Author + information
- Received July 27, 2005
- Revision received September 14, 2005
- Accepted September 26, 2005
- Published online March 7, 2006.
- J. Susie Woo, MD⁎,
- Christina Derleth, MD⁎,
- John R. Stratton, MD, FACC†,‡ and
- Wayne C. Levy, MD, FACC†,⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. Wayne C. Levy, University of Washington, Box 356422, 1959 NE Pacific Street, Seattle, Washington 98195
Objectives The aim of this study was to determine whether changes in oxygen efficiency occur with aging or exercise training in healthy young and older subjects.
Background Exercise capacity declines with age and improves with exercise training. Whether changes in oxygen efficiency, defined as the oxygen cost per unit work, contributes to the effects of aging or training has not yet been defined.
Methods Sixty-one healthy subjects were recruited into four groups of younger women (ages 20 to 33 years, n = 15), younger men (ages 20 to 30 years, n = 12), older women (ages 65 to 79 years, n = 16), and older men (ages 65 to 77 years, n = 18). All subjects underwent cardiopulmonary exercise testing to analyze aerobic parameters before and after three to six months of supervised aerobic exercise training.
Results Before training, younger subjects had a much higher exercise capacity, as shown by a 42% higher peak oxygen consumption (Vo2) (ml/kg/min, p < 0.0001). This was associated with an 11% lower work Vo2/W (p = 0.02) and an 8% higher efficiency than older subjects (p = 0.03). With training, older subjects displayed a larger increase in peak W/kg (+29% vs. +12%, p = 0.001), a larger decrease in work Vo2/W (−24% vs. −2%, p < 0.0001), and a greater improvement in exercise efficiency (+30% vs. 2%, p < 0.0001) compared to the young.
Conclusions Older age is associated with a decreased exercise efficiency and an increase in the oxygen cost of exercise, which contribute to a decreased exercise capacity. These age-related changes are reversed with exercise training, which improves efficiency to a greater degree in the elderly than in the young.
Older age, female gender, and the untrained state all are known to be associated with a decrement in exercise capacity, as reflected by a decreased peak oxygen consumption (Vo2) or peak workload. There are multiple potential contributors to a reduced exercise capacity, including a reduction in maximal cardiac output, which is common to aging, female gender, and the untrained state. This reduction in maximal cardiac output can in turn be due to multiple mechanisms, such as a reduced peak heart rate, as seen with aging, or a reduced stroke volume, which is seen in females and the untrained.
Another possible contributor to a reduced exercise capacity is a reduction in exercise efficiency, which can be crudely defined as the energy output/energy input or Watts (W)/Vo2. Little is known about the potential role of exercise efficiency in the decline of exercise capacity with aging, female gender, or the untrained state. We have recently shown that patients with heart failure have a substantially reduced exercise efficiency compared to age-matched control subjects (1). Moreover, measures of efficiency had better correlation with heart failure symptoms than did the peak Vo2(1).
The purpose of this study was to determine whether a reduction in exercise efficiency occurs with aging, whether exercise efficiency differs in men versus women, and whether increases in exercise capacity from several months of supervised exercise training would be associated with improvements in exercise efficiency.
Sixty-one healthy, sedentary adult volunteers were recruited from the Seattle area into groups of younger women (ages 20 to 33 years, n = 15), younger men (ages 20 to 30 years, n = 12), older women (ages 65 to 79 years, n = 16), and older men (ages 65 to 77 years, n = 18). Subjects were excluded if they had performed regular exercise in the past year. Other exclusion criteria included any prior history of angina, myocardial infarction, stroke, hypertension, chronic pulmonary disease, diabetes, current medication use other than hormone or thyroid replacement therapy, current smoking, or exercise-limiting orthopedic impairment. Entry requirements included a normal hematocrit, creatinine, fasting blood glucose, total cholesterol, resting electrocardiogram, M-mode and two-dimensional echocardiogram, and Bruce protocol maximal exercise test, including immediate post-exercise tomographic sestamibi imaging for all older subjects to rule out occult coronary disease. All older females were required to be on hormone replacement therapy. All subjects signed an informed consent form approved by the Human Subjects Committee at the University of Washington.
