the cardiopulmonary response to incremental exercise test: the effect of aging
TRANSCRIPT
Aging Clin. Exp. Res. 6: 267-275, 1994
The cardiopulmonary response to incremental exercise test: The effect of aging L. Fuso1, R. Antonelli Incalzi2, R. Muzzolonl, M. Di Gennaro2, F. Gliozzil, R. Pistellil, and P.U. Carbonin2
lOepartment of Respiratory Physiology, 20epartment of Geriatrics, Catholic University, Rome, Italy
ABSTRACT The aims of the present study were to define the respective roles of the cardiac and respiratory response to exercise as determinants of the age-related physiological decrease in exercise performance, and to assess the relationship between aging and interindividual variability in the response to effort. We studied.91 normal subjects recruited in three age-groups: Group A (42 children, aged 10±2 years); Group B (29 young adults, aged 27±5 years); Group C (20 elderly, aged 74±9 years). All the subjects underwent an incremental cycle ergometer exercise test with a work load increase of 15 W every 2 minutes in groups A and C, and 25 W every 2 minutes in group B, until they achieved 80% of the predicted maximal heart rate. Ventilatory equivalent changes during exercise were significantly lower in group A than in the other two groups, and in group B compared to group C. Exercise-induced changes in oxygen pulse were significantly higher in group A, but no difference was found between groups Band C. Thus, gasexchange function and overall exercise performance decrease with advancing age, whereas cardiovascular performance is well maintained in normal elderly subjects. Discriminant analysis showed that the exercise response conformed to the group-specific model in 74% and 79% of subjects in groups A and B, but only in 50% of the group C subjects; 5% and 45% of the elderly subjects were functionally classified in groups A and B, respec-
tively. On the basis of these data, it may be concluded that aging accounts for a dramatic increase in interindividual variability in adaptation to physical effort, and that the inverse relationship between age and exercise performance is mainly related to the declining efficacy of the respiratory response to effort with age. (Aging Clin. Exp. Res. 6: 267-275, 1994)
INTRODUCTION
The maximal oxygen uptake (V02max) represents the maximal ability of the cardiorespiratory system to deliver oxygen to the periphery and the capacity of muscles to extract oxygen from the blood. Thus, both central and peripheral factors are the determinants of V02max (1). The age-related physiological decrease in V02max and exercise capacity has long been attributed to a progressively decreasing maximal cardiac output (2-4). However, it was recently recognized that advancing age affects the maximal cardiac output only marginally, becau,se the Frank-Starling mechanism compensates for the age-related decrease in chronotropic response to effort (5-7). Accordingly, the previously reported negative effect of advancing age on maximal cardiac output is likely to depend on silent coronary artery disease and/or physical deconditioning (6). Therefore, attention should be focused on the other determinants of V02max, Le., pulmonary and muscular factors, whose relative
Key words: Aging, exercise test, interindividual variability, maximal oxygen uptake. Correspondence: Dr. Leonello Fuso, Fisiopatologia Respiratoria, Universita. Cattolica S. Cuore, Largo A. Gemelli 8, 00168 Roma, Italy. Received December 16, 1992; accepted October 5, 1993.
Aging Clin. Exp. Res., Vol. 6, No.4 267
L. Fuso, R. Antonelli Incolzi, R. Muzzolon, et 01.
importance has not been well defined in older people.
That interindividual variability also increases with age is true for many physiological functions (8) and exercise capacity as well (9, 10). This finding may be relevant to the actual value and use of normal standards and reference equations for the incremental exercise test (11). However, the proportion of elderly normal subjects whose exercise performance does not conform to a model considered typical for their age is not clear from the literature.
The present study was designed to clarify the effects of aging on the respiratory and cardiac response to exercise, and the relationship between aging and interindividual variability in adaptation to effort.
