To determine the mechanisms underlying increased aerobic power in response to exercise training in octogenarians, we studied mildly frail elderly men and women randomly assigned to an exercise group (n = 22) who participated in a training program of 6 mo of physical therapy, strength training, and walking followed by 3 mo of more intense endurance exercise at 78% of peak heart rate or a control sedentary group (n = 24). Peak O2 consumption (V̇o2 peak) increased 14% in the exercise group (P < 0.0001) but decreased slightly in controls. Training induced 14% increase (P = 0.027) in peak exercise cardiac output (Q̇), determined via acetylene re-breathing, and no change in arteriovenous O2 content difference. The increase in Q̇ was mediated by increases in heart rate (P = 0.009) and probably stroke volume (P = 0.096). Left ventricular stroke work also increased significantly. In the men, the increase in V̇o2 peak was exclusively due to a large increase in peak Q̇ (22%). In the women, the gain in V̇o2 peak was due to small increases in Q̇ and O2 extraction from skeletal muscles. Pulse pressure normalized for stroke volume and arterial elastance during peak effort did not change with training. Controls showed no changes. The results suggest that, although frail octogenarians have a diminished capacity for improvement in aerobic power in response to exercise training, this adaptation is mediated mostly by an increase in Q̇ during peak effort. Furthermore, Q̇ likely plays a greater role in the adaptive increase in V̇o2 peak in old men than old women.
- cardiovascular changes
- exercise training
maximal aerobic power (V̇o2 max) decreases progressively with age (3, 9, 12, 18). A lower V̇o2 max in the elderly is due to a compromised cardiac reserve reflected in a lower cardiac output and a reduction in O2 extraction from skeletal muscles during maximal effort (14). Among the factors contributing to the decline in V̇o2 max in advanced age are structural and functional deteriorations of the cardiovascular system resulting in an increase in arterial stiffness and diminished left ventricular (LV) function, decreased physical activity, and chronic degenerative diseases. Physical inactivity appears to play a role in the decrease in V̇o2 max and deterioration of LV performance with advancing age. This notion is supported by studies showing that endurance exercise training is partially effective in compensating for the age-associated decline in V̇o2 max and impairment of LV systolic reserve and diastolic filling (22–25, 27). In a recent randomized study, we found that the subjects over the age of 78 yr with mild-to-moderate frailty can also adapt to exercise training, but the magnitude of the increase in their aerobic capacity was modest (2). However, the mechanisms underlying an increase in aerobic exercise capacity in the octogenarians are not known. The increase in aerobic power in response to endurance exercise training can be mediated by an increase in cardiac output, enhanced O2 extraction from skeletal muscles, or both. Therefore, the purpose of this study was to explore the physiological basis for the adaptive increase in aerobic power in response to training that we reported recently in these octogenarians (2). Specifically, we sought to determine whether the gain in peak O2 consumption (V̇o2) in these octogenarians is due to an increase in maximal cardiac output, arteriovenous O2 content difference [(a-v)O2], or both. To achieve this objective, we selected those individuals from our laboratory's study (2) who 1) increased their aerobic capacity in response to training and 2) were able to tolerate vigorous exercise long enough to make it possible to measure cardiac output reliably. Our hypothesis was that in mildly frail octogenarians the increase in peak V̇o2 in response to exercise training is mediated by increases in both cardiac output and O2 extraction.
Subjects. We studied 22 elderly men and women, 83.0 ± 3.6 yr old (mean ± SD), selected from the exercise group (n = 69), and 24 men and women, 84.0 ± 4.2 yr old, selected from the control group (n = 50), that we have reported recently elsewhere (2). The selection was based on the following criteria to meet the objectives of this study: 1) ability to perform a peak exercise cardiac output test, which is a physically demanding procedure because it requires the volunteers to exercise at or near their peak effort for at least 1 min to measure V̇o2 followed by cardiac output, and 2) an increase in aerobic power in response to training. The subjects' recruitment and enrollment criteria have been described in detail previously (2). In brief, these were: 1) age ≥ 78 yr, 2) mild-to-moderate frailty, and 3) absence of symptoms or conditions severe enough to prevent subjects from exercising. Patients with heart failure were excluded. The volunteers' clinical characteristics are listed in Table 1. All participants gave written, informed consent, and the study protocol was approved by the Human Studies Committee of Washington University School of Medicine.
