patients with heart disease in patients with and without beta-adrenergic receptor blockersin Inﬂuence of stroke volume and exercise tolerance on peak oxygenpulse Journal of Cardiology

(1)

Original article

Inﬂuence of stroke volume and exercise tolerance on peak oxygen pulse in patients with and without beta-adrenergic receptor blockers in patients with heart disease

Makoto Murata (MD)

^a

, Hitoshi Adachi (MD)

^b,

*, Shigeru Oshima (MD, FJCC)

^b

, Masahiko Kurabayashi (MD, FJCC)

^a

aDepartmentofMedicineandBiologicalScience,GraduateSchoolofMedicine,GunmaUniversitySchoolofMedicine,Maebashi,Gunma,Japan

bGunmaPrefecturalCardiovascularCenter,DepartmentofCardiology,Maebashi,Gunma,Japan

Introduction

Thequotientofoxygenuptake(V˙O2)andheartrateHRiscalled theO2pulse(V˙O2=HR).Itisthevolumeofoxygentakenupbythe pulmonarybloodduringtheperiodofaheartbeatanddependson thevolume ofoxygen extracted by theperipheral tissues. This measurementisaproductofstrokevolume(SV)andthearterial- mixedvenousbloodO₂difference[C(a-v)O₂].

In normal subjects and in patients with heart failure, the maximumC(a-v)O2reachesanalmostconstantvalueof13–14mL/

dL[1,2].Therefore,theO2pulseatpeakexercisecanbeexpressed asfollows:peakV˙O2=HR¼peakSVk,wherekisa constantas mentionedabove,and thepeakO₂ pulsecanberegardedasan indicatorofcardiacpumpfunction.Inclinicalsettings,theO2pulse canbeusedtodeterminecardiacoutputduringexercise[3]andto detecttheonsetofmyocardialischemia[4].

Beta-adrenergicreceptorblockersarerecommendedforvari- oustypesofheartdiseases,suchasischemicheartdisease,heart failure,andhypertension[5–7].Bysuppressingsympatheticnerve activity,theydiminishheartrateatrestaswellasduringexercise [8]. That is, the O2 pulse may be higher than expected, and estimatingcardiacfunctionduringexerciseusingthisvaluewith beta-adrenergic receptor blocker usage may be misleading.

ARTICLE INFO

Articlehistory:

Received24November2015

Receivedinrevisedform10February2016 Accepted13February2016

Availableonlinexxx

Keywords:

Oxygenpulse Strokevolume

Beta-adrenergicreceptorblocker Cardiopulmonaryexercisetest

ABSTRACT

Background: Inagivenindividual,aconsistentrelationshipexistsbetweenoxygenuptake(V˙O2)and heartrate(HR)duringexercise.ThequotientofV˙O₂andHR(V˙O₂=HR)iscalledtheoxygenpulse(O2

pulse),anditsvalueisdependentonstrokevolume(SV).However,itisdifﬁculttobelievethattheO2

pulsewouldindicatetheSVwhenHRhasbeenmodiﬁedaswiththeuseofbeta-adrenergicreceptor blockers(BB).Untilnow,theeffectofBBonpeakO2pulsehasnotbeenpreciselystudied.Wetriedto clarifytheeffectofBBontherelationshipbetweenO2pulseandSV.

Methods:Of699consecutiveheartdiseasesubjectswhoperformedcardiopulmonaryexercisetests (CPX)from2012to2014,weenrolled430subjectswhohadsinusrhythmandcouldperformCPXuntil exhaustion.Onehundredandﬁfty-sevensubjectsweretakingBB.SVwasevaluatedduringCPXusing impedancecardiography,andwecomparedthepeakO2pulsewithpeakSVbetweenpatientswithoutBB (GroupA)andwithBB(GroupB).

Results:TheHRsatrestandpeakexerciseinGroupAweregreaterthanthoseinGroupB(74.413.0/

min vs. 71.811.3/min, p<0.01, 134.921.7/min vs. 124.923.6/min, p<0.01, respectively). The regressionlineofthepeakO2pulseagainstthepeakSVwassteeperinGroupBthaninGroupA.Whenwe dividedthepatientsintotwogroupsaccordingtotheaveragevaluesofthepeakSVandpeakV˙O2,O2pulse/

SVratioinGroupBabovetheaveragewasgreaterthanthatinGroupA,whereasitwassimilarinthetwo groupsthatwerebelowaverage.

Conclusion:WefoundthattheincreaseintheO2pulsewasdisproportionatelygreaterthantheSVthat wasmeasuredbyimpedancecardiographywhenaBBwasusedinpatientswithpreservedSVand exercisetolerance.

ß2016PublishedbyElsevierLtdonbehalfofJapaneseCollegeofCardiology.

