- Email: [email protected]
Resting diastolic function and left ventricular mass are related to exercise capacity in hypertensive men but not in women
AJH
1998;11:1252–1257
Resting Diastolic Function and Left Ventricular Mass Are Related to Exercise Capacity in Hypertensive Men But Not in Women Ali G. Gharavi, Joseph A. Diamond, Adam Y. Goldman, Neil L. Coplan, Jeffrey S. Jhang, Marilyn Steinmetz, Rochelle Goldsmith, and Robert A. Phillips
We evaluated the impact of diastolic function and gender on exercise capacity in sedentary, untreated hypertensive subjects (34 men, 23 women) using echocardiography and a bicycle ergometry with measurement of oxygen consumption (VO2). In men, peak (A) mitral inflow velocity and left ventricular (LV) mass were inversely related to peak VO2 (r 5 20.64) and maximal workload (r 5 20.57) and were the sole independent determinants of exercise capacity. In women, there was no relationship between any echocardiographic measure and exercise capacity.
L
eft ventricular (LV) diastolic filling is an important determinant of the cardiac response to dynamic exercise. Diastolic function measured by Doppler echocardiography predicts maximal oxygen consumption (VO2max) in normotensive subjects.1 In hypertensive subjects, resting
Received September 10, 1997. Accepted April 27, 1998. From the Hypertension Section, Cardiovascular Institute, Mount Sinai School of Medicine (AGG, JAD, AYG, JSJ, MS, RG, RAP), and Division of Cardiology, Lenox Hill Hospital (NLC), New York City, New York. This study was presented in part at the 45th Scientific Session of the American College of Cardiology, March 1996. This study was supported in part by a grant from Pfizer, Inc. Address correspondence and reprint requests to Ali Gharavi, MD, Mount Sinai School of Medicine, Hypertension Section, Cardiovascular Institute, One Gustave L. Levy Place, Box 1085, New York, NY 10029-6574.
© 1998 by the American Journal of Hypertension, Ltd. Published by Elsevier Science, Inc.
Thus, LV mass and Doppler-determined diastolic function predict maximal VO2 in hypertensive men but not in women. This finding may be related to gender differences in the contribution of diastolic filling to exercise capacity or may reflect limitations of resting Doppler echocardiography to predict exercise diastolic filling in hypertensive women. Am J Hypertens 1998;11:1252–1257 © 1998 American Journal of Hypertension, Ltd. KEY WORDS:
Gender, men, women, diastolic function, exercise capacity, echocardiography.
peak diastolic filling rates are directly related to the magnitude of change in ejection fraction during exercise.2 Subjects with reduced ejection fraction response show decreased diastolic filling rates and increased LV mass.2 Presumably, LV hypertrophy increases myocardial stiffness and impairs relaxation, leading to poor diastolic filling and inadequate increase in ejection fraction. Gender is also a determinant of exercise capacity and the cardiac response to exercise. Compared to men, women generally achieve lower VO2max and maximal workloads during dynamic exercise.1,3 Women also have lower LV mass index but similar Doppler indices of diastolic function at rest.4 In addition, some studies have suggested that compared to men, women have little or no increase in stroke volume index and ejection fraction during exercise.3,5–7 These data suggest that there are gender differences in the relative contribution of systolic and diastolic func0895-7061/98/$19.00 PII S0895-7061(98)00138-1
AJH–OCTOBER 1998 –VOL. 11, NO. 10, PART 1
GENDER, DIASTOLIC FUNCTION, AND EXERCISE CAPACITY
tion to exercise capacity. The purpose of this study was to evaluate the relationship between resting echocardiographic variables and exercise performance in hypertensive subjects. We also examined the effects of gender in the relationship between resting diastolic function and exercise capacity. MATERIALS AND METHODS We enrolled hypertensive subjects (34 men and 23 women) who were referred for evaluation because of newly diagnosed or uncontrolled hypertension. They all had office blood pressures .140/90 mm Hg on at least two separate visits; most had office diastolic blood pressures .100 mm Hg. All antihypertensive medications (if any) were discontinued for at least 2 weeks before all studies. All subjects were sedentary and none had engaged in any regular aerobic training within a year of the study. Other significant concomitant illnesses (eg, coronary artery disease, congestive heart failure) were excluded by a standard history, physical examination, routine laboratory test, and additional diagnostic studies as indicated. Informed, signed consent was obtained and the protocol was approved by the Institutional Review Board of the Mount Sinai School of Medicine. To accurately characterize blood pressure, 24-h blood pressure monitoring was performed on a day of typical weekly activity using the SpaceLabs 90202 and 90207 devices (SpaceLabs, Redmond, WA).8 Two-dimension guided M-mode echocardiography was performed with an ATL Ultramark 6 scanner (Advanced Technology Laboratories, Inc., Bothell, WA) using a 2.5- or 3.5-MHZ transducer. Echocardiographic tracings were coded and analyzed with a commercially available analyzer (Nova MicroSonics, Advanced Technology Laboratories) in a random order by two independent observers blinded to the subjects’ identity and clinical characteristics.9 