832 lacc Vol. 5, No.4 April!985:832-9 Relative Attenuation of Sympatheti Drive During Exerise in Patients With Congestive Heart Failure GARY S. FRANCIS, MD, FACC, STEVEN R. GOLDSMITH, MD, FACC, SUSAN ZIESCHE, RN, HISAYOSHI NAKAJIMA, MD, JAY N. COHN, MD, FACC Minneapolis. Minnesota Patients with ongestive heart failure have been onsidered to have augmented sympatheti drive both at rest and during dynami exerise. The augmentation observed during exerise may be related to the state of near exhaustion experiened by patients with heart failure at relatively low work loads. To ompare the response of the sympatheti nervous system to exerise in normal subjets and patients with heart failure when they are working in a omparable physiologiframe of referene, the data for both groups an be expressed as perent peak oxygen onsumption ahieved (perent peak V0 2 ) rather than as a funtion of absolute oxygenonsumption (V0 2 ). Ten healthy ontrol subjets and 31 patients with hroni linial lass II and III heart failure were studied during upright maximal biyle exerise. Eighteen of the 31 patients had primary ardiomyopathy and 13 had ishemi ardiomyopathy. The average ejetion fration at rest was 24 ± 10% (±SD) in the group with heart failure. Heart rate, systoli blood pressure, V0 2 and plasma norepinephrine levels were measured at rest and throughout exerise. When the data were expressed as a funtion of perent peak V0 2 ahieved, patients with heart failure demonstrated a flatter slope (p = 0.004) than normal in the response of plasma norepinephrine to exerise, indiating a relative blunting of sympatheti drive. This was aompanied by attenuated heart rate (p = 0.001) and blood pressure (p < 0.001) responses. These differenes were not apparent when the data are expressed as a funtion of absolute V0 2 Thus, patients with heart failure may have a relative attenuation rather than augmentation ofsympathetidrive during exerise when omparisons are made using a omparable physiologi frame of referene. Redued sympatheti drive might partially explain the blunted heart rate and blood pressure responses to exerise and is onsistent with the broader hypothesis that heart failure is haraterized by a generalized inability of various stimuli to maximally ativate the sympatheti nervous system despite inreased sympatheti drive in the basal state. (J Am Coil CardioI1985;5:832-9) Patients with hroni ongestive heart failure have an altered responsiveness of the sympatheti nervous system to ertain forms of stress. For example, during orthostasis suh patients have a redued heart rate response (1,2) and a less than normal inrease in plasma norepinephrine and plasma renin ativity (3,4). The usual derease in forearm blood flow in response to upright tilt is also attenuated in heart failure (5,6), and patients with severe heart failure do not From the Department of Mediine, University of Minnesota Medial Shool, Minneapolis, Minnesota and the Veterans Administration Medial Center, Minneapolis, Minnesota. This study was supported by Grants HL22977-03 and HL07184 from the National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, Maryland and a grant from the Veterans Administration Researh Servie, Washington, D.C. Dr. Goldsmith is a reipient of a Clinial Investigator Award from the National Institutes of Health, Bethesda, Maryland. Manusript reeived August 20, 1984; revised manusript reeived Otober 24, 1984, aepted November 6, 1984. Address for reprints: Gary S. Franis, MD, Cardiovasular Researh, Veterans Administration Medial Center, Minneapolis, Minnesota 55417. 1985 by the Amerian College of Cardiology exhibit an inrease in plasma norepinephrine in response to nitroprusside-indued vasodilation (7). In addition to abnormal responses to baroreeptor ativation initiated by tilt or pharmaologi vasodilation, patients with heart failure have an abnormal sympatheti response to upright dynami exerise (8). This abnormality is haraterized by a redued hronotropi response to exerise (8), a derease in stroke volume despite exessive inrements in end-diastoli pressure (9), alterations in the redistribution of regional blood flow (10) and a redution in maximal oxygen onsumption (\'0 2 ) (II). Early work by Chidsey et al. (12) suggested that patients with heart failure demonstrated an augmented sympatheti response to exerise, as measured by plasma ateholamines, when ompared with a normal ontrol group. This relative augmentation of sympatheti drive ourred despite a redued heart rate-blood pressure response, suggesting a lak of responsiveness of the failing myoardium to ateholamine 0735-1097/851$3.30