* Correspondingauthor at: Gunma Prefectural CardiovascularCenter, 3-12, Kameizumi,Maebashi,Gunma3710004,Japan.Tel.:+81272697455;

fax:+81272691492.

E-mailaddress:[email protected](H.Adachi).

ContentslistsavailableatScienceDirect

Journal of Cardiology

j our na l ho me pa g e : w ww . e l se v i e r . com / l oca t e / j j cc

http://dx.doi.org/10.1016/j.jjcc.2016.02.017

0914-5087/ß2016PublishedbyElsevierLtdonbehalfofJapaneseCollegeofCardiology.

(2)

However,theinﬂuenceofbeta-adrenergicreceptorblockersonthe peakO2pulsehasnotbeenpreciselystudied.Theinﬂuenceofbeta- adrenergicreceptorblockerson peakO2 pulsecanbeestimated usinganequationof O2pulse dividedbythemeasured SV(O2

pulse/SV ratio). If this parameter is greater than expected in patients taking beta-adrenergic receptor blockers, this ﬁnding wouldbeattributedtoadecreasedheartrate.Thus,weplannedto quantifytheeffectofbeta-adrenergicreceptorblockersontheO2

pulseusingtheO₂pulse/SVratio.

Method Subjects

Weperformed699cardiopulmonaryexercise(CPX)testsfrom thelatterhalfof2012toearly2014.Oftheseconsecutive696CPX tests,430patientswhohadsinusrhythmandcouldperformCPX untilexhaustionwereenrolled.Patientswithresidualmyocardial ischemiawere excluded. Onehundred and ﬁfty-seven subjects weretakingbeta-adrenergicreceptorblockers.Patientstakingan insufﬁcient dose (carvedilol<5mg/day or bisoprolol<2.5mg/

day)ofbeta-adrenergicreceptorblockers,andpatientswithlung emphysemaormoderate-to-severeanemia(Hb<10mg/dL)were excluded. Patients, in whom the effect of the beta-adrenergic receptorblockerwasnotsufficient,thatis,theheartrateatrestdid notdecrease5beats/min,werealsoexcluded.Patientswhodid nottakebeta-adrenergicreceptorblockerswereassignedtoGroup A,andpatientswhotookbeta-adrenergicreceptorblockerswere assignedtoGroupB.Patients’profilesareshowninTables1and 2.Totestwhethertheinfluenceofthebeta-adrenergicreceptor blockervariedbasedonexercisetoleranceandpeakSVatpeak exercise,eachgroupwasdividedintotwogroupsaccordingtothe average values of the %peak V˙O2, peak SV, or left ventricular ejectionfraction(LVEF).

Cardiopulmonaryexercisetest

The anaerobic threshold (AT) and peak V˙O2 wereevaluated usingsymptom-limitedCPXtestingonanupright,calibratedcycle ergometer(StrengthErgo8,MitsubishiElectricEngineering,Tokyo,

Japan)withanelectrocardiograph(ML-9000,FukudaDenshiLtd., Tokyo,Japan).CPXwasperformed2–4hafteralightmeal.Thistest beganwith3minofrestand3minofwarm-upat0Wfollowedby continuous increase of the work rate by 1W every 6s until exhaustion, as recommended by Buchfuhrer et al. [9] and as reportedbyus[10].TocertifythatpatientsperformedCPXwith enough vigor, they were forced to keep pedaling until the respiratory quotient(R) reached>1.10.The workrateincrease levelswerechosenonthebasisoftheﬁtnessofthesubjectstokeep theexerciseperiodbetween8and15min[9].V˙O2,carbon-dioxide production(V˙CO2),andminuteventilation(V˙E)weremeasuredon a breath-by-breath basis using a gas analyzer (MINATO 300S, Minato Science Co. Ltd., Osaka, Japan). The peak V˙O2 was determinedasthehighestV˙O2achievedduringexercise.ATwas measuredbytheV-slopemethod[11].

Impedancecardiography

TheSVwasevaluatedduringCPXusingimpedancecardiogra- phy(PhysioFlowLab-1,ManatecBiomedical,Paris,France).The PhysioFlow device is a range of non-invasive hemodynamic monitors.Ithasbeenreported toprovidecontinuous,accurate, reproducible,andsensitivemeasurementsforcardiacoutputand other parameters [12,13]. It has shown non-inferiority to the predicatedevicethermodilutionSwan-Ganzcatheter[14,15]and superiority to a standard impedance cardiography[16]. Before startingtheexerciseprotocol,thepatientswereattachedtothe impedancecardiographelectrodesofimpedance cardiographas previously described [17,18]. In brief, a constant sinusoidal alternating current (1.8mA, 75kHz) was applied between the couplesofelectrodesplacedonthesupraclavicularfossaattheleft baseoftheneckandalongthexiphoid.Theassociatedvoltagewas detectedbytwoinnerelectrodepairspositioned5cmapartfrom thecorrespondingcouplesofelectrodesthatwereparalleltothe currentpath.Thisvoltagewastransmittedtoanampliﬁerandan impedancesignal(z)wasproduced.TheSVwascalculatedusing the following formula by Sramek–Bernstein: SV=volume of electricallyparticipatingintrathoracictissueventricularejection timeindexofcontractility,whichwastheratioofthepeakrate of change in the thoracic bio-impedance (dZ/dtmax) and the thoracicﬂuidindexortotalthoracicimpedance.

