Cardiac Frequency: Relationship to Body Composition

Transcription

1 Quarterly Journal of Experimental Physiology (1980) 65, 9-17 Radiographic Heart Volume, Stroke Volume and Exercise Cardiac Frequency: Relationship to Body Composition and Other Factors in Healthy Adult Males J. E. COTES, D. L. EVANS, G. R. JOHNSON, I. M. MACINTYRE and M. J. SAUNDERS MRC Pneumoconiosis Unit, Nr. Cardif and National Coal Board, Mine Rescue Station, Dinas Porth, Mid Glamorgan (RECEIVED FOR PUBLICATION 15 JUNE 1979) Amongst 288 healthy male workers in heavy industry the radiographic heart volume was related to the fat-free mass and the percentage body fat which between them explained 27 % of the variance. Exercise was performed on a treadmill using a belt speed of 80 m. min-1 (3 mph) and an incline which was increased progressively to 14 % when the oxygen uptake was on average 61 % of the maximum for these subjects. The exercise cardiac output was related to the uptake of oxygen, the fat-free mass and the ambient temperature; these factors accounted for 520% of the variance. The exercise stroke volume was related to the heart volume, the fat-free mass, the thigh muscle width, the ambient temperature, the Harvard Pack Index and other measurements which between them accounted for 42 % of the variance. The exercise cardiac frequency was correlated negatively with the heart volume, the fat-free mass, the age and the Harvard Pack Index. The variates between them described 41 % of the variance. The findings provide reference values for heart volume, stroke volume and exercise cardiac frequency in similar subjects. A person's body size contributes materially to the size of the lung and hence to the distribution of ventilation minute volume as between breathing frequency and tidal volume. The same is also true for the heart. Thus Cotes, Berry, Burkinshaw, Davies, Hall, Jones and Knibbs [1973] observed an inverse correlation between indices of body muscle and the cardiac frequency at any specified level of oxygen uptake during exercise; since in these circumstances cardiac output varies less between subjects than does the cardiac frequency the correlation implies an association between cardiac stroke volume and body muscle. This has recently been demonstrated for New Guineans [Patrick and Cotes, 1978]. An association of heart volume with body muscle, which was suggested by Evans [1969] was also observed in that study. The present paper explores the relationships in greater detail using as subjects workers in heavy industry in the U.K. The results provide a basis for reference values for resting heart volume and for stroke volume and cardiac frequency during treadmill Address for correspondence: Dr. J. E. Cotes, University Department of Occupational Health and Hygiene, 21 Claremont Place, Newcastle upon Tyne NE2 4AA. 9

2 10 J. E. COTES et al. exercise against which to compare the effects of physical activity, inactivity, drugs and diseases of the circulation, heart and other organs. Methods The subjects were 378 coal miners who were part-time members of mine rescue teams in South Wales. They were investigated at the time of their annual medical examination which included a questionnaire on respiratory symptoms, a brief medical examination with measurement of systemic blood pressure, recording of the forced expiratory volume and vital capacity, simple anthropometry and the Harvard pack test and a vertical standing jump. With a view possibly to replacing the pack test by one which was entirely submaximal the subjects, before doing the pack test, performed progressive exercise on a treadmill for 12min; the results of that comparison are reported elsewhere [Cotes, Dicken, Evans, Johnson, Kalinowska, Maclntyre and Saunders, 1980]. For the present study additional anthropometric measurements were carried out, cardiac output was measured by a CO2 rebreathing method during the last minute of the exercise and the heart volume at rest was determined by a radiographic method. These procedures and their purpose were described to the subjects prior to their taking part and informed consent was freely given by those who participated. The body measurements comprised body mass using a beam balance, stature recorded with a Harpenden stadiometer during slight vertical traction of the head which was held in the Frankfurt Plane and four skinfold measurements. The latter, which were obtained using Harpenden skin calipers (Holtain) were made at the mid-points of the left upper arm in the mid-line anteriorally and posteriorally, below the angle of the left scapula and above the left anteriorsuperior iliac spine. The sum of the skinfold thicknesses was used to estimate the percentage of body mass which is fat [Durnin and Wolmersley, 1974]. In addition the cifcumference (C) and anterior skinfold thickness (AST) of the thigh were obtained at a point one-third of the subischeal height (i.e. stature minus sitting height) above the lower femoral condyles. These results were used to calculate the thigh muscle width (TMW) using the following empirical relationship: TMW (cm)=0-24c (cm) -1 07AST (mm) This relationship was obtained in a subsidiary study on men in whom thigh radiographs were also obtained for direct measurement of thigh muscle width [see Cotes, Davies, Edholm, Healy and Tanner, 1969, for details]. The heart volume was obtained by the method of Asmussen [1959] from posterio-anterior and lateral chest radiographs taken during cardiac diastole with the subject in an upright posture. The 'posterior border of the heart was outlined by barium gruel. The treadmill exercise was performed using a belt speed of 80 m. min-' (3 mph). The treadmill surface was initially horizontal and was inclined