lacc Vol. 5. No.4 FRANCIS ET AL. EXERCISE ANDHEART FAILURE 833 stimulation. In that study neither the patients with heart failure nor the ontrol patients were exerised maximally. Using maximal exerise, as defined by a failure of oxygen onsumption to inrease further during ontinued exerise, we have also noted that patients with heart failure have a relative augmentation of sympatheti drive early during submaximal exerise when ompared with normal ontrol subjets. They do not ahieve the extremely high levels of plasma norepinephrine ahieved by normal subjets at maximal exerise, perhaps beause they do not exerise as long (13). In the previous report from our laboratory (13), data were represented as a funtion of absolute work intensity as measured by total body oxygen uptake rather than as a funtion of relative work intensity. However, omparing the plasma norepinephrine in study patients with that of normal subjets when they are exerising at the same absolute VOz is somewhat misleading beause all subjets (patients and ontrols) are at varying levels of work intensity. It is therefore possible that the observed " augmented" sympatheti response to exerise in heart failure ompared with normal subjets is an artifat related to the fat that patients with heart failure are very near their maximal exerise tolerane while exerising at a low level of absolute oxygen onsumption. Patients with heart failure and normal subjets are examined in an unomparable physiologi frame of referene. If the oxygen onsumption data are expressed as perent of peak VOz ahieved rather than as absolute VOz, the omparison between patients with heart failure and normal subjets may be more relative. In this study we report on an expanded series and representthe ardiovasularresponses to exerise in two ways: I) as a funtion of absolute work intensity measured as total body oxygen uptake, and 2) as a funtion of relative work intensity, expressed as perent individual maximal or peak oxygen uptake. The latter analysis allows a omparison of patients with ongestive heart failure with normal subjets when all individuals are working at omparable intensities of exerise, and highlights the importane of making suh omparisons at relative work intensities. Methods Patients. Thirty-one male outpatients 32 to 71 years of age (mean 56) with hroni. stable New York Heart Assoiation linial lass II to III heart failure were seleted for study. The range of ardia disease inluded idiopathi or aloholi ardiomyopathy in 18 patients and oronary artery disease in 13 patients. The presene of oronary disease was established by oronary arteriography or a previous history of doumented myoardial infartion, No patient had sustained a myoardial infartion in the 4 months preeding the study. Two patients had heart failure after aorti or mitral valve replaement, respetively, at least 5 years before entering the study. All patients had been in linially stable ondition for at least 4 months before exerise evaluation. One patient had hroni atrial fibrillation, one had a hroni juntional rhythm; all others had sinus rhythm. All but one patient had a standard hest X-ray film obtained before exerise. and the average ardiothorai ratio for the group was 0.55 :±: 0.06 (:±: SO). Eah patient had a radionulide left ventriulogram performed at rest before exerise, and the average ejetion fration was 24 :±: 10%. Twenty-four of the 31 patients had a tehnially satisfatory M-mode ehoardiogram before exerise to assess left ventriular internal diastoli dimension. and the average value for the group was 69 :±: 10 mm. Patients were exluded who had exerise-indued angina petoris, alf laudiation or exerise-indued arrhythmias. Those with intrinsi pulmonary disease that might limit exerise were exluded by history. physial examination and routine spirometry. All patients exept two were reeiving long-term digoxin therapy at the time of exerise testing. and all but five were reeiving long-term diureti therapy. None was reeiving vasodilator therapy at the time of the study. A ontrol group was omposedof10 healthy men without ardiovasular disease or major medial illness (mean age 38 years, range 15 to 54). The normal subjets did not routinely engage in a regular exerise program. All patients and subjets gave written informed onsent whih was approved by the Institutional Review Board of the hospital. Method of exerise testing. A alibrated eletroni biyle ergometer (Collins) was used for the exerise studies. The exerise program onsisted of a starting work load of 25 watts for patients with heart failure