwell S. Effect of antioxidant vitamin supplementation on muscle function after eccentric exercise. Eur J Appl Physiol. 1993;67:426-430.

[29] Thomas TR, Londeree BR, Lawson DA, et al. Physiological and psychological responses to eccentric exercise. Can J Appl Physiol. 1994;19:91-100.

[30] Saxton JM, Donnelly AE, Roper HP. Indices of free-radical-mediated damage following maximal voluntary eccentric and concentric muscular work. Eur J Appl Physiol. 1994;68:189-193.

[31] Teague BN, Schwane JA. Effect of intermittent eccentric contractions on symptoms on muscle microinjury. Med Sci Sports Exerc. 1995; 27:1378-1384.

[32] Hill AV. Production and absorption of work by muscle. Science. 1960;131:897-903.

[33] Benedict FG, Parmenter HS. The energy metabolism of women while ascending and descending stairs. Am J Physiol. 1928;84:675-698.

[34] Kamon E. Negative and positive work in climbing a laddermill. J Appl Physiol. 1970;29:1-5.

[35] Seliger V, Dolejs L, Karas V, Dachlopnikova I. Adaptation of trained athletes' energy expenditure to repeated concentric and eccentric muscle contractions. Int Z Angew Physiol. 1968;26:227-234.

[36] Davies C, Barnes C. Negative (eccentric) work, II: physiological responses to walking uphill and downhill on a motor-driven treadmill. Egonomics. 1972;15:121-131.

[37] Knuttgen HG. Human performance in high-intensity exercise with concentric and eccentric muscle contractions. Int J Sports Med. 1986; 7(suppl):6-9.

[38] Dean E, Ross J. Downhill walking induces rapid shallow breathing. Eur J Physiol. 1989;415:351-354.

[39] Abbott BC, Aubert XM. The force exerted by active striated muscle during and after change in length. J Physiol (Lond). 1952;117:77-86.

[40] Banister EW, Jackson RC. The effect of speed and load changes on oxygen intake for equivalent power output during bicycle ergometry. Int Z Angew Physiol. 1967;24:284-290.

[41] Seabury JJ, Adams WC, Ramey MR. Influence of pedaling rate and power output on energy expenditure during bicycle ergometry. Ergonomics. 1977;20:491-498.

[42] Davies C, Barnes C. Negative (eccentric) work, I: effects of repeated exercise. Ergonomics. 1972;15:3-14.

[43] Dick RW, Cavanagh PR. An explanation of the upward drift in oxygen uptake during prolonged submaximal downhill running. Med Sci Sports Exerc. 1987;19:310-317.

[44] Pandolf KB, Kamon E, Noble BJ. Perceived exertion and physiological responses during negative and positive work in climbing a laddermill. J Sports Med Phys Fitness. 1978;18:227-236.

[45] Pimental NA, Shapiro Y, Pandolf KB. Comparison of uphill and downhill walking and concentric and eccentric cycling. Ergonomics. 1982;25:373-380.

[46] Shephard RJ. Physical Activity and Aging. Rockville, Md: Aspen Publishers Inc; 1987:3-77, 100-113.

[47] Skinner JS. Exercise Testing and Exercise Prescription for Special Cases. Philadelphia, Pa: Lea & Febiger; 1987:67-75.

[48] Vandervoort AA, Hayes KC, Belanger AY. Strength and endurance of skeletal muscle in the elderly. Physiotherapy Canada. 1986;38: 167-173.

[49] Bigland-Ritchie B, Graichen H, Woods J. A variable-speed motorized bicycle ergometer for positive and negative work exercise. J Appl Physiol. 1973;35:739-740.

[50] Borg GAV. Psychophysical basis of perceived exertion. Med Sci Sports Exerc. 1982;14:377-381.

[51] Hill AV. Heat of shortening and dynamic constants of muscle. Proc R Soc Lond B Biol Sci. 1938;126:136-195.