End-diastolic measurements of LV structures were made according to the Penn convention and LV mass was indexed to body surface area (grams/meter squared).10,11 LV hypertrophy was defined using gender-specific criteria.9 Two-dimension guided pulsed Doppler interrogation of the LV was done from the apical two- and fourchamber view. Recordings of LV inflow were made where maximal velocity of early filling was obtained. At least five spectral tracings of the mitral inflow were digitized and averaged to obtain peak velocity of early and late diastolic filling and their ratio. The following measurements were also made: isovolumic relaxation time, time to peak early diastolic filling, deceleration time of peak early diastolic filling, and slope (descent) of early diastolic filling. All subjects underwent a symptom-limited graded exercise test on a bicycle ergometer in the upright position, with simultaneous electrocardiogram and re-
1253
spiratory gas analysis. Subjects performed 2 min of unloaded exercise followed by increasing workloads, based on age and weight, of 10 to 30 W/min on an electronically braked cycle ergometer (Mijnhardt). Patients were encouraged to give their maximal effort and exercised until exhaustion developed. Simultaneous 12-lead electrocardiographic monitoring (Case 12, Marquette) was performed and blood pressure was recorded every minute by manual sphygmomanometry. Oxygen consumption (VO2), carbon dioxide production, minute ventilation, and respiratory exchange ratio were computed from measurements made every eighth breath (2001 System, Medical Graphics Corp., St. Paul, MN). Maximal VO2 (VO2max) was defined as the highest VO2 reached during exercise or as the value at which the rate of increase in VO2 did not exceed 0.75 mL/kg/min over a 2-min period. Atrial natriuretic peptide levels were measured from extracted plasma by radioimmunoassay as previously described.12 Data are expressed as mean value 6 SD. Gender differences were evaluated by two-way analysis of variance followed by Bonferroni’s correction method. The relationship between VO2max and resting echocardiographic variables were analyzed by linear correlations. Variables with P , .1 were included in a stepwise linear regression analysis to detect independent determinants of exercise capacity. The impact of left ventricular hypertrophy (LVH) and more severe diastolic dysfunction on exercise capacity was evaluated by analysis of covariance using VO2max and maximal workload as dependent variables and age as a covariate independent variable. Severity of diastolic dysfunction was defined as age-adjusted peak A . median value for each gender (66 cm/sec in men, 70 cm/sec in women). P , .05 was considered significant. RESULTS Subject characteristics were similar between sexes except for the expected higher LV mass, maximal workload, and VO2max in men (Table 1). Both genders gave similar effort in the exercise test as demonstrated by equivalent peak exercise respiratory exchange ratios. Compared to age-specific normal values,13 our subjects had reduced E/A ratios, prolonged isovolumetric relaxation times, shortened deceleration times, and decreased deceleration slope of the E wave. Atrial natriuretic peptide levels (a marker of left atrial pressure) were related directly to peak A wave (r 5 0.39, P , .05) and inversely to VO2max (r 5 20.34, P , .05). In the entire group, multivariate analysis showed that gender and age were independent predictors of VO2max (r2 5 0.46, P , .001). Gender, age, and peak systolic BP were each independent predictors of maximal workload (r2 5 0.68, P , .001) (Table 2). Because gender was the strongest determinant of exercise ca-
1254
AJH–OCTOBER 1998 –VOL. 11, NO. 10, PART 1
GHARAVI ET AL
TABLE 1. CHARACTERISTICS OF STUDY SUBJECTS
Age (years) Body mass index (kg/m2) Ambulatory BP (mm Hg) Heart rate, rest (beats/min) Peak exercise BP (mm Hg) Peak exercise heart rate (beats/min) VO2max (mL/kg/min) Maximal workload (W) Peak respiratory exchange ratio ANP (pg/mL) LV mass (g/m2) LV hypertrophy (%) Ejection fraction (%) Fraction shortening (%) Peak A wave (cm/sec) Peak E wave (cm/sec) E/A ratio Isovolumetric relaxation time (msec) Time to peak E wave (msec) Deceleration time of E wave (msec) Slope of deceleration of E wave (cm/sec2)
Women (n 5 23)
Men (n 5 34)
49 6 10 30.8 6 3.9 154 6 12/99 6 11 72 6 9 226 6 30/118 6 23 157 6 18 17.5 6 4.4 83 6 26 1.1 6 0.1 29 6 14 93 6 21 6 (26%) 66 6 5 37 6 6 67 6 18 66 6 14 0.98 6 0.4 119 6 27 70 6 20 129 6 53 307 6 120
49 6 8 29.1 6 3.4 157 6 16/103 6 12 69 6 9 242 6 29/121 6 20 165 6 14 22.9 6 9.4* 133 6 36* 1.2 6 0.1 23 6 10 116 6 39* 12 (35%) 62 6 11 34 6 8 63 6 13 60 6 13 0.95 6 0.3 124 6 23 83 6 22 138 6 44 281 6 119
*P , .001 men v women. Ambulatory BP, average 24-h ambulatory blood pressure; ANP, atrial natriuretic peptide; BP, blood pressure; LV mass, left ventricular mass indexed to body surface area; E/A, ratio of peak early to late transmitral flow velocities; Peak A wave, peak late transmitral flow velocity; Peak E wave, peak early transmitral flow velocity; VO2max, maximal exercise oxygen consumption.
pacity, the analyses were performed within each gender group. In men, VO2max was inversely related to peak A wave, age, LV mass, and atrial natriuretic peptide (ANP) levels (Table 2, Figure 1). In addition, LV mass directly correlated with ANP level (r 5 0.39, P , .05).