and 50 watts for normal subjets. The work load was inreased every 4 minutes by 25 watts until symptomati maximum. All patients stopped beause of dyspnea or fatigue, and no one was limited by hest pain, alf laudiation, arrhythmia or eletroardiographi hanges. Heart rate and rhythm were monitored ontinuously from the eletroardiogram, and blood pressure was measured by standard uff tehnique at rest and during the seond minute of eah 4 minute work load. Heart rate and blood pressure were also measured at symptomati maximum. All patients were studied while sitting upright at least 2 hours after a light meal in an aironditioned laboratory. Metaboli measurements ofthe expiredfrations ofoz, COz and the minute volume were obtainedat rest and during eah work load. Expired air was olleted through a Hans Rudolph nonrebreathing valve into a mixing hamber and sampled ontinuously on line by fast responding oxygen and arbon dioxide gas analyzers (Bekman Metaboli Measurement Cart). Calibration ofthe gas analyzers with known preise gas onentrations and alibration of the volume transduer with a preisely alibrated I liter volume pump were performed just before and after exerise. Peak oxygen onsumption (peak VOz) was taken to represent the value
834 FRANCIS ET AL. EXERCISE AND HEART FAILURE lacc Vol. 5. No.4 April 1985:832-9 during maximal exhaustion, just before stopping. This value usually orresponded to the value at whih ontinued exerise failed to bring about a further inrease in \'0 2, The perent peak \'0 2 was alulated retrospetively by dividing the absolute \'0 2 at any point in time by the peak \'0 2 ahieved and multiplying by 100. Plasma norepinephrine. Eah patient and normal subjet had a small atheter seured in an anteubital vein the morning of the exerise test for purposes of drawing blood. The atheter was plaed in the arm opposite to where blood pressure measurements were made. Blood was drawn after a period of at least 15 minutes of upright sitting on the biyle during the ontrol basal state, and during the third minute of eah exerise work load and during symptomati maximum just before stopping the test. In 6 of the 3I patients, a sample ould not be obtained beause of tehnial diffiulties at some time during the exerise test. Nevertheless, eah patient and normal subjet had at least three data points, whih allowed us to haraterize their individual urves during exerise. The blood samples were immediately spun in a refrigerated entrifuge and analyzed later using a radioenzymati assay (CAT-A-KIT, Upjohn Diagnostis). Dupliate measurements of plasma norepinephrine by this method in our laboratory have a oeffiient of variation of less than 4%. Statistial analysis. The intent of the analysis was to quantitatively desribe the differene between the group with heart failure and the ontrol group with respet to hanges in heart rate, systoli blood pressure and plasma norepinephrine during upright dynami exerise. Beause of the unusual heterogeneity of the data in the heart failure group, the patients were further lassified into those with only modest exerise intolerane (able to ahieve a work load of at least 100 watts) and those with severe exerise intolerane (able to ahieve a work load of 50 or 75 watts). This distintion, although arbitrary, orresponds roughly with New York Heart Assoiation linial lasses II and III. A total ofthree groups were analyzed as follows: group I = normal ontrol subjets (n = 10); group 2 = patients with heart failure able to exerise to a work load of at least 100 watts (n = II); and group 3 = patients with heart failure able to exerise to only 50 or 75 watts (n = 20). Beause of the omplexity of relating multiple variables in two groups over time, several two and three variable models were fitted to the log of eah individual's data points. Heart rate, systoli blood pressure and plasma norepinephrine were used as the dependent variables, and absolute \'0 2 and perent peak \'0 2 ahieved were used as the independent variables. For eah subjet the relation between \10 2 or perent peak \10 2 and eah of the three variables (heart rate, systoli blood pressure and plasma norepinephrine) was summarized by two regression oeffiients. Multivariate analysis (MANOYA, SPSS version 8.3, Health Siene Computer System, University of Minnesota, with Hotelling's t test as riterion) of the pair of oeffiients was performed, whih allowed omparison of the three groups. The group response was estimated by the averages of the oeffiients. Subsequent analysis of variane (ANOY A) and t tests of eah of the oeffiients were performed to