3 HEART SIZE. RELATION TO BODY COMPOSITION 11 progressively at a rate of 1 4% min1 for 10 min up to 140%; this was maintained for at least 2 min. For each minute of exercise measurements were made of oxygen uptake and carbon dioxide output. To this end the subjects breathed through a low resistance valve box of deadspace 145ml which was supported on a gantry. Inspired ventilation minute volume at ambient temperature was recorded using a digital gas meter [Saunders and James, 1978] and the concentrations of oxygen and carbon dioxide in the mixed expired gas were obtained using a Servomex OA150 paramagnetic meter for oxygen and a Hartmann Braun (URAS 3) infra-red analyser for carbon dioxide. The oxygen analyser was calibrated using oxygen-free nitrogen and air. The carbon dioxide analyser was calibrated using a Wosthoff pump and set before each period of exercise on a standard gas mixture. The cardiac output was measured in the 12th minute of exercise by the indirect Fick method with carbon dioxide as the test gas. The quantities required for the calculation included CO2 output and the concentrations of CO2 in the arterial and mixed venous blood. The CO2 output (f'co2) was that measured during the 11th minute of exercise increased by a factor of 1 02 to allow for the rise which occurred into the 12th minute [cf. Cotes, Allsopp and Sardi, 1969]. The factor was obtained in a subsidiary group of six subjects in whom the fico2 in the 11th minute was on average 66-4mmol.min-1 and in the 12th minute at the time of the first measurement of cardiac output 67 7mmol.min-. The corresponding R values were and The arterial CO2 concentration was obtained from the end tidal CO2 concentration which was recorded for six breaths at the start of the 12th minute of exercise, using an Elema Schonander mingograf 34 recorder. After converting to CO2 tension using the measured barometric pressure, the correction factors of Jones, Campbell, McHardy, Higgs and Clode [1966] were applied and the blood concentration was then read from the tables of McHardy [1967]. The mixed venous CO2 concentration was obtained by the rebreathing method; this was performed in duplicate from a rubber bag of capacity 61 which was initially two-thirds filled with a gas mixture comprising 14% CO2 in 2. Rebreathing was done deeply and rapidly in time to a verbal cue until either the CO2 concentration reached a steady level (plateau) or for 10 seconds whichever occurred sooner; the interval between duplicates was approximately 1 minute. The mixed venous tension was obtained from either the plateau concentration during rebreathing or, if there was no plateau, by extrapolation to the concentration at 20 seconds of rebreathing [Denison, 1967]. In either event the downstream correction of Jones, McHardy, Naimark and Campbell [1967] was applied. The stroke volume was obtained by dividing cardiac output by the cardiac frequency during the 11th minute of exercise since in the subsidiary study referred to this was also found to apply during the 12th minute [cf. Cotes, Allsopp and Sardi, 1969]. The frequency was obtained off the electrocardiogram from two electrodes applied respectively over the cardiac apex and above the right shoulder (modified lead CR5). An earth lead was also used. Further details of the physiological procedures and calculations, also of the