Cardiopulmonary exercise testing was performed at baseline, after three months, and after six months of exercise training. Subjects sat in a chair for at least 2 min before the start of exercise to obtain a resting Vo2. Subjects were assigned to one of five exercise treadmill protocols (with maximum speeds of 3.5, 4, 4.5, 5, and 6 mph), based on an estimation of their level of fitness. All protocols included 2 min of walking at a 0° incline at 3.5 mph at the beginning of exercise to allow comparison of Vo2across all subjects at a matched workload. Subjects then walked at a 0° incline at the maximum speed of their treadmill protocol before the initiation of ramp exercise. Exercise protocols were terminated at the point of volitional fatigue. Almost all subjects reached a peak respiratory exchange ratio (RER) of >1.1, indicating maximal effort. Peak RER was 1.22 ± 0.09 before training and 1.17 ± 0.08 after training. After termination of exercise, subjects sat quietly in a chair. Measurements of Vo2were continued for at least 6 min of recovery to allow estimation of the oxygen debt.
After baseline testing, all subjects underwent a six-month supervised training program. Exercise was initiated at a target intensity of 50% to 60% heart rate (HR) reserve, increased to 80% to 85% by the third or fourth month and continued at that level for the remainder of the study. The exercise program consisted of walking/jogging, bicycling, and stretching, each for 30 min, for a total of 90 min per session, three times per week.
Data were obtained with a metabolic cart (Medical Graphics, St. Paul, Minnesota) coupled to a Quinton Q65 treadmill. Gas and volume calibrations were performed before each test. Resting Vo2was defined as the lowest average Vo2for 2 min of rest before exercise. Peak Vo2and peak workload were defined as the highest 60-s averages for Vo2and W, respectively. Recovery was defined as the first 6 min after termination of the treadmill protocol.
Recovery respiratory kinetics were calculated from the fitting of Vo2and carbon dioxide production (Vco2) data to a monoexponential curve using Microsoft Excel 2000 Solver add-in (Microsoft Corp., Redmond, Washington), as previously described by Mitchell et al. (2).
Measures of oxygen cost and efficiency
Three-month exercise data were used only when six-month data were not available. Six-month exercise data were available in 53 of the 61 patients. Watts were estimated from the speed, grade, and weight of each subject, using the American College of Sports Medicine’s guidelines and equations for energy expenditure during graded walking and running (3). The oxygen cost of exercise during exercise and recovery, oxygen debt, and efficiency were calculated as follows (1,2):
In the efficiency equation, 1,435 = constant by which W were converted to calories, and k = 5,000 calories/ml of Vo2(4). For efficiency at the matched workload of 3.5 mph at 0 grade, k = 3,840 + 1,180 × RER (1).
Data are expressed as mean ± SD, unless otherwise noted. Group differences were evaluated by paired ttests and analysis of variance for repeated measures, using Statview 5 (Abacus Concepts, Berkeley, California). Significance was defined as p ≤ 0.05.
Effects of gender
At baseline, male subjects weighed 11 kg more (p = 0.0001) than their female counterparts. The men had an 8% higher weight-adjusted resting Vo2(ml/kg) than the women (p < 0.0001). Resting HR was 8% higher in women than in men (p = 0.02), but peak HR was higher in men than in women (p = 0.03). The O2pulse (Vo2/HR), a surrogate for stroke volume, was 40% higher at rest (p < 0.0001) and 49% higher at peak exercise (p < 0.0001) in men than in women. As expected, males had a 32% higher peak Vo2(ml/kg/min, p < 0.0001) and were able to exercise to a 21% higher peak workload (W/kg, p < 0.0001) than their female counterparts. There was no significant difference in O2debt (p = 0.31), work Vo2/W (p = 0.56), recovery Vo2/W (p = 0.76), or exercise efficiency (p = 0.57) between men and women (Table 1).Separate analyses of the young and older groups similarly revealed no gender differences in O2debt, work Vo2/W, or exercise efficiency (data not shown).
Effects of aging
Younger and older subjects were well-matched in weight (p = 0.76). There was no difference in baseline resting HR (p = 0.58) or resting O2pulse (p = 0.11). At peak exercise, however, younger subjects had a 42% higher baseline peak Vo2(ml/kg/min, p < 0.0001), a 21% higher peak HR (p < 0.0001), and a 14% higher peak O2pulse than older subjects (p = 0.0002). The young patients achieved a 28% higher peak workload (W/kg, p < 0.0001) compared to the older subjects. The total oxygen cost of exercise (work Vo2/W) was 11% higher in the elderly group (p = 0.02), primarily because of a 34% higher recovery Vo2/W (p = 0.002) and an 18% higher O2debt (p = 0.012) compared to the young. Older subjects also displayed slower recovery kinetics, with increased Vo2and Vco2ramp recovery time constants compared to the young (p < 0.0001). The increased oxygen cost of exercise in the elderly subjects translated to an 8% lower exercise efficiency (p = 0.03) compared to the young before training. These age-related differences at baseline and after training are shown in Table 2.