indiate how the groups differed, if at all. The multipliity of omparisons was taken into aount by means of Bonferroni 's inequality method, and the onlusion s were not affeted for a signifiane level of 0.05. Results Hemodynami response to exerise. When heart rate response to exerise is plotted in relation to absolute \'0 2 (Fig. I), patients with heart failure demonstrate a signifiantly greater (p = 0.01) heart rate response for any given omparable oxygen onsumption. That is, for any given V0 2, patients with either mild or severe heart failure have a higher than normal heart rate. There is no differene in heart rate response between the groups with mild and severe heart failure. However, when the oxygen onsumption data are expressed as perent peak \'0 2 so that all groups are at equivalent intensities of exerise, this relation reverses and is onsistent with the well known onept that patients with heart disease display a form of hronotropi inompetene (Fig. 2). There is a signifiant (p = 0.(01) differene be- Figure 1. The heart rate response to dynami upright exerise is plotted as a funtion of absolute work intensity, expressed by total body oxygen onsumption (\'0 2 ), The two groups of patients with heart failure (mild [group 2] and severe [group 3]) are ompared with a normal ontrol group of subjets and show a greater (p = 0.01) inremental inrease in heart rate during exerise ompared with normal subjets (group I). Group I versus group 2 (slope), p = 0.01; group I versus group 3 (slope), p = 0.01 ; group 2 versus group 3 (slope), p = NS. _ 180 IJ) '0 4>.0 --- 2 120... e CIl I 160 140 5 7 9 II 13 15 17 19 V 02 (mllkg/min) 21 23 25 27 29
JACC Vol. 5. No.4 FRANCIS ET AL. EXERCISE AND HEART FAILURE 835 I'" E t80 l!! 160 :> <n <n yoatbx GroupI G'P2 l!! a. 140 GrOUP3 "0 00 iii 120.!.! <n -, 100 CIl 20 40 60 80 100 "10 PeakV 02 Figure 2. The heart rate response to dynami upright exerise is plottedas a funtionof relativeworkintensity, expressedas perent of individual peak oxygen uptake (% peak V0 2 ). The statistiaily signifiant differenein slopesbetweenthe ongestiveheart failure groups (groups 2 and 3) and the normal group (group I) indiate that patients with heart failure are unable to raise their heart rate in response to upright dynami exerise to the same extent as normal ontrol subjets when funtioning at relative work intensities. Group I versus group 2 (slope), p = 0.001; group 1 versus group 3 (slope), p < O.00I; group 2 versus group 3 (slope), P = NS. tween patients with heart failure and normal ontrol subjets, indiating that those with heart failure have a smaller inremental inrease in heart rate in response to dynami upright exerise. In Figure 3 the systoli blood pressure response to exerise is plotted in relation to absolute individual V0 2. For any given individual V0 2, there is no signifiant differene between the three groups with regard to systoli blood pressure during exerise. However, when the three groups are ompared at equivalent levels of exerise (that is, perent peak V0 2 ahieved), there is a lear and signifiant separation between the groups, with the patients having heart failure showing a redued blood pressure response to ex- Figure 3. The systoliblood pressureresponseto dynamiupright exerise is plotted as a funtion of absolute work intensity, expressed by total body oxygen onsumption (V0 2 ). Although patients with mild (group 2) and more severe (group 3) heart failure ahieve a muh lower V0 2 than do normal ontrol subjets (group I) during exerise, the slopes of the urves are not different, indiating that there is no differene in the inremental hange in systoli blood pressure between groups during upright biyle exerise when the response is represented as a funtion of absolute work intensity. 200 180 160 140 y=a+bx Figure 4. The systoli blood pressure responseto dynami upright exerise is plotted as a funtion of relative work intensity, expressedas perentof individualpeakoxygen uptake (% peakv0 2 ). AIl slopes are signifiantly different, indiating that at relative work intensities, there is less inrease in systoli blood pressure during exerise in those patients with more severe heart failure. Group 1 versus group 2 (slope), p < O.OO!; group 1 versus group 3 (slope), p < 0.001; group 2 versus group 3 (slope), p = 0.03. erise (p < 0.001) (Fig. 4). Those patients with more severe heart failure have a smaller inrease in blood pressure during exerise (p = 0.03). Plasma norepinephrine response to exerise. The response of plasma norepinephrine to dynami upright exerise is shown in Figures 5 and 6. At any given