Previous studies have demonstrated that frail octogenarians have an attenuated capacity for cardiovascular adaptations to endurance exercise training. In the present study, we determined the magnitude of cardiovascular and metabolic adaptations to high-intensity endurance exercise training in healthy, nonfrail elderly subjects. Ten subjects [8 men, 2 women, 80.3 yr (SD2.5)] completed 10–12 mo (108 exercise sessions) of a supervised endurance exercise training program consisting of 2.5 sessions/wk (SD 0.2), 58 min/session (SD 6), at an intensity of 83% (SD 5) of peak heart rate. Primary outcomes were maximal attainable aerobic power [peak aerobic capacity (V̇o2peak)]; serum lipids, oral glucose tolerance, and insulin action during a hyperglycemic clamp; body composition by dual-energy X-ray absorptiometry, and energy expenditure using doubly labeled water and indirect calorimetry. The training program resulted in an increase in V̇o2peak of 15% (SD 7) [22.9 (SD 3.3) to 26.2 ml·kg−1·min−1 (SD 4.0); P < 0.0001]. Favorable lipid changes included reductions in total cholesterol (−8%; P = 0.002) and LDL cholesterol (−10%; P = 0.003), with no significant change in HDL cholesterol or triglycerides. Insulin action improved, as evidenced by a 29% increase in glucose disposal rate relative to insulin concentration during the hyperglycemic clamp. Fat mass decreased by 1.8 kg (SD 1.4) (P = 0.003); lean mass did not change. Total energy expenditure increased by 400 kcal/day because of an increase in physical activity. No change occurred in resting metabolism. In summary, healthy nonfrail octogenarians can adapt to high-intensity endurance exercise training with improvements in aerobic power, insulin action, and serum lipid and lipoprotein risk factors for coronary heart disease; however, the adaptations in aerobic power and insulin action are attenuated compared with middle-aged individuals.
- oxygen uptake
- cardiovascular adaptation
- hyperglycemic clamp
- doubly labeled water
aging is associated with a reduction in maximal aerobic power (V̇o2max) (4, 9), predominantly due to structural and functional deteriorations of the cardiovascular system, chronic degenerative diseases, and decreased physical activity. The latter factor has been supported by data suggesting an attenuation of the age-associated decline in V̇o2max with habitual endurance exercise training (17, 26). The capacity of elderly individuals to adapt to endurance exercise training may be limited by biological aging effects and the ability to withstand physiological overload (i.e., frequency, intensity, and duration) due to orthopedic and morbidity-related limitations. Indeed, similar to others (22), our laboratory recently reported that, although frail elderly women and men undergo cardiovascular adaptations to exercise, the effect of training on aerobic power was attenuated compared with middle-aged individuals (6).
Unfavorable changes in body composition also occur with aging, including a reduction in lean mass and an increase in fat mass, specifically abdominal adipose tissue (18). Body composition changes are related to the deterioration in glucose tolerance (33) and an increase in insulin resistance (5) that commonly occur with aging. Although the aging process undoubtedly impacts body composition and carbohydrate metabolism, reductions in physical activity play a key role. Whereas numerous studies have demonstrated the benefits of physical activity on glucose tolerance and/or insulin action (10, 13, 16, 27, 31), there is a paucity of data on individuals beyond 70 yr of age, especially regarding the response to high-intensity endurance exercise. Moreover, because earlier data suggest that endurance exercise training results in a compensatory reduction in physical activity throughout the rest of the day in elderly people (8), an additional aim of the study was to explore the impact of a rigorous exercise regimen on the components of daily energy expenditure.