Exercise test and determination of highest attainable (peak) V̇O2. After a familiarization trial, a maximal treadmill exercise test was performed using the following protocol: after a warm-up during which the participants walked on a tread-mill at 0% grade for 3–4 min at speeds ranging from 0.5 to 1.2 miles/h, speed was kept constant and the grade was increased 1 or 2% every 1 or 2 min until the subjects either could no longer continue to exercise because of exhaustion or cardiac symptoms or they showed ECG changes or other abnormalities that made it unsafe for them to continue to exercise. V̇o2 was measured continuously with open-circuit spirometry, as previously described (11). V̇o2 max was defined by using the following criteria (11): 1) an attainment of plateau of V̇o2 despite increasing exercise intensity, 2) a respiratory exchange ratio (RER) ≥ 1.1, and 3) measured V̇o2 lower than predicted for the work rate. Because many participants did not meet the criteria for V̇o2 max, the highest V̇o2 attained is designated as peak V̇o2. Peak O2 pulse was calculated as peak V̇o2/peak heart rate (ml/beat).
Cardiac output. One or 2 wk after the peak V̇o2 test, cardiac output was measured noninvasively with the use of the acetylene-rebreathing (C2H2) method during submaximal and peak exercise, as described previously (17). Cardiac output was measured (29) from the disappearance rate of the C2H2 (calculated by a computer) and the final system volume (sum of the volumes of anesthesia bag, respiratory tract, rebreathing valve, dead space, and alveolar volume). The final system volume was calculated from the volume and the ratio of the initial-to-final concentrations of the concurrently administered helium during rebreathing (29). A capillary line was inserted into a three-way valve attached to the mouthpiece for continuous monitoring of concentration of gases by the mass spectrometer throughout the 20-s re-breathing procedure (29).
The volunteers exercised on a treadmill at two exercise intensities: 1) an intensity requiring 50% of their previously determined highest attainable V̇o2. During exercise, V̇o2 was measured by using an “on-line” system for 4 min, blood pressure (BP) was measured with the use of a sphygmomanometer during the last minute, and then cardiac output was measured with the use of the acetylene rebreathing for 10 breaths. 2) After 10 min of rest, the participants exercised again on the treadmill at the lowest intensity required to elicit 100% of the previously determined peak V̇o2, and the C2H2 rebreathing was repeated. We used 7 instead of 10 breaths for determination of cardiac output during peak effort because, in our experience, it can be tolerated better, particularly by octogenarians, and still provide reliable data. Total peripheral resistance (TPR) was calculated as mean BP divided by cardiac output. The procedure was repeated at completion of the study with a similar treadmill exercise protocol using the same absolute exercise intensity (50% of initial peak V̇o2) and also during peak exercise. (a-v)O2 was calculated from peak exercise cardiac output and peak V̇o2.
Arterial elastance, used as a index of effective arterial load (28), was calculated as the ratio of end-systolic pressure to stroke volume. End-systolic pressure was estimated as (2 × SBP + DBP)/3, where SBP and DBP are systolic and diastolic BPs (10). TPR and LV stroke work were calculated with the use of the standard formulas. Pulse pressure (13, 19) and the ratio of pulse pressure to stroke volume were used as a surrogate for arterial stiffness (7, 16, 17). The effects of training on these variables were evaluated during submaximal exercise at an absolute intensity and during peak effort.
Body composition. Lean body mass (LBM) was determined with the use of dual-energy X-ray absorptiometry as described earlier (30).
Exercise training program. Because of frailty, our volunteers did not have the physical capacity to participate in a vigorous training program. Therefore, we designed an exercise training program that included three phases, each consisting of 36 sessions of exercise to be completed in 3 mo: 1) physical therapy, 2) strength training, and 3) endurance exercise program. In the first 3 mo, participants underwent physical therapy exercises. After 3 mo, strength training was added to physical therapy. During the initial 6 mo (physical therapy and strength training), the volunteers participated also in an endurance exercise program consisting of walking that was increased progressively in duration and speed as tolerated by each volunteer. After 6 mo, the participants then began more vigorous endurance-exercise training made possible by improvements in strength, flexibility, balance, and endurance gained by the initial 6 mo of training preparations. The endurance-exercise training program included walking on a treadmill, cycling, and rowing, which were tailored to each participant's ability and limitations. The participants were expected to exercise three times per week. The endurance exercise was at an intensity that elicited 70–75% of peak heart rate. The duration of the exercise sessions was increased progressively from 20 to 60 min. After 4–6 wk of the endurance exercise program, interval training was added to the exercise regimen. It consisted of several 3- to 5-min endurance exercise bouts requiring 85–90% of the subjects' peak heart rate interspersed with 3–4 min of light exercises. Transportation was provided to the volunteers to attend the exercise facility.