On multivariate analysis, peak A wave was the only independent predictor of VO2max (r2 5 0.41, P , .001), whereas LV mass and peak A remained independently related to maximal workload (r2 5 0.42, P , .001) (Table 2). Similar results were found when LV mass was included in the model as a categorical vari-
TABLE 2. CORRELATION OF EXERCISE CAPACITY WITH SELECTED CLINICAL VARIABLES Overall Group (n 5 57)
Women (n 5 23)
Variable
VO2max
Maximal Workload
VO2max
Maximal Workload
Gender Age BMI Ambulatory SBP Peak SBP LV mass EF Peak A Peak E E/A ANP
0.53*\ 20.42†\ 0.01 20.03 0.31‡ 0.10 20.12 20.39† 20.20 0.25 20.28
0.61*\ 20.42†\ 0.10 0.10 0.43†\ 0.10 20.10 20.32† 20.21 0.16 20.29‡
— 20.45‡ 0.08 20.04 0.40§ 0.37§ 0.17 20.13 0.29 0.34 20.17
— 20.65*\ 0.01 0.37 0.53‡\ 0.38§ 0.16 0.01 0.34 0.29 0.09
All values using Pearson correlation coefficient. See Table 1 for other abbreviations. * P , .001; † P , .01; ‡ P , .05; § P , .1. \ Independent determinant of exercise capacity on multivariate analysis.
Men (n 5 34) VO2max
— 20.48† 0.20 20.16 20.01 20.34‡ 20.06 20.64*\ 20.38‡ 0.27 20.35‡
Maximal Workload
— 20.46† 0.01 20.17 0.18 20.37‡\ 0.04 20.57*\ 20.26 0.28 20.40‡
AJH–OCTOBER 1998 –VOL. 11, NO. 10, PART 1
GENDER, DIASTOLIC FUNCTION, AND EXERCISE CAPACITY
1255
FIGURE 1. Relationship between resting peak A wave and maximal oxygen consumption (VO2max) in women (A) and in men (B).
able based on the presence or absence of LVH. Therefore, men with LVH had a 21% reduction in maximal workloads compared to their counterparts without LVH (114 6 11 v 144 6 14 W, P 5 .02). Similarly, men with more severe diastolic dysfunction (peak A waves .66 cm/sec) had a 19% reduction in VO2max (20.5 6 0.8 v 25.3 6 0.8 mL/kg/min, P , .001) and maximal workloads (118 6 8 v 146 6 8 W, P , .05) after adjustment for age. In women, the relationship between VO2max and echocardiographic measures of systolic or diastolic function was weak and did not approach statistical significance (Table 2, Figure 1). In addition, LV mass did not correlate with ANP or any Doppler parameter. Only age and peak systolic BP correlated with exercise capacity. On multivariate analysis, peak systolic BP was independently related to VO2max (r2 5 0.34, P , .05), whereas age and peak exercise systolic blood pressure remained independently related to maximal workload (r2 5 0.67, P , .001) (Table 2). There was no difference in exercise capacity between women according to the presence of LVH or more severe diastolic dysfunction (defined as peak A wave .70 cm/ sec) even after adjustment for age and other clinical variables. DISCUSSION In this study of sedentary hypertensive subjects, we found significant gender differences in the relation between resting echocardiographic variables and exercise capacity. LV mass and peak A wave were in-
versely related to exercise capacity in men but not in women. The presence of LVH or worse diastolic function was associated with significantly reduced exercise tolerance in hypertensive men, even after adjustment for age and other clinical variables. In addition, we found a direct correlation between ANP levels and LV mass in male subjects. These findings in men are compatible with previous studies showing that hypertension alters chamber compliance, causes LV hypertrophy and diastolic dysfunction, and ultimately leads to poor stroke volume response and diminished exercise capacity.2,12,14 In contrast to men, LV mass and diastolic filling at rest were not significantly related to exercise capacity in hypertensive women. Our women had lower LV mass index and exercise capacity compared to men but other clinical characteristics were similar. The lower exercise capacity in women could not be explained by differences in exercise effort, blood pressure level, degree of obesity, diastolic filling parameters, or prevalence of LV hypertrophy. These gender differences in cardiac mass and exercise tolerance have been previously described but have not been adequately explained.1,3,4 In addition, many studies assessing the role of cardiac function in exercise have not performed gender-specific analyses.1,2,15 Although the interpretation of Doppler variables is generally more complicated in hypertensive subjects, there are no reported gender limitations of this technique for the evaluation of diastolic function.4,12 Thus, our findings cannot be solely attributed to a limitation of the
1256
GHARAVI ET AL
Doppler echocardiographic technique. Rather data suggest that our observations stem from gender effects on the cardiac response to exercise. Multiple studies have demonstrated that the relative contribution of Frank-Starling mechanisms and enhanced contractility to the increase in stroke volume during exercise differs between the sexes.3,5,6,14 Several but not all studies have indicated that women achieve little or no increase in stroke volume or ejection fraction during dynamic exercise. Faggard et al3 showed that hypertensive women achieved a smaller increase in stroke volume and cardiac output from rest to exercise in the upright position. Using radionuclide angiography during upright5 and supine6,16 exercise, other studies showed that healthy male and female volunteers have similar increases in stroke volume but achieve it through different mechanisms: men have an increase in ejection fraction without an increase in end-diastolic counts, whereas women have an increase in end-diastolic counts without an increase in ejection fraction. Using radionuclide angiography and right-sided catheterization in normal subjects, Sullivan et al7 did not observe any gender differences in the stroke volume response to upright exercise in normal adults, although there was a trend toward lower ejection fraction response in women. Taken together, the published data suggest that women are able to augment diastolic filling significantly beyond resting values during exercise. Thus, resting measures of diastolic filling may not necessarily correlate with exercise diastolic function in women and would not be an accurate indicator of stroke