Therefore, the primary purpose of this study was to determine whether healthy elderly subjects aged 77–87 yr maintain the ability to adapt to a rigorous endurance exercise training program with measurable changes in V̇o2max, insulin action, and serum lipids. The secondary purpose was to evaluate the impact of the exercise regimen on body composition, total energy expenditure (TEE), and resting metabolic rate (RMR). The aims were to determine 1) whether healthy elderly people are able to complete a long-term high-intensity (>80% of peak heart rate) exercise training regimen, 2) the magnitude of the adaptive capacity for improved V̇o2max, and 3) the effects of the exercise training on lipid and lipoprotein cardiovascular disease risk factors and glucose tolerance and/or insulin action. We hypothesized that healthy elderly would be able to complete the endurance program and experience adaptations in primary and secondary outcomes; however, their adaptations would be more robust compared with our laboratory’s previous results in frail octogenarians (6) and attenuated compared with our laboratory’s previous results in middle-aged individuals (17).
MATERIALS AND METHODS
Participants included 10 nonfrail healthy individuals (8 men, 2 women) aged 80.3 yr (SD 2.5) (range 77–87 yr). Participants were screened with a medical history, physical examination, standard blood chemistries, graded exercise stress test to identify those with ischemic heart disease, and a physical performance test (PPT) which evaluated ability to perform 9 physical tasks (e.g., stair climbing, walk 50 ft, put on and remove a laboratory coat, lift a 7-lb. book to a shelf 12 in. above shoulder height, pick up a penny from standing height, static and dynamic balance tests). To be eligible for this study, participants had to score >32 (maximum score = 36) on the PPT (1), obtain a peak oxygen consumption (V̇o2peak) of ≥18 ml·kg−1·min−1, and be capable of and willing to participate in a supervised endurance exercise program. No subjects had diabetes mellitus, and none were being treated with oral hypoglycemic agents. All subjects were on stable medications, defined as no change in medication or dosage at least 3 mo before the study and throughout the duration of the study. Although well controlled, 40% had been diagnosed with hypertension and were being treated with a diuretic (n = 1), α-adrenoreceptor blocker (n = 2), or β-blocker (n = 1). Informed consent was obtained from all subjects before enrollment in the study, which was approved by the Washington University Human Studies Committee and the General Clinical Research Center Scientific Advisory Committee.
The endurance exercise training program consisted of 108 supervised endurance exercise sessions, designed to be performed 3 days/wk for 9 mo. With inclusion of the preconditioning phase and allowing for minor illness or travel, the entire program was completed in 10–12 mo. The program began with ∼4 wk of flexibility and balance exercises that were designed to prevent injury during the endurance exercise training. The endurance exercise training program began at an intensity of 60–75% V̇o2peak, and it progressed gradually to 85% of V̇o2peak within 4–6 wk depending on individual tolerance. Similarly, exercise duration was increased to 60 min in a similar time frame. Participants were permitted to rest when changing exercise mode; however, rest periods were discouraged after initial conditioning occurred. A variety of exercise modes were used, including a 4-lane 17-lap per mile indoor track, treadmills, rowing ergometers, and stair climbing ergometers. The exercise modes were individualized based on subject preferences, abilities, and orthopedic capabilities. Participants were encouraged to utilize more than one mode to minimize joint pain, muscle fatigue, and boredom.
A graded treadmill exercise test was conducted at baseline, 3 mo, and 9 mo to determine V̇o2max and the heart rate, blood pressure, and ECG responses to exercise. Subjects walked on a treadmill at the fastest comfortable pace at 0% grade for 3–4 min, after which the grade increased by 1–2% every 1 or 2 min. The test continued until the subjects were unable to continue due to volitional exhaustion, ECG changes, or other abnormalities that rendered it unsafe to continue to exercise. Oxygen consumption (V̇o2) was measured continuously by using open-circuit spirometry as described previously (17). V̇o2max is conventionally determined by the highest V̇o2 that corresponds to at least two of the following criteria: 1) a plateau in V̇o2 values despite an increase in exercise intensity, 2) maximal heart rate within 10 beats/min of the age-predicted maximum (220 − age), and 3) a respiratory exchange ratio (RER) >1.10. Because most but not all participants (9 of 10 at baseline and 10 of 10 at final) attained a true maximal value at all testing time points, V̇o2max was determined as the highest attained V̇o2 and was operationally defined as V̇o2peak.
Serum lipids and lipoproteins.
