Relation of Etiology of Heart Failure (Ischemic Versus Nonischemic) Before Transplantation to Delayed Pulmonary Oxygen Uptake Kinetics After Heart Transplantation

Relation of Etiology of Heart Failure (Ischemic Versus Nonischemic) Before Transplantation to Delayed Pulmonary Oxygen Uptake Kinetics After Heart Transplantation Corey R. Tomczak, MSca Nicholas G. Jendzjowsky, BSca, Kenneth J. Riess, MSca, Wayne Tymchak, MDb, Daniel Kim, MDb, Robert Haennel, PhDa, and Mark J. Haykowsky, PhDa,b,* The effect that pretransplantation heart failure cause has on pulmonary oxygen uptake (V̇O2p) kinetics and peak aerobic power (V̇O2peak) in heart transplant recipients (HTRs) has not been studied. We examined V̇O2p kinetics and V̇O2peak in HTRs with previous ischemic heart failure (I-HTRs; n ⴝ 16, mean age 64 ⴞ 6 years) or nonischemic heart failure (NI-HTRs; n ⴝ 13, mean age 50 ⴞ 12 years). HTRs performed an incremental exercise (V̇O2peak) test and a constant work rate submaximal exercise (V̇O2p kinetics) test. A monoexponential model was used to determine the phase II V̇O2p time constant (␶). Phase II V̇O2p ␶ was slower in I-HTRs (49 ⴞ 10 seconds) than in NI-HTRs (34 ⴞ 10 seconds) (p <0.001). No significant difference was found between I-HTRs and NI-HTRs for V̇O2peak (19.0 ⴞ 6.4 vs 23.0 ⴞ 8.2 ml · kgⴚ1 · minⴚ1, respectively), change in heart rate from rest to steady-state exercise (11 ⴞ 8 vs 9 ⴞ 9 beats · minⴚ1, respectively), or peak exercise heart rate (140 ⴞ 22 vs 144 ⴞ 22 beats · minⴚ1, respectively). In conclusion, the prolonged phase II V̇O2p ␶ in I-HTRs compared with NI-HTRs suggests that the magnitude of alteration in V̇O2p kinetics after heart transplantation may be dependent on previous heart failure cause. © 2007 Elsevier Inc. All rights reserved. (Am J Cardiol 2007;99:1745–1749) Pulmonary oxygen uptake (V̇O2p) kinetics are prolonged in heart transplant recipients (HTRs).1–3 The slower V̇O2p kinetics may be due to reduced oxygen delivery1,4,5 secondary to cardiac denervation,6 diastolic dysfunction4,6 and abnormal vascular function.7,8 However, Grassi et al2 found that improving cardiac output kinetics in HTRs does not alter V̇O2p kinetics, suggesting that skeletal muscle oxidative metabolism abnormalities may prolong V̇O2p kinetics. A limitation of previous investigations was the failure to examine the impact of pretransplantation heart failure cause. Clark et al9 and others10 –12 have shown that the magnitude of decrease in peak aerobic power (V̇O2peak) was dependant on heart failure cause. Furthermore, Patel et al7 demonstrated that abnormal peripheral vascular endothelial function was reversible after transplantation in those with nonischemic but not ischemic heart failure. A consequence of persistent impairments in peripheral vascular function found in HTRs with antecedent ischemic cardiomyopathy is that it may alter oxygen delivery to the muscles, thus prolonging a Cardiovascular Therapeutic Exercise Laboratory, Faculty of Rehabilitation Medicine, and bDivision of Cardiology, Faculty of Medicine, University of Alberta, Edmonton, Alberta, Canada. Manuscript received December 22, 2006; revised manuscript received and accepted January 17, 2007. Dr. Haykowsky is funded by the Heart and Stroke Foundation of Canada and a Canadian Institutes of Health Research (CIHR) New Investigator Award. Mr. Tomczak is a CIHR Strategic Training Fellow in Tomorrow’s Research Cardiovascular Health Professionals (TORCH) and is supported by a doctoral Canada Graduate Scholarship from the Natural Sciences and Engineering Research Council of Canada. *Corresponding author: Tel: 780-492-5970; fax: 780-492-4429. E-mail address: mark.haykowsky@ualberta.ca (M.J. Haykowsky). 