SUBJECTS AND METHODS Study population Three groups of normal subjects were studied: - Group A consisted of 42 male children, aged
10.3± 1.8 years, range 8-14 years; - Group B was made up of 29 young adults,
14 males and 15 females, aged 26.9±4.8 years, range 19-40 years; and
- Group C comprised 20 elderly persons, 10 males and 10 females, aged 73.6±9.1 years, range 61-90 years.
Three subjects in group B and six in group C had stopped smoking for more than 5 years; two subjects in group B were still smoking less than 10 cigarettes a day.
A written informed consent was obtained from all study participants, or their parents. All subjects underwent a complete cardiac and respiratory evaluation based on clinical history, physical examination, ECG recording, 20 echocardiogram, and spirometry. For these last two tests, normal reference values were obtained from Feigenbaum (12) and Knudson (13), respectively. The folloWing parameters were also determined in groups Band C: hemoglobin, red and white blood cell counts, hematocrit, corpuscular values, serum glucose, nitrogen, creatinine, sodium, potassium, calcium, phosphorus, albumin, aminotransferases, ammonia, bilirubin, cholesterol, and triglycerides.
Exclusion criteria were:
268 Aging Clin. Exp. Res., Vol. 6, No.4
1) any evidence of cardiac and/or pulmonary diseases according to history, clinical findings, ECG, 20 echocardiogram, spirometry;
2) any abnormality in one or more of the biochemical or hematologic parameters;
3) historical evidence of a completely sedentary life-style or heavy physical activity. Both B and C subjects walked daily for about two miles in 1 hour;
4) historical data and/or symptoms or physical findings of neurologic or orthopedic disease;
5) signs of malnutrition according to Mitchell et al. (14).
Exercise test An incremental work test was performed on a
calibrated, electrically braked cycle ergometer Dynavit Conditronic 30 (Keiper Oynavit, Kaiserslautern, Germany). The initial work rate of 25 Watts was increased by 15 Watts every 2 minutes in groups A and C, and by 25 Watts every 2 minutes in group B. Subjects exercised until they achieved 80% of the calculated maximal heart rate; the corresponding V02 (V02-80%HR) rather than V02max was employed as the last stage of the exercise in an attempt to obtain linear relationships between work power and physiological variables and, thus, improve discrimination among groups.
Additional criteria to stop the test were serious cardiac arrhythmias, hypotension, severe exercise-related symptoms, and muscular exhaustion.
Data collection and analysis Subjects breathed through a low resistance,
unidirectional valve (Hans Rudolph Inc, Kansas City, MO, U.S.A.) connected to a 5-L mixing chamber, and to a Fleisch No.3 pneumotachograph for respiratory flow measurement. Expired gas samples from the mixing chamber were analyzed for oxygen (02) and carbon dioxide (C02) content by a quadrupole mass spectrometer Airspec 2000 (Airspec Ltd, Biggin Hill, Kent, u.K.). Measurements were taken before exercise, while each subject was comfortably seated on the cycle ergometer, and during the last 30 seconds of each work step. Automatically acquired data were processed on line by an HP-9825A computer, running an original software
program for the measurement of pulmonary ventilation (VE) , frequency of breathing (f), tidal volume (VT), alveolar ventilation (VA), inspiratory to total respiratory cycle duration ratio (TIff TOT), mean inspiratory flow (VTffI), O2 uptake (V02), CO2 output (VC02), and VC02/V02 ratio, i.e., respiratory quotient (QR). Blood pressure was recorded at each work step; the heart rate (HR) and the electrocardiogram were continuously monitored.
The exercise-related changes in ventilatory equivalent for oxygen (VE/V02) and oxygen pulse (V02/HR) were calculated from the previous variables, and were both corrected for body weight. It is noteworthy that a relatively low VE/V02 value states that a given V02 can be achieved without a very large increase in VE,
which is consistent with an efficient respiratory response to exercise (15-18). The efficiency of the cardiac response to exercise is directly proportional to V02/HR, since high V02/HR values mean that no large increase in HR is required to achieve a given V02 (15-18).