4 12 J. E. COTES et al. anthropometric, radiographic and other measurements are given elsewhere [Cotes, 1979]. The statistical procedures included multiple regression analyses by the stepdown method in which terms were eliminated progressively until only those remained which contributed significantly to the multiple regression analysis. The 5 % level of probability was accepted as significant. Results Measurements of cardiac output were obtained in 353 subjects but in only 288 of these were complete results obtained, including the Harvard pack and other tests. The mean results and ranges for the 288 subjects are given in Table I. The equations arising from the multiple regression analysis are given in Table II and the findings of the analysis are summarized below. l'ahle I. W1ean values and ranges for- the anthropometric and physiological indices obtained on the 288 subjects Abbreviation and Index units Mean value Range Age (A, a) Stature (St, m) Body mass (M, kg) Body fat (F, %) Fat-free mass (FFM, kg) Thigh muscle width (TMW, cm) Heart volume (HV, ml) Subischial height (SIH, cm) Exercise 02 uptake (602, mmol.min-') Cardiac-frequency at i602 = 67 mmol. min - (fc6, min -) Cardiac output (0t, I. min ) Stroke volume* (SV, ml) Harvard pack index (HPI) Laboratory temperature t, OC *During I 1th or 12th minute of exercise, for details see text. The heart volume was found to be correlated positively with anthropometric indices including fat-free mass, thigh muscle width, body mass, stature and percentage body mass which is fat. It was also correlated with indices which reflect the physiological response to exercise including exercise stroke volume and cardiac frequency and the maximal oxygen uptake obtained by extrapolation to an assumed maximal heart rate. The heart volume was independent of age, systolic and diastolic blood pressure and whether or not the subjects smoked. The best description of heart volume was in terms of the fat-free mass, and the percentage body fat (equation 1), and these terms accounted for 270% of the variance. Cardiac output during the last minute of exercise was significantly correlated with the uptake of oxygen (equation 2). It was also correlated with the fat-free mass, the thigh muscle width, the heart volume and the ambient temperature;

5 HEART SIZE, RELATION TO BODY COMPOSITION 13 C.)= CtQ~~~~~~~~~C CO~~~~~~~~~~C 0&'Je Co C' i C'4 r~- 00 t_ C._ U a.) to > -o o - Co -C~~~~~~ - C~~~~~~~~~~~~~~~~~~~' -C D 0 ) 2 O:0 WI + =o o Q )o r- 9_ 0 *e o ri - o C +' > 00S 2 CCbO ' CZ L0 r- C 0) ) 3a T C' u ei _ FWV > u a) u u C ~CZ VzC oi LiUQo 0)_1 -Q C O Nis-N N)q>1 1. _ e r \ ~~0 +.. Y. X,y >4 b c5 ~ Q ' ^ n Q

6 14 J. E. COTES et al. the additional inclusion of age, stature, body mass and whether or not the subjects smoked, did not further reduce the variance about the regression equation. The relationship with the lowest residual standard deviation was on oxygen uptake, fat-free mass and ambient temperature as independent variates (equation 3); there were no significant interaction terms. However, fat-free mass could be replaced by thigh muscle width and heart volume without material loss of sensitivity. The exercise stroke volume was correlated with the fat-free mass, thigh muscle width, heart volume, forced vital capacity standardized for age and stature, uptake of oxygen and ambient temperature. The regression equation of stroke volume on most of these variables without and with the interaction terms (equations 4 and 5) accounted for approximately 280% of the total variation. Adding the Harvard pack index increased the percentage variance explained by the regression to 420%. The stroke volume was independent of age, systemic blood pressure and performance during the standing jump. The exercise cardiac frequency at an oxygen uptake of 67 mmol. min- 1 (1 51.min-') was significantly negatively correlated with age, heart volume standardized for body fat and the Harvard pack index and positively correlated with the reciprocal of the fat-free mass (equation 6). In the equation the alternative use of the heart volume standardized for fat-free mass as well as percentage of fat did not materially affect the relationship. Discussion The present subjects embraced a wide range of levels of physical fitness as assessed by the Harvard pack index. Thus whilst the measurements of stroke volume were made at a mean uptake of oxygen of 86mmol.min-' equivalent to 61 % of the estimated maximum, the range was wide (31 %-99 %). For most subjects the rate of work was that associated with a maximal stroke volume [Chapman, Fisher and Sproule, 1960] but in the subjects with the highest capacity for exercise the oxygen uptake as a percentage of maximum was lower than necessary to achieve a maximal value. Nonetheless for the group as a whole there was a weak negative correlation between stroke volume and oxygen uptake as a percentage of maximum, reflecting the large stroke volumes of the subjects with high aerobic capacities. The measurement of cardiac output by the indirect carbon dioxide method is sensitive to fluctuation in the quantity of CO2 dissolved in lung tissue [Farhi, Nesarajah, Olszowska, Metildi and Ellis, 1976]. However, the absolute levels were similar to those obtained by the direct method [Reeves, Grover, Blount and Filley, 1961]; in addition any error was unlikely to have affected the relationship of stroke volume to the variates considered in the present paper. But it should be noted that the present analysis relates to only the first measurement on each subject; the second result was consistently lower by on average (4%) and subsequent investigation showed this to be an error of