volume response and ultimately, exercise capacity. In contrast, diastolic filling changes little during exercise in men, and therefore, LV filling at rest may be a good predictor of diastolic contribution to exercise capacity. This ability to augment diastolic filling during exercise may potentially be related to gender differences in venous capacitance and the capacity to maintain preload in the upright posture. Thus, a larger trial using invasive hemodynamic measurements would be required to provide further clarification on this topic. In the present study, both genders showed reduced E/A ratios accompanied by prolonged isovolumetric relaxation times, shortened deceleration times, and decreased deceleration slope of E wave (compared to age specific normal values).12 These combined abnormalities suggest diastolic dysfunction beyond the early stage characterized by reversal of E/A wave ratios. Our subjects had probably begun to “pseudonormalize” their LV inflow patterns, leading to elevation of the E wave and shortening of the deceleration time.17 This is further supported by the direct correlation between the peak A wave and atrial natriuretic peptide levels, which reflect cardiac filling pressures. Hence, in the presence of moderately advanced dia-
AJH–OCTOBER 1998 –VOL. 11, NO. 10, PART 1
stolic dysfunction, only peak A waves in men remained the most reliable indicator of LV diastolic properties and exercise capacity. The findings of this study apply to sedentary, untreated or poorly controlled hypertensive subjects only; the effect of aerobic exercise training or specific pharmacologic intervention on the relation between resting echocardiographic data and exercise capacity was not evaluated. However, our study has useful clinical implications; namely that the finding of LVH and abnormal diastolic function on resting echocardiography will correlate with diminished exercise capacity in hypertensive men but not in women. Thus, echocardiographic findings may explain exercise intolerance and the symptoms of diastolic heart failure in sedentary hypertensive men. In contrast, resting echocardiographic data may not be adequate to explain reduced exercise capacity in hypertensive women. This may be attributable to gender differences in the relative contribution of diastolic filling to cardiac performance during exercise; alternatively, echocardiographic measurements of resting diastolic function may not be related to exercise diastolic function in hypertensive women. Hence, investigation of a cardiac cause of exercise intolerance in sedentary hypertensive women may necessitate an evaluation of cardiac function during exercise. Our data also suggest that future studies on the cardiac physiology of exercise should perform gender-specific analyses and focus on defining clinically useful cardiac predictors of exercise capacity in hypertensive women. ACKNOWLEDGMENT We thank Mr. Howard Printz for his encouragement and generous support of the Hypertension Section.
REFERENCES 1. Vanoverschelde JLJ, Essamri B, Vanbutsele R, et al: Contribution of left ventricular diastolic function to exercise capacity in normal subjects. J Appl Physiol 1993;74(5): 2225–2233. 2. Cuocolo A, Sax FL, Brush JE, et al: Left ventricular hypertrophy and impaired diastolic filling in essential hypertension: diastolic mechanisms for systolic dysfunction during exercise. Circulation 1990;81:978 –986. 3. Faggard RH, Thijs LB, Amery AK: The effects of gender on aerobic power and exercise hemodynamics in hypertensive adults. Med Sci Sports Exercise 1995;27(1):29 –34. 4. Hinderliter AL, Light KC, Willis PW: Gender differences in left ventricular structure and function in young adults with normal and marginally elevated blood pressure. Am J Hypertens 1992;5:32–36. 5. Higginbotham MB, Morris KG, Coleman E, Cobb FR: Sex-related differences in the normal cardiac response to upright exercise. Circulation 1984;70:357–366. 6. Hanley PC, Zinsmeister AR, Clements IP, et al: Genderrelated differences in cardiac response to supine exercise
AJH–OCTOBER 1998 –VOL. 11, NO. 10, PART 1
GENDER, DIASTOLIC FUNCTION, AND EXERCISE CAPACITY
assessed by radionuclide angiography. J Am Coll Cardiol 1989;624 – 629. 7. Sullivan MJ, Cobb FR, Higginbotham MB: Stroke volume increases by similar mechanisms during upright exercise in normal men and women. Am J Cardiol 1991; 67:1405–1412. 8. Moore CR, Krakoff LR, Phillips RA. Comfirmation or exclusion of stage I hypertension by ambulatory blood pressure monitoring. Hypertension 1997;29:1109 –1113. 9. Jhang JS, Diamond JA, Phillips RA. Interobserver variability of left ventricular measurements in a population of predominantly obese hypertensives using simultaneously acquired and displayed M-mode and 2-D cine echocardiography. Echocardiography 1997;14:9 –15. 10. Devereux RB, Reichek N. Echocardiographic determination of left ventricular mass in man. Circulation 1977;55: 613– 618. 11. Lauer MS, Anderson KM, Larson M, Levy D. A new method for indexing left ventricular mass for differences in body size. Am J Cardiol 1994;74:487– 491.
1257
12. Eison HB, Rosen, MJ, Phillips RA. Krakoff LR. Determinants of atrial natriuretic factor in the adult respiratory distress syndrome. Chest 1988;94:1040 –1045. 13. Benjamin EJ, Levy D, Anderson KM, et al: Determinants of Doppler indexes of left ventricular diastolic function in normal subjects (the Framingham Heart study). Am J Cardiol 1992;70:508 –515. 14. Lim PO, MacFadyen R, Clarkson PB, McDonald TM. Impaired exercise tolerance in hypertensive patients. Ann Intern Med 1996;124(1, pt1):41–55. 15. Poliner LR, Dehmer GJ, Lewi SE, et al: Left ventricular performance in normal subjects: a comparison of the responses to exercise in the upright and supine positions. Circulation 1980;62:528 –534. 16. Adams KF, Vincent LM, McAllister SM, et al. The influence of age and gender on left ventricular response to supine exercise in asymptomatic subjects. Am Heart J 1987;113:732–742. 17. Nishimura R, Appleton C. “Diastology”: beyond E and A. J Am Coll Cardiol 1996;27:372–374.