Control group. The volunteers randomized to the control group were given a home exercise program that consisted of stretching, relaxation, and yoga types of exercises, three times per week for 9 mo. Furthermore, the control subjects were transported to our exercise facility to exercise under supervision once a month as described in detail elsewhere (2).
Statistics. We used t-tests for comparison of initial and final data. Wilcoxon's test was used when appropriate. Least-square regression analyses were performed to examine relationships between physiological variables. Multiple regression analysis was also performed on those variables that showed a significant association in the univariate analysis. Data are expressed as means ± SD.
Resting data. There was a small weight gain of borderline significance most likely due to a significant but small (3%) increase in total LBM (Table 2). There were no changes in heart rate or BP in response to training (Table 2). Pulse pressure did not change with training (Table 2). There were also no changes in the controls (Table 2). Cardiac history and medications were similar in the two groups (Table 1).
Exercise training. All participants completed 108 sessions of exercise training. The volunteers were able to exercise an average of 2.61 ± 0.5 days/wk, with an intensity averaging 87%, ranging from 73 to 100%, of peak heart rate in the third phase (endurance exercise) of their training program. The duration of exercise averaged 69.4 ± 3.7 min per exercise session, including warm-up, cool-down, and rest periods during exercise session.
Maximal attainable O2 uptake capacity. Peak V̇o2 expressed in liters per minute (14%) or normalized for weight or LBM increased significantly but modestly (2 ml · kg-1 · min-1) in response to training (Table 3). Peak V̇o2 obtained during peak exercise cardiac output determination was also increased significantly (Table 4). Eleven of the subjects before training and 18 after training met the criteria for V̇o2 max. Those subjects who attained V̇o2 max during their initial exercise test showed an 11% increase in maximal aerobic capacity (before: 1.32 ± 0.39 l/min, after: 1.46 ± 0.39 l/min, P < 0.001; RER: 1.17 ± 0.07 vs. 1.19 ± 0.08, P = 0.5; maximal heart rate: 137.3 ± 16.8 vs. 139.1 ± 13.7 beats/min, P = 0.66). The difference in the increases (before vs. after, P < 0.001) in peak V̇o2 between the participants who exercised 3 days/wk (n = 11) and those who did not (n = 11) was not significant (3 days/wk: 2.1 ± 1.9 ml · kg-1 · min-1, <3 days/wk: 1.9 ± 1.3 ml · kg-1 · min-1, P = 0.76). The RER during peak effort was higher after training. O2 pulse increased 10% in the exercise group but not in the controls (Table 3).
In contrast, control subjects did not display any changes.
Submaximal exercise. Heart rate during submaximal exercise at a comparable absolute exercise intensity (V̇o2 pre: 0.71 ± 0.25 l/min, post: 0.70 ± 0.23 l/min, P = 0.50) was lower (pre: 93.4 ± 14 beats/min, post: 87.4 ± 7 beats/min, P = 0.042) in response to training. However, stroke volume (pre: 85 ± 24 ml, post: 81 ± 24 ml, P = 0.4), systolic (pre: 168 ± 20 mmHg, post: 170 ± 20 mmHg, P = 0.7) and diastolic (pre: 85 ± 8 mmHg, post: 83 ± 12 mmHg, P = 0.5) BP, and arterial elastance (pre: 1.8 ± 0.5 mmHg/ml, post: 1.9 ± 0.9 mmHg/ml, P = 0.3) during submaximal exercise were not affected by training. The ratio of pulse pressure to stroke volume also did not change.
Heart rate and BP during peak exercise. Heart rate, SBP, and the rate-pressure product during peak effort increased significantly in response to training (Table 4). Training had no significant effect on DBP. Pulse pressure was greater after training. However, pulse pressure normalized for stroke volume did not change with training. There were no significant changes in the controls.