7 HEART SIZE, RELATION TO BODY COMPOSITION 15 which the magnitude varied inversely with the time interval between determinations. The error disappeared at 2 5min but this exceeded the 1-min interval which obtained in the present study. The relationship of heart volume to fat-free mass was in accord with the predictions of Evans [1969] and with our previous findings for New Guineans [Patrick and Cotes, 1978]. The additional association with body fat was also logical and suggested that both variables should be taken into account when compiling reference values for heart size. The heart size of the present subjects was larger than that of New Guineans of similar fat-free mass but the percentage of fat was also greater. However, allowing for body fat only marginally reduced the discrepancy (mean values for present subjects 872 ml, for New Guinean males observed 564 ml predicted from equation ml). The difference might have been due to the New Guinea subjects not swallowing barium prior to the lateral chest X-ray. The relationship of cardiac output to the other variables was similarly in line with expectation, though the regression coefficient on ambient temperature of min'- C1- was rather large. It indicated a contribution by the circulation to thermal regulation at ambient temperatures within the normal range, in this instance C, mean 201 C. The change was mediated mainly through an affect on stroke volume (2-56 ml C- 1) but the work of Miller and Martin [1975] shows that cardiac frequency may also contribute. The nature of the association is unclear and merits further investigation. The association of stroke volume with fat-free mass, heart volume and thigh muscle width was to be expected since the heart is a muscle and likely to be commensurate in size with the muscles to which it supplies blood. The associations with the vital capacity standardized for age and stature, and with the Harvard pack index probably reflect an association with the force of contraction of the heart. However, there was no association with muscle power as assessed by the standing jump. The existence of significant interaction terms in the prediction equation (equation 5) is consistent with stroke volume being determined by a combination of anthropometric and physiological factors. These are approximated to a useful extent by the present combination of variates though further refinement might be achieved by including blood volume or other related variable. Meanwhile, the coefficient of variation about the regression equation of 140% is of the same order as that which has been found useful for the description of reference values for indices of lung function and should have practical applications. The cardiac frequency at a standard uptake of oxygen has been suggested as an index of the cardiac response to exercise [Cotes, 1972]; in adults and children of both sexes it has been shown to be inversely related to the fat-free mass [Cotes, Berry, Burkinshaw, Davies, Hall, Jones and Knibbs, 1973] and, in the case of adult females and males to age [Cotes, Hall, Johnson, Jones and Knibbs, 1974]. Using an index based on walking speed Bassey, Bryant, Clark, Fentem, Jones, MacDonald and Patrick [1979] came to a similar conclusion for males. Independent of fat-free mass the frequency in the present subjects is

8 16 J. E. COTES et al. negatively correlated with heart volume, age and the Harvard pack index. These associations are in line with expectations; that with heart volume reflects the negative correlation between frequency and stroke volume. The association with age probably reflects the age-related decline in sympathetic vasomotor tone [Conway, Wheeler and Sannerstedt, 1971]. However, this has so far only been demonstrated over a wider range of ages than obtained in the present study, so an association with some other variate whilst not amenable for demonstration using the present data cannot be ruled out. The association with the Harvard pack index is further evidence that the cardiac frequency standardized for uptake of oxygen and for anthropometric factors provides a measure of what is colloquially called physical fitness. The present results relate to men in a physically demanding job but their circulatory responses to exercise vary over a wide range. Thus the associations which have been demonstrated with body composition and other factors are probably not artifacts due to selection of subjects. Indeed the relationship of heart volume on fat-free mass and body fat has recently been found to obtain also for other coal miners, [Cotes et al., work in progress]. Further studies are needed to extend these observations. Meanwhile the results suggest that the physiological anthropometry of the cardiovascular system may be described with a precision similar to that of the respiratory system; the findings provide a basis for reference values which may be of use for study of physical condition and the effects of disorders of the heart and other organs. Acknowledgments We are indebted to Mr. E. 0. Evans and other members of staff at the NCB Dinas Porth Mine Rescue Centre for their assistance in this study and to the National Coal Board and Dr. J. S. McLinctock for the provision of facilities. Mr. W. P. Audsley took the chest radiography, Mr. K. Dicken and Miss L. Rowbotham provided technical assistance, and Mr. G. Berry contributed statistical advice. We are also indebted to our subjects who made this study possible and to Dr. J. M. Patrick who kindly commented on the manuscript. Mr. D. Thomas designed the gantry to support the valve box. References AsMUSSEN, E. (1959). Heart size by X-ray. Acta Radiology, 31, BASSEY, E. J., BRYANT, J. C., CLARK, E., FENTEM, P. H., JONES, P. R. M., MACDONALD, I. A. and PATRICK, J. M. (1979). Factors affecting cardiac frequency during self paced walking: Body composition, age, sex and habitual activity. Journal of Physiology in the press. CHAPMAN, C. B., FISHER, J. N. and SPROULE, B. J. (1960). Behaviour of stroke volume at rest and during exercise in human beings. Journal of Clinical Investigation, 39, CONWAY, J., WHEELER, R. and SANNERSTEDT, A. (1971). Sympathetic nervous activity during exercise in relation to age. Cardiovascular Research, 5, COTES, J. E. (1972). Response to progressive exercise: A three index test. British Journal of Diseases of the Chest, 66, COTES, J. E. (1979). Lun,g FunIctimiI: Assesslnent (an(1 Application in Medicine. 4th edition. Oxford: Blackwells Scientific Publicationls.