1998;11:1252–1257
Resting Diastolic Function and Left Ventricular Mass Are Related to Exercise Capacity in Hypertensive Men But Not in Women Ali G. Gharavi, Joseph A. Diamond, Adam Y. Goldman, Neil L. Coplan, Jeffrey S. Jhang, Marilyn Steinmetz, Rochelle Goldsmith, and Robert A. Phillips
We evaluated the impact of diastolic function and gender on exercise capacity in sedentary, untreated hypertensive subjects (34 men, 23 women) using echocardiography and a bicycle ergometry with measurement of oxygen consumption (VO2). In men, peak (A) mitral inflow velocity and left ventricular (LV) mass were inversely related to peak VO2 (r 5 20.64) and maximal workload (r 5 20.57) and were the sole independent determinants of exercise capacity. In women, there was no relationship between any echocardiographic measure and exercise capacity.
L
eft ventricular (LV) diastolic filling is an important determinant of the cardiac response to dynamic exercise. Diastolic function measured by Doppler echocardiography predicts maximal oxygen consumption (VO2max) in normotensive subjects.1 In hypertensive subjects, resting
Received September 10, 1997. Accepted April 27, 1998. From the Hypertension Section, Cardiovascular Institute, Mount Sinai School of Medicine (AGG, JAD, AYG, JSJ, MS, RG, RAP), and Division of Cardiology, Lenox Hill Hospital (NLC), New York City, New York. This study was presented in part at the 45th Scientific Session of the American College of Cardiology, March 1996. This study was supported in part by a grant from Pfizer, Inc. Address correspondence and reprint requests to Ali Gharavi, MD, Mount Sinai School of Medicine, Hypertension Section, Cardiovascular Institute, One Gustave L. Levy Place, Box 1085, New York, NY 10029-6574.
© 1998 by the American Journal of Hypertension, Ltd. Published by Elsevier Science, Inc.
Thus, LV mass and Doppler-determined diastolic function predict maximal VO2 in hypertensive men but not in women. This finding may be related to gender differences in the contribution of diastolic filling to exercise capacity or may reflect limitations of resting Doppler echocardiography to predict exercise diastolic filling in hypertensive women. Am J Hypertens 1998;11:1252–1257 © 1998 American Journal of Hypertension, Ltd. KEY WORDS:
Gender, men, women, diastolic function, exercise capacity, echocardiography.
peak diastolic filling rates are directly related to the magnitude of change in ejection fraction during exercise.2 Subjects with reduced ejection fraction response show decreased diastolic filling rates and increased LV mass.2 Presumably, LV hypertrophy increases myocardial stiffness and impairs relaxation, leading to poor diastolic filling and inadequate increase in ejection fraction. Gender is also a determinant of exercise capacity and the cardiac response to exercise. Compared to men, women generally achieve lower VO2max and maximal workloads during dynamic exercise.1,3 Women also have lower LV mass index but similar Doppler indices of diastolic function at rest.4 In addition, some studies have suggested that compared to men, women have little or no increase in stroke volume index and ejection fraction during exercise.3,5–7 These data suggest that there are gender differences in the relative contribution of systolic and diastolic func0895-7061/98/$19.00 PII S0895-7061(98)00138-1
AJH–OCTOBER 1998 –VOL. 11, NO. 10, PART 1
GENDER, DIASTOLIC FUNCTION, AND EXERCISE CAPACITY
tion to exercise capacity. The purpose of this study was to evaluate the relationship between resting echocardiographic variables and exercise performance in hypertensive subjects. We also examined the effects of gender in the relationship between resting diastolic function and exercise capacity. MATERIALS AND METHODS We enrolled hypertensive subjects (34 men and 23 women) who were referred for evaluation because of newly diagnosed or uncontrolled hypertension. They all had office blood pressures .140/90 mm Hg on at least two separate visits; most had office diastolic blood pressures .100 mm Hg. All antihypertensive medications (if any) were discontinued for at least 2 weeks before all studies. All subjects were sedentary and none had engaged in any regular aerobic training within a year of the study. Other significant concomitant illnesses (eg, coronary artery disease, congestive heart failure) were excluded by a standard history, physical examination, routine laboratory test, and additional diagnostic studies as indicated. Informed, signed consent was obtained and the protocol was approved by the Institutional Review Board of the Mount Sinai School of Medicine. To accurately characterize blood pressure, 24-h blood pressure monitoring was performed on a day of typical weekly activity using the SpaceLabs 90202 and 90207 devices (SpaceLabs, Redmond, WA).8 Two-dimension guided M-mode echocardiography was performed with an ATL Ultramark 6 scanner (Advanced Technology Laboratories, Inc., Bothell, WA) using a 2.5- or 3.5-MHZ transducer. Echocardiographic tracings were coded and analyzed with a commercially available analyzer (Nova MicroSonics, Advanced Technology Laboratories) in a random order by two independent observers blinded to the subjects’ identity and clinical characteristics.9 End-diastolic measurements of LV structures were made according to the