CLINICAL INVESTIGATION.

Low to Moderate Intensity Endurance Training in Healthy Oldei Adults: Physiological Responses after Four Months Joel D.Posner, KevinM. Gorman, Lisa Windsor-Landsberg, James Larsen, Michael Bleiman, Carl Shaw, Beth Rosenberg, and Janice Knebl Objective: To determine the physiological adaptations in previously sedentary healthy older men and women (mean age = 68) to a 16-week low-to-moderate-intensity exercise program. Design: Randomized, controlled trial. Setting: An exercise facility and testing laboratory in a gerontological research institute. Participants:Two-hundred forty-seven community-dwelling older persons free of significant cardiovascular, pulmonary, or uncontrolled metabolic disease, anemia, electrolyte abnormality, resting BP of 165/90 or greater, or chronic disease affecting the ability to exercise on a bicycle. Intervention: Subjects were randomly assigned to either an exercise (n = 166) or attention control group ( n = 81). Exercisers trained thrice weekly for 40 minutes on a cycle ergometer (5-minute warm u p , 30 minutes at training heart rate (THR), 5-minute cool down). THR was set at 70% of

peak heart rate attained on a maximal exercise test (mean = 115 2 15). Control subjects attended weekly group talks. Testing took place before and after the program. Results: Peak attained oxygen uptake (V02max) increased 8.5% in exercisers and decreased slightly in controls (p < .001) and oxygen uptake at ventilatory threshold (VeT V02) increased by 3.5% in exercisers and decreased by 3% in controls (p < .001). This pattern of a greater increase in V02max than VeT V 0 2 is different from that seen in young and middle-aged subjects. Conclusion:This study demonstrated that a large scale training program is feasible for healthy older people, that physiologic improvements can be measured after 16 weeks of lowto-moderate-intensity training, and that mechanisms of adaptation to exercise may be different in elderly subjects from those in younger ones. J Am Geriatr SOC401-7,1992

he physiologic benefits of exercise training in the young have long been known; recently the benefits among older subjects have been demonstrated. Seals et al’ exercised 11 healthy older subjects (mean age = 62) who increased their maximum attained oxygen uptake ( V 0 2 max) by 30% after a year of training, Thomas et a12 studied 44 healthy older males (mean age = 62) who showed an 18% increase in VOZ max after a year of training, Blumenthal et a13 exercised 33 subjects (mean age = 63) who increased their V 0 2 max by 11.6% after 4 months of training, and-in the largest study published to date-Cunningham et a14 trained 100 males at retirement (mean age = 63) for a year and found a 10% increase in V 0 2 max. In these studies, a moderately high to high intensity of training was used. Seals’ trainees worked at pulse rates of up to 156 beats per minute (BPMs),’ Thomas’ at 129 BPMs,~and Blumenthal’s at 70% of the maximum heart rate reserve (MHRR).3 Cunningham’s subjects were reported to have worked at pulse rates that ranged from 80 to 167 BPM.4 Though lower levels of

exercise may be better suited to exercise programs for large groups of older people, few investigators have examined the results of an exercise program of low intensity. Foster5 exercised 16 older women (mean age = 78.4) for 10 weeks, 8 women at 40% MHRR and 8 at 60% MHRR. She found increases of 12.6% in VOz max in the low intensity group and 15.4% in the higher intensity group. During the first 6 months of Seals’ study, his 11 subjects exercised with 20 to 30 minutes of unsupervised walking at least three times a week with an average heart rate of 107 BPM; they increased their VOz max by 12% after this portion of the training. To examine physiologic adaptations to a training program that could be followed at home by a large group of older subjects, we conducted a randomized controlled trial of 4 months of supervised low to moderate intensity exercise (prescribed training heart rate = 115 BPM & 15 or about 45% of MHRR) that would lend itself to a home program among 247 previously sedentary healthy older (mean age = 68; range = 6086) men and women, two-thirds of whom were exercised for 4 months and one-third of whom joined a once weekly discussion group. Since changes in physiological functioning may be confounded by extraneous environmental factors, we incorporated the non-exercising ”attention”control group in our design to determine what changes occurred as a result of the exercise program. To do this, we measured V 0 2 max and ven-

T

From the Gertrude and Jacob Arronson Foundation Laboratory of the Philadelphia Geriatric Center and the Division of Geriatric Medicine of the Medical College of Pennsylvania, Philadelphia, Pennsylvania. Supported in part by the National Institute on Aging Teaching Nursing Home Grant #I AG03934. Address correspondence to Joel D. Posner, MD, Medical College of Pennsylvania, 3300 Henry Avenue, Philadelphia, PA 19129.