Statistical analysis Linear regression analysis was used to assess
the relationship between the physiological variables and the work load. The validity of the linear models was estimated on the basis of r-squared values (19).
Differences between groups were assessed by one-way analysis of variance (ANOVA) and Tukey's multiple range test (19).
Effect of aging on exercise response
We used the discriminant analysis in order to define the correct classification of the subjects into the three age groups (19,20).
This technique proVided a cross-tabulation of the actual group membership us the predicted group membership, as identified by the effects of exercise on independent variables recorded during the test. Linear combinations of these variables are formed in discriminant functions, and serve as the basis for classifying cases into one of the groups. Wilks' lambda coefficient was used to test the significance of the discriminant functions. The homogeneity of the exercise response of each group was expressed by the percentage of concordance between actual and predicted group membership (20).
RESULTS Table 1 reports the anthropometric variables
and respiratory functional parameters in each group of subjects.
The test was stopped before completing the 2-minute last stage in 2 subjects of group B, in 4 of group C because of muscular exhaustion, and in 1 subject of group C because of repetitive ventricular premature complexes. However, these subjects were not excluded from the study because they reached 80% of the calculated maximal HR.
The results of linear regression analysis are summarized in Table 2. The physiological response to work load of almost all the variables was accurately described by the linear models (r-
Table 1 - Anthropometric and respiratory functional data of the population studied.
Group A Group B Group C Children Adults Elderly
(mean±SD) (mean±SD) (mean±SD)
No. 42 29 20
Height (em) 14S.6±13.7 169.1±10.8 161.3±8.7
Weight (kg) 40.6±11.3 63.7±11.2 69.9±9.1
FVC (% pred) 94.9±9.S 106.3±13.6 109.1±21.6
FEVI (% pred) 89.6±8.9 100.4±13.1 101.7±21.3
FEVl/FVC (%) 86.2±S.4 81.4±S.8 74.6±9.7
FVC=forced vital capacity; FEVl=forced expiratory volume in 1 second.
Aging Clin. Exp. Res., Vol. 6, No.4 269
L. Fuso, R. Antonelli Incalzi, R. Muzzolon, et al.
Table 2 - Mean values of r-squared derived from linear re-gressions having work load as independent variable.
r-squared Group A Group B Group C Children Adults Elderly
VE 0.96 0.96 0.96
VT 0.86 0.89 0.86
0.74 0.70 0.71
VTff, 0.95 0.95 0.95
T,ffTOT 0.42 0.22 0.65
V02 0.97 0.98 0.95
VC02 0.96 0.96 0.94
HR 0.96 0.94 0.94
VE=pulmonary ventilation; VT=tidal volume; f=frequency of breathing; VTffI=mean inspiratory flow; TIff TOT=inspiratory to total respiratory cycle duration ratio; VOz=Oz uptake; VCOz=COz output; HR=heart rate.
LminfmL x kg-1/Walt
2.5
2
1.5
0.5
O...l...---
ANOVA: 71.2; p<O.001
_ Children c::J Adults DElderly
Values expressed as mean ± SD
Figure 1 - Comparison between groups in the exercise-related changes of ventilatory equivalent.
270 Aging Clin. Exp. Res., Vol. 6, No.4
mUHR x kg-1/Watt
0.5
0.4
NS
0.3
0.2
0. 1
0--'------
ANOVA: 17.8; p<O.001
Children o Adults o Elderly
Values expressed as mean ± SD
Figure 2 - Comparison between groups in the exercise-related changes of oxygen pulse. NS: not significant difference between adults and elderly (Tukey's multiple range test).
squared >0.70). Only the exercise-related changes in T,/TToT were poorly correlated with work load. The facial mask supporting the Hans Rudolph valve may have contributed to increase the well-known interindividual variability in T,/TTOT (21, 22), and to weaken the correlation between this parameter and work load (23). For this reason, this variable was not further used in the discriminant analysis. The linear relationship between almost all variables and the work load (Table 2), and the achievement of 80% of the predicted maximal HR in all subiects, are reliable indicators that effective tests have be~· performed (15). ~I
Figure 1 shows that the exercise-induced changes in ventilatory equivalent were signifi~ cantly lower in group A than in group B, and in group B than in group C (group A: 1.16±0.3; group B: 1. 73±0.3; group C: 2.11±0.4 Lmin/mL x kg-l/Watt; ANOVA: 71.2, p<O.OOl).