Penn convention and LV mass was indexed to body surface area (grams/meter squared).10,11 LV hypertrophy was defined using gender-specific criteria.9 Two-dimension guided pulsed Doppler interrogation of the LV was done from the apical two- and fourchamber view. Recordings of LV inflow were made where maximal velocity of early filling was obtained. At least five spectral tracings of the mitral inflow were digitized and averaged to obtain peak velocity of early and late diastolic filling and their ratio. The following measurements were also made: isovolumic relaxation time, time to peak early diastolic filling, deceleration time of peak early diastolic filling, and slope (descent) of early diastolic filling. All subjects underwent a symptom-limited graded exercise test on a bicycle ergometer in the upright position, with simultaneous electrocardiogram and re-
1253
spiratory gas analysis. Subjects performed 2 min of unloaded exercise followed by increasing workloads, based on age and weight, of 10 to 30 W/min on an electronically braked cycle ergometer (Mijnhardt). Patients were encouraged to give their maximal effort and exercised until exhaustion developed. Simultaneous 12-lead electrocardiographic monitoring (Case 12, Marquette) was performed and blood pressure was recorded every minute by manual sphygmomanometry. Oxygen consumption (VO2), carbon dioxide production, minute ventilation, and respiratory exchange ratio were computed from measurements made every eighth breath (2001 System, Medical Graphics Corp., St. Paul, MN). Maximal VO2 (VO2max) was defined as the highest VO2 reached during exercise or as the value at which the rate of increase in VO2 did not exceed 0.75 mL/kg/min over a 2-min period. Atrial natriuretic peptide levels were measured from extracted plasma by radioimmunoassay as previously described.12 Data are expressed as mean value 6 SD. Gender differences were evaluated by two-way analysis of variance followed by Bonferroni’s correction method. The relationship between VO2max and resting echocardiographic variables were analyzed by linear correlations. Variables with P , .1 were included in a stepwise linear regression analysis to detect independent determinants of exercise capacity. The impact of left ventricular hypertrophy (LVH) and more severe diastolic dysfunction on exercise capacity was evaluated by analysis of covariance using VO2max and maximal workload as dependent variables and age as a covariate independent variable. Severity of diastolic dysfunction was defined as age-adjusted peak A . median value for each gender (66 cm/sec in men, 70 cm/sec in women). P , .05 was considered significant. RESULTS Subject characteristics were similar between sexes except for the expected higher LV mass, maximal workload, and VO2max in men (Table 1). Both genders gave similar effort in the exercise test as demonstrated by equivalent peak exercise respiratory exchange ratios. Compared to age-specific normal values,13 our subjects had reduced E/A ratios, prolonged isovolumetric relaxation times, shortened deceleration times, and decreased deceleration slope of the E wave. Atrial natriuretic peptide levels (a marker of left atrial pressure) were related directly to peak A wave (r 5 0.39, P , .05) and inversely to VO2max (r 5 20.34, P , .05). In the entire group, multivariate analysis showed that gender and age were independent predictors of VO2max (r2 5 0.46, P , .001). Gender, age, and peak systolic BP were each independent predictors of maximal workload (r2 5 0.68, P , .001) (Table 2). Because gender was the strongest determinant of exercise ca-
1254
AJH–OCTOBER 1998 –VOL. 11, NO. 10, PART 1
GHARAVI ET AL
TABLE 1. CHARACTERISTICS OF STUDY SUBJECTS
Age (years) Body mass index (kg/m2) Ambulatory BP (mm Hg) Heart rate, rest (beats/min) Peak exercise BP (mm Hg) Peak exercise heart rate (beats/min) VO2max (mL/kg/min) Maximal workload (W) Peak respiratory exchange ratio ANP (pg/mL) LV mass (g/m2) LV hypertrophy (%) Ejection fraction (%) Fraction shortening (%) Peak A wave (cm/sec) Peak E wave (cm/sec) E/A ratio Isovolumetric relaxation time (msec) Time to peak E wave (msec) Deceleration time of E wave (msec) Slope of deceleration of E wave (cm/sec2)
Women (n 5 23)
Men (n 5 34)
49 6 10 30.8 6 3.9 154 6 12/99 6 11 72 6 9 226 6 30/118 6 23 157 6 18 17.5 6 4.4 83 6 26 1.1 6 0.1 29 6 14 93 6 21 6 (26%) 66 6 5 37 6 6 67 6 18 66 6 14 0.98 6 0.4 119 6 27 70 6 20 129 6 53 307 6 120
49 6 8 29.1 6 3.4 157 6 16/103 6 12 69 6 9 242 6 29/121 6 20 165 6 14 22.9 6 9.4* 133 6 36* 1.2 6 0.1 23 6 10 116 6 39* 12 (35%) 62 6 11 34 6 8 63 6 13 60 6 13 0.95 6 0.3 124 6 23 83 6 22 138 6 44 281 6 119
*P , .001 men v women. Ambulatory BP, average 24-h ambulatory blood pressure; ANP, atrial natriuretic peptide; BP, blood pressure; LV mass, left ventricular mass indexed to body surface area; E/A, ratio of peak early to late transmitral flow velocities; Peak A wave, peak late transmitral flow velocity; Peak E wave, peak early transmitral flow velocity; VO2max, maximal exercise oxygen consumption.
pacity, the analyses were performed within each gender group. In men, VO2max was inversely related to peak A wave, age, LV mass, and atrial natriuretic peptide (ANP) levels (Table 2, Figure 1). In addition, LV mass directly correlated with ANP level (r 5 0.39, P , .05).