Echocardiography

Cardiacfunctionatrestwasevaluatedusingechocardiography withinaweekoftheCPXbyastandardprocedureforrecording Table1

Characteristicsofstudypopulation.

Parameters GroupA GroupB p

N 281 149

Male/female 223/58 125/24 0.25

Age(years) 61.516.4 63.412.9 0.99

BMI 23.03.0 23.53.8 0.37

Underlyingheartdisease(n,(%))

Previousmyocardialinfarction 71(25.3) 48(32.2) 0.13

PreviousPCI 115(40.9) 36(24.2) <0.01

Previousopenheartsurgery 61(21.7) 17(11.4) 0.01

DCM/HHD 8(2.8) 38(25.5) <0.01

Greatvesseldisease 7(2.5) 3(2.0) 0.75

Others 19(6.8) 7(4.7) 0.39

Comorbidities Diabetesmellitus

HbA1c(%) 6.190.80 6.210.67 0.73

Hypertension(n,(%)) 184(65.5) 98(65.8) 0.95 Dyslipidemia(n,(%)) 143(50.9) 94(63.1) 0.02

Hb(mg/dL) 13.51.5 14.01.5 0.02

Echocardiographicﬁndings

EF(%) 62.312.7 46.519.4 <0.01

E/A 1.170.52 1.170.73 0.57

DcT(ms) 216.564.7 206.356.0 0.24

E/E⁰ 8.182.76 10.605.54 <0.01

BMI,bodymassindex;PCI,percutaneouscoronaryintervention;DCM,dilated cardiomyopathy; HHD, hypertensive heart disease; Hb, hemoglobin; EF, ejectionfraction;DcT,decelerationtime.

Table2

Hemodynamicandmetabolicresponsesinthetwogroups.

GroupA GroupB p

N 281 149

RestHR(/min) 74.413.0 71.811.3 <0.01

PeakHR(/min) 134.921.7 124.923.6 <0.01 DHR/DWR 0.630.21 0.620.28 0.54 Peak ˙VO2(mL/min/kg) 19.85.6 17.85.1 <0.01

Peak ˙VO2(%) 80.119.6 72.821.1 <0.01

AT(mL/min/kg) 13.03.3 11.93.1 <0.01

AT(%) 81.720.3 75.619.8 <0.01

˙VEvs. ˙VCO2slope 34.228.6 34.69.7 <0.01 PeakO2pulse(mL/beat) 9.342.34 9.152.75 0.45

Ratpeakexercise 1.200.08 1.200.09 0.38

SVatpeak(mL/beat) 89.217.4 85.822.3 0.08 O2pulse–SVratio

(PeakO2pulse/peakSV100)

10.72.8 10.92.8 0.44

HR,heartrate;WR,workrate;AT,anaerobicthreshold; ˙VO2,oxygenuptake;AT, anaerobicthreshold; ˙VE,minuteventilation; ˙VCO2,carbondioxideproduction;

O2pulse,oxygenpulse;R,gasexchangeratio;SV,strokevolume.

(3)

images and making measurements [19,20]. The ultrasound equipmentthatwasusedwaseitherVivid5or7(GeneralElectric MedicalSystems,Milwaukee,WI,USA).TheLVEFwascalculated usingthemodiﬁedSimpsonmethod. Thediastolicfunctionwas evaluatedusing pulsedDoppler recordingsof themitral inﬂow velocitiesEandAwaves,decelerationtime(DcT),andthetissue Doppler-derived early diastolic mitral annular motion at the septum(E⁰),andtheratioofEandE⁰ (E/E⁰).

Dataanalysis

All data were expressed as meanstandard deviation. The differencebetweentwogroupswasassessedusingStudent’st-test.

Chisquareanalyseswerealsousedwhereapplicable.Thedifferences betweenthe three groups were assessed by one-way analysis of variancewitha Bonferronianalysis asa posthocanalysis. These analyseswereperformedusingSPSSversion18.(SPSSInc.,Chicago,IL, USA).Ap-valueof<0.05wasconsideredsigniﬁcant.Therelationship betweenthepeakO2pulseandthepeakSVwascalculatedbyalinear regressionanalysis.ToquantifythediscrepancybetweenthepeakO2

pulseandtheSV atpeakexercise, theratioofthepeak O2 pulse against the peak SV (the ‘O2 pulse–SV ratio’) was calculated as follows:peakO2pulse/peakSV100.Forexample,ifthepeakO2

pulsewas100mL/beatandpeakSVwas10mL,theO2pulse–SVratio was10.0.