0002-9149/07/$ – see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2007.01.058 V̇O2p kinetics and reducing V̇O2peak. We tested the hypothesis that V̇O2p kinetics would be significantly slower and V̇O2peak would be lower in HTRs with previous ischemic heart failure (I-HTRs) than in those with nonischemic heart failure (NI-HTRs). Methods Participants included clinically stable I-HTRs (n ⫽ 16, mean age 64 ⫾ 6 years) and NI-HTRs (n ⫽ 13, mean age 50 ⫾ 12 years). I-HTRs had angiographic evidence of coronary artery disease and histories consistent with previous myocardial injury. The NI-HTRs did not meet these criteria. All participants were receiving standard posttransplantation pharmacologic therapy (Table 1). This investigation received ethical approval from the University of Alberta Health Research Ethics Board. The incremental exercise test was performed on an electrically braked cycle ergometer (Corival; Lode Diagnostics BV, Groningen, The Netherlands). After a 2-minute rest period, the initial power out was set at 15 W and increased by 15 W every 2 minutes until volitional exhaustion. During the test, expired gases were collected and analyzed with a computerized metabolic system (TrueOne 2400 Metabolic Measurement System; ParvoMedics, Salt Lake City, Utah) to determine the ventilatory threshold and V̇O2peak. On a separate day, subjects performed a constant work rate protocol that began with 3 to 5 minutes of rest followed by an abrupt step increase in work rate to approximately 80% of the V̇O2p at the ventilatory threshold for 5 minutes. V̇O2p data from the constant work rate exercise session were averaged into 5-second time bins, and aberrant breaths considered to not reflect V̇O2p kinetic responses were rewww.AJConline.org 1746 The American Journal of Cardiology (www.AJConline.org) Table 1 Participant characteristics and medications Variable I-HTRs (n ⫽ 16) NI-HTRs (n ⫽ 13) Age (yrs) Time after transplantation (yrs) Body mass index (kg/m2) Medications† Cyclosporine Tacrolimus Sirolimus Mycophenolate mofetil Azathioprine Prednisone Angiotensin-converting enzyme inhibitor Angiotensin II receptor blocker ␣ blocker Angiotensin-converting enzyme inhibitor ⫹ thiazide diuretic ␤ blocker Calcium channel blocker Diuretic Aspirin Lipid-lowering agent 64 ⫾ 6* 4.8 ⫾ 4.2 30 ⫾ 4 50 ⫾ 12 5.4 ⫾ 5.4 28 ⫾ 5 13 2 2 9 1 4 9 7 5 1 9 3 6 8 — 1 1 — — 1 1 9 4 12 11 2 9 8 11 9 * p ⬍0.05 versus NI-HTRs. Values are numbers of patients per respective group. Figure 1. Features of the biphasic response observed in V̇O2p after a step change in work rate. Phase I and phase II V̇O2p responses are illustrated. Phase II parameters correspond to those used in the first-order mathematical model for determining V̇O2p kinetics. Time 0 indicates exercise onset, A indicates amplitude response, ␶ indicates the V̇O2p time constant, and TD indicates the time delay for the onset of the phase II response and reflects the duration of the phase I response. The inset graph illustrates the squarewave protocol used for the exercise perturbation. † moved.13 Data were smoothed using a 5-point moving average to reduce measurement noise and enhance the underlying characteristics of the V̇O2p responses.1 After the phase I “cardiodynamic” response, the kinetic parameters of phase II V̇O2p responses were determined using a first-order (monoexponential) mathematical model of the