Effects of training
All age and gender groups lost 1% to 2% of their baseline weight with exercise training (p = 0.005). There was no change in resting Vo2with training (p = 0.62). Resting HR decreased significantly with training (p < 0.0001). Resting and peak O2pulse increased to a significant degree after training (p = 0.026 and p < 0.0001, respectively). As expected, training resulted in an overall 13% increase in peak Vo2(p < 0.0001) and a 19% increase in peak W/kg (p < 0.0001). Training also led to a 24% decrease in O2debt (p < 0.0001), a 14% decrease in the oxygen cost of exercise (work Vo2/W, p < 0.0001), and a 17% increase in exercise efficiency (p < 0.0001). The results of training are summarized in Table 3.
There was a significant age effect in response to training, as older subjects showed an increase in exercise efficiency while the young did not (p < 0.0001 for the training × age effect). Older subjects had a larger decrease in O2debt (−29% vs. −17%, p = 0.01), as well as a marked decrease in recovery Vo2/W (−46% vs. −18%, p = 0.0001) and work Vo2/W (−24% vs. −2%, p < 0.0001) with training compared to the young. This translated into a 30% increase in exercise efficiency for the elderly subjects (p < 0.0001), as opposed to a 2% increase in the young (p = 0.42). The effect of training on efficiency for each individual subject is displayed in Figure 1.The elderly patients also showed a larger increase in peak W/kg (+29% vs. +12%, p = 0.001) after training than the young. Although the younger subjects made no significant improvement in exercise efficiency, they had a relatively larger increase in peak Vo2(+15% vs. +12%, p = 0.003) and peak O2pulse (+18% vs. +10%, p = 0.002) than the older subjects.
This age-associated difference in response to training was also seen when directly comparing Vo2between the young and the old for a matched submaximal workload. When performing the same workload (walking at 3.5 mph at a 0° grade), younger subjects had a 16% lower Vo2(ml/kg/min), and therefore a 16% lower Vo2/W than older subjects before exercise training (p < 0.0001). The young exercised more efficiently than the older subjects at baseline (20.9% vs. 17.3%, p < 0.0001). With training, the young decreased their Vo2(ml/kg/min) at 3.5 mph by 7% (p = 0.03), whereas the older patients decreased their Vo2by 16% (p = 0.0002). Efficiency trended toward improvement with a 6% increase in the young (p = 0.09), as compared to a 21% increase after training in the older subjects (p = 0.0003). In summary, the older subjects responded to training with a greater improvement in efficiency than their younger counterparts, as measured both at a fixed matched workload and over the entire exercise period (Fig. 2).
There was no measurable effect of gender on the response to training, as men and women showed almost identical decreases in exercise Vo2/W, recovery Vo2/W, and work Vo2/W. Men and women also showed no difference in their relative increase in peak Vo2(p = 0.06), decrease in O2debt (p = 0.66), and increase in exercise efficiency (p = 0.26) in response to training (Table 1).
Our study characterized the effects of age and gender on exercise efficiency, and prospectively examined the effect of training on efficiency in both younger and older men and women.
The older subjects in this study displayed the expected reductions in peak Vo2and peak workload compared to their young counterparts. Aging was also associated with a decreased exercise efficiency, increased oxygen debt, and increased recovery Vo2/W compared to those in the young. Our findings contradict those of some studies that have suggested no change in efficiency with age (5–7). Often, however, calculations of efficiency did not account for the total oxygen cost of exercise, including the Vo2during recovery. Patients with heart failure were reported to have a decreased oxygen cost during exercise compared to control subjects, implying an increased efficiency, until the oxygen cost during recovery was considered (1,2). Similarly, the exclusion of recovery data from our analysis would have led to the very different result of finding no difference in efficiency with age. It seems clear, however, that the oxygen debt should be included in calculating the cost of performing a given amount of work. A higher Vo2and decreased efficiency have previously been described in older versus younger subjects at fixed work rates of cycle ergometer exercise (8,9). Our study confirms these findings in a larger population and shows that older subjects can improve their efficiency with exercise training.
The age-related decrement in exercise efficiency is likely multifactorial. Older age has been associated with an approximately 25% decrease in muscle capillarization and mitochondrial enzyme activity (10). Reduced skeletal muscle oxidative capacity may then lead to premature or excessive lactate accumulation and an increase in the oxygen cost of exercise. Chisari et al. (11) measured serum lactate levels before and at multiple time points after graded treadmill exercise in a group of 34 older and 10 younger subjects. Resting lactate levels were not significantly different, and all subjects exercised until they reached 75% of their maximum HR to try to ensure primarily aerobic metabolism. Older subjects showed significantly higher levels of lactate at all time points, which were ascribed to an age-associated fall in mitochondrial oxidative metabolism. This decline in metabolic capacity may be due to mitochondrial disease or degeneration, as “ragged red” muscle fibers and cytochrome C oxidase-deficient myofibers, both markers for mitochondrial disease, have been shown to be increased in the elderly population (12,13).