ThisstudywasapprovedbytheEthicsCommitteeofGunma PrefecturalCardiovascular Centerand was conducted in accor- dancewiththeDeclarationofHelsinki.

Results

TherewerenosigniﬁcantdifferencesinageandHbA1camong theparticipants;thesetypesofdifferenceshavebeenknown to have an effect on heart rate response. LVEF and E/E⁰ were signiﬁcantlylowerinGroupB(Table1).

Asshown inTable2,heartratesatboth restand peakwere lowerinGroupBalthoughtherewasnodifferenceintheheartrate responsetotheexercisebetweentwogroups.Asfortheexercise tolerance,thepeakV˙O2,AT,andV˙Evs.V˙CO2 relationshipwere lowerinGroupBthaninGroupA.Therewasnodifferenceinthe peakO2 pulsebetween twogroups. TheO2 pulse–SVratio was similarintwogroups.ThepeakRwas>1.15inbothgroups,which indicatedthatthemaximumexercisetestwasachieved.

Therelationships betweenthepeak SVandpeakO2 pulsein GroupsAandBareshowninFig.1.Theregressionlineofthepeak O2pulseagainstthepeakSVinGroupBwassteeperthanthatin GroupA.The regressionline of thelowerpartof GroupBwas almostthesameasorsmallerthanthatofGroupA.However,the twolinesgraduallydivergedfromeachother,andthelineofGroup BbecamegreaterthanthatofGroupAasthepeakSVincreased.

Whenwedividedeachgroupintotwocategoriesaccordingto theaveragevalueofthepeakSV(88.0mL),impairedcategoryand preservedcategory,andcomparedtheO2pulse–SVratiobetween GroupsAandBwithineachcategory,theO₂pulse–SVratiosfor GroupsAandBweresimilarintheimpairedcategory.Ontheother hand,orthepreservedcategory,theO2pulse–SVratiowasgreater inGroupBthaninGroupAasshowninFig.2.Thecardiacfunction ofthesesubjectsisshowninTable3.InthepreservedSVgroup, althoughtherewasasigniﬁcantdifferenceinEFbetweenGroupsA and B, both sets of data werewithin normal limits. When we dividedeachgroupaccordingtotheaverage valueofpeakV

˙O2

(19.0mL/min/kg), similar resultswereobtained as is shown in Fig.3.

We also evaluated the effect of beta-adrenergic receptor blockersontheO2pulse–SVratioinpatientswithpreservedLVEF [350%,n=250(GroupA),65(GroupB)].Asshownintheright

panelofFig.4,althoughittendedtobehigherinGroupBcompared withGroupA,therewasnosigniﬁcantdifferencebetweenthetwo groups(p=0.058).

The differences among Group A, the carvedilol group, and bisoprolol group are shown in Table 4. The average doses of carvedilol and bisoprolol were 9.26.3mg and 2.92.8mg, respectively.TherewasnodifferenceinthepeakO₂pulse–SVratio amongthethreegroups.

Becausethereisapossibilitythattheeffectofbeta-adrenergic receptorblocker ontheO₂ pulse–SVratio isgreaterinpatients Fig.2.Differenceof‘O2pulse–SVratio’betweentwogroups.Inthecategoryof preservedstrokevolume,O2pulse–SVratioofGroupBisgreaterthanthatofGroup A.SV,strokevolume.

Table3

CardiacfunctionofGroupsAandBinimpairedorpreservedstrokevolume.

ImpairedSV PreservedSV

GroupA GroupB GroupA GroupB

N 126 65 155 84

EF(%) 63.012.8 43.919.7 62.79.2^** 51.117.4^**

E/A 0.90.4 1.30.9 1.20.6^* 1.00.5

DcT(ms) 231.055.5 193.053.6 208.857.4 224.552.7 E/E⁰ 8.52.4 10.45.6 7.72.6^** 10.95.4^**

EF,ejectionfraction;E,earlydiastolicvelocity;A,peaklatediastolicvelocity;

DcT,decelerationtime;E⁰,earlydiastolicvelocityatthemitralannulus.

* p<0.05vs.matchedimpairedSVgroup.

** p<0.01vs.matchedimpairedSVgroup.

Fig.1.RelationshipbetweenpeakstrokevolumeandpeakO2pulse.Itisshownthat theregressionlineofGroupBissteeperthanthatofGroupA,andasthepeakstroke volumebecomesgreater,thedifferencebetweentwolinesbecomeslarger.SV, strokevolume.