form Y(t) ⫽ Y(b) ⫹ A · [1 ⫺ e⫺(t ⫺ TD)/␶)], where Y is V̇O2p at a given time (t), b is the baseline value of V̇O2p over the last 60 seconds previous to exercise onset, A is the amplitude change in the phase II V̇O2p response, ␶ is the time constant or time for phase II V̇O2p to reach 63% of A, and TD is the time delay of the phase II V̇O2p onset (Figure 1). The time delay of phase II reflects the duration of the phase I response, and we further calculated the amplitude change (A) of the phase I response as the difference between V̇O2p at phase II onset and b, the baseline. We then determined the steady-state V̇O2p as the V̇O2p at b ⫹ phase I V̇O2p A ⫹ phase II V̇O2p A. V̇O2p kinetic parameters were determined using nonlinear regression. The iterative procedure of the computer program (Origin 7.5, OriginLab Corporation, Northampton, Massachusetts) used a Levenberg-Marquardt algorithm whereby the best fit was defined by minimization of the residual sum of squares. The data-fitting window was extended from the phase II onset to approximately 240 seconds into the exponential response to allow the accurate determination of the entire phase II evolution. The datafitting window was lengthened and shortened iteratively until the exponential fit departed from the measured response profile as determined by (1) visual inspection of the curve for appropriateness of fit, (2) visual inspection of the residuals for clustering and systematic deviations from the x-axis, (3) a sudden increase in ␶, and (4) demonstration of a local threshold in the reduced chi-square value.14 The onset of the exponential increase in phase II was further verified using similar criteria. Statistical analysis was performed using SPSS version 11.0 (SPSS, Inc., Chicago, Illinois). Given a significant difference between groups for age and previously reported age-related differences in V̇O2peak15,16 and V̇O2p kinetics,17 analysis of covariance was performed on all outcome variables using age as the covariate. We calculated the 95% confidence intervals for the estimates of ␶ values for phase II V̇O2p using the method of Lamarra et al13: K1 ⫽ L ⫻ (SD/A), where the true value of ␶ is within K1 seconds of the obtained value with 95% confidence, based on the SD of breath-by-breath V̇O2p fluctuations over a 2-minute steadystate period, the phase II V̇O2p amplitude (A) response, and a constant (L). Correlation regression was used to determine the relation between V̇O2p kinetics, the duration of time after transplantation, and V̇O2peak. Data are presented as mean ⫾ SD, and p values ⬍0.05 were considered statistically significant. Results V̇O2p at the ventilatory threshold and at peak exercise, peak heart rate, peak power output, peak respiratory exchange ratio, and the minute ventilation/carbon dioxide production slope was not significantly different between I-HTRs and NI-HTRs (Table 2). No significant difference was found between I-HTRs and NI-HTRs for steady-state exercise heart rate minus heart rate at rest (11 ⫾ 8 vs 9 ⫾ 9 beats · min⫺1, respectively), steady-state heart rate (102 ⫾ 12 vs 102 ⫾ 16 beats · min⫺1, respectively), steady-state power Heart Failure/V̇O2p Kinetics and Heart Transplantation 1747 Table 2 Cardiopulmonary parameters from peak exercise testing Variable Rest Heart rate (beats · min⫺1) Mean arterial pressure (mm Hg) Ventilatory threshold V̇O2p (L · min⫺1) V̇O2p (ml · kg⫺1 · min⫺1) Peak exercise Power output (W) Heart rate (beats · min⫺1) V̇O2p (L · min⫺1) V̇O2p (ml · kg⫺1 · min⫺1) V̇E/V̇CO2 slope Respiratory exchange