Peak exercise cardiac output and stroke volume. Peak exercise cardiac output increased 14% (P = 0.027) in response to training (Table 4). In contrast, (a-v)O2 during peak effort did not change. Stroke volume during peak exercise increased by 10% (P = 0.097) and peak heart rate by 5% (P = 0.009) in response to training. Training induced a 10% decrease in TPR during peak effort, which was marginally significant (P = 0.076). The control group did not exhibit any significant changes in cardiac output, stroke volume, (a-v)O2, or stroke work.
LV systolic performance. LV stroke work during peak exercise increased 16% (P = 0.037) in response to training (Table 4). Arterial elastance during peak effort did not change in the exercise group. There were no changes in these variables in the controls. We found a strong and significant inverse relationship between the changes (before vs. after) in arterial elastance and cardiac output during peak exercise in both the exercise (r = 0.79, slope = -4.44, P < 0.001) and control (r = 0.855, slope = -3.8, P < 0.001) groups, suggesting that those who were able to decrease vascular load demonstrated a larger cardiac output during peak effort.
Hemodynamic factors influencing the adaptive increase in peak V̇O2. We found a significant direct relationship between the increases in peak V̇o2 and cardiac output in response to training (r = 0.46, P = 0.033, Fig. 1A). In contrast, the increase in peak V̇o2 did not correlate with the changes in the (a-v)O2 (r = 0.011, P = 0.96, Fig. 1B). The increase in peak heart rate also correlated significantly with the increase in peak V̇o2 (r = 0.512, P = 0.015), whereas the changes in arterial elastance (r = 0.14, P = 0.5), LV stroke work (r = 0.37, P = 0.10), or stroke volume (r = 0.27, P = 0.23) did not. Peak O2 pulse correlated strongly with peak stroke volume both in the exercise group before (r = 0.73, P < 0.001) and after (r = 0.90, P < 0.001) training, and in the controls during the initial (r = 0.85, P < 0.001) and final (r = 0.68, P < 0.001) evaluations (Fig. 2). However, the changes in O2 pulse and stroke volume in response to training did not correlate significantly (r = 0.28, P = 0.21).
Multiple regression analysis showed that, between the increases in cardiac output and heart rate, the latter contributed significantly to the training-induced increase in aerobic capacity in these octogenarians (r = 0.61; change in cardiac output: P = 0.095; change in heart rate: P = 0.043). Furthermore, multivariate analysis showed that both of the physiological determinants of cardiac output (i.e., heart rate and stroke volume) contributed significantly and equally to the training-induced increase in peak exercise cardiac output (r = 0.99, P < 0.001 for both changes in heart rate and stroke volume).
In those participants from the exercise group that our laboratory had recently reported on (2) who had a cardiac output measurement and whose peak V̇o2 did not increase in response to training (n = 8) (before: 1.22 ± 0.19 l/min, after: 1.18 ± 0.18 l/min, P = 0.031), peak exercise cardiac output (before: 11.0 ± 2.8 l/min, after: 10.7 ± 2.4 l/min, P = 0.60), LV stroke work (before: 152.2 ± 40.9, after: 156.9 ± 42.2 gram meter (g · m), P = 0.64), heart rate (before: 122.0 ± 13.6 beats/min, after: 122.0 ± 21.1 beats/min, P = 0.95), (a-v)O2 (before: 10.6 ± 1.8 ml O2/dl, after: 10.5 ± 1.4 ml O2/dl, P = 0.84), and SBP (190 ± 36 vs. 201 ± 28 mmHg, P = 0.44) did not increase after training.
Effect of gender. There were 12 women and 10 men in the exercise group and 15 women and 9 men in the control group. We found differences in adaptive responses to training between men and women, but they did not reach statistical significance because of the small number of participants. The men increased their peak V̇o2 15.4% (from 1.36 ± 0.30 l/min to 1.57 ± 0.28 l/min, P = 0.007) and their peak exercise cardiac output 21.6% (before: 11.6 ± 3.0 l/min, after: 14.1 ± 3.4 l/min, P = 0.06). There was a small but insignificant decrease in the (a-v)O2 (before: 12.2 ± 3.0 ml O2/100 ml, after: 11.4 ± 1.6 ml O2/100 ml, P = 0.43). The increase (15%) in stroke volume (before: 92.9 ± 17 ml, after: 107.0 ± 20 ml) did not reach statistical significance (P = 0.11). Peak heart rate was 124.1 ± 17 beats/min before and 130.9 ± 13 beats/min (P = 0.14) after training. Training induced a 26% increase in LV stroke work in the men, which was of borderline significance (before: 155 ± 38 g · m, after: 195 ± 47 gm, P = 0.067, Fig. 3). TPR during peak exercise did not decrease significantly (before: 903 ± 323 dyn/cm-5, after: 744 ± 136 dyn/cm-5, P = 0.14).