Blood samples were obtained in the fasting state at baseline and after completion of the training program for the determination of serum total cholesterol (TC), LDL cholesterol (LDL-C), HDL cholesterol (HDL-C), and triglycerides (TG). Samples were obtained ∼24 h after the last bout of exercise. TC and glycerol-blanked TG were measured by automated enzymatic methods by using commercial kits, HDL-C was measured in plasma after precipitation of apolipoprotein B-containing lipoproteins, and LDL-C was calculated by using the Friedewald equation (7).
Oral glucose tolerance test.
A 75-g, 2-h oral glucose tolerance test (OGTT) was conducted in the morning after a 12- to 14-h fast ∼24 h after the last exercise session. Diet was monitored for 3 days before each OGTT to ensure that a minimum of 150 g of carbohydrate was consumed. Blood samples for the determination of glucose and insulin concentrations were obtained before and 30, 60, 90, and 120 min after ingestion of the glucose beverage. Plasma glucose concentrations were measured by the glucose oxidase method (Beckman Instruments, Fullerton, CA), and serum insulin concentrations were measured by a double-antibody radioimmunoassay (24). The total areas under the glucose and insulin curves were calculated by using the trapezoidal rule. The insulin sensitivity index (ISI) of Matsuda and DeFronzo (23) was calculated as [10,000/square root of (fasting glucose × fasting insulin) × (mean OGTT glucose × mean OGTT insulin concentration)].
A modified hyperglycemic clamp procedure with a superimposed arginine infusion and fat meal (14) was performed to assess changes in glucose disposal rate (GDR) and insulin secretory responses to intravenous glucose, intravenous arginine, and oral fat stimuli. Participants were admitted as inpatients to the General Clinical Research Center (GCRC) the evening before each test. To avoid the acute effects of the last exercise bout on RMR, which was measured before the clamp, subjects did not exercise the day of admission, thus enabling a minimum of 36 h since the last exercise session. Dinner was prepared in the GCRC metabolic kitchen, with the same energy level and composition during the baseline and final studies. The clamp was performed in the morning after a 12-h fast.
Modifications to the previously described clamp procedure (14) included raising the plasma glucose concentration to 14 mM (250 mg/dl) and extending the period of hyperglycemia to 3 h. Clamp stage 1 (0–90 min) involved hyperglycemia alone, stage 2 (90–135 min) was hyperglycemia with superimposed arginine infusion, and stage 3 (135–180 min) was hyperglycemia with arginine and a 25-g liquid-fat meal administered orally. A potassium phosphate solution (20 meq potassium, 13.5 mM phosphate in 500 ml 0.9% NaCl) was infused (100 ml/h) to prevent a decline in blood potassium concentration secondary to hyperinsulinemia. Plasma glucose concentrations were measured at bedside by using a YSI glucose analyzer (Yellow Springs Instrument, Yellow Springs, OH), and insulin concentrations were measured by a double-antibody radioimmunoassay (24). GDR and mean insulin concentration were calculated for each clamp stage and for the entire period of hyperglycemia (10–180 min). GDRs are expressed relative to fat-free mass.
Dual-energy X-ray absorptiometry (DXA; model 1000/W, Enhanced Whole Body software version 5.64, Hologic, Waltham, MA) was used to assess changes in whole body fat mass and lean body mass.
TEE was measured during 2-wk periods at baseline and at the end of the intervention using the doubly labeled water (DLW) method (29) in a subgroup of participants (n = 5). Subjects were admitted to the GCRC at 1700, and they received a standard meal prepared in the GCRC metabolic kitchen at 1800. A baseline urine sample was collected at 2000, after which an oral dose of DLW was administered, containing 0.25 g of H218O and 0.12 g of 2H2O per kilogram of total body water. Total body water was calculated from fat-free mass assuming a constant hydration factor of 0.73. Postdose urine samples were collected at 0900 the next morning and at the same time of day 1 and 2 wk later for the measurement of TEE. Samples were cleaned with dry carbon black, stored in 2-ml cryovials at −20°C, and analyzed for H218O and 2H2O abundance by isotope ratio mass spectrometry (VG-Fisons SIRA Series II IRMS) in the Washington University School of Medicine Mass Spectrometry Resource. Isotope elimination rates, carbon dioxide production, and TEE were calculated as described previously (30) but using the slope-intercept method.