(4)

withchronicheartfailure,wecompareditbetweenpatientswith and without beta-adrenergic receptor blocker in patients with chronicheartfailure[lowSV(<88.0mL),low exercisetolerance (peakV

˙O2<19:0mL=min=kg)andlowEF(<50%)].However, the O2pulse–SVratioinGroupsAandBwas10.63.4and10.83.4, respectively,andtherewasnosigniﬁcantdifferencebetweenthem.

Discussion

Thisstudyprovidednovelinformationdemonstratingthatthe effectofbeta-adrenergicreceptorblockerwasnotconstantduring exercise. When a subject is taking beta-adrenergic receptor

blockers, his/her exercisetolerance is preserved and/or he/she canincreaseSVduringexercise,thepeakO2pulsedoesnotindicate thepeakSV.However,itindicatedarelativelyhighervalue.This phenomenonbecamemoreapparentastheexercisetolerancewas greater.Tothebestofourknowledge,thisstudywastheﬁrstto investigatetheeffectofbeta-adrenergicreceptorblockersusing theO2pulse–SVratio.ThereweretwopatientsinwhomtheSVand O2 pulsevalueswere191mLand 54mL,and20.1mL/beatand 16.5mL/beat respectively, as shown in Fig. 1, and these two numericalvalueswereprominent.Itappearedasifthesepersons steepenedor shallowedtheregressionlinesofGroupsBand A.

However, even ifthese persons wereomitted as inappropriate samples,theinclinationsoftheregressionlinesdidnotchange.

Therefore,theseprominentpatientsdidnotaffecttheconclusion.

Usually,asthepeakvalueforC(a-v)O2isapproximately14[1,2], peak SV can be calculated by the following equation: peak SV=peakO2pulse/14100.BecauseO2pulseisaquotientofV˙O2

andheartrate(V˙O2=heartrate),theO2pulseisinﬂuencedbyheart rate. When the heart rate slows because of beta-adrenergic receptorblockeradministration,theO2pulsewouldbehigherthan expected. On the other hand, the SV should be constant or diminishedwhenbeta-adrenergicreceptorblockersareadminis- tered[21].Therefore,thepeakO2pulse/peakSVwouldbecome greater.However,inthisstudy,theO2pulse–SVratiowassimilar inpatientswithlowerexercisetolerancealthoughbeta-adrenergic receptorblockerswereused.Onereasonforthiswouldbethat,in thesepatients,heartratedecreasedandthepeakV˙O2diminished becauseofmulti-factorialfactorssuchastheonsetofmyocardial ischemia, increase of mitral regurgitation, or aggravation of diastolicdysfunction.Thatis,intheequationforV˙O2=heartrate, both denominator and numerator decreased simultaneously.

Hence,V

˙O2=heartratedidnotincrease.BecauseO2pulse–SVratio

isaC(a-v)O2,increaseofregressionlineinFig.1wouldbeaffected byincreaseofC(a-v)O2althoughthevariationofthepeakvalueof C(a-v)O2is notsogreat. However,in GroupB, steepnessofthe regression line was greater than Group A. This difference in steepness cannot be explained by the difference of C(a- v)O₂.Rather,itisthoughtthatthis isbecausethechangeofthe sympatheticnerveactivityisdifferentbetween GroupsAandB whenexerciseintensityincreased.

TheoxygenpulsehasbeenshownnottopredictSVinheart failure patients [22]. Actually, when the oxygen pulse was observed at a submaximal exercise intensity, it would not representtheSVif theexerciseintensitywasnotpresented as

%peakexercisebecausetheC(a-v)O2variedatvariousintensities.

However,althoughtherewasaslightvarietyinthepeakC(a-v)O2

accordingtothesubjects’background[23],italmostconvergedat approximately the constant value (11–16mL/dL) [1–3]. In this study, we used the peak value. Therefore, there should be no problemtreatingtheoxygenpulseasarepresentativeofSV.

On the other hand, in patients with preserved exercise tolerance,thediscrepancyinthepeakO2pulseandpeakSVwas apparent.Inthesepatients,theimpairmentoftheV˙O2 wasless than that of the heart rate. This may be the reason why the discrepancywasgreater.Usually,thelinearincreaseintheSVand ejectionfractionwasbluntedatthe50–60%ofpeakexerciseduring arampprotocol[24,25].Fromthatperiod,increaseofheartrate enhancesinordertomaintainthelinearincreaseofcardiacoutput.

