Myocardial Injury and Ventricular Dysfunction Related to Training Levels Among Nonelite Participants...

Myocardial Injury and Ventricular Dysfunction Relatedto Training Levels Among Nonelite Participants in the

Boston MarathonTomas G. Neilan, MD; James L. Januzzi, MD; Elizabeth Lee-Lewandrowski, PhD;

Thanh-Thao Ton-Nu, MD; Danita M. Yoerger, MD; Davinder S. Jassal, MD;Kent B. Lewandrowski, MD; Arthur J. Siegel, MD; Jane E. Marshall, RDCS; Pamela S. Douglas, MD;

David Lawlor, MD; Michael H. Picard, MD; Malissa J. Wood, MD

Background—Multiple studies have individually documented cardiac dysfunction and biochemical evidence of cardiacinjury after endurance sports; however, convincing associations between the two are lacking. We aimed to determinethe associations between the observed transient cardiac dysfunction and biochemical evidence of cardiac injury inamateur participants in endurance sports and to elicit the risk factors for the observed injury and dysfunction.

Methods and Results—We screened 60 nonelite participants, before and after the 2004 and 2005 Boston Marathons, withechocardiography and serum biomarkers. Echocardiography included conventional measures as well as tissueDoppler–derived strain and strain rate imaging. Biomarkers included cardiac troponin T (cTnT) and N-terminalpro-brain natriuretic peptide (NT-proBNP). All subjects completed the race. Echocardiographic abnormalities after therace included altered diastolic filling, increased pulmonary pressures and right ventricular dimensions, and decreasedright ventricular systolic function. At baseline, all had unmeasurable troponin. After the race, �60% of participants hadincreased cTnT �99th percentile of normal (�0.01 ng/mL), whereas 40% had a cTnT level at or above the decision limitfor acute myocardial necrosis (�0.03 ng/mL). After the race, NT-proBNP concentrations increased from 63(interquartile range [IQR] 21 to 81) pg/mL to 131 (IQR 82 to 193) pg/mL (P�0.001). The increase in biomarkerscorrelated with post-race diastolic dysfunction, increased pulmonary pressures, and right ventricular dysfunction (rightventricular mid strain, r��0.70, P�0.001) and inversely with training mileage (r��0.71, P�0.001). Compared withathletes training �45 miles/wk, athletes who trained �35 miles/wk demonstrated increased pulmonary pressures, rightventricular dysfunction (mid strain 16�5% versus 25�4%, P�0.001), myocyte injury (cTnT 0.09 versus �0.01 ng/mL,P�0.001), and stress (NT-proBNP 182 versus 106 pg/mL, P�0.001).

Conclusions—Completion of a marathon is associated with correlative biochemical and echocardiographic evidence ofcardiac dysfunction and injury, and this risk is increased in those participants with less training. (Circulation. 2006;114:2325-2333.)

Key Words: echocardiography � exercise � natriuretic peptides

The risk of death from coronary heart disease is decreased2-fold in physically active people.1 Regular exercise has

beneficial effects on blood pressure, lipid profile, insulinresistance, and overall risk of death.2 Although prolongedexercise has been associated with pathological changes suchas increased platelet aggregation,3 left ventricular (LV) wallmotion abnormalities,4 and, infrequently, acute myocardialinfarction, the overall risk is low5,6 and multiple studies havedemonstrated short- and long-term beneficial effects of en-durance exercise.7,8 Multiple studies have also independentlydemonstrated biochemical9 and structural evidence4 of tran-

Editorial p 2306Clinical Perspective p 2333

sient cardiac injury and dysfunction after endurance sports,but these studies have not linked these abnormalities and mayhave been limited by assay specificity and the sensitivity ofimaging modalities. In fact, to date, convincing associationsbetween the cardiac dysfunction and biochemical evidence ofcardiac injury with prolonged exercise have not been dem-onstrated, and risk factors for the development of signs ofmyocardial damage or dysfunction are unknown. Therefore,

Received June 21, 2006; revision received August 23, 2006; accepted August 25, 2006.From the Cardiac Ultrasound Laboratory (T.G.N., J.L.J., T.-T.T.-N., D.M.Y., D.S.J., J.E.M., M.H.P., M.J.W.), Division of Cardiology, Pathology

(E.L.-L., K.B.L.), and Pediatric Surgery (D.L.), Massachusetts General Hospital, and the Department of Medicine (A.J.S.), McLean Hospital, HarvardMedical School, Boston, Mass; and the Division of Cardiovascular Medicine (P.S.D.), Duke University Medical Center, Durham, NC.

Correspondence to Dr Malissa J. Wood, Cardiac Ultrasound Laboratory, Division of Cardiology, Massachusetts General Hospital, 55 Fruit St, VBK508, Boston, MA 02115-2696. E-mail [email protected]

© 2006 American Heart Association, Inc.