RMR was measured by indirect calorimetry (Deltatrac II Metabolic Monitor, Sensormedics, Yoma Linda, CA) in the morning after an overnight fast in the GCRC. A 1-h measurement was performed, with the subject supine in bed, awake but motionless. After a 20- to 30-min equilibration period, the final 30–40 min of steady V̇o2 values were averaged to calculate RMR.
Energy expended in physical activity was calculated from TEE by subtracting the measured RMR and an estimated value for the thermic effect of meals (TEM). TEM was calculated as 10% of TEE.
The effect of the exercise training intervention on all major outcomes was assessed by using paired t-tests. All data are expressed as means (SD). An α level of 0.05 was considered significant.
This study was powered on the two primary outcomes: V̇o2max and insulin action. The power to detect a statistically significant improvement in V̇o2peak was based on our laboratory’s results in the frail elderly (6), in whom a 14% increase was observed in response to less intense and shorter duration exercise. With a minimal effect size of 1.0 (6) and an α level of 0.05, a sample size of 10 would provide a power of 0.90 to detect a significant improvement in aerobic power should it exist. The power to detect a statistically significant improvement in insulin action during the hyperglycemic clamp was based on the reductions in plasma insulin concentration of 13 and 23% during the early (0–10 min) and late (15–180 min) phases of hyperglycemia, respectively, observed previously in response to training (15). A sample size of 10 would enable the detection of a 15% reduction in plasma insulin concentration at an α level of 0.05 with a power of 0.80.
Thirteen nonfrail, nonobese individuals enrolled in the supervised exercise study; three were lost to follow-up because of unrelated health problems (n = 2) or lack of compliance with the exercise regimen (n = 1). Ten subjects [8 men and 2 women; body mass index = 23.5 kg/m2 (SD 2.6)], ranging in age from 77–87 [80.3 (SD 2.5)] yr, completed the full training program (108 sessions) in 10–12 mo. After the initial 4–6 wk of endurance training, subjects averaged 2.5 sessions/wk (SD 0.2), with 58 min (SD 6) of exercise per session at an intensity of 83% (SD 5) of peak heart rate. The primary exercise modes included inclined treadmill walking, stair-climbing ergometer, and rowing ergometer.
V̇o2max and body composition.
V̇o2peak increased by 12.3% (SD 6.4) in absolute terms [+0.20 l/min (SD 0.10)] and 15.0% (SD 7.6) in relative terms [+3.4 ml·kg−1·min−1 (SD 1.7)] in response to the training program (Table 1). Of the subjects who completed the full exercise training program, 9 of the 10 attained V̇o2max at baseline, and all 10 subjects attained true V̇o2max during final testing. Body weight decreased by 1.6 kg (SD 1.9) (P = 0.025), due to a reduction in fat mass (−1.8 kg (SD 1.4); P < 0.003) with no change in lean body mass (Table 1).
Favorable changes occurred in serum lipids and lipoproteins (Table 2). Specifically, TC decreased by 8% (−0.40 mM (SD0.30); P = 0.002) and LDL-C decreased 10% (−0.33 mM (SD0.25); P = 0.003); however, no significant changes occurred in HDL-C (P = 0.69) or TG (P = 0.11). Subjects were specifically instructed to maintain usual dietary habits, and no changes in dietary intake were apparent from 3-day diet records obtained at baseline and final testing (data not shown).
Oral glucose tolerance.
Both fasting plasma glucose and oral glucose tolerance were normal at baseline, and they did not change with exercise training (Table 2). However, fasting insulin concentration decreased significantly [−19.6 pmol/l (SD 21.1); P = 0.017], and there was a trend for a reduction in insulin area under the curve [−19.4% (SD 28.9); P = 0.06]. The ISI, which incorporates all glucose and insulin values during the OGTT, improved by 32 % (SD 35) (P = 0.018).