Theincreaseofthecardiacpumpfunctionisessentialtoincrease thepeakV˙O2[26].Therefore,alackofnormalheartrateresponse duringexercisemayleadtoadepressedpeakV

˙O2.Inourstudy,the peakheartratewaslowerinGroupB,theenhancedincreaseinthe heart rate was blunted. However, the O2 pulse–SV ratio was increasedinthisstudy.Thereasonthattheincreaseinthepeak V˙O2 was relatively maintained remained unclear. But it was probablethat anincrease in theC(a-v)O2would occur because Fig.3.Differencein‘O2pulse–SVratio’betweenthetwogroups.Inthecategoryof

preservedpeakV˙O2,O2pulse–SVratioofGroupBisgreaterthanthatofGroupA.SV, strokevolume;V˙O2,oxygenuptake.

Fig.4. Difference in ‘O2 pulse–SV ratio’between thetwo groups. LVEF,left ventricularejectionfraction;SV,strokevolume.

Table4

Hemodynamicdifferenceamongthethreegroups.

GroupA GroupB

Carvedilol Bisoprolol

n 281 89 56

RestHR 74.413.0 74.210.0 69.611.8⁺

PeakHR 134.921.7 130.324.7 115.720.0^++,**

PeakSV 89.217.4 83.621.0^** 86.024.1

PeakO2pulse 9.32.3 8.72.6 9.83.2

PeakO2pulse–SVratio 10.72.8 10.62.5 11.73.4 n,number;HR,heartrate;SV,strokevolume;O2pulse,oxygenpulse.

** p<0.01vs.carvedilol.

+ p<0.05vs.GroupA.

++ p<0.01vs.GroupA

(5)

patientswithpreservedexercisetoleranceusuallydidnotreach thenadirofcriticalcapillaryPO2duringthepeakexercise.

IntherightpanelofFig.2,weshoweddifferencesoftheO2

pulse–SVratiobetweenGroupsAandBinpatientswithpreserved SV.InTable3,wepresentedthecardiacfunctionofthesesubjects.

ThereweresigniﬁcantdifferencesintheEFandE/E⁰betweenthe twogroupswithpreservedSV.However,whenwecheckedifthese datawhethertheyarewithinnormallimitsornot,EFinGroupB was within normal limits and E/E⁰ was on the border line.

Therefore,itmaybeacceptablethatthispanelwasarepresentative of the comparison of beta-blockers in almost normal cardiac function.

We added the examination concerning thedifference of O2

pulse–SVratiobetweenGroupsAandBinpatientswithpreserved LVEF (Fig. 4). As a result, there was no signiﬁcant difference between them. This may be because of the physiological characteristicsoftheLVEF.Although theLVEFhasbeenwidely usedasatentativemarkerforseverityofheartfailure,ithasbeen widelyknownthatthere wasa lackofconnectionbetween the LVEFand peak oxygen uptake [27,28],and that LVEF does not necessarilyrepresentheartfailureseverity.Aswell,althoughthe cardiac uptake of metaiodobenzylguanidine (MIBG) has been reportedtobepositivelyassociatedwithLVEF,theMIBGuptake wasunevenandhadnorelationshipwithLVEFinpatientswith preservedEF[29].Fromthesedata,weassumedthattheeffectsof beta-adrenergicreceptorblockerswerevariable.Thesewouldbe thereasonthattherewasnoapparentdifferenceintheO2pulse–

SVratio.

Inthisstudy,theplasmahemoglobinlevelwaslowerinGroupA thaninGroupB,althoughpatientswithmoderatetosevereanemia wereexcluded.Althoughanemiacanincreasetheheartrate,and lowerhemoglobinmayenhancetheheartrateresponsethevalue ofplasmahemoglobininGroupAwas13.5mg/dL.Thiswasnotthe rangeof anemia.Therefore, theeffectof thelowerhemoglobin levelinGroupAmaybeminimal.

As for the different hemodynamic effects of carvedilol and bisoprolol, it was revealed that there was no hemodynamic differencebetweenthetwodrugs.However,becausethedrugdose wasdifferentbetweenthetwogroups,itwasnotclearwhether thisphenomenonwasaclasseffectordrugeffect.

Studylimitations

In this present study, SV was measured using impedance cardiography.Thisrangehasbeenreportedtobeaccurate[12–

16].However,insomecases,signalsmaynotbedetectedandnot measured.Because suchcasescannotbeused, itdecreased the numberof studyparticipants.Furtherinvestigations withmore participantswillberequired.

Nowadays,anumberofbeta-adrenergicreceptorblockerscan be used for heart disease. In this study, four types of beta- adrenergic receptor blockers were used. The effect of beta- adrenergic receptor blockers on heart rate response differs by beta-adrenergic receptor selectivity and/or co-existent alpha- adrenergic receptor blocking action. In this study, because the numberofpatientswithbeta-adrenergicreceptorblockerswasnot sufﬁcient, we could not clarify the difference in the beta- adrenergic receptor blockers. More precise evaluation will be needed.