In the women, the increase in peak V̇o2 was smaller, 10.4% (before: 0.872 ± 0.17 l/min, after: 0.963 ± 0.17 l/min, P < 0.001). The increases in cardiac output (before: 8.2 ± 2.2 l/min, after: 8.8 ± 2.1 l/min, P = 0.27, Fig. 3), (a-v)O2 (before: 11.0 ± 1.9, after: 11.3 ± 2.2 ml O2/dl, P = 0.6), stroke volume (before: 64.6 ± 16.0, after 66.3 ± 16.4 ml, P = 0.65), and LV stroke work (before: 107 ± 35 g · m, after: 114 ± 39 g.m., P = 0.34) were small. Peak heart rate increased significantly in the women (before: 126.1 ± 18.3, after: 133.4 ± 18.6, P = 0.033). The frequency (men: 2.66 ± 0.5 days/wk, women: 2.53 ± 0.4 days/wk, P = 0.60) of the training stimulus was similar between men and women.
The findings of this study provide evidence that suggests that octogenarians with mild to moderate frailty can adapt to exercise training with improvements in aerobic power, cardiac output, and LV stroke work. However, the magnitude of these adaptive responses is small compared with those in 60- to 72-yr-olds (6, 25). The reasons for the attenuated adaptations to training are not clear. One possibility is an inadequate training stimulus because of frailty. Although the training intensity relative to peak heart rate in these octogenarians was similar to that which induced a larger increase (22%) in V̇o2 (6 ml · kg-1 · min-1 vs. 2 ml · kg-1 · min-1) in 60- to 72-yr-old subjects (6, 25), the absolute exercise intensity was low because of their low peak V̇o2 values, limiting their ability to exercise vigorously. It is likely that the frequency of training was also low to induce major cardiovascular adaptations in most of the octogenarians. The frequency of exercise was limited by the volunteers' low energy level and by residual fatigue. Another possible factor is the presence of concurrent hypertension or subclinical coronary artery disease. However, the presence of coronary artery disease does not generally prevent attainment of cardiovascular adaptations to exercise training (1, 5, 8, 21), and, furthermore, the majority of the volunteers did not have significant myocardial ischemia. It is also possible that a marked reduction in cardiovascular reserve in old age may limit the biological capacity to adapt to training.
Our data suggest that the mechanism responsible for the increase in peak V̇o2 in response to training was a larger peak exercise cardiac output, as evidenced by a significant increase in cardiac output without a concurrent increase in (a-v)O2. The positive correlation between the increases in peak V̇o2 and cardiac output and the absence of any association between the changes in peak (a-v)O2 and peak V̇o2 provide further evidence that the primary mechanism for the gain in peak V̇o2 was an increase in cardiac output instead of enhanced O2 extraction by skeletal muscles. The finding of a larger cardiac output and LV stroke work during peak effort suggests an improvement in LV systolic function. Absence of decreases in effective arterial load and the ratio of pulse pressure to stroke volume during peak exercise suggests that training had no significant effect on arterial stiffness in these octogenarians.
Our findings suggest that an increase in peak heart rate played a significant role in improving peak V̇o2. The increase in peak cardiac output was in part due to a higher heart rate during peak effort. However, it is also likely that a larger stroke volume may have contributed to the increase in cardiac output even though the increase in stroke volume was statistically of borderline significance. In fact, the magnitude of increase in stroke volume was twice as much as the increase in heart rate (10 vs. 5%) in these octogenarians. The reasons for a larger stroke volume are not clear. It is possible that increased blood volume could have contributed to this adaptation, as others have shown in older subjects (4, 20). However, Okazaki et al. (15) have recently reported an absence of an increase in blood volume in response to training in older men. The increase in LBM in response to strength training was probably too small to induce a larger O2 extraction in exercising skeletal muscle during peak effort, as evidenced by an absence of an increase in (a-v)O2.