ratio I-HTRs NI-HTRs 91 ⫾ 10 95 ⫾ 8 93 ⫾ 16 90 ⫾ 10 1.07 ⫾ 0.26 12.5 ⫾ 3.2 1.17 ⫾ 0.33 14.5 ⫾ 5.5 98 ⫾ 35 140 ⫾ 22 1.63 ⫾ 0.50 19.0 ⫾ 6.4 38 ⫾ 4 1.1 ⫾ 0.1 115 ⫾ 35 144 ⫾ 22 1.82 ⫾ 0.54 23.0 ⫾ 8.2 39 ⫾ 5 1.1 ⫾ 0.1 Values are expressed as mean ⫾ SD. V̇CO2 ⫽ carbon dioxide production. V̇E ⫽ minute ventilation. Table 3 Pulmonary oxygen uptake kinetic parameters for ischemic heart transplant recipients and nonischemic heart transplant recipients Parameter ⫺1 Y(b) (L · min ) Phase I A (L · min⫺1) TD (s) Phase II A (L · min⫺1) Phase II ␶ (s) Steady-state V̇O2p (L · min⫺1) I-HTRs NI-HTRs 0.34 ⫾ 0.07 0.23 ⫾ 0.12 61 ⫾ 25 0.28 ⫾ 0.10 49 ⫾ 10* 0.85 ⫾ 0.16 0.30 ⫾ 0.10 0.27 ⫾ 0.17 56 ⫾ 17 0.26 ⫾ 0.12 34 ⫾ 10 0.83 ⫾ 0.22 Values are expressed as mean ⫾ SD. * p ⬍0.001 versus NI-HTRs. A ⫽ amplitude change in V̇O2p for phase I and phase II responses; TD ⫽ phase II time delay; Y(b) ⫽ baseline V̇O2p; ␶ ⫽ time constant. output (24 ⫾ 9 vs 30 ⫾ 6 W, respectively), or steady-state V̇O2p during the square-wave V̇O2p kinetics protocol (Table 3). Figure 2 illustrates the V̇O2p response, phase II monoexponential curve fits, and corresponding residuals for a representative I-HTR and NI-HTR. V̇O2p during the pretransition baseline was not significantly different between groups (Table 3). Phases I and II V̇O2p amplitude responses were similar between groups, but phase II V̇O2p ␶ was 31% slower in I-HTRs than in NI-HTRs (p ⬍0.001; Table 3 and Figure 3). The 95% confidence interval of ␶ for phase II V̇O2p was ⫾4.7 seconds for I-HTRs and ⫾5.1 seconds for NI-HTRs. These values are based on the phase II V̇O2p amplitude responses reported in Table 3, and SDs of steadystate breath-by-breath V̇O2p fluctuations of 0.025 and 0.023 L · min⫺1 for I-HTRs and NI-HTRs, respectively. Finally, V̇O2p kinetics were not related to the duration of time after transplantation (r ⫽ 0.23, p ⫽ 0.23) or V̇O2peak (r ⫽ 0.16, p ⫽ 0.42). Discussion The principal new finding of this investigation is that V̇O2p kinetics are significantly slower in I-HTRs than in NIHTRs. This finding extends the results of previous studies by demonstrating that the extent of the impairment in V̇O2p Figure 2. V̇O2p responses to a step change from rest to a work rate approximating V̇O2p at 80% of the ventilatory threshold for a I-HTR and NI-HTR. Data points represent 5-second averages. Monoexponential curve fits and corresponding residuals are also shown. Note that the time to reach 63% of the V̇O2p asymptote (␶) is markedly slower in the representative I-HTR compared with the representative NI-HTR. Figure 3. Vertical point plots illustrating individual phase II V̇O2p time constants (␶) for I-HTRs (n ⫽ 16) and NI-HTRs (n ⫽ 13). Overlapping data points for respective groups are shown parallel to their point of origin. Group means are also indicated. kinetics found in HTRs may be related to previous heart failure cause. The prolonged V̇O2p kinetics1,2,4,18 found in HTRs are attributed in part to the impaired cardiac output kinetics.19,20 In turn, the abnormal cardiac output kinetics are related to blunted heart rate kinetics associated with cardiac denervation, because the rapid increase in stroke volume during the transition to submaximal constant load exercise is greater in 1748 The American Journal of Cardiology (www.AJConline.org) HTRs than in healthy subjects.21 On the basis of these findings, the divergent V̇O2p kinetics between our I-HTRs and NI-HTRs are unlikely due to differences in cardiac output kinetics, because the change in heart rate from rest to steady-state