Exercise-induced changes in oxygen pulse were higher in group A than in groups Band C (group A: 0.37±0.1; group B: 0.27±0.1; group C: 0.28±0.1 mL/HRxkg-1/Watt; ANOVA: 17.8, p<O.OOl) , but no significant difference was found between groups Band C (Fig. 2).
Figure 3 shows that the breathing pattern of group A subjects during the test was characterized by a significantly higher increase in f compared to groups Band C (group A: 0.18±0.04; group B: 0 .04±0.05; group C: 0 .09±0.05 f-min/Watt; ANOVA: 20.7, p<O .OOl) . In these two latter groups, the exercise-induced increase in VT was significantly higher than in group A (group A: 0.0051±0.001, group B: 0.0086±0.003; group C: 0.0083±0.003 L/Watt; ANOVA: 23.8 , p<O.OOl). Groups Band C did not differ either in f or in VT effort-related changes.
FREQUENCY OF BREATHING
f-minlWatt
0.25
0.2
0.15 NS
0 .1
005
0 ...1....---
ANOVA: 20.7; p<0.001
Effect of aging on exercise response
Figures 4 and 5 show the linear relationships between VE and V02 and between VE and VT in the elderly group, where these relations were significant.
Results of discriminant analysis are reported in Tables 3 and 4. Two discriminant functions were obtained, functions 1 and 2, accounting for 70% and 30% of the total between group variability, respectively. The Eigenvalue, i.e., the ratio of intergroup to intragroup sums of squares, was higher for function 1 (0.87 us 0.37), which is consistent with function 1 achieving a better discrimination. Wilks ' lambda, i.e. , the ratio of the intragroup sum of squares to the total sum of squares, was lower for function 1. This finding means that the intragroup variability was small compared to total variability, the latter depending mostly upon differences among groups (20) .
TIDAL VOLUME
lJWatt
0.010
0.008
0.006
0.004
0 .002
0...1....---
ANOVA: 23.8; p<0.001
_ Children c:::::=J Adu Its c:::::=J Elderly
Values expressed as mean ± SD
Figure 3 - Comparison between groups in the exercise-related changes of breathing pattern: frequency of breathing (left panel) and tidal volume (right panel). NS: not significant difference between adults and elderly (Tukey 's multiple range test).
Aging Clin. Exp. Res. , Vol. 6 , No. 4 271
L. Fuso, R. Antonelli Incalzi, R. Muzzolon, et af.
r=O.88 2.1
0
1.5
~ '" 0
> 0.9
0
0.3 (;I
0.13 0.21 0.29 0.37 0.45
VEIW
Figure 4 - Experimental points, regression line, and correlation coefficient of the linear relationship between VE and V02 exercise-related changes, in the elderly group.
Both functions discriminated the groups significantly, although function 1 had a better discriminant power (Table 3). Exercise-related
Table 3 - Variables and discriminant functions used to discriminate groups.
Function Function 1 2
Coefficients
V E exercise-related changes 0.568 0.701
VTffj exercise-related changes 0.125 0.089
VOz exercise-related changes 0.491 -0.165
VCOz exercise-related changes -1.045 -0.494
HR exercise-related changes -0.623 -0.315
f exercise-related changes -0.618 0.658
Eigenvalue 0.87 0.37
Relative percentage of the total between groups variability 70.37 29.63
Wilks'lambda 0.39 0.73
p <0.0001 <0.0001
VE=pulmonary ventilation; VTifj=mean inspiratory flow; V02=02 uptake; VC02=C02 output; HR=heart rate; f=frequency of breathing.