During the hyperglycemic clamps, plasma glucose concentrations averaged 13.6 (SD 0.2) and 13.7 mM (SD 0.2) for baseline and final measurements, respectively. Insulin concentrations were lower during the final clamp for stage 1 [246.1 (SD 154.6) to 172.9 pmol/l (SD 132.7); P = 0.008] but not for stage 2 [1,890.4 (SD 1,109.8) to 1,517.5 pmol/l (SD 971.4); P = 0.137] or stage 3 [3,407.2 (SD 2,240.2) to 2,528.1 pmol/l (SD 1,493.2); P = 0.093]. GDR was higher after the training program during the third period of hyperglycemia [18.6 (SD 5.1) to 21.1 mg·kg fat-free mass−1·min−1 (SD 5.2), +13.4%; P = 0.017], when insulin concentrations were the highest (135–180 min). When expressed relative to plasma insulin concentration, however, GDR improved by 30.1 (SD 30.8) and 37.9% (SD 42.6) during stages 2 and 3, respectively, (P < 0.05; Fig. 1).
TEE and RMR.
TEE increased 404 kcal/day (SD 324) P = 0.049) due to an increase in physical activity energy expenditure [+425 kcal/day (SD 326) P = 0.042], with no change in RMR [preexercise: 1,496 kcal/day (SD 268) postexercise: 1,435 kcal/day (SD 152); P = 0.53].
Although much of the disability in old age is due to disease and the aging process itself, loss of function due to physical inactivity also appears to play a role (9). The primary findings of this study were that a long-term high-intensity endurance exercise program confers favorable changes in cardiovascular fitness, as evidenced by increases in V̇o2max and insulin action, as measured by oral glucose tolerance and hyperglycemic clamp, in healthy elderly >77 yr of age. However, the magnitude of the adaptations in V̇o2max and insulin action were attenuated compared with individuals in the seventh decade of life (17, 21, 35).
The mechanistic reasons regarding the blunting of the adaptation of V̇o2max among octogenarians are not known. V̇o2max was assessed after 24 wk of training in individuals aged 79–91 yr (22). A significant 13% increase in absolute maximal aerobic power occurred in women (n = 12); however, in contrast to our study, there was no change in men (n = 9, −1%). Discrepancies in findings may be related to the fact that training duration was in months and training time per exercise session were approximately one-half of that in our study and the inability of their older subjects to achieve a true V̇o2max as reflected by average RER values of ∼1.06.
Previously, our laboratory also reported that frail elderly subjects were able to respond to 9 mo of exercise training (3 mo of which included high-intensity endurance exercise) with a 14% increase in absolute aerobic power (6). It should be noted, however, that the 14% increase may be attributable in part to the subjects’ ability to give a better effort on the maximal treadmill test after the training program, as evidenced by an increase in maximal heart rate and RER at posttest. Furthermore, these individuals were selected from a larger study (2) on the basis of their ability to complete a demanding cardiac output assessment. In the present study, on average, our nonfrail subjects achieved their age-predicted maximal heart rate and an RER of >1.2 at baseline, indicative of a maximal effort; these parameters did not change after the training program.
It was previously found that the adaptive increase in V̇o2max in response to 9 or 12 mo of strenuous endurance exercise training ranged from 24 to 30% in healthy 60- to 71-yr-old women and men (17, 32). The magnitude of this increase is similar, in relative terms, to that found in response to intense training in younger individuals (3, 12, 21, 25). Thus it appears that the capacity to adapt to training with an increase in aerobic power is reduced in healthy individuals beyond 77 yr of age (see Fig. 2), because the increase in their V̇o2peak averaged only 15% when expressed as milliliters of O2 per kilogram per minute and 11% when expressed as millilters of O2 per kilogram lean body mass per minute. The differences are considerably larger when expressed in absolute terms relative to body weight: 3.4 ml O2·kg−1·min−1 in the octogenarians, 6–7.5 ml O2·kg−1·min−1 in the 60–71 yr olds (17, 32), and ∼11 ml O2·kg−1·min−1 in young individuals (12).