9 HEART SIZE, RELATION TO BODY COMPOSITION 17 COTES, J. E., ALLSOPP, D. and SARDI, F. (1969). Human cardio-pulmonary responses to exercise. Quarterly Journal of Experimental Physiology, 54, COTES, J. E., BERRY, G., BURKINSHAW, L., DAVIES, C. T. M., HALL, A. M., JONES, P. R. M. and KNIBBS, A. V. (1973). Cardiac frequency during submaximal exercise in young adults: Relation to lean body mass, total body potassium and amount of leg muscle. Quarterly Journal of Experimental Physiology, 58, COTES, J. E., DAVIES, C. T. M., EDHOLM, 0. G., HEALY, M. J. R. and TANNER, J. M. (1969). Factors relating to the aerobic capacity of 46 healthy British males and females ages years. Proceedings of the Royal Society of London (B), 174, COTES, J. E., DICKEN, C., EVANS, D. L., JOHNSON, G. R., KALINOWSKA, E. B., MACINTYRE, I. M. and SAUNDERS, M. J. (1980). Prediction of Harvard pack index from the result of an 11 min progressive exercise test and anthropometric measurements in coal miners. Ergonomics (in the press). COTES, J. E., HALL, A. M., JOHNSON, G. R., JONES, P. R. M. and KNIBBS, A. V. (1974). Decline with age of cardiac frequency during submaximal exercise in healthy women. Journal of Physiology, 238, 24P-25P. DENISON, D. (1967). The measurement of mixed venous blood gas tensions by mass spectrometry. Bulletin Physio-pathologie Respiratoire, 3, DURNIN, J. V. G. A. and WOLMERSLEY, J. (1974). Body fat assessed from total body density and its estimation from skinfold thickness: Measurements on 481 men and women aged from years. British Journal of Nutrition, 32, EVANS, D. W. (1969). Heart volume-normal values and practical applications. Malattie Cardiovascolari, 10, FARHI, L. E., NESARAJAH, M. S., OLSZOWSKA, A. J., METILDI, L. A. and ELLIS, A. K. (1976). Cardiac output determination by simple one-step rebreathing technique. Respiration Physiology, 28, JONES, J. L., CAMPBELL, E. J. M., MCHARDY, G. J. R., HIGGs, B. E. and CLODE, M. (1967). The estimation of carbon dioxide pressure of mixed venous blood during exercise. Clinical Science, 32, JONES, N. L., MCHARDY, G. J. R., NAIMARK, A. and CAMPBELL, E. J. M. (1966). Physiological dead space and alveolar-arterial gas pressure differences during exercise. Clinical Science, 31, MCHARDY, G. J. R. (1967). The relationship between the differences in pressure and content of carbon dioxide in arterial and venous blood. Clinical Science, 32, MILLER, G. J. and MARTIN, H. de V. (1975). Effect of ambient temperatures between 21 C and 35 C on the responses to progressive submaximal exercise in partially acclimated man. Ergonomics, 18, PATRICK, J. M. and COTES, J. E. (1978). Cardiac output during submaximal exercise in New Guineans: The relation with body size and habitat. Quarterly Journal of Experimental Physiology, 63, REEVES, J. T., GROVER, R. F., BLOUNT, S. G. JR. and FILLEY, G. F. (1961). Cardiac output response to standing and treadmill walking. Journal of Applied Physiology, 16, SAUNDERS, M. J. and JAMES, P. J. (1978). Digital respiratory rate and ventilation recorder. Medical Biological Engineering Computing, 16, B