Circulation is available at http://www.circulationaha.org DOI: 10.1161/CIRCULATIONAHA.106.647461

2325

Exercise Physiology

by guest on January 29, 2016http://circ.ahajournals.org/Downloaded from

http://circ.ahajournals.org/

the aim of the present study was to determine the associationsbetween the observed transient cardiac dysfunction and injuryamong amateur participants in endurance sports and to elicitthe factors associated with the development of this dysfunc-tion and injury.

MethodsScreening and Approval ProcessThe protocol was approved by the Partners Healthcare SystemHuman Subjects Review Committee and The Boston Athletic Asso-ciation. Sixty amateur runners scheduled to participate in the 2004and 2005 Boston Marathons were recruited by open e-mail invitationsent to local running clubs. We did not approach any individualrunner directly; rather, a general e-mail was sent to all local runningclubs requesting them to forward it to all their members. Weaccepted all responders up to 30 subjects per year. The sample sizewas limited by the combination of the extensive finish line resourcesrequired in arranging such a study and the desire to limit anyinconvenience and discomfort that a prolonged delay after a mara-thon may have on subjects. We excluded subjects with a history ofcardiovascular disease. Athletes volunteered for the study andprovided written consent. No subject refused consent. Comprehen-sive screening consisted of a general questionnaire to ascertainpersonal and marathon history, as well as a log of marathon trainingfor the previous 4 months. Assessment included measurement ofheart rate, blood pressure, serum biomarkers, and a completeechocardiographic evaluation. Runners were screened �1 weekbefore and immediately (approximately 20 minutes) after completionof the marathon. Liberal oral intake of fluid was encouraged after themarathon to minimize volume depletion.

EchocardiographySubjects underwent 2-dimensional pulsed-Doppler and color tissue-Doppler (TD) imaging by use of a commercial system (Vivid 7, GEHealthcare, Milwaukee, Wis) and a phased-array transducer. Stan-dard measurements were performed according to American Societyof Echocardiography guidelines.10 Peak pulmonary artery pressurewas estimated from the jet of tricuspid regurgitation with theBernoulli equation,11 and the mean pulmonary artery pressure(mPAP) was estimated from the pulmonary outflow Doppler accel-eration time.12 Pulsed annular Doppler and TD were used to quantifyregional and global systolic and diastolic function.13,14 From thepulsed-Doppler mitral annular velocities (peak early [E�] and late[A�]), indices of LV diastolic function were recorded.

From the apical 4-chamber view, color TD images were acquiredwith the sample volume placed at the basal (tricuspid annulus), mid,and apical levels (moderator band) of the right ventricle (RV). Fromthese images, RV myocardial wall velocities (VENDO), strain (�), andstrain rate (SR), indices of RV systolic function, were measuredoffline (EchoPac, Version 6.3, GE Healthcare).

Biochemical StudiesBlood was collected into tubes containing EDTA and into serumseparator tubes and processed immediately. Assays for all cardiacbiomarkers were performed before the first freeze-thaw. None of thespecimens demonstrated signs of hemolysis. Quantitative determi-nation of cardiac troponin T (cTnT; Stat T, Roche Diagnostics,Indianapolis, Ind) was measured on a Roche Elecsys 1010 platform.The 99th percentile for normal subjects is 0.01 ng/mL, whereas thecutpoint providing 10% coefficient of variation with this assay is0.03 ng/mL, thus representing the conventional upper limit of normalfor this assay.15

N-terminal pro-brain natriuretic peptide (NT-proBNP) levels weremeasured with an electrochemiluminescence sandwich immunoas-say (Elecsys ProBNP, Roche Diagnostics) with the Roche 2010system. For subjects �75 years of age, the upper limit of normal isconsidered 125 pg/mL. Ischemia-modified albumin (Inverness Med-ical Innovations, Waltham, Mass) was measured by the albumin

cobalt binding test on a Roche Hitachi 911 platform. We applied�95 U/mL (97.5th percentile of healthy patients) as the upper limitof normal. In addition, serum sodium was also measured.