272 Aging Clin. Exp. Res., Vol. 6, No.4
r=O.71 1.4
0
0 0
1.0 0 0 0 0 0
~ >
0.6
0
0 0.2
0
0.13 0.21 0.29 0.37 0.45
VEIW
Figure 5 - Experimental points, regression line, and correlation coefficient of the linear relationship between VE
and VT exercise-related changes, in the elderly group.
changes in VC02, HR, f and VE were the main determinants of the discriminant functions, i.e., of the differences among groups, as reflected by the magnitude of their respective discriminant function coefficients (20).
Table 4 shows a cross-tabulation of the actual group membership us the group membership predicted by the comparative analysis of exerciserelated changes in VE, VTffJ, V02, VC02, HR, and f. The concordance between actual and predicted group membership was high in both groups A (73.8%) and B (79.4%). On the other hand, the response to exercise conformed to
Table 4 - Percentages of concordance between actual and predicted group membership.
Actual No. of Predicted group group cases membership
Group A Group B Group C
Group A 42 31 (73.8%) 10 (23.8%) 1 (2.4%) (Children)
Group B 29 3 (10.3%) 23 (79.4%) 3 (10.3%) (Adults)
Group C 20 1 (5%) 9(45%) 10 (50%) (Elderly)
the group-specific pattern in only SO% of the group C subjects, whereas in the remaining 4S% and S% it conformed to group B and A pattern, respectively.
DISCUSSION
The Widely recognized inverse relationship between maximal exercise performance and age is generally attributed to a progressive decline in the functional capacity of the cardiovascular and respiratory systems and to a loss of muscle mass. Our results show a significant decline in overall respiratory efficiency with advancing age. In fact, the exercise test was performed by elderly subjects with a maximum increase in ventilatory equivalent (Fig. 1). On the contrary, in children, the srri~ller increase in ventilation needed to achieve a given increase in oxygen consumption, clearly indicates low resistances to oxygen flow, and a better exercise performance. Thus, the ventilatory equivalent can also be considered as an indicator of overall exercise capacity (IS, 18).
The decreased efficacy of pulmonary gas-exchange is the most likely explanation for the Significant increase with aging of the ventilatory equivalent changes during exercise. Indeed, the age-related decrease in the elastic properties of the lungs can explain an increased closing capacity of the airways and a lower arterial oxygen tension in elderly normal subjects (24, 2S). Our results confirm that aging affects the efficacy of the pulmonary gas-exchange mechanisms (26), and indicate that aging plays an important role in work capacity decline.
However, peripheral factors may partly account for the direct relationship between aging and ventilatory equivalent. It is known that the age-related changes in muscle power and metabolism, as well as oxygen transport mechanisms are important factors in decreasing the maximal work capacity with aging (6) because they reduce the arteriovenous O2 difference and diminish the capacity of exercising muscles to extract O2 from the blood (1). The oxygen supply pathway may be altered with advancing age in this way, and consequently cause a ventilation excess per unit of oxygen consumption, i.e., an increase in the ventilatory equivalent during exer-
Effect of aging on exercise response
cise. However, our results do not allow us to assess the role of the peripheral determinants of the ventilatory equivalent.
Our data exclude that a deficiency in stroke volume is able to explain the decline in exercise performance with aging. Stroke volume is related to the oxygen pulse during exercise (IS, 27) and, in our sample, no significant difference in oxygen pulse was found between young adult and elderly subjects during exercise (Fig. 2). These data confirm recent reports (S-7), and suggest that the cardiac response to exercise is fairly good in healthy elderly subjects.
Finally, it is likely that the ventilatory response, as defined by f and VT , is uninvolved in the decrease in exercise performance with aging, as no difference was found between adult and elderly subjects in the breathing pattern changes during exercise (Fig. 3).