absolute V0 2, patients with heart failure have a greater level of plasma norepinephrine. As with heart rate and blood pressure data, however, it may be inappropriate to onsider all three groups to be at equivalent levels of intensity of exerise even though they are at equivalent absolute values for V0 2, Figure 5. The plasmanorepinephrineresponseto dynami upright exerise is plotted as a funtion of absolute work intensity, expressed by total body oxygen onsumption (V0 2 ). AIl slopes are signifiantly different (p = 0.004), indiating a higher plasma norepinephrine leve! during exerise in patients with more severe heart failure when the response is represented as a funtion of absolute work intensity. 2400 2200 Y=eAtBx+x2 2000 e- o. 1800 GROUP 3 (l) : ';: s: 0. Q) '0. Q) 5 z EV> o a::: 10 15 20 V 0 2 (ml/kg/min) 25 30 7 to t5 20 25 30 \1 02 (mllkg/minl
836 FRANCIS ET AL. EXERCISE AND HEART FAilURE JACC Vol. 5. No, 4 2300 2100 1900 Cl 0-1700 Q) ';: 1500 s: 0- Q) t300 '0. f:! 0 Z 1100 0 E 900 0"' 0:: 700 y:ea+blt+cil.2 300 l:...:::::...l...j...l-.l...>...l...j...j...j 20 40 60 80 too % Peak V 0 2 Figure 6. The plasmanorepinephrine response to dynami upright exerise is plotted as a funtion of relative work intensity, expressed by perent peak oxygen uptake (% peak V0 2 ). There is no differene in slope between group 2 (mild heart failure) and group3 (severeheartfailure). Group I (normal) has a steeperslope than that in groups 2 and 3 (p = 0.(02), indiating that at relative workintensities, normal subjets haveaugmented sympatheti drive. sine patients with severe heart failure (group 3) are at near maximum at a V0 2 of 10 mllkg per min, while patients with milder heart failure (group 2) are near maximum at a V0 2 of 17 mllkg per min and normal ontrol subjets (group l) are only at about 50% of peak V0 2 when exerising at a V0 2 of 17 mllkg per min. To irumvent this problem, the data are again expressed as perent peak V0 2 ahieved (Fig. 6), whih indiates distintly different shaped urves for the patients with heart failure and the normal ontrol subjets (p == 0.(02). Patients with mild or more severe forms of heart failure begin with higher rest levels of plasma norepinephrine and respond to upright dynami exerise with a more gradual inrease at relative work intensities. In ontrast, normal subjets have a somewhat flatter response during submaximal exerise but show an impressive exponential rise in plasma norepinephrine when they reah roughly 50% of their peak V0 2 and are able to generate a muh higher plasma norepinephrine at peak V0 2 ompared with patients with heart failure. The maximal responses of heart rate and systoli blood pressure to dynami upright biyle exerise are shown in Figure 7. The data indiate that maximal ahievable heart rate and systoli blood pressure responses are blunted as the severity of heart failure inreases. Figure 8 indiates that the peak V0 2 ahieved varies, as expeted, among the three groups. Although maximal plasma norepinephrine ahieved at peak V0 2 is greater in normal subjets than in patients with heart failure, the measurement fails to separate mild from severe heart failure (Fig. 8). Disussion The study of various hemodynami and metaboli responses to exerise in patients with heart failure is important beause exerise limitation is what ultimately disables the patient. The mehanisms responsible for limiting the aerobi exerise apaity of patients with heart failure are unknown but are possibly related to inadequate blood flow to working skeletal musle (14,15) or inreased pulmonary apillary pressure (16). A better understanding of the biologi events that our during exerise is needed beause it has beome inreasingly lear that simple measurements of left ventriular performane at rest do not usually relate to funtional apaity in patients with heart failure (17-20). The funtional lassifiation of suh patients is best assessed by exerise testing (21). New pharmaologi therapy for heart failure is ommonly assessed by measuring exerise tolerane (22-26). Knowledge of the sympatheti nervous system response during exerise may allow for a learer understanding of its potential role in either enhaning or possibly inhibiting exerise tolerane. -- 20 0 200 I isem Cl J: 160 E 160 E Q) "' Figure 7. The maximal heartrate and systoli blood 0 120 :; 120 III.D pressureahievable during upright biyleexerise '" III '" appear related to the severity of the heart failure. It 0'" 80 '8 80 Valuesin all three groupsare differentby p < 0.01. a:: 0 (; J: '" 40 iii l:! "0 in 40 >- en I i SEM Group I Group 2 Group :3 Groupl Group 2 Group :3