[52] Wilkinson L. SYSTAT: The System for Statistics. Evanston, Ill: SYSTAT Inc; 1988:581-582.

[53] American Thoracic Society. Standardization of spirometry: 1987 update. Am Rev Respir Dis. 1987;136:1285-1298.

[54] Gaesser GA, Brooks GA. Muscular efficiency during steady-rate exercise: effects of speed and work rate. J Appl Physiol. 1975;38:1132-1139.

[55] Astrand PO, Rodahl K. Textbook of Work Physiology. New York, NY: McGraw-Hill Inc; 1986:13-53, 188-191.

[56] McArdle WD, Katch FI, Katch VL. Exercise Physiology: Energy, Nutrition, and Human Performance. Philadelphia, Pa: Lea & Febiger; 1986: 563-590.

[57] Cunningham DJC, Robbins PA, Wolff CB. Integration of respiratory responses in alveolar partial pressures of [CO.sub.2] and [O.sub.2] and in arterial pH. In: Macklem PT, ed. Handbook of Physiology, Section 3--Volume III: The Respiratory System, Part 2: Mechanics of Breathing. Baltimore, Md: Williams & Wilkins; 1986:475-527.

[58] Hey EN, Lloyd BB, Cunningham DJC et al. Effects of various respiratory stimuli on the depth and frequency of breathing in man. Respir Physiol. 1966;1:193-205.

[59] Whipp BJ, Pardy RL. Breathing during exercise. In: Macklem PT, ed. Handbook of Physiology, Section 3--Volume III: The Respiratory System, Part 2: Mechanics of Breathing. Baltimore, Md: Williams & Wilkins; 1986:605-629.

[60] Bellemare F, Grassino A. Evaluation of human diaphragm fatigue. J Appl Physiol. 1982;53:1196-1206.

[61] Clark FJ, von Euler C. On the regulation of depth and rate of breathing. J Physiol (Lond). 1972;222:267-295.

[62] Bechbache RR, Duffin J. The entrainment of breathing frequency by exercise rhythm. J Physiol (Lond). 1977;272:553-561.

[63] Kay JDS, Petersen ES, Vejby-Christensen H. Mean and breath-by-breath pattern of breathing in man during steady-state exercise. J Physiol (Lond). 1975;251:657-669.

[64] Paterson DJ, Wood GA, Morton AR, Henstridge JD. The entrainment of ventilation frequency to exercise rhythm. Eur J Appl Physiol. 1986;55:530-537.

[65] Shock NW, Norris AH. Neuromuscular coordination as a factor in age changes in muscular exercise. In: Brunner D, Jokl E, eds. Physical Activity and Aging. New York, NY: Karger; 1970:92-99.

[66] Grimby G, Saltin B. The aging muscle. Clin Physiol. 1983;3:209-218.

[67] Larsson L, Grimby G, Karlsson J. Muscle strength and speed of movement in relation to age and muscle morphology. J Appl Physiol. 1979;46:451-456.

[68] Dean E. Physiology and therapeutic implications of negative work: a review. Phys Ther. 1988;68:233-237.

F Chung, PT, is Section Head, Surgical and Critical Care Physiotherapy, Physiotherapy Department, Burnaby Hospital, Burnaby, British Columbia, Canada. This study was completed in partial fulfillment of the requirements of Mr Chung's Master of Science degree.

E Dean, PhD, PT, is Professor, School of Rehabilitation Sciences, University of British Columbia, T325-2211 Wesbrook Mall, Vancouver, British Columbia, Canada V6T 1Z3 ([e-mail]). Address all correspondence to Dr Dean.

J Ross, PT, is Section Head, Critical Care, Rehabilitation Services, Vancouver General Hospital, Vancouver, British Columbia, Canada.

This study was approved by the Human Research Ethics Committee of the University of British Columbia.

This study was supported, in part, by funding from the Canadian Lung Association.

This article was submitted December 4, 1997, and was accepted January 11, 1999.