The second goal of this study was to assess interindividual variability in response to effort in older people. For this reason, although the sample studied was not very large, care was taken to recruit subjects with a wide age distribution to obtain a good separation between groups. Moreover, subject fitness and level of habitual activity were quite homogeneous within each age-group, as was sex distribution, except for the pediatric group in which all subjects were male. However, this absolute predominance of males in group A did not affect the comparison between groups, as it is known that the response to exercise is unrelated to sex until puberty (IS). Our data confirm the increase in variability with aging, and show that the percentage of variability is considerable in the geriatric population. The results of discriminant analysis clearly indicated that the children and young adults were correctly classified in the proper functional group with quite comparable accuracy. On the contrary, elderly subjects suffered from a significant misclassification; only SO% of these subjects showed a response to effort which conformed to that considered typical for this age group, whereas 4S% performed like young adults and, in one case, as children (Table 4). A careful analysis of the basic anatomic and physiologic features and life-style of the misclassified elderly subjects could not detect any difference between these and the correctly classified elderly persons. Thus, the individual response of
Aging Clin. Exp. Res., Vol. 6, No.4 273
L. Fuso, R. Antonelli Incalzi, R. Muzzolon, et al.
the healthy elderly person to exercise seems to be unpredictable.
Exercise-related changes in VC02, HR, f and V E were the main determinants of intergroup variability (Table 3). The well-known inverse relationship between aging and exercise-related increase in VC02 (18), and the age-related differences between cardiac and respiratory adaptation to exercise account for the high discriminant power of these variables.
The increase in interindividual variability with aging lends support to the findings of Jones et al. (9) that the prediction equations for maximal exercise capacity lose their value when used on populations at age and height extremes. Considering these results, these workers derived new reference equations based on an interactive and nonlinear influence of age on exercise capacity (9).
Some limitations in our study deserve mention. Firstly, silent coronary artery disease could not be completely excluded because myocardial scintigraphy was not performed. However, cardiac performance, as indicated by the oxygen pulse, corresponded to a normal heart function. Secondly, the cardiac response to exercise was not assessed in as much detail as the respiratory response, because only noninvasive measurements were taken.
Despite these limitations, the present study confirms that age is characterized by a large increase in interindividual variability and a decreased efficacy of the respiratory response, and thus helps to clarify the relationship between age and exercise capacity. These findings should be considered when exercise test results are interpreted. Future studies should aim at confirming the role possibly played by muscle factors and the O2 transport mechanisms in influencing the age-related decrease in V02max and exercise capacity.
ACKNOWLEDGEMENT The authors wish to thank Miss Julie A. Karimi for her
help in preparing the manuscript.
REFERENCES 1. Wagner p.o., Hoppeler H., Saltin B.: Determinants of
maximal oxygen uptake. In: Crystal R.G., West J.B. (Eds.), The lung: scientific foundations. Raven Press, New York, 1991, pp. 1585-1593.
274 Aging Clin. Exp. Res., Vol. 6, No.4
2. Gerstenblith G., Lakatta E.G., Weisfeldt M.L.: Age changes in myocardial function and exercise response. Prog. Cardiovasc. Dis. 19: 1-21, 1976.
3. Strandell T.: Circulatory studies on healthy old men. Acta Med. Scand. 175 (Suppl. 414): 1-44, 1964.
4. Conway J., Wheeler R., Sannerstedt R.: Sympathetic nervous activity during exercise in relation to age. Cardiovasc. Res. 5: 577-581,1971.
5. Rodeheffer R.J., Gerstenblith G., Becker L.C, Fleg J.L., Weisfeldt M.L., Lakatta E.G.: Exercise cardiac output is maintained with advancing age in healthy human subjects: cardiac dilatation and increased stroke volume compensate for a diminished heart rate. Circulation 69: 203-213, 1984.
6. Fleg J.L., Goldberg A.P.: Exercise in older people: cardiovascular and metabolic adaptations. In: Hazzard W.R., Andres R., Bierman E.L., Blass J.P. (Eds.), Principles of geriatric medicine and gerontology, ed. 2. McGraw-Hill, New York, 1990, pp. 85-100.