Statistical AnalysisData for cTnT and NT-proBNP are presented as medians withinterquartile range (IQR). Data for all other variables are presentedas mean�SD. A paired Student t test was used to compare changesbefore and after the marathon. Pearson and Spearman correlationcoefficients were calculated as standard. In order to better understandthe relationship of post-race NT-proBNP and cTnT concentrationswith structural abnormalities on echocardiography, we performedunivariable and multivariable logistic regression analyses. For theevaluation of NT-proBNP, a log-transformed NT-proBNP in thehighest tertile was used as the dependent variable. Covariatesexamined for inclusion in the multivariable models of NT-proBNPand structural abnormalities were age, E� lateral, and A� lateral. Toexamine the relationship between cTnT and RV strain, cTnT was logtransformed and used as the dependent variable in a manner similarto NT-proBNP, and the covariates age and strain at the apex andmid-ventricle were examined. Multivariate statistical analyses wereperformed with the use of SPSS software (SPSS Inc, Chicago, Ill).Odds ratios (ORs) with 95% confidence intervals (CIs) were gener-ated. Runners were stratified into 3 groups according to trainingmileage: group A, �35 miles/wk; group B, 36 to 45 miles/wk; andgroup C, �45 miles/wk. Parameters of interest were compared byusing a 1-way analysis of variance (ANOVA). If the ANOVAshowed an overall difference, post hoc comparisons were performedwith a Student t test. Comparisons between biomarker concentrationsbefore and after the race were performed with nonparametric tests,whereas differences in post-race values across the 3 training groupswere compared with the Kruskal-Wallis test. A 2-sided probabilityvalue of �0.05 was considered significant.

The authors had full access to and take responsibility for theintegrity of the data. All authors have read and agree to themanuscript as written.

ResultsBaseline ParametersThe participants included 41 men and 19 women (Table 1)with an average age of 41 years (range 21 to 65). The meantraining mileage was 42�9 miles/wk, and the average finishtime was approximately 4 hours (4 hours 5 minutes, range 2hours 55 minutes to 5 hours 55 minutes). All runnersdemonstrated weight loss (158�26 versus 155�26 lb,P�0.001). Heart rate increased (61�12 versus 101�15 bpm,P�0.001) and systolic blood pressure decreased (114�12versus 101�15 mm Hg, P�0.001) after the marathon. Norunner required medical attention.

EchocardiographyAt baseline, all echocardiographic indices, including chamberand wall dimensions, were within normal limits (Table 1).The time from cessation of running to acquisition of echo-cardiographic images was similar among all runners (20minutes) and independent of training status or finish time.There were no significant changes in left atrial dimensions,left atrial area, or right atrial area before versus after themarathon. LV and RV size were normal at baseline (Table 2).Although LV dimensions (not shown), volumes, and ejectionfraction remained unchanged, RV dimensions and area in-creased and the RV percentage area change decreased(41�7% versus 33�7%, P�0.001). Although heart rateswere increased after the marathon, we did not observe fusionof E and A waves either by pulsed-annular Doppler or by

2326 Circulation November 28, 2006



transmitral Doppler. There was a reduction in early transmi-tral diastolic filling velocities and an increase in late trans-mitral filling, which led to a decreased E/A ratio. There weresimilar alterations in diastolic function as assessed by TD-derived annular velocities (Table 2). The alterations in LVdiastolic filling were associated with the increase in NT-proBNP (E� lateral versus NT-proBNP, r��0.67, P�0.001;A� lateral versus NT-proBNP, r�0.55, P�0.001).

There was an increase in both the peak pulmonary arterysystolic pressure (PASP) (20�3 versus 41�7 mm Hg,P�0.001, n�29/60) and mPAP (12�3 versus 25�9,P�0.001, n�59/60). This increase in mPAP was positivelyassociated with the increase in cTnT (r�0.70, P�0.001) andinversely associated with the average training mileage(r��0.71, P�0.001). Despite the increase in heart rate, therewas a reduction in TD-derived indices of RV systolic func-tion (Table 2). The RV endocardial velocity and � decreased,whereas the SR remained unchanged. The reduction in RVendocardial velocities and strain correlated with the increasein cTnT (RV basal �, r��0.68; mid �, r��0.70; and apical�, r��0.72; P�0.001 for all).

Runners were stratified according to training mileage(Table 3). At baseline, runners who had higher trainingmileage had increased LV dimensions, reduced ejectionfraction, and improved diastolic indices (E� before the racefor those who trained �45 miles/wk 12�2 cm · s�1 versus10�2 cm · s�1 for those who trained �35 miles/wk,P�0.001). Upon completion of the marathon, those partici-pants who averaged �35 miles/wk had the greatest increases

in PASP, mPAP, RV dimensions, and RV area and thegreatest changes in LV diastolic and RV systolic function(Figure 1). Overall, there were no associations between thealterations in PASP, mPAP, RV dimensions, RV area, LVdiastolic and RV systolic function, and subject’s age, changein body weight, or gender.

Biochemical Markers

cTnTSerum cTnT concentrations were �0.01 ng/mL in all partic-ipants at baseline. After the race, the median cTnT value was0.022 ng/mL (IQR 0.03 to 0.06 ng/mL; P�0.001 for differ-ence from before the race), with a post-race range of cTnTvalues from �0.01 to 0.82 ng/mL. An increase in cTnT�0.01 ng/mL was recorded in 38 subjects (63%), whereas 28subjects (47%) had cTnT values �0.03 ng/mL, and 8 subjectshad post-race cTnT values �0.10 ng/mL.