Additional sources of inefficiency may include changes in cardiac function, skeletal muscle blood flow from a decrease in both capillary density and capillary-to-fiber ratio with age (10), nutrition, and hormone levels. In young women, plasma epinephrine levels are significantly correlated with both the magnitude and duration of excess post-exercise oxygen consumption (14). The increased recovery Vo2/W seen in the elderly subjects in our study may thus be related to circulating levels of catecholamines, which have been shown to be increased with age (15).
Older subjects in our study also showed a significant slowing in both Vo2and Vco2recovery kinetics. Several studies have shown a slowing of Vo2kinetics in the elderly at the onset of cycling or ramp exercise (16–18). Post-exercise oxygen kinetics, however, have only been examined in one previous study (19), which found a tendency for Vo2recovery to be faster in the young. The faster Vo2kinetics in the young was associated with greater capillarization per muscle fiber area and shorter O2diffusion distances. The association was weaker in the elderly group, suggesting that other factors, such as mitochondrial density, mitochondrial enzyme activity, or vascular function may play a larger role in controlling O2kinetics in the older patients.
As expected, the women in our study showed a lower peak Vo2(ml/kg/min) than their male counterparts. This gender difference in aerobic capacity is well-described in published data and is attributed to the higher body fat composition, lower hemoglobin content, and smaller heart size of women. The lower resting and peak O2pulses (Vo2/HR) in the women in our study reflect their decreased Vo2at rest and with exercise and are consistent with prior findings of decreased maximal stroke volume associated with female gender (20). Interestingly, this gender difference in peak Vo2was not associated with a difference in exercise efficiency, either at baseline or after training. Thus, the decreased exercise capacity seen in women is unlike that of the elderly subjects in that it is not associated with a decreased exercise efficiency, but may largely be explained by gender-related differences in maximal heart rate, stroke volume, and peripheral oxygen extraction.
Our study confirms that aging does not preclude a response to training, as the elderly subjects were able to improve in all the same exercise parameters as their younger counterparts. However, the elderly patients were not able to increase their peak Vo2to the same degree as the young. Although some longitudinal studies have demonstrated a slowing in the rate of decline of maximal aerobic capacity from continued years of regular vigorous endurance exercise (21,22), more recent analyses suggest that training may have no effect (23), or may even increase the rate of decline of Vo2max with age (24). Thus, our data suggest that improvement of peak Vo2in the elderly subjects may be limited by other factors that decline with age despite activity or exercise training, including maximal HR and diastolic filling rates (25,26).
All groups showed a decrease in oxygen debt, a decrease in recovery Vo2/W and improved Vo2, and Vco2recovery kinetics with training. These findings are consistent with previous cross-sectional and longitudinal studies that have shown a decrease in the magnitude and duration of post-exercise Vo2, an increase in Vo2and Vco2recovery rates, and a decrease in blood lactate response in both men and women associated with training (27–30). These changes with training also translated into an overall 17% increase in exercise efficiency. In a previous study (7), mean gross efficiency was similarly shown to increase across all age groups of adult males aged 23 to 63 years after they underwent an 8-month training program.
What was new and unexpected in our study was the disproportionately greater response to training in the elderly subjects, with complete reversal of age-related decrements in oxygen debt, recovery Vo2/W, and exercise efficiency. After training, the elderly subjects showed a lower oxygen debt, lower recovery Vo2/W, and higher efficiency than the untrained young. Babcock et al. (31) had similar findings when they demonstrated that training of older individuals could result in improvement of Vo2on-kinetics to levels approaching those of the fit young. Using regression analysis, Chilibeck et al. (32) found that for small differences in Vo2max, older subjects showed a larger difference in Vo2kinetics as compared to the young.
These changes in exercise efficiency and Vo2kinetics with age and training may be explained on the cellular level by a disproportionately larger degree of mitochondrial dysfunction (11) in older people. Evidence suggests that there is a progressive decline in mitochondrial respiratory rate and enzyme activity with age (33). Fortunately, capillary density and mitochondrial enzyme activity have been shown to increase with training in older persons, to levels similar to those seen in young individuals (10,34). Meredith et al. (35) found that sedentary older subjects increased their muscle oxidative capacity by 128%, compared to only 28% in the sedentary young, such that levels were similar between the two groups after training. These improvements in oxidative efficiency likely contribute to the marked lowering of oxygen debt and oxygen cost of exercise in response to training seen in the elderly subjects in our study.