7. Weisfeldt M.L., Lakatta E.G., Gerstenblith G.: Aging and the heart. In: Braunwald E. (Ed.), Heart disease, ed. 4. WB Saunders, Philadelphia, 1992, pp. 1656-1669.
8. Rowe J.W.: Aging and geriatric medicine. In: Wyngaarden J.B., Smith L.H. Jr, Bennett J.C (Eds.), Cecil Textbook of medicine, ed. 19. WB Saunders, Philadelphia, 1992, pp. 21-26.
9. Jones N.L., Summers E. , Killian KJ.: Influence of age and stature on exercise capacity during incremental cycle ergometry in men and women. Am. Rev. Respir. Dis. 140: 1373-1380, 1989.
10. Astrand P.O., Rodahl K: Textbook of work physiology, ed. 3. McGraw-Hill, New York, 1986.
11. Jones N.L., Makrides L., Hitchcock C, Chypchar T., McCartney N.: Normal standards for an incremental progressive cycle ergometer test. Am. Rev. Respir. Dis. 131: 700-708,1985.
12. Feigenbaum H.: Echocardiography, ed. 4. Lea & Febiger, Philadelphia, 1986.
13. Knudson R.J., Slatin R.C, Lebowitz MD., Burrows B.: The maximal expiratory flow-volume curve. Normal standards, variability and effects of age. Am. Rev. Respir. Dis. 113: 587-600, 1976.
14. Mitchell CO., Lipschitz D.A.: Detection of protein-calorie malnutrition in the elderly. Am. J. Clin. Nutr. 35: 398-406, 1982.
15. Jones N.L.: Clinical exercise testing, ed. 3. WB Saunders, Philadelphia, 1988.
16. Wasserman K, Whipp B.J.: Exercise physiology ill health and disease. Am. Rev. Respir. Dis. 112: 219-249, 1975.
17. Spiro S.G.: Exercise testing in clinical medicine. Br. J. Dis. Chest 71: 145-172, 1977.
18. Mellerowicz H., Smodlaka V.N.: Ergometry. Basics of medical exercise testing. Urban-Schwarzenberg, Baltimore, 1981.
19. Snedecor G.W., Cochran W.G.: Statistical methods. The Iowa State University Press, Ames, Iowa, 1967.
20. Norusis M.J.: Predicting cure and credit: discriminant analysis. In: SPSS/PC+ Advanced Statistics V.2.0. SPSS Inc., Chicago, 1988, pp. B1-B39.
21. Jones N.L.: Physiological basis of exercise testing. In: Fishman AP. (Ed.), Pulmonary diseases and disorders, ed. 2. McGraw-Hill, New York, 1988, pp. 235-250.
22. Gallagher c.G.: Exercise and chronic obstructive pulmonary disease. Med. Clin. North Am. 74: 619-641, 1990.
23. Askanazi J., Silverberg P.A, Foster RJ., Hyman AI., Milic-Emili J., Kinney J.M.: Effects of respiratory ap-
Effect of aging on exercise response
paratus on breathing pattern. J. Appl. Physio!.: Respirat. Environ. Exercise Physiol. 48: 577-580, 1980.
24. Anthonisen N.R, Danson J., Robertson P.c., Ross W.RD.: Airway closure as a function of age. Resp. Physio!. 8: 58-65, 1969.
25. Nunn J.F.: Applied respiratory physiology, ed. 3. Butterworths, London, 1987.
26. Knudson RJ.: Physiology of the aging lung. In: Crystal RG., West J.B. (Eds.), The lung: scientific foundations. Raven Press, New York, 1991, pp. 1749-1759.
27. Wait J.: Cardiopulmonary stress testing: a review of noninvasive approaches. Chest 90: 504-510, 1986.
Aging Clin. Exp. Res., Vol. 6, No.4 275