’

JAGS 4O:l-7, 1992 0 1992 bu the American Geriatrics Societu

0002-&614/92/$3.50

2

POSNER ET AL

JAGS-JANUARY 1992-VOL. 40, NO. 1

tilatory threshold (VeT)in both groups at the beginning and the end of the 4-month training period.

specific health promotion techniques were discussed. Control subjects had blood pressures checked at each session. All subjects underwent repeat physiologic testMETHODS ing upon completion of the 4-month exercise or control Subjects and Study Design Subjects past their program. Subjects without V 0 2 max data on the second 60th birthday (mean age 68.6 5.1, range 60-86) were testing occasion ( n = 19) were excluded from this recruited from among a number of senior citizens cen- analysis. Reasons for incomplete data in these subjects ters in the greater Philadelphia area. We advertised for include non-compliance, acute illness, travel, and "people aged 60 and over in good health who do not scheduling conflicts. Complete data were obtained on exercise." Activity histories were obtained and we ex- 166 subjects in the exercise group and 81 subjects in cluded those volunteers who regularly engaged in aero- the control group (total n = 247). Table 1 shows the bic exercise (swimming,bicycling, aerobic dancing, etc.) age and sex distribution of included subjects. Exercise Test Protocol All subjects had been to the three or more times per week; all others underwent a multi-stage screening protocol including history and laboratory for one familiarizationsession with the exphysical examination, resting electrocardiogram,exten- ercise testing protocol prior to exercise testing. Exercise sive blood work including a fasting lipid profile on testing was conducted no sooner than 2 hours after a blood drawn in the morning after a fast of at least 12 light meal, and subjects were requested to refrain from hours (total cholesterol by enzymatic method, high cigarette smoking and consumption of caffeinated bevdensity lipoproteins by magnesium phosphate precip- erages on the day of the test. Three electrocardiitation, low density lipoproteins, triglycerides by en- ographic leads (11, V3, V5) were monitored continuzymatic method, very low density lipoproteins by cal- ously, and blood pressure was recorded every minute culation), full pulmonary function tests, and a progres- during exercise and for 10 minutes during recovery. sive exercise test with continuous monitoring of An electromagnetically-braked ergometer (Meinhart electrocardiogram, ventilation, and gas exchange data. KEM 11) was used with the subject pedaling at a conExercise tests were terminated at subject exhaustion stant rate of 60 rpm throughout the test. Seat height of unless an ischemic response to exercise testing or other the ergometer was adjusted to a comfortable position exercise induced abnormality occurred,6in which case according to each subject's height. After 1 minute of the test was terminated and the subject eliminated from rest, exercise was initiated by 2 minutes of unloaded the study. Volunteers with any of the following abnor- cycling and subsequent uniform increases in resistance malities were excluded from the study: significant car- every minute until exhaustion. Work rate was increased diovascular or pulmonary disease, uncontrolled meta- by 15 watts per minute for all subjects. This work rate bolic disease (diabetes, thyroid disease), anemia, elec- increment was selected so that subjects would reach trolyte abnormalities, resting blood pressure of 165/90 exhaustion within approximately 7-17 minutes of inor greater, or chronic diseases affecting the ability to cremental work.' All subjects did so. Equipment and Measurements Ventilatory and exercise on a bicycle. Forty percent of the original volunteers were excluded by the screening process. gas exchange variables were measured breath-byAfter testing, subjects were matched for age and sex, breath using a computerized system (Medical Graphics, then randomized into either an exercise or a control System 2000). Expired gases were sampled at the group. Two-thirds of all subjects entered a supervised mouthpiece (18 mL/sec) and analyzed for C02 (Datex 4-month aerobic exercise program while one-third of CD 102 capnograph) and 0 2 (Applied Electrochemistry subjects were assigned to a non-exercising "attention" S-3A O2analyzer) content. Calibration of gas analyzers control group. Subjects in the attention control group and flow module has been previously described.' We participated in a 16-week lifestyle enrichment and defined V 0 2 max as the maximal attained oxygen discussion program with a group leader. The 1-hour uptake calculated as the average of all breaths within weekly small group meetings were designed to achieve a 20-second period surrounding the highest recorded equivalence in personalized attention, social interac- V 0 2 at volitional fatigue.' As in other studies using tion, and staff support offered in the exercise training cycle ergometry in older subjects, very few subjects groups. Discussions centered around social, economic, demonstrated a "plateau" of oxygen uptake. We exand psychological subjects; neither exercise nor other amined a subset of subjects ( n = 56) who exhibited r _+