As in previous studies on individuals with normal glucose tolerance (10, 15, 31), the exercise training resulted in an improvement in insulin action, reflected in a higher ISI during the OGTT and an increase in GDR relative to insulin concentration during the hyperglycemic clamp. However, as with V̇o2max, the improvement in insulin action was smaller than that observed in young adults (34) or middle-aged individuals (15). For example, Short et al. (34) reported that insulin sensitivity assessed with an intravenous glucose tolerance test did not improve among men and women aged 60–87 yr in response to 16 wk of stationary cycling, whereas a 72% improvement was observed among younger (20–39 yr) individuals. The reduction in fasting insulin concentrations and improvements in insulin sensitivity, despite normal values at baseline, has important health implications. Fasting hyperinsulinemia and insulin resistance are components of the metabolic syndrome, which increases mortality from cardiovascular diseases and all causes among nondiabetic individuals (11).
The deterioration in the capacity to adapt to exercise with advancing age is not surprising, because a reduction in the capacity to adapt to stresses and maintain homeostasis is a common definition of aging. Among the factors responsible for this decline are a loss and/or deterioration in the function of cells, including heart (19) and skeletal muscle cells (28). Notably, there is a deterioration of the extracellular matrix with loss of elastin and increased calcification, resulting in stiffening of the arteries (19). Osteoarthritis, degeneration of the joints, and musculoskeletal pain promote mobility limitations (20). In addition to the decrease in the capacity of the skeletal muscles and cardiovascular system to adapt to exercise training, the deterioration in structure and function limits the capacity to generate an adaptive exercise stimulus in terms of magnitude, duration, and frequency. This limitation was very evident in the healthy and well-motivated elderly subjects in this study, who had to take rest breaks during the exercise and who were unable to average more than 2.5 exercise sessions per week because of joint discomfort and persistent fatigue.
Despite these limitations, the participants in this study were able to significantly improve their exercise capacity, insulin action, cholesterol levels, and body composition, providing evidence that healthy elderly subjects beyond 77 yr of age can still benefit from endurance exercise training. This finding is in keeping with the results of a previous study showing that even frail octogenarians can adapt to exercise training with a reduction in frailty and a small increase in aerobic power (6).
It has been reported that exercise training results in a compensatory decrease in physical activity during the rest of the day so that there is no increase in TEE (8). We were not able to confirm the existence of such a phenomenon. In fact, TEE increased by ∼400 kcal/day as the result of an increase in physical activity. Thus, despite reporting some persistence of fatigue, the octogenarians in this study actually increased their physical activity even on the days in which they did not participate in the formal exercise program.
A limitation of the present study was the small sample size and the absence of a control group. All of these factors are related to the stringent inclusion criteria and the difficulty of finding healthy, nonfrail elderly subjects who are willing and able to enroll in a vigorous exercise program. The absence of a control group has little implication for our study conclusions because spontaneous improvement in our primary and secondary outcomes in the absence of an intervention in this population is highly unlikely. This study was conducted concurrently with studies investigating the effects of exercise on frailty in the same-aged population. Results from these studies indicate a slight reduction in V̇o2max, physical function, and lean body mass throughout the same 10- to 12-mo period in a randomized control group (2, 6).
In conclusion, despite a reduced capacity to exercise and to adapt to endurance exercise-training, our results show that well-motivated healthy elderly subjects aged 77 yr or more can benefit from endurance exercise training. The benefits include improvements in aerobic power, insulin action, body composition, and lipid and lipoprotein risk factors for coronary artery disease.
This study was supported by the Washington University Claude D. Pepper Older American Independence Center Grant AG-13629, General Clinical Research Center Grant RR-00036, and Diabetes Research and Training Center Grant DK-20579. E. M. Evans was supported by Institutional National Research Service Award AG-00078 and Individual National Research Service Award AG-05874.
Present address of E. M. Evans: Dept. of Kinesiology, University of Illinois at Urbana-Champaign, 906 S. Goodwin Avenue, Urbana, IL 61801.
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.
- Copyright © 2005 the American Physiological Society