It is plausible that a portion of the increase in aerobic power in response to training was due to the inability of many (50%) of the participants to attain true V̇o2 max during their initial test because of muscle weakness and fatigue, whereas most (78%) were able to attain their maximal O2 uptake capacity at the conclusion of the study. However, we do not believe that this can entirely account for the increase in aerobic capacity because we found a significant but a smaller gain (11%) even in those who were able to attain a V̇o2 max. Furthermore, the absence of an increase in peak V̇o2 in the controls and a significant rise in O2 pulse in response to training suggest that our volunteers had actually undergone physiological adaptations to training. A significantly lower heart rate during submaximal exercise at the same absolute exercise intensity provides further evidence for physiological adaptations to exercise training. Therefore, it seems likely that the participants adapted to training even though the observed 14% increase in aerobic capacity may overestimate the actual magnitude of the physiological adaptations.
The number of volunteers in the exercise group is too small to provide definitive information regarding whether gender can play a role in cardiovascular adaptations to training in the octogenarians, as has been demonstrated in 60- to 72-yr-old adults (24, 25). Nevertheless, on the basis of our data, it is plausible that the mechanisms underlying the increase in aerobic capacity in response to training may differ between the male and female octogenarians. It is likely that the male octogenarians increased their peak V̇o2 exclusively by a large increase in cardiac output and an improvement in LV systolic performance during peak exercise, whereas in the female octogenarians the increase in peak V̇o2 in response to training was due to small increases in cardiac output and (a-v)O2. Spina et al. (24, 25) have reported that the increase in V̇o2 in response to training in older men is mediated by a large increase in maximal cardiac output, stroke volume, and LV stroke work and a small increase in (a-v)O2 (25). In contrast, the gain in V̇o2 in older women is, instead, due to enhanced O2 extraction by exercising skeletal muscles during maximal exercise, leading to the conclusion that postmenopausal women not on hormone replacement therapy do not undergo an adaptive increase in stroke volume in response to exercise training (25). The present small increase in peak cardiac output in women was mostly due to an increase in peak heart rate, instead of stroke volume.
We did not observe any significant decrease in resting BP in response to training in our subjects. The absence of BP-lowering effect of exercise is probably due to a small number of subjects with elevated BP (only half of the volunteers were hypertensive). The level of BP may also have been too high (SBP: 160–190 mmHg) to respond favorably to exercise training. It is also possible that, similar to other adaptations, the BP-lowering effect of exercise training may be attenuated in the octogenarians, particularly in those with moderate-to-severe hypertension.
Limitations of this study include the following. 1) A relatively small number of subjects were studied; however, we and others have shown significant adaptations to training with a smaller sample size in 60- to 75-yr-old subjects (6, 26, 27). 2) It may be argued that the small cardiovascular adaptations to exercise training in our subjects are because of chronic cardiovascular disorders, i.e., coronary artery disease (CAD) rather than because of aging per se. Arguing against this interpretation are: 1) most of our subjects did not have clinical CAD, and 2) we and others have shown that patients with proven CAD can markedly improve cardiovascular function with training (5, 8).
In conclusion, our data suggest that although some octogenarians with mild-to-moderate frailty can adapt to exercise training with improvements in aerobic power, cardiac output, and LV stroke work during peak effort, these adaptations are attenuated and modest in magnitude. Our observations suggest, however, that, in those octogenarians who retain the capacity to improve their aerobic power in response to training, the increase in peak V̇o2 is mediated by an increase in cardiac output, due to increases in heart rate and probably stroke volume, without any improvement in O2 extraction by skeletal muscles. Our findings also suggest that gender may play a role in the mechanisms underlying the increase in aerobic capacity in response to training in the octogenarians. It appears that cardiac output plays a larger role in the gain in peak V̇o2 in the old frail men than in the women.
This study was supported by the Washington University Claude D. Pepper Older Americans Independence Center Grant AG-13629 from the National Institute on Aging and General Clinical Research Center Grant 5-MO1-RR-00036. D. T. Villareal is supported by National Center for Research Resources Grant K23 RR-16191.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Present address of R. J. Spina: Dept. of Kinesiology, San Francisco State Univ., 1600 Holloway Ave., San Francisco, CA 94132-4161.
- Copyright © 2003 the American Physiological Society