With cTnT release after completion of the marathonconsidered as a function of training adequacy, those partici-pants who averaged �35 miles/wk in training had the greatestincrease in cTnT (P�0.001 across groups; Figure 2) after therace. Among those training �35 miles/wk, the medianpost-race cTnT value was 0.09 ng/mL, whereas all subjectshad a post-race cTnT �0.03 ng/mL, with a range from 0.03to 0.82 ng/mL. Also, 50% of subjects in this category had apost-race cTnT �0.10 ng/mL, representing all subjects withcTnT at or above this magnitude in the study. Among thosetraining 36 to 45 miles/wk, there was significantly lowerrelease of cTnT, with a median value of 0.02 ng/mL(P�0.008 versus baseline) and a range of �0.01 to 0.07ng/mL. Of these subjects, 45% had a post-race cTnT �0.03ng/mL, and none were �0.10 ng/mL. Finally, among thosetraining �45 miles/wk, no appreciable change in mediancTnT from baseline was noted (median �0.01; P�0.87compared with baseline), with only 2 subjects (8.7%) dem-onstrating post-race cTnT values �0.03 ng/mL.

NT-proBNPAt baseline, overall median NT-proBNP concentrations were106 pg/mL (IQR 65 to 175 pg/mL); after the marathon,NT-proBNP levels were significantly higher (182 pg/mL,IQR 112 to 219 pg/mL, P�0.001 between before and afterthe race), with a post-race range of NT-proBNP concentra-tions from 14 to 506 pg/mL. Overall, 54% of participants hadlevels of NT-proBNP above the upper limit of normal forexclusion of heart failure. Females were more likely thanmales to show a post-race increase in NT-proBNP concen-trations (63% versus 17%, P�0.01), and the increase inNT-proBNP was independently associated with the attenua-tion in LV early diastolic filling (E� lateral, OR 8.4, 95% CI1.3 to 56.0, P�0.026).

In a fashion similar to cTnT, those participants whoaveraged �35 miles/wk had the greatest increase in NT-proBNP (P�0.03 across all 3 training groups) (Figure 3) afterthe race. Among participants who trained �35 miles/wk, themedian post-race NT-proBNP value was 182 pg/mL (IQR112 to 219 pg/mL; P�0.001 compared with before the race),with an overall range of 82 to 506 pg/mL; of this group, 75%had a post-race NT-proBNP �125 pg/mL. Among those

TABLE 1. Baseline Parameters Among the AmateurStudy Participants

Age, y 41�11

Gender, n (%)

Male 41 (68)

Female 19 (32)

Body mass index, kg/m2 23�3

Heart rate, bpm 61�12

Blood pressure, mm Hg

Systolic 114�12

Diastolic 69�8

Average trainingmileage, miles/wk

42�9

Left atrial size, mm 34�4

LV posterior wallthickness, mm

10�1

LV diastolicdimensions, mm

47�5

LV ejection fraction, % 60�6

Indexed LV mass, g/m2 104�16

PASP, mm Hg 20�3

mPAP, mm Hg 12�3

RV diastolicdimensions, mm

35�4

RV fractional areachange, %

41�7

Neilan et al Marathon Running and the Heart 2327



training 36 to 45 miles/wk, there was lower release ofNT-proBNP after the race, with a median value of 94 pg/mL(IQR 56 to 211 pg/mL; P�0.004 versus baseline) and anoverall range of 14 to 271 pg/mL after the race. Of thesesubjects, 45% had a post-race NT-proBNP �125 pg/mL. Onmultivariate analysis, the increase in cTnT was independentlyassociated with a reduction in RV contractility (RV mid �,OR 14.0, 95% CI 1.37 to 142.8, P�0.026). Finally, amongthose training �45 miles/wk, median NT-proBNP concentra-tions were similarly lower at 106 pg/mL (IQR 64 to 175pg/mL; P�0.001 versus baseline), with an overall range of 28to 385 pg/mL and 48% of subjects �125 pg/mL after therace.

Other BiomarkersLevels of ischemia-modified albumin decreased from base-line (94�6 before the race versus 73�10 U/mL after the race,P�0.001), whereas serum sodium was unchanged (139�2versus 139�3 mEq/L, P�0.76).

DiscussionAlthough prior studies involving endurance athletes afterexercise have demonstrated cardiac dysfunction, and others

have demonstrated biochemical evidence of possible cardiacinjury, none have correlated these findings and the riskfactors for the development of such abnormalities are un-known. In recreational athletes completing a marathon, wehave suggested and correlated echocardiographic and bio-chemical evidence of cardiac dysfunction and injury. Withprolonged exercise, we found increased PAPs, increased RVdimensions, and decreased RV function (which correlatedwith release of cTnT), as well as alterations in LV diastolicfunction (which correlated with an increase in NT-proBNPvalues). Most strikingly, these changes were strongly influ