JACC Vol. 5, No.4 FRANCIS ET AL. EXERCISE ANDHEART FAILURE 837 35 I! SEM 2800 2400 I!SEM 30 - C7' 25 <E N.;? 20 0Ql a.. 15 10 I GroupI Group 2 Group 3 E 2000 a. - Figure 8. The individual peak \10 2 obtained Ql.s.: a. Ql 'Q. Ql oz o E III o a:: 1600 1200 800 40 0 ::t. Groupl Group2 Group 3 during dynami upright exerise is different (p < 0.01) amongall three groups. The peak plasma norepinephrine ahieved during exerise is different between the normal and the two heart failure groups (p < 0.04), but not between the groups with mild and severe heart failure, Response of plasma norepinephrine to exerise in heart failure. Current data indiate that patients with heart failure have an attenuated sympatheti nervous system response to stimuli suh as orthostasis (3) and aute vasodilation with nitroprusside (7). It is therefore somewhat surprising that the response of the sympatheti nervous system to submaximal exerise has been reported to be relatively augmented rather than attenuated (12,13). Suh augmentation of sympatheti nervous system ativity during submaximal exerise would seem to play a useful role by inreasing heart rate and myoardial ontratility. However, the heart rate, blood pressure and stroke volume responses to exerise are usually redued in heart failure. In fat, patients with heart failure are unable to ahieve very high irulating levels of norepinephrine during maximal exerise (13). Blood levels of at least 1,500 to 2,000 pg/ml (or nearly 10 times the basal level) of norepinephrine are neessary to ahieve a measurable hemodynami effet when norepinephrine is infused by vein (27). Patients with heart failure typially ahieve maximal plasma norepinephrine levelsjust above this threshold (Fig. 8) and, in some individual ases, below it during dynami exerise. Comparison of ontrol subjets and patients in similar physiologi ontext. The present study has attempted to examine the sympatheti nervous system response to exerise over the full spetrum of exerise from rest to exhaustion. The response of the plasma norepinephrine relative to absolute oxygen onsumption, as in previous reports (12,13), has been reexamined and onfirms the apparent augmentation of the sympatheti nervous system response to submaximal exerise in patients with heart failure (Fig. 5). Although patients with heart failure, when ompared with healthy ontrol subjets, do have a greater plasma norepinephrine at any given absolute oxygen onsumption during submaximal exerise, it is lear that they annot do as muh exerise and are therefore loser to peak \'0 2 or maximal tolerable exerise at a relatively low oxygen onsumption. To analyze the data in a omparable physiologi frame of referene, the response of plasma norepinephrine has been plotted as a funtion of perent peak \'0 2 ahieved rather than absolute \'0 2 (Fig. 6) (28). Examination of the relation between plasma norepinephrine and perent peak \'0 2 ahieved indiates that the two groups with heart failure (moderate and advaned) have similar slopes whih are less steep over the duration of exerise ompared with those of normal ontrol subjets. Therefore, in patients with heart failure plasma norepinephrine an be augmented during exerise, but the augmentation is somewhat attenuated ompared with that in normal subjets at similar work intensities. This interpretation is opposite to the report of Chidsey et al. (12) and is also ontrary to our own previously published work (13). Possible importane of redued sympatheti drive during exerise in heart failure. The potential biologi importane of the relatively attenuated sympatheti nervous system response to exerise in patients with heart failure deserves omment. The fators that limit exerise tolerane in patients are not entirely lear, but may be dependent to some extent on the ability to inrease ardia output (15), dilate the peripheral vasulature in response to metaboli demands (29) and tolerate high left ventriular filling pressures (20). The hanges in heart rate, stroke volume and