On multivariate analysis, peak A wave was the only independent predictor of VO2max (r2 5 0.41, P , .001), whereas LV mass and peak A remained independently related to maximal workload (r2 5 0.42, P , .001) (Table 2). Similar results were found when LV mass was included in the model as a categorical vari-
TABLE 2. CORRELATION OF EXERCISE CAPACITY WITH SELECTED CLINICAL VARIABLES Overall Group (n 5 57)
Women (n 5 23)
Variable
VO2max
Maximal Workload
VO2max
Maximal Workload
Gender Age BMI Ambulatory SBP Peak SBP LV mass EF Peak A Peak E E/A ANP
0.53*\ 20.42†\ 0.01 20.03 0.31‡ 0.10 20.12 20.39† 20.20 0.25 20.28
0.61*\ 20.42†\ 0.10 0.10 0.43†\ 0.10 20.10 20.32† 20.21 0.16 20.29‡
— 20.45‡ 0.08 20.04 0.40§ 0.37§ 0.17 20.13 0.29 0.34 20.17
— 20.65*\ 0.01 0.37 0.53‡\ 0.38§ 0.16 0.01 0.34 0.29 0.09
All values using Pearson correlation coefficient. See Table 1 for other abbreviations. * P , .001; † P , .01; ‡ P , .05; § P , .1. \ Independent determinant of exercise capacity on multivariate analysis.
Men (n 5 34) VO2max
— 20.48† 0.20 20.16 20.01 20.34‡ 20.06 20.64*\ 20.38‡ 0.27 20.35‡
Maximal Workload
— 20.46† 0.01 20.17 0.18 20.37‡\ 0.04 20.57*\ 20.26 0.28 20.40‡
AJH–OCTOBER 1998 –VOL. 11, NO. 10, PART 1
GENDER, DIASTOLIC FUNCTION, AND EXERCISE CAPACITY
1255
FIGURE 1. Relationship between resting peak A wave and maximal oxygen consumption (VO2max) in women (A) and in men (B).
able based on the presence or absence of LVH. Therefore, men with LVH had a 21% reduction in maximal workloads compared to their counterparts without LVH (114 6 11 v 144 6 14 W, P 5 .02). Similarly, men with more severe diastolic dysfunction (peak A waves .66 cm/sec) had a 19% reduction in VO2max (20.5 6 0.8 v 25.3 6 0.8 mL/kg/min, P , .001) and maximal workloads (118 6 8 v 146 6 8 W, P , .05) after adjustment for age. In women, the relationship between VO2max and echocardiographic measures of systolic or diastolic function was weak and did not approach statistical significance (Table 2, Figure 1). In addition, LV mass did not correlate with ANP or any Doppler parameter. Only age and peak systolic BP correlated with exercise capacity. On multivariate analysis, peak systolic BP was independently related to VO2max (r2 5 0.34, P , .05), whereas age and peak exercise systolic blood pressure remained independently related to maximal workload (r2 5 0.67, P , .001) (Table 2). There was no difference in exercise capacity between women according to the presence of LVH or more severe diastolic dysfunction (defined as peak A wave .70 cm/ sec) even after adjustment for age and other clinical variables. DISCUSSION In this study of sedentary hypertensive subjects, we found significant gender differences in the relation between resting echocardiographic variables and exercise capacity. LV mass and peak A wave were in-
versely related to exercise capacity in men but not in women. The presence of LVH or worse diastolic function was associated with significantly reduced exercise tolerance in hypertensive men, even after adjustment for age and other clinical variables. In addition, we found a direct correlation between ANP levels and LV mass in male subjects. These findings in men are compatible with previous studies showing that hypertension alters chamber compliance, causes LV hypertrophy and diastolic dysfunction, and ultimately leads to poor stroke volume response and diminished exercise capacity.2,12,14 In contrast to men, LV mass and diastolic filling at rest were not significantly related to exercise capacity in hypertensive women. Our women had lower LV mass index and exercise capacity compared to men but other clinical characteristics were similar. The lower exercise capacity in women could not be explained by differences in exercise effort, blood pressure level, degree of obesity, diastolic filling parameters, or prevalence of LV hypertrophy. These gender differences in cardiac mass and exercise tolerance have been previously described but have not been adequately explained.1,3,4 In addition, many studies assessing the role of cardiac function in exercise have not performed gender-specific analyses.1,2,15 Although the interpretation of Doppler variables is generally more complicated in hypertensive subjects, there are no reported gender limitations of this technique for the evaluation of diastolic function.4,12 Thus, our findings cannot be solely attributed to a limitation of the
1256
GHARAVI ET AL
Doppler echocardiographic technique. Rather data suggest that our observations stem from gender effects on the cardiac response to exercise. Multiple studies have demonstrated that the relative contribution of Frank-Starling mechanisms and enhanced contractility to the increase in stroke volume during exercise differs between the sexes.3,5,6,14 Several but not all studies have indicated that women achieve little or no increase in stroke volume or ejection fraction during dynamic exercise. Faggard et al3 showed that hypertensive women achieved a smaller increase in stroke volume and cardiac output from rest to exercise in the upright position. Using radionuclide angiography during upright5 and supine6,16 exercise, other studies showed that healthy male and female volunteers have similar increases in stroke volume but achieve it through different mechanisms: men have an increase in ejection fraction without an increase in end-diastolic counts, whereas women have an increase in end-diastolic counts without an increase in ejection fraction. Using radionuclide angiography and right-sided catheterization in normal subjects, Sullivan et al7 did not observe any gender differences in the stroke volume response to upright exercise in normal adults, although there was a trend toward lower ejection fraction response in women. Taken together, the published data suggest that women are able to augment diastolic filling significantly beyond resting values during exercise. Thus, resting measures of diastolic filling may not necessarily correlate with exercise diastolic function in women and would not be an accurate indicator of stroke