Conclusion

It was shown that the increase of the O2 pulse was disproportionately greater than the measured SV when beta- adrenergicreceptorblockerswereusedinpatientswithpreserved SVandexercisetolerance.Carefulconsiderationwillbenecessary

whenthepeakO2pulseisusedtoevaluatethecardiacfunction duringexercise.

Funding

Thisresearchreceivednograntfromanyfundingagencyinthe public,commercial,ornot-for-proﬁtsectors.

Conﬂictsofinterest

Theauthorsdeclarethattherearenoconﬂictsofinterest.

Acknowledgments

TheauthorsthankMrMasanoriUeda,MrYasuyukiKobayashi, Miss IzumiTeshima, and other medical technologists fromthe physiologicallaboratoryfortheirtechnicalassistance.

References

[1]SullivanMJ,KnightJD,HigginbothamMB,CobbFR.Relationbetweencentral andperipheralhemodynamicsduringexerciseinpatientswithchronicheart failure.Circulation1969;80:769–81.

[2]WeberKT, JanickiJS.Cardiopulmonaryexercise testing forevaluation of chronicheartfailure.AmJCardiol1985;55:22A–31A.

[3]StringerWW,HansenJE,WassermanK.Cardiacoutputestimatednoninva- sivelyfromoxygenuptakeduringexercise.JApplPhysiol1997;82:908–12.

[4]BelardinelliR,LacalapriceaF,CarlebF,MinnuccibA,CianciaG,PernaaGP, D’EusanioaG.Exercise-inducedmyocardialischaemiadetectedbycardiopul- monaryexercisetesting.EurHeartJ2003;24:1304–13.

[5]MatsuzakiM,IshiiM,IzumiT,ItohH,ItohH,UedaY,OkumuraK,OtujiY, KiharaY,KyoS,SaitoY,ShimokawaH,TsutsuiY,TsutamotoT,NakataniT,etal.

Guidelinesfortreatmentofchronicheartfailure(JCS2010).http://www.j-circ.

or.jp/guideline/pdf/JCS2010_matsuzaki_h.pdf.

[6]HuntSA,AbrahamWT,ChinMH,FeldmanAF,FrancisGS,GaniatsTG,JessupM, KonstamMA,ManciniDM,MichlK,OatesJA,RahkoPS,SilverMA,Stevenson LW,YancyCW,etal.ACC/AHA2005GuidelineUpdatefortheDiagnosisand ManagementofChronicHeartFailureintheAdult:areportoftheAmerican College ofCardiology/AmericanHeartAssociation TaskForce onPractice Guidelines(WritingCommitteetoUpdatethe2001GuidelinesfortheEvalu- ationandManagementofHeartFailure):developedincollaborationwiththe AmericanCollegeofChestPhysiciansandtheInternationalSocietyforHeart andLungTransplantation:endorsedbytheHeartRhythmSociety.Circulation 2005;112:e154–235.

[7]McMurrayJJ,AdamopoulosS,AnkerSD,AuricchioA,Bo¨hmM,DicksteinK,Falk V,FilippatosG,FonsecaC,Gomez-SanchezMA,JaarsmaT,KøberL,LipGYH, MaggioniAP,ParkhomenkoA,etal.ESCGuidelinesforthediagnosisand treatmentofacuteandchronicheartfailure2012.TheTaskForceforthe DiagnosisandTreatmentofAcuteandChronic HeartFailure2012ofthe EuropeanSocietyofCardiology.DevelopedincollaborationwiththeHeart FailureAssociation(HFA)oftheESC.EurHeartJ2012;33:1787–847.

[8]MignattiA,SimsDB,ColomboPC,GarciaLI,BijouR,DengMC,JordeUP.Resting heartratedoesnotreﬂectthedegreeofbeta-blockadeinsubjectswithheart failureonchronicbeta-blockertherapy.CardiovascTher2009;27:42–8.

[9]BuchfuhrerML,HartsenJE,RobinsonTE,SueDY,WassermanK,WhippBJ.

Optimizing theexerciseprotocolfor cardiopulmonaryassessment.J Appl Physiol1983;55:1558–64.

[10]TaguchiT,AdachiH,HoshizakiH,OshimaS,KurabayashiM.Effectofphysical trainingonventilatorpatternsduringexerciseinpatientswithheartfailure.J Cardiol2014;65:343–8.

[11]BeaverWL,WassermanK,WhippBJ.Anewmethodfordetectinganaerobic thresholdbygasexchange.JApplPhysiol1986;60:2020–7.

[12]HetherringtonM,TeoKK,HaenneiR,GreenwoodP,RossallRE,KappagodaT.

Useofimpedancecardiographyinevaluatingtheexerciseresponseofpatients withleftventriculardysfunction.EurHeartJ1985;12:1016–24.