and the steady-state exercise heart rate were similar in the 2 groups. The slower V̇O2p kinetics found in I-HTRs may be linked to impaired vascular function or oxidative metabolism and a concomitant decrease in skeletal muscle blood flow or use.8 Patel et al7 showed that the peripheral vascular endothelial dysfunction associated with nonischemic cardiomyopathy is reversed after heart transplantation, but persistent abnormalities in vascular function remain in HTRs with antecedent ischemic cardiomyopathy. Schaufelberger et al22 and Stratton et al23 also demonstrated that the altered skeletal morphology (i.e., lower percentage type I fibers, capillary density, and oxidative enzyme activity) and abnormal oxidative metabolism associated with the heart failure syndrome is not reversed after heart transplantation. Currently, the effect that previous heart failure cause has on posttransplantation skeletal morphology and bioenergetics remains unknown. Braith et al24 recently examined the effect that resistance training has on skeletal morphology and enzymatic activity in HTRs. Fortunately, these investigators provided individual baseline data for HTRs with previous ischemic cardiomyopathy (i.e., I-HTRs) or idiopathetic dilated cardiomyopathy (i.e., NI-HTRs). Reanalysis of their baseline data based on pretransplantation heart failure cause reveals no significant difference between I-HTRs and NIHTRs for percentage type I (24.9 ⫾ 9.1% vs 17.7 ⫾ 5%, respectively, p ⬎0.05), type IIa (31.7 ⫾ 7.4% vs 35.9 ⫾ 6%, respectively, p ⬎0.05), and type IIx (43.4 ⫾ 12% vs 46.5 ⫾ 6%, respectively, p ⬎0.05) myosin heavy chain isoforms or lactate dehydrogenase (106 ⫾ 34 vs 111 ⫾ 40 mmol · g⫺1 wet weight · min⫺1, respectively, p ⬎0.05), citrate synthase (11.1 ⫾ 2.6 vs 9.4 ⫾ 1.2, mmol · g⫺1 wet weight · min⫺1, respectively, p ⬎0.05), or 3-hydroxyacyl coenzyme-A dehydrogenase (3.6 ⫾ 0.9 vs 3.7 ⫾ 0.9 mmol · g⫺1 wet weight · min⫺1, respectively, p ⬎0.05). On the basis of this analysis, we suspect that the slower V̇O2p kinetics in I-HTRs may be secondary to persistent abnormalities in vascular function and a concomitant reduction in skeletal muscle blood flow, because the impairment in skeletal morphology and biochemistry appear similar between I-HTRs and NI-HTRs in the study by Braith et al.24 However, a role for impaired skeletal muscle oxygen use cannot be entirely ruled out in the present investigation. Several previous investigators have found that the extent of the decrease in peak aerobic power is related to cause of heart failure, with patients with ischemic cardiomyopathy having V̇O2peak values 13% to 21% lower than those with nonischemic cardiomyopathy. Consistent with these results, we found that V̇O2peak was approximately 17% lower in I-HTRs than in NI-HTRs. However, there was no significant relation between V̇O2peak and V̇O2p kinetics in our study group. This finding suggests that the factors limiting V̇O2peak and V̇O2p kinetics differ. Notably, the mean V̇O2peak for our I-HTRs (19 ml · kg⫺1 · min⫺1) is just higher than the threshold level required for independent living (i.e., 15 to 18 ml · kg⫺1 · min⫺1).25 Finally, the phase II V̇O2p ␶ for our I-HTRs (49 seconds) is similar to that reported by Koike et al26 for subjects with ischemic heart disease and impaired left ventricular systolic function. Taken together, these findings suggest that our I-HTRs will quickly fatigue when performing activities of daily living that require abrupt increases in metabolic demand, such as transitioning from rest to a