volume response and ultimately, exercise capacity. In contrast, diastolic filling changes little during exercise in men, and therefore, LV filling at rest may be a good predictor of diastolic contribution to exercise capacity. This ability to augment diastolic filling during exercise may potentially be related to gender differences in venous capacitance and the capacity to maintain preload in the upright posture. Thus, a larger trial using invasive hemodynamic measurements would be required to provide further clarification on this topic. In the present study, both genders showed reduced E/A ratios accompanied by prolonged isovolumetric relaxation times, shortened deceleration times, and decreased deceleration slope of E wave (compared to age specific normal values).12 These combined abnormalities suggest diastolic dysfunction beyond the early stage characterized by reversal of E/A wave ratios. Our subjects had probably begun to “pseudonormalize” their LV inflow patterns, leading to elevation of the E wave and shortening of the deceleration time.17 This is further supported by the direct correlation between the peak A wave and atrial natriuretic peptide levels, which reflect cardiac filling pressures. Hence, in the presence of moderately advanced dia-
AJH–OCTOBER 1998 –VOL. 11, NO. 10, PART 1
stolic dysfunction, only peak A waves in men remained the most reliable indicator of LV diastolic properties and exercise capacity. The findings of this study apply to sedentary, untreated or poorly controlled hypertensive subjects only; the effect of aerobic exercise training or specific pharmacologic intervention on the relation between resting echocardiographic data and exercise capacity was not evaluated. However, our study has useful clinical implications; namely that the finding of LVH and abnormal diastolic function on resting echocardiography will correlate with diminished exercise capacity in hypertensive men but not in women. Thus, echocardiographic findings may explain exercise intolerance and the symptoms of diastolic heart failure in sedentary hypertensive men. In contrast, resting echocardiographic data may not be adequate to explain reduced exercise capacity in hypertensive women. This may be attributable to gender differences in the relative contribution of diastolic filling to cardiac performance during exercise; alternatively, echocardiographic measurements of resting diastolic function may not be related to exercise diastolic function in hypertensive women. Hence, investigation of a cardiac cause of exercise intolerance in sedentary hypertensive women may necessitate an evaluation of cardiac function during exercise. Our data also suggest that future studies on the cardiac physiology of exercise should perform gender-specific analyses and focus on defining clinically useful cardiac predictors of exercise capacity in hypertensive women. ACKNOWLEDGMENT We thank Mr. Howard Printz for his encouragement and generous support of the Hypertension Section.
REFERENCES 1. Vanoverschelde JLJ, Essamri B, Vanbutsele R, et al: Contribution of left ventricular diastolic function to exercise capacity in normal subjects. J Appl Physiol 1993;74(5): 2225–2233. 2. Cuocolo A, Sax FL, Brush JE, et al: Left ventricular hypertrophy and impaired diastolic filling in essential hypertension: diastolic mechanisms for systolic dysfunction during exercise. Circulation 1990;81:978 –986. 3. Faggard RH, Thijs LB, Amery AK: The effects of gender on aerobic power and exercise hemodynamics in hypertensive adults. Med Sci Sports Exercise 1995;27(1):29 –34. 4. Hinderliter AL, Light KC, Willis PW: Gender differences in left ventricular structure and function in young adults with normal and marginally elevated blood pressure. Am J Hypertens 1992;5:32–36. 5. Higginbotham MB, Morris KG, Coleman E, Cobb FR: Sex-related differences in the normal cardiac response to upright exercise. Circulation 1984;70:357–366. 6. Hanley PC, Zinsmeister AR, Clements IP, et al: Genderrelated differences in cardiac response to supine exercise
AJH–OCTOBER 1998 –VOL. 11, NO. 10, PART 1
GENDER, DIASTOLIC FUNCTION, AND EXERCISE CAPACITY
assessed by radionuclide angiography. J Am Coll Cardiol 1989;624 – 629. 7. Sullivan MJ, Cobb FR, Higginbotham MB: Stroke volume increases by similar mechanisms during upright exercise in normal men and women. Am J Cardiol 1991; 67:1405–1412. 8. Moore CR, Krakoff LR, Phillips RA. Comfirmation or exclusion of stage I hypertension by ambulatory blood pressure monitoring. Hypertension 1997;29:1109 –1113. 9. Jhang JS, Diamond JA, Phillips RA. Interobserver variability of left ventricular measurements in a population of predominantly obese hypertensives using simultaneously acquired and displayed M-mode and 2-D cine echocardiography. Echocardiography 1997;14:9 –15. 10. Devereux RB, Reichek N. Echocardiographic determination of left ventricular mass in man. Circulation 1977;55: 613– 618. 11. Lauer MS, Anderson KM, Larson M, Levy D. A new method for indexing left ventricular mass for differences in body size. Am J Cardiol 1994;74:487– 491.
1257
12. Eison HB, Rosen, MJ, Phillips RA. Krakoff LR. Determinants of atrial natriuretic factor in the adult respiratory distress syndrome. Chest 1988;94:1040 –1045. 13. Benjamin EJ, Levy D, Anderson KM, et al: Determinants of Doppler indexes of left ventricular diastolic function in normal subjects (the Framingham Heart study). Am J Cardiol 1992;70:508 –515. 14. Lim PO, MacFadyen R, Clarkson PB, McDonald TM. Impaired exercise tolerance in hypertensive patients. Ann Intern Med 1996;124(1, pt1):41–55. 15. Poliner LR, Dehmer GJ, Lewi SE, et al: Left ventricular performance in normal subjects: a comparison of the responses to exercise in the upright and supine positions. Circulation 1980;62:528 –534. 16. Adams KF, Vincent LM, McAllister SM, et al. The influence of age and gender on left ventricular response to supine exercise in asymptomatic subjects. Am Heart J 1987;113:732–742. 17. Nishimura R, Appleton C. “Diastology”: beyond E and A. J Am Coll Cardiol 1996;27:372–374.