[13]RichardR,Lonsdorfer-WolfE,CharlouxA,DoutreleauS,BuchheitM,Oswald- MammosserM,LampertE,MettauerB,GenyB,LonsdorferJ.Non-invasive cardiacoutputevaluationduringamaximalprogressiveexercisetest,usinga newimpedancecardiographdevice.EurJApplPhysiol2001;85:202–7.

[14]DonovanKD,DobbGJ,WoodsWP,HockingsBE.Comparisonoftransthoracic electrical impedanceand thermodilution methodsfor measuring cardiac output.CritCareMed1986;14:1038–44.

[15]BelardinelliR,CiampaniN,CostantiniC,BlandiniA,PurcatoA.Comparisonof impedancecardiographywiththermodilutionanddirectFickmethodsfor noninvasivemeasurementofstrokevolumeandcardiacoutputduringincre- mental exerciseinpatientswith ischemiccardiomyopathy. AmJCardiol 1996;77:1293–301.

[16]KinnenE.Cardiacoutputfromtransthoracicimpedancevariations.AnnNY AcadSci1970;170:747–56.

(6)

[17]WhiteSW,QuailAW,DeLeeuwPW,TraugottFM,BrownWJ,PorgesWL,Cottee DB.Impedancecardiographyforcardiacoutputmeasurement:anevaluation ofaccuracyandlimitations.EurHeartJ1990;11(Suppl.1):79–92.

[18]BernsteinDP.Anewstrokevolumeequationforthoracicelectricalbioimpe- dance:theoryandrationale.CritCareMed1986;14:904–9.

[19]LangRM,BierigM,DevereuxRB,FlachskampfFA,FosterE,PellikkaPA,Picard MH,RomanMJ,SewardJ,ShanewiseJ,SolomonS,SpencerKT,StJohnSutton M,StewartW,AmericanSocietyofEchocardiography’sNomenclatureand StandardsCommittee.etal.Recommendationsforchamberquantiﬁcation.

EurJEchocardiogr2006;7:79–108.

[20]NaguehSF,AppletonCP,GillebertTC,MarinoPN,OhJK,SmisethOA,Waggoner AD,FlachskampfFA,PellikkaPA,EvangelisaA.Recommendationsforthe evaluationofleftventriculardiastolicfunctionbyechocardiography.EurJ Echocardiogr2009;10:165–93.

[21]RobertsH,TsaoY,GrimmerSFM,WinstanleyPA,OrmeML,BreckenridgeAM.

Haemodynamiceffectsofatenolol,labetalol,pindololandcaptopril:acom- parisoninhypertensivepatientswithspecialreferencetochangesinlimb blood ﬂow,heart rateand leftventricular function.Br JClin Pharmacol 1987;24:163–72.

[22]McConnellTR,ShearnWM,KingerTA, StroheckerK.Oxygenpulseisnot predictive of stroke volume in heart failure. J Sports Med PhysFitness 2006;46:286–92.

[23]Dhakal BH,MalhotraR, MurphyRM,Pappagianopoulos PPP, BaggishAL, WeinerRB,HoustisNE,Eisman AS,HoughSS,LewisGD.Mechanismsof

exerciseintoleranceinheartfailurewithpreservedejectionfraction.Therole ofabnormalperipheraloxygenextraction.CircHeartFail2015;8:286–94.

[24]BoucherCA,AndersonMD,SchneiderMS,MurphyJH,OkadaRD,KanarekDJ.

Leftventricularfunctionbeforeandafterreachingtheanaerobicthreshold.

Chest1985;87:145–50.

[25]KoikeA,ItohH,TaniguchiK,HiroeM.Detectingabnormalitiesinleftventric- ular function during exercise by respiratory measurement. Circulation 1989;80:1737–46.

[26]DeMarcoT,WolfelE,FeldmanAM,LowesB,HigginbothamMB,GhaliJK, WagonerL,KirlinPC,KennettJD,GoelS,SaxonLA,BoehmerJP,MannD,Galle E,EcklundF,etal.Impactofcardiacresynchronizationtherapyonexercise performance,functionalcapacity,andqualityoflifeinsystolicheartfailure withQRSprolongation:COMPANIONtrialsub-study.JCardFail2008;14:9–

18.

[27]FroelicherVF.Interpretationofspeciﬁcexercisetestresponses.In:Froelicher VF,editor. Exerciseandthe heart. 2nded.,Chicago: Year BookMedical PublishersInc.;1987.p.83–145.

[28]FranciosaJA,ParkM,LevineTB.Lackofcorrelationbetweenexercisecapacity andindexesofrestingleftventricularperformanceinheartfailure.AmJ Cardiol1981;47:33–9.

[29]NakaeI,MitsunamiK,HayashiH,HorieM.Associationofiodine-123BMIPP andMIBGcardiacuptakeswithleftventricularfunctionalparametersassessed bygatedmyocardialperfusionSPECT.JPublicHealthFront2013;2:179–88.