moderate walking speed, as evidenced by the impaired phase II V̇O2p time constant (␶) in the I-HTR group. Our findings also highlight the clinical importance of determining V̇O2p kinetics for identifying patients with persistent impairments in metabolic readjustment ability. A limitation of our investigation is that V̇O2p kinetics were determined during a single constant load exercise bout. However, our phase II V̇O2p ␶ values are similar to those previously reported for HTRs.4,5 Moreover, we observed a relatively narrow 95% confidence interval for phase II V̇O2p ␶, which is notably improved on previous reports for HTRs (13.6 to 20.4 seconds).1 Averaging further repetitions of the constant load exercise bout would only serve to narrow our observed 95% confidence interval for phase II V̇O2p ␶ values and would not alter our findings. A second limitation is that V̇O2p kinetics were measured on average 5.1 years (range 0.4 to 18.2) after transplantation. However, we did not find a significant relation between V̇O2p kinetics and time after transplantation. Thus, the slower V̇O2p kinetics found in I-HTRs do not appear to improve after transplantation. A final limitation is that we did not examine the underlying mechanisms for the divergent V̇O2p kinetics between I-HTRs and NI-HTRs. However, the primary objective of this investigation was to determine if the abnormal V̇O2p kinetics found in HTRs are related to pretransplantation heart failure cause. Accordingly, future investigations are required to determine the role that impairments in oxygen delivery or use play in the abnormal V̇O2p kinetics found in I-HTRs. 1. Paterson DH, Cunningham DA, Pickering JG, Babcock MA, Boughner DR. Oxygen uptake kinetics in cardiac transplant recipients. J Appl Physiol 1994;77:1935–1940. 2. Grassi B, Marconi C, Meyer M, Rieu M, Cerretelli P. Gas exchange and cardiovascular kinetics with different exercise protocols in heart transplant recipients. J Appl Physiol 1997;82:1952–1962. 3. M’Bouh S, Bellmont S, Lampert E, Epailly E, Zoll J, N’Guessan B, Ribera F, Geny B, Oyono S, Arnold P, et al. An impaired cardiodynamic phase contributes to the abnormal VO(2) kinetics at exercise onset in both congestive heart failure and heart transplant patients but results from differing mechanisms. Transplant Proc 2001;33:3543– 3545. 4. Mettauer B, Zhao QM, Epailly E, Charloux A, Lampert E, HeitzNaegelen B, Piquard F, di Prampero PE, Lonsdorfer J. VO(2) kinetics reveal a central limitation at the onset of subthreshold exercise in heart transplant recipients. J Appl Physiol 2000;88:1228 –1238. 5. Lanfranconi F, Borrelli E, Ferri A, Porcelli S, Maccherini M, Chiavarelli M, Grassi B. Noninvasive evaluation of skeletal muscle oxidative metabolism after heart transplant. Med Sci Sports Exerc 2006;38: 1374 –1383. 6. Kao AC, Van Trigt P III, Shaeffer-McCall GS, Shaw JP, Kuzil BB, Page RD, Higginbotham MB. Central and peripheral limitations to upright exercise in untrained cardiac transplant recipients. Circulation 1994;89:2605–2615. 7. Patel AR, Kuvin JT, Pandian NG, Smith JJ, Udelson JE, Mendelsohn ME, Konstam MA, Karas RH. Heart failure etiology affects peripheral vascular endothelial function after cardiac transplantation. J Am Coll Cardiol 2001;37:195–200. 8. Haykowsky M, Eves N, Figgures L, McLean A, Koller M, Taylor D, Tymchak W. Effect of exercise training on VO2peak and left ventric- Heart Failure/V̇O2p Kinetics and Heart Transplantation 9. 10. 11. 12. 13. 14. 15. 16. 