838 FRANCIS ET AL. EXERCISE ANDHEART FAILURE lacc Vol. 5, No.4 arteriovenous oxygen differenes are the most important preditors of maximal oxygen onsumption (30). Patients who are most severely limited by heart failure an inrease ardia output only by inreasing heart rate beause they are unable to improve stroke volume with exertion (14). Very high filling pressures may be well tolerated during exerise (20), and it is likely that the ability to inrease ardia output and speifially to inrease heart rate is of major importane in determining exerise tolerane in the syndrome of heart failure. Left ventriular funtion at rest is only a weak determinant of exerise apaity (31). Heart rate response is learly attenuated during exerise in patients with heart failure when plotted as funtion of relative work intensity (Fig. 2). It is possible, although not proved, that the attenuated sympatheti nervous system response is partly responsible for the blunted heart rate response. The inability to maximize the sympatheti response during exerise and at peak \.'0 2 ould aount for the relative hronotropi inompetene observed in these patients. Similarly, the blunted systoli blood pressure response to exerise might also be related to an attenuated sympatheti response whih ould ontribute to a redued stroke volume. It is therefore possible that an attenuated inrease in sympatheti nervous system ativity during exerise is in part responsible for the exerise intolerane of heart failure. This attenuated sympathetidrive during exerisein patients with heart failure oupled with redued density of beta-adrenergi reeptors in the failing myoardium (32) may at together to redue the hronotropi and inotropi response to exerise, leading eventually to exerise intolerane. This hypothesis is onsistent with experimental data demonstrating a dereased myoardial responsiveness to isoproterenol (33) and a redution of the ardia response to postganglioni sympatheti nerve stimulation (34) in dogs with heart failure. Impliations. The mehanism of the relatively attenuated sympatheti response to exerise in heart failure is unknown. However, the attenuation is qualitatively similar to that observed after orthostasis and nitroprusside infusion in patients with heart failure, although different mehanisms may be involved. It is possible in some ases to restore sympatheti responsiveness with pharmaologi therapy (35) and with more definitive therapy suh as heart transplantation (36). Sympatheti responsiveness to nitroprusside infusion, as with orthostasis and exerise, appears to range from a near normal response in patients with mild heart failure to a virtual absene of responsiveness in patients with severe heart failure (7). Patients who respond to nitroprusside with near normal ativation of the sympatheti nervous system appear to have a more favorable long-term prognosis (7). It is therefore likely that the attenuated responses of plasma norepinephrine to orthostasis, nitroprusside and exerise reflet to some extent the overall severity of ongestive heart failure. In the ase of exerise, the blunted sympatheti response may possibly be ontributing in the exerise intolerane. We are grateful to Ada Simon, PhD for performing the plasma norepinephrine assays. The statistial help of Dorothy Aeppli, PhD, Department of Biometris, University of Minnesota and the tehnial support of Therese Noel, RN are greatly appreiated. The seretarial assistane of Sandy Thiesse is gratefully aknowledged. Referenes I. Abelmann WH, Fareeduddin K. Cirulatory response to upright tilt in patients with heart disease. Aerospae Med 1967;38:60-5. 2. Abelmann WH, Fareeduddin K. Inreased tolerane of orthostati stress in patients with heart disease. Am J Cardiol 1969;23:354-63. 3. Levine TB, Franis GS, Goldsmith SR, Cohn IN. The neurohumoral and hemodynami response to orthostati tilt in patients with ongestive heart failure. Cirulation 1983;67:1070-5. 4. Kubo SH, Cody RJ. Cirulatory autoregulation in hroni ongestive heart failure: responses to head-up tilt in 4 1 patients. Am J Cardiol 1983;52:512-8. 5. Goldsmith SR, Franis GS, Levine TB, Cohn IN. Regional blood flow response to orthostasis in patients with ongestive heart failure. J Am Coli Cardiol 1983;1:1391-5. 6. Brigden W, Sharpey-Shafer EP. Postural hanges in peripheral blood flow in ases with left heart failure. Clin Si 1950;9:93-100. 7. Olivari MT, Levine TB, Cohn IN. Abnormal neurohumoral response to nitroprusside infusion in ongestive heart failure. J Am Coli Cardiol 1983;2:411-7. 8. Brue RA, Kusumi F, Hosmer D. Maximal oxygen intake and normographi assessment of funtional aerobi impairment in ardiovasular disease. Am Heart J 1973;85:546-62. 9. Ross J Jr, Gault JH, Mason DT, Linhart JW, Braunwald E. Left ventriular performane during musular exerise in patients with and without ardia dysfuntion. Cirulation 1966;34:597-608. 10. Zelis R, Mason DT, Braunwald E. Partition of blood flow to the utaneous and musular beds of the forearm at rest and during leg exerise in normal subjets and in patients with heart failure. Cir Res 1969;24:799-806. II. Patterson JA, Naughton J, Pietras RJ, Gunnar RM. Treadmill exerise in assessment of the funtional apaity of patients with ardia disease. Am J Cardiol 1972;30:757-62. 