In general, the reversibility of these parameters with training suggests that a significant portion of the changes that are seen with aging may in fact be due to lower fitness levels in the sedentary elderly as compared to the sedentary young. With only moderate changes in cardiorespiratory fitness, the elderly appear to achieve greater relative gains in exercise efficiency and other exercise-responsive aerobic parameters compared to the young.
Both O2consumption and W were adjusted for weight, but not for fat-free mass, which may have allowed a more accurate comparison between older and younger subjects of varying weight and body composition. The work performed by each subject was calculated according to American College of Sports Medicine guidelines, but still may not accurately reflect the wide variability in how efficiently individual people exercise. Direct measurement of stroke volume or cardiac output, as well as biochemical or muscle biopsy data, would be helpful to delineate the relative contribution of central versus peripheral factors to the changes seen in oxygen cost of exercise with aging and training. Without data from subjects that are aged 30 to 65 years or a more longitudinal study, it is also impossible to determine exactly when the changes that we associate with aging actually occur.
Our findings suggest that the decline in aerobic capacity seen with older age is associated with a decreased exercise efficiency, an increased oxygen cost of exercise and O2debt, and slower recovery kinetics. These changes may in large part be due to inactivity, with an associated decline in mitochondrial oxidative efficiency and a greater reliance on anaerobic metabolism in older persons. Exercise training results in an improvement in these parameters, likely by inducing increased O2delivery through increased stroke volume and muscle capillarization, as well as improved O2utilization from an increase in mitochondrial enzyme activity. The importance of efficiency in aerobic performance was recently underscored by the finding that muscular efficiency and reduced body fat contributed equally to an impressive 18% improvement in steady-state power over a seven-year period in a super-elite athlete (36). Even with relatively low levels of exercise training, our subjects made significant improvements in efficiency, oxygen debt, and recovery Vo2/W that were even greater in the elderly subjects than in the young. As in the older population, female gender is associated with a decreased aerobic capacity, but was not associated with a similar difference in exercise efficiency.
This research was supported by the National Institutes of Health grant no. AG 15462 and by the Medical Research Service of the Department of Veterans Affairs.
- Abbreviations and Acronyms
- heart rate
- respiratory exchange ratio
- carbon dioxide production
- oxygen consumption
- Received July 27, 2005.
- Revision received September 14, 2005.
- Accepted September 26, 2005.
- American College of Cardiology Foundation
- American College of Sports Medicine,
- Franklin B.A.,
- Whaley M.H.,
- Howley E.T.
- McConnell A.K.,
- Davies C.T.
- Coggan A.R.,
- Spina R.J.,
- King D.S.,
- et al.
- Chisari C.,
- Bresci M.,
- Licitra R.,
- Stampacchia G.,
- Rossi B.
- Chilibeck P.D.,
- Paterson D.H.,
- Smith W.D.,
- Cunningham D.A.
- Scheuermann B.W.,
- Bell C.,
- Paterson D.H.,
- Barstow T.J.,
- Kowalchuk J.M.
- Chilibeck P.D.,
- Paterson D.H.,
- Cunningham D.A.,
- Taylor A.W.,
- Noble E.G.
- Proctor D.N.,
- Beck K.C.,
- Shen P.H.,
- Eickhoff T.J.,
- Halliwill J.R.,
- Joyner M.J.
- Trappe S.W.,
- Costill D.L.,
- Vukovich M.D.,
- Jones J.,
- Melham T.
- Rogers M.A.,
- Hagberg J.M.,
- Martin W.H. 3rd.,
- Ehsani A.A.,
- Holloszy J.O.
- Wilson T.M.,
- Tanaka H.
- Pimentel A.E.,
- Gentile C.L.,
- Tanaka H.,
- Seals D.R.,
- Gates P.E.
- Marti B.,
- Howald H.
- Levy W.C.,
- Cerqueira M.D.,
- Abrass I.B.,
- Schwartz R.S.,
- Stratton J.R.
- Cannon E.W.R.,
- Rhodes E.C.,
- Langill R.H.
- Short K.R.,
- Sedlock D.A.
- Hagberg J.M.,
- Hickson R.C.,
- Ehsani A.A.,
- Holloszy J.O.
- Coggan A.R.,
- Spina R.J.,
- King D.S.,
- et al.
- Meredith C.N.,
- Frontera W.R.,
- Fisher E.C.,
- et al.
- Coyle E.F.