17. ular systolic function in recent cardiac transplant recipients. Am J Cardiol 2005;95:1002–1004. Clark AL, Harrington D, Chua TP, Coats AJ. Exercise capacity in chronic heart failure is related to the aetiology of heart disease. Heart 1997;78:569 –571. Arena R, Myers J, Abella J, Peberdy MA. Influence of heart failure etiology on the prognostic value of peak oxygen consumption and minute ventilation/carbon dioxide production slope. Chest 2005;128: 2812–2817. Arena R, Tevald M, Peberdy MA. Influence of etiology on ventilatory expired gas and prognosis in heart failure. Int J Cardiol 2005;99:217– 223. De Feo S, Franceschini L, Brighetti G, Cicoira M, Zanolla L, Rossi A, Golia G, Zardini P. Ischemic etiology of heart failure identifies patients with more severely impaired exercise capacity. Int J Cardiol 2005; 104:292–297. Lamarra N, Whipp BJ, Ward SA, Wasserman K. Effect of interbreath fluctuations on characterizing exercise gas exchange kinetics. J Appl Physiol 1987;62:2003–2012. Rossiter HB, Ward SA, Howe FA, Kowalchuk JM, Griffiths JR, Whipp BJ. Dynamics of intramuscular 31P-MRS P(i) peak splitting and the slow components of PCr and O2 uptake during exercise. J Appl Physiol 2002;93:2059 –2069. Fitzgerald MD, Tanaka H, Tran ZV, Seals DR. Age-related declines in maximal aerobic capacity in regularly exercising vs. sedentary women: a meta-analysis. J Appl Physiol 1997;83:160 –165. Wilson TM, Tanaka H. Meta-analysis of the age-associated decline in maximal aerobic capacity in men: relation to training status. Am J Physiol Heart Circ Physiol 2000;278:H829 –H834. Babcock MA, Paterson DH, Cunningham DA, Dickinson JR. Exercise on-transient gas exchange kinetics are slowed as a function of age. Med Sci Sports Exerc 1994;26:440 – 446. 1749 18. Borrelli E, Pogliaghi S, Molinello A, Diciolla F, Maccherini M, Grassi B. Serial assessment of peak VO2 and VO2 kinetics early after heart transplantation. Med Sci Sports Exerc 2003;35:1798 –1804. 19. Cerretelli P, Marconi C, Meyer M, Ferretti G, Grassi B. Gas exchange kinetics in heart transplant recipients. Chest 1992;101:199S–205S. 20. Warburton DE, Sheel AW, Hodges AN, Stewart IB, Yoshida EM, Levy RD, McKenzie DC. Effects of upper extremity exercise training on peak aerobic and anaerobic fitness in patients after transplantation. Am J Cardiol 2004;93:939 –943. 21. Braith RW, Plunkett MB, Mills RM Jr. Cardiac output responses during exercise in volume-expanded heart transplant recipients. Am J Cardiol 1998;81:1152–1156. 22. Schaufelberger M, Eriksson BO, Lonn L, Rundqvist B, Sunnerhagen KS, Swedberg K. Skeletal muscle characteristics, muscle strength and thigh muscle area in patients before and after cardiac transplantation. Eur J Heart Fail 2001;3:59 – 67. 23. Stratton JR, Kemp GJ, Daly RC, Yacoub M, Rajagopalan B. Effects of cardiac transplantation on bioenergetic abnormalities of skeletal muscle in congestive heart failure. Circulation 1994;89:1624 –1631. 24. Braith RW, Magyari PM, Pierce GL, Edwards DG, Hill JA, White LJ, Aranda JM Jr. Effect of resistance exercise on skeletal muscle myopathy in heart transplant recipients. Am J Cardiol 2005;95: 1192–1198. 25. Paterson DH, Govindasamy D, Vidmar M, Cunningham DA, Koval JJ. Longitudinal study of determinants of dependence in an elderly population. J Am Geriatr Soc 2004;52:1632–1638. 26. Koike A, Hiroe M, Adachi H, Yajima T, Yamauchi Y, Nogami A, Ito H, Miyahara Y, Korenaga M, Marumo F. Oxygen uptake kinetics are determined by cardiac function at onset of exercise rather than peak exercise in patients with prior myocardial infarction. Circulation 1994; 90:2324 –2332.