12. Chidsey CA, Harrison DC, Braunwald E. Augmentation of the plasma norepinephrine response to exerise in patients with ongestive heart failure. N Engl J Med 1962;267:650-4. 13. Franis GS, Goldsmith SR, Zieshe SM, Cohn IN. Response of plasma norepinephrine and epinephrine to dynami exerise in patients with ongestive heart failure. Am J Cardiol 1982;49: 1152-6. 14. Weber KT, Kinasewitz GT, Janiki JS, Fishman AP. Oxygen utilization and ventilation during exerise in patients with hroni ardia failure. Cirulation 1982;65:1213-23. 15. Wilson JR, Ferraro N. Exerise intolerane in patients with hroni left heart failure: relation to oxygen transport and ventilatory abnormalities. Am J Cardio1 1983;51:1358-63. 16. Christie RV, Meakins JC. The intrapleural pressure in ongestive heart failure and its linial signifiane. J Clin Invest 1934;13:323-45. J7. Franis GS, Goldsmith SR, Cohn IN. Relationship of exerise apaity to resting left ventriular performane and basal plasma norepinephrine levels in patients with ongestive heart failure. Am Heart J 1982;104:725-31. 18. Franiosa JA, Park M, Levine TB. Lak of orrelation between ex-
lacc Vol. 5, No.4 FRANCIS ET AL. EXERCISE AND HEARTFAILURE 839 erise apaity and indexes of resting left ventriular performane in heart failure. Am J Cardiol 1981;47:33-9. 19. Benge W, Lithfield RL, Marus M. Exerise apaity in patients with severe left ventriular dysfuntion. Cirulation 1980;61:955-9. 20. Lithfield RL, Kerber RE, Benge JW, et al. Normal exerise apaity in patients with severe left ventriular dysfuntion: ompensatory mehanisms. Cirulation 1982;66:129-34. 21. Franiosa JA, Zieshe S, Wilen M. Funtional apaity of patients with hroni left ventriular failure: relationship of biyle exerise performane to linial and hemodynami haraterization. Am J Med 1979;67:460-6. 22. Franiosa JA, Cohn IN. Immediate effets of hydralazine-isosorbide dinitrate ombination on exerise apaity and exerise hemodynamis in patients with left ventriular failure. Cirulation 1979;59:1085-91. 23. Franiosa JA, Cohn IN. Effet of isosorbide dinitrate on response to submaximal and maximal exerise in patients with ongestive heart failure. Am J Cardiol 1979;43:1009-14. 24. Franiosa JA, Goldsmith SR, Cohn IN. Contrasting immediate and long-term effets of isosorbide dinitrate on exerise apaity in ongestive heart failure. Am J Med 1980;69:559-66. 25. Rubin SA, Chatterjee KJ, Parmley WW. Metaboli assessment of exerise in hroni heart failure patients treated with short-term vasodilators. Cirulation 1980;61:543-8. 26. Clausen JP. Cirulatory adjustments to dynami exerise and effet of physial training in normal subjets and in patients with oronary artery disease. Prog Cardiovas Dis 1976;18:459-95. 27. Silverberg AB, Shah SD, Haymond MW, Cryer PE. Norepinephrine hormone and neurotransmitterin man. Am J Physiol I978;234:E252-6. 28. Lewis SF, Taylor WF, Graham RM, Pettinger WA, Shutte JE, Blomqvist CG. Cardiovasular responses to exerise as funtions of absolute and relative workload. J Appl Physiol 1983;54:1314-23. 29. Mason DT, Zelis R, Longhurst J, Lee G. Cardio-irulatory responses to musular exerise in ongestive heart failure. Prog Cardiovas Dis 1977;19:475-89. 30. Higginbotham MB, Morris KG, Conn EH, Coleman RE, Cobb FR. Determinants of variable exerise performane among patients with severe left ventriular dysfuntion. Am J Cardiol 1983;51:52-60. 31. MKiman MD, Sullivan M, Jensen D, Froeliher VF. Treadmill performane and ardia funtion in seleted patients with oronary heart disease. J Am Coil Cardiol 1984;3:253-61. 32. Bristow MR, Ginsburg R, Minobe W, et al. Dereased ateholamine sensitivity and f3-adrenergi-reeptor density in failing human hearts. N Engl J Med 1982;307:205-11. 33. Newman WHo A depressed response on left ventriular ontratile fore to isoproterenol and norepinephrine in dogs with ongestive heart failure. Am Heart J 1977;93:216-21. 34. Covell JW, Chidsey CA, Braunwald E. Redution of the ardia response to postganglioni sympatheti nerve stimulation in experimental heart failure. Cir Res 1966;19:51-6. 35. Cody RJ, Franklin KW, Kluger J, Laragh IH. Mehanisms governing the postural response and baroreeptor abnormalities in hroni ongestive heart failure: effets of aute and hroni long-term onverting enzyme inhibition. Cirulation 1982;66:135-42. 36. Levine TB, Olivari MT, Carlyle P, Franis GS, Cohn IN. Reversibility by heart transplantation of abnormal neurohumoral ontrol mehanisms in hroni ongestive heart failure. Cirulation 1982;66(suppl 11):11-134