AGES

TABLE 1. AGE AND SEX DISTRIBUTION OF SUBJECTS MALE FEMALE TOTAL

60-64 65-69 70-74 75-79 80+

Totals *Does nof equal 100% due to rounding.

18 33 30 8 5 94

42 66 35 8 2 153

60 99 65 16 7 247

PERCENT 24 40 26 6 3 99*

3

IANUARY 1992-VOL. 40, NO. 1

PHYSICAL TRAINING IN HEALTHY OLDER ADULTS

values greater than 1.1and maximal heart rates no less than 2 BPM below age-predicted maximum heart rate. We used 2 x 2 analysis of covariance to examine group differences in physiological measurements between exercisers and controls in this subset of subjects versus the total sample at time 2 with baseline values partialled out. The results of these analyses indicate that subjects in the experimental and control conditions did not differ with regard to membership in this subset versus the total sample in VO2 max lf(1,242) = 0.09. P > 0.051 or V 0 2 at VeT lf(1,235) = 0.12, P > 0.051. To determine ventilatory threshold (VeT) both of the following criteria were used: (1) A systematic increase in the ventilatory equivalent for 0 2 (minute ventilation/ O2uptake) without an increase in the ventilatory equivalent for C02 (minute ventilation/C02 output); and (2) A systematic increase in the end tidal 0 2 partial pressure (Pet 02) without a decrease in end tidal C02partial pressure (Pet C O ~ ) lo .~, Methods for assuring interobserver and test-to-test reproducibility have been previously described as has the correlation of VeT with blood lactate levels in our laboratory.’ A further study was conducted to determine the test-to-test variability of maximal exercise responses: twelve healthy elderly subjects performed three maximal exercise tests in a span of no greater than 3 weeks which resulted in similar measurements of O2 uptake. (Hotelling T2’-’O = 1.46, P > 0.05). Pearson correlation coefficients indicate that measurements of O2 uptake among the testing occasions were highly correlated (Y > 0.97, P < 0.001). Blood pressures and heart rates were measured sitting, and results were reported as mean of values obtained during the first three sessions ( pre”) and last three sessions (“post”). Body composition measurements (weight, estimated percent body fat, estimated lean body mass) were assessed by weighing subjects and using a Lange skinfold caliper by methods described by Durnin and Fardy.*l,12 The same two observers performed all measures on all subjects.

the exercising and non-exercising subjects following 4 months of activities, a series of ANCOVAs were performed on resting, submaximal, and maximal physiological variables in which the dependent variable was the post-test score and the covariate was the pre-test score. The SPSSX procedure for ANCOVA was used which included an analysis of covariance and a test for the assumption of parallel slopes. If the f test for the test for the assumption of parallel slopes was significant, then the analysis of covariance to fit separate (nonparallel) slopes was reported. Otherwise, the f statistic reported pertains to the ANCOVA for parallel slopes. A dagger indicates f ratios from the separate slope analyses (Table 4). Violation of the assumption of the parallel slopes is of special interest due to the experimental design of the study because it indicates that change occurred at differing rates in the experimental and control groups. Means are reported -t- standard deviation.

Training Program Following completion of the screening protocol and baseline testing, subjects in the exercise program were assigned to small groups ( n = 4-6) for thrice weekly supervised exercise sessions for 16 weeks. Training was conducted on Monark cycle ergometers (Model 868) and monitored by a physical trainer. Each training session began with measurement of resting heart rate and sitting blood pressure and was followed by 5 minutes of unloaded cycling, 30 minutes of loaded cycling at training heart rate (THR), and 5 minutes of unloaded cycling. THR was the heart rate recorded at 70% of maximal oxygen uptake on the baseline exercise test. This heart rate was selected to ensure that subjects could stay within the program and exercise safely at home. Mean prescribed THR was 115 k 15. Heart rate during training sessions was monitored continuously with a heart rate monitor (Exersentry 11). Workload was adjusted during each exercise session to maintain training heart rate. Statistical Analysis To assess differences between

RESULTS Attrition rates of 8% were noted for the control group and 5% for the exercise group over the course of the 4-month experimental period. The mean participation rate (defined as the number of sessions attended divided by the number of sessions that it was possible to attend) was 67% for control subjects and 81% for the exercise group. During approximately 75 % of the training sessions, subjects exercised at or above their prescribed THR. Mean attained THR over the 4-month training period was 112 & 12. Only one subject was forced to stop training because of injury. Resting Parameters (Table 2) Significant group differences were demonstrated following training in body composition measurements (weight, estimated percent body fat, and estimated lean body mass). In the exercise group there was a slight increase in body weight, a small decrease in estimated percent body fat, and an increase in estimated lean body mass. Small but significant differences were found in resting systolic and diastolic blood pressures; there were declines in the exercise group and slight increases in the control group. No significant differences between groups were noted for resting heart rate or blood lipid values (total cholesterol, high density lipoprotein, low density lipoprotein, very low density lipoprotein, or triglycerides). Responses to Submaximal Exercise Throughout the 16-week supervised exercise program, work rate was adjusted to the level necessary to achieve the prescribed training heart rate (THR). Training work rate increased 20% from the beginning to the end of the study period. Training blood pressure, which was measured at the 20th minute of loaded cycling during each training session, declined in the exercise group by the 16th week of training as did 5-minute recovery blood pressure (Table 3). Both female and male exercisers had significant decreases in systolic (4.5 mm Hg f 2.2, mean +- SD) and

4

POSNER ET AL

Variable

Weight (kg) % Body fat Lean body mass (kg) Resting systolic blood pressure (mmHg) Resting diastolic blood pressure (mmHg)

JAGS-JANUARY 1992-VOL. 40, NO. 1

TABLE 2. CHANGES IN SELECTED VARIABLES AT REST Exercise Group (n = 166) Control Group (n = 81)

f*

Significance

68.9 f 14.3 28.6 f 8.6 49.5 k 12.0 129.9 k 11.2

4.03 5.93 8.62 24.38

P < 0.05

76.1 f 4.9

9.23

P < 0.01

Pre

Post

Pre

Post

68.9 f 11.6 28.3 f 7.7 49.5 k 10.7 128.6 k 12.1

69.4 f 11.7 27.5 -+ 7.7 50.5 f 10.9 125.3 & 12.3

69.1 f 14.6 28.5 f 8.1 49.4 k 12.2 127.6 k 10.4

75.1 f 6.4

73.6 k 6.2

74.9 f 5.9

P C 0.05 P < 0.01 P < 0.001

Data are expressed as mean + standard deviation. * f values indicate group differences at the post test with baseline values partialled out using analysis of variance.

TABLE 3. SUBMAXIMAL RESPONSES AND RECOVERY BLOOD PRESSURE IN EXERCISE GROUP Week 1 Week 16

by the control group after the training period (f = 42.88, P C 0.001). There was no significant increase in maximal heart rate after training and no significant difference in maximal heart rate between the exercise Training Work rate 47.1 f 18.6 56.4 f 22.8** and the control groups. Exercisers exhibited significant (watts) post-training improvements compared to controls in Training systolic blood 160.7 & 18.1 157.6 f 20.1* maximal oxygen pulse (P C 0.001) and maximal minute pressure ventilation (P C 0.001) and significant decreases in Training diastolic blood 75.7 f 7.4 71.8 f 6.5** maximal diastolic blood pressure (P < 0.001) (Table 5). pressure 5-Minute recovery sys132.5 f 13.4 127.8 k 12.9** A significant difference between groups was found tolic blood pressure for maximal oxygen uptake in mL/min (P < 0.001) as 5-Minute recovery dia74.4 f 6.3 71.9 f 6.5** well as when normalized for body weight (mL/kg/min, stolic blood pressure P < 0.001). Exercisers increased their VOz max after 16 t test probability: weeks of training by an average of 8.5% while the VOn P < 0.01 max of control subjects decreased slightly during the ** P < 0.001 same period (Table 6). Training work rate (watts) represents the work load necessary to maintain Relation of V e T to V 0 2 max The amount of work training heart rate throughout the 16-week program. Training systolic and diastolic blood pressure are measured at the 20th performed between VeT and V 0 2 max at the postminute of loaded cycling during each training session. training test differed significantly for exercisers and 5-minute recovery blood pressures are recorded 5 minutes after cessation controls (f(1,236) = 17.69, P < 0.001). Exercisers inof cycling following each exercise session. Values are means k SD with each weekly mean W e e k 1, Week 16) creased from 53.4 watts (k24.6) before training to 63.7 representing the average of 3 values. watts (k28.4) after training (P C 0.001) while controls showed essentially no change (50.3 k 24.9 to 51.0 +24.0). VOz max increased 8.5% in the exercisers while VeT diastolic (5.3 mm Hg f 3.4) blood pressures at sub- increased 3.3% so that the relation VeT/V02 max demaximal exercise levels after training. The decreases creased in the exercise group (P < 0.001); it did not were consistent over the different submaximal work change in the control group. The group interaction for levels. There were no changes at the second observa- VeT/V02 max was significant (f = 8.11, P < 0.001). tion in the control subjects. The changes were more DISCUSSION pronounced in the females. While downward trends were noted in heart rate at submaximal work rates in In this study, a large group of healthy older (mean both female and male exercisers, the interactions be- age = 69) subjects completed a 16-week low to modtween the exercise groups and control groups over time erate intensity exercise program with few injuries and were not significant (data not shown). high compliance. A second group completed a 16-week Ventilatory Threshold There were significant dif- period with once weekly sessions designed to render ferences between the exercise and control groups in only social contact. This attention control group had measures of VeT after the program: work at VeT in an attrition rate of only 8% and maintained an overall watts increased by 5% in exercisers and decreased by compliance of 67% demonstrating the feasibility of an 8% in controls; VOz VeT increased by 3.5% in exercis- exercise study necessitating a large contact control ers while it decreased by 3% in controls; similar group. changes were seen when VOz VeT was normalized for The intensity of exercise was fixed so that the preweight (Table 4). scribed training heart rate was the heart rate at 70% of Maximal Work Maximal work rate in watts in- the VOn max. This was approximately 45% of the creased by about 11% in the exercise group after train- MHRR, indicating a relatively low training inten~ity.'~ ing and was significantly different from that attained The use of a moderate intensity program was designed 73

PHYSICAL TRAINING IN HEALTHY OLDER ADULTS

IANUARY 1992-VOL. 40, NO. 1

TABLE 4. SELECTED MEASURES OF VENTILATORY THRESHOLD Control Group (n = 81) Exercise Group (n = 166) f* Pre Post Pre Post

Ventilatory threshold Work rate (watts) Ventilatory threshold oxygen-uptakemL/ min mL/kg/min

Ventilatory threshold oxygen uptake % VO

Significance

60.9 f 16.5

64.1 f 19.2

59.8 f 18.5

55.0 k 20.8

20.73

P < 0.001

964.7 f 242.3

997.4 f 231.1

964.7 k 262.7

935.4 f 244.9

15.89

P < 0.001

14.0 f 2.7 64.6 k 10.7

14.4 k 2.5 61.6 f 9.9

13.9 f 2.6 64.9 f 10.2

13.5 f 2.8 64.9 f 10.4

7.68t 8.11

5

P < 0.01 P < 0.01

2 max Data are expressed as mean + standard deviation. * f values indicate group differences at the post test with baseline values partialled out using analysis of covariance. t Fit separate slopes test.

TABLE 5. CHANGES IN SELECTED VARIABLES AT MAXIMAL EXERCISE Control Group ( n = 81) Exercise Group (n = 166) f* Significance Pre Post Pre Post Maximum workrate

113.9 k 34.8

126.8 k 38.4

(watts) Maximal 0 2 pulse (ml/o2/

beat) Maximal systolic blood pressure Maximal diastolic blood pressure Maximal minute ventilation (L/min)

10.7 f 3.6

11.1

* 3.5

108.2 f 35.3

105.2 zk 36.5

42.88

P < 0.001

10.6 f 3.7

10.0 f 3.4

16.06

P < 0.001

195.0 f 21.6

198.0 f 22.1

199.3 f 22.8

199.9 f 21.3

0.31

NS

91.2 f 12.2

8 6 . 2 3 12.3

92.5 f 10.4

91.8 zk 11.1

14.12

P < 0.001

54.6 f 19.5

62.1 f 22.8

53.6 k 19.7

54.6 f 19.3

13.49

P < 0.001

Data are listed as means + standard deviation. * f values indicate group differences at the post test with baseline values partialled out using analysis of covariance.

to make it easy for a large number of older subjects eventually to follow the program at home safely. The changes in resting systolic and diastolic blood pressures following the exercise program are similar to those previously described in studies of exercise programs of different intensities.14-16There was a slight decrease in estimated percent body fat which, coupled with the lack of change in body weight, suggests an increase in lean body ma5s.l’ Our finding that neither total cholesterol nor the high density lipid fraction changed a ees with previous studies in this aged populationgs and was not surprising considering the length of our exercise program and its intensity. In one study of previously sedentary middle aged men, for example, Wood et all9showed that a year long running program of less than 8 miles per week was ineffective in changing cholesterol levels, whereas one greater than 8 miles per week increased plasma HDL-cholesterol levels. Our subjects experienced an 8% increase in V 0 2 max. This is, not surprisingly, less than was experienced by similarly aged subjects studied by Thomas et a12who, after a year’s training at a much higher intensity program (exercise at 80% to 85% of V 0 2 max) experienced an 18% increase in V 0 2 max and only slightly less than the change experienced by Seals’ 11 subjects after 6 months of light training.’ Our subjects

showed a small (3%) change in VeT as did Thomas’ (5%). In both this study and Thomas’, V 0 2 max increased more than VeT (the ratio VeT/V02 max decreased). This is in contrast to findings in younger subjects; a group of female college students exercised by Gibbons et aI2’ at an intensity comparable to the one used in this study showed an increase of 16% in V 0 2 max and 23% in VeT; a group of middle aged men (mean age = 43) exercised by Davis21at an intensity comparable to those in Thomas’ study showed an increase of 25% in V 0 2 max and 44% in VeT. In both studies, the ratio VeT/V02max increased after exercise. Our results and Thomas’ are in contrast to those of Blumentha13 whose older subjects experienced an 11.6%increase in V 0 2 max and a 13% increase in VeT after 4 months of an exercise program of an intensity closer to Thomas’ than to ours.3 Our finding of a decrease in VeT/V02 max instead of an increase is important as it raises the question of mechanisms of adaptations to training which may be different in older exercisers than in younger ones. VeT is the point at which measurable glycolytic metabolism begins. Work done up to VeT is essentially oxidative, whereas work done between the onset of VeT and VOz max is both oxidative and glycolytic (see Figure 1).This would mean that improvements purely in oxygen delivery and utilization (changes in cardiac output, pul-

6

POSNER ET AL

IAGS-IANUARY 1992-VOL.40, NO. 1

TABLE 6. CHANGES IN MAXIMUM ATTAINED OXYGEN UPTAKE Exercise Group (n = 166) Control Group (n = 81) mL/min mL/kg/min

~~

Pre

Post

Pre

Post

P

Significance

1538.9 f 501.8 22.2 f 5.7

1669.9 f 550.8 23.9 f 6.1

1518.2 k 504.6 21.8 f 5.3

1466.4 & 505.7 21.1 f 5.2

34.95 (1,244) 28.18 (1,244)

P < 0.001 P < 0.001

Data are expressed as mean + standard deviation. * f values indicate group differences at the post test with baseline values partialled out using analysis of covariance.

s

Oxidative and

.-

&

Q,

E C 3 .-

glycolytlc

30

metabolism

20

Primarily oxidative metabolism

EE

O P V Y

e

l

X

0

0

Pre

Post

FIGURE 1. Oxygen consumption in older subjects before and after exercise training. Work performed up to ventilatory threshold (VeT) is supplied primarily from oxidative metabolism. Work performed from VeT to maximum oxygen consumption (VOzmax) is supplied by oxidative and glycolytic metabolism.

(aged 60-72) who increased their VOz max by cycle ergometry with no change in VOz max by arm ergometry. Such training does not increase V 0 2 max in young and middle aged men.’6-28 It is also known that a program like ours can have a beneficial effect on peripheral muscles in older subjects but not younger ones; Meredith et a129found that 10 elderly subjects (mean age = 65) responded to a training program similar to ours with increased glycogen muscle stores and increased muscle oxidative capacity, while young subjects (mean age = 24) did not exhibit these peripheral changes. Thus, it is possible that a low to moderate intensity training program like ours causes an adaptation in muscle in the elderly which modifies the ability of that muscle to deal with the byproducts of glycolysis and therefore increases the work that can be performed above VeT. This increase in ability to perform work would then result in an increase in V 0 2 max. Whatever the mechanism, the changes we have seen, albeit small, may have significance. We have shown that our exercise subjects not only exhibited the physiologic changes we have described here, including a 20% increase in training work rate, but also exhibited, over the two years following training, a significant reduction in new-onset cardiovascular diagnoses compared to our control subjects. This was true not only of a group that trained during the entire 2 years, but also of a subgroup that stopped after the 4-months program herein de~cribed.~’ This study showed that a large-scale moderate intensity training program for older people is practical and results in measurable physiologic improvements and that these improvements differ from those seen in younger adults, suggesting different mechanisms of adaptation between the two age groups.

monary mechanics, ventilation/perfusion matching in the lung, oxygen carrying capacity of the blood, shunting of blood to exercising muscle, ability of muscle to utilize oxygen) would improve VeT proportionately more than VOz max (and increase the ratio VeT/V02 max). This is the kind of improvement seen in young and middle aged subjects following an endurance exercise program. Our older subjects and Thomas’ increased VOn max to a greater degree than VeT (and decreased the ratio VeT/V02 max). This adaptation can not be purely one of improvement in oxygen delivery. On the other hand, an increase in the ability of the quadriceps, hamstrings, and calves to perform work could account for the improvements we have seen. The byproducts of glycolysis (hydrogen ions and inorganic phosphate) interfere with muscle metabolism.” The ACKNOWLEDGMENTS buildup of these byproducts limits the amount of total work that muscles can perform. We postulate that our We would like to thank Kevin McCully for reading program increased the ability of the peripheral muscles the manuscript. to buffer the by-products of glycolysis, allowing these muscle to perform more total work. The increase in REFERENCES total work performed would then increase the maximal 1. Seals DR, Hagberg JM,Hurley BF et al. Endurance training in older men oxygen consumed. and women: 1. Cardiovascular responses to exercise. J Appl Physiol This would be consistent with certain other obser1984;57(4):1024-1029. vations. It is known that glycolytic fast-twitch muscle 2. Thomas SG, Cunningham DA, Thompson J, Rechnitzer PA. Exercise fibers atrophy with age,23and we8 and C~nningham’~ training and ‘ventilation threshold“ in elderly. J Appl Physiol 1985;59(5):1472-1476. have shown that age is marked by a loss of ability to 3. Blumenthal JA,Emery CF, Madden DJ et al. Cardiovascular and behavioral perform work after VeT. It has also been shown that effects of aerobic exercise training in healthy older men and women. J Gerontol 1989;44(5):M147-157. changes in peripheral muscle can have an effect on Donner AP. Exercise train4. Cunningham DA, Rechnitzer PA, Howard JH, VOz max in older subjects. Frontera et alZ5provided ing of men at retirement: A clinical trial. J Gerontol 1987;42(1):17-23. strength training for the quadriceps of 12 older subjects 5. Foster VL, Hume GJE, Bymes WC et al. Endurance training for elderly

IANUARY 1992-VOL. 40, NO. I women: Moderate vs low intensity. J Gerontol 1989;44(6):M184-178. 6. Posner JD, Gorman KM, Klein HS, Woldow A. Exercise capacity in the elderly. Am J Cardiol 1986;57:52C-58C. 7. Buchfuhrer MJ, Hansen JE, Robinson TE. Optimizing the exercise protocol for cardiopulmonary assessment. J Appl Physiol 1983;55(5):1558-1564. 8. Posner JD, Gorman KM, Klein HS, Cline CJ.Ventilatory threshold Measurement and variation with age. J Appl Physiol 1987;63(4):1519-1525. 9. Wasserman K. The anaerobic threshold measurement to evaluate exercise performance. Am Rev Respir Dis Suppl 1984;129:S35-S40. 10. Whipp BJ, Davis JA, Torres F, Wasserman K. A test to determine parameters of aerobic function during exercise. J Appl Physiol 1981;50:217-221. 11. Dumin JVGA, Rahaman MM. The assessment of the amount of fat in the human body from measurements of skinfold thickness. Br J Nutr 1967;21:681. 12. Fardy PS, Hellerstein HK. National Exercise and Heart Disease Project: Procedures for obtaining anthropometric measurements. Cleveland: CWRU School of Medicine. 1974. 13. Karvonen MJ, Kentala E, Mustala 0. The effects of training on heart rate. Ann Med Exp Biol Fenn 1957;35:305-307. 14. Stamford BA. Physiologic effects of training upon institutionalizedgeriatric men. J Gerontol 1972;27:451-455. 15. Barry AJ, Daly JW, Pruett EDR et al. The effects of physical conditioning on older individuals. I. Working capacity, circulatory-respiratoryfunction and work electrocardiogram.J Gerontol 1966;21:182-191. 16. deVries HA. Physiological effects of an exercise training regimen upon men aged 52-88. J Gerontol 1970;25:325-336. 17. Sidney KH, Shephard RJ, Hamson JE. Endurance training and body composition of the elderly. Am J Clin Nutr 1977;30:326-333. 18. Van Der Eems K, Ismail AH. Serum lipids: Interactions between age and moderate intensity exercise. Br J Sports Med 1985;19(2):112-114. 19. Wood PD, Haskell WL, Blair SN et al. Increased exercise level and plasma

PHYSICAL TRAINING IN HEALTHY OLDER ADULTS

20. 21. 22.

23. 24. 25. 26. 27. 28. 29. 30.

7

lipoprotein concentrations: A one-year, randomized, controlled study m sedentary, middle-aged men. Metabolism 1983;32(1):31-39. Gibbons ES, Jessup GT, Wells TD, Werthmann DA. Effects of various training intensity levels on anaerobic threshold and aerobic capacity in females. J Sports Med 1983;23:315-318. Davis JA, Frank MH, Whipp BJ, Wasserman K. Anaerobic threshold alterations caused by endurance training in middle-aged men. J Appl Physiol 1979;46(6):1039-1046. Wilkie DR. Muscular fatigue: Effects of hydrogen ions and inorganic phosphate. Fed Proc 1986;2921-2923. Tomonaga M. Histochemical and ultrashuctural changes in senile human skeletal muscle. J Am Geriatr SOC1977;25:125-131. Cunningham DA, Nancekievill EA, Paterson DH et al. Ventilation threshold and aging. J Gerontol 1985;40(6):703-707. Frontera WR, Meredith CN, OReilly KP, Evans WJ. Strength training and determinants of VO2 max in older men. J Appl Physiol 1990;68(1):329333. Hickson RC, Rosenkoetter MA, Brown MM. Strength training effects on aerobic power and short term endurance. Med Sci Sports Exerc 1980;12:336-339. Hurley BF, Seals DR, Ehsani AA et al. Effects of hizh-intensitv s t r e n ~ h training on cardiovascular function. Med Sci SportsExerc 1984;16:48y3488. Rutherford OM, Greig CA, Sargeant AJ, Jones DA. Strength training and power output: Transference effects in the human quadriceps muscle. J Sports Sci 1986;4:101-107. Meredith CN, Frontera WR, Fisher EC. Peripheral effects of endurance training in young and old subjects. J Appl Physiol 1989;66(6):2844-2849. Posner JD, Gorman KM, Gitlin LN et al. Effects of exercise training in the elderly on the occurrence and time to onset of cardiovascular diagnoses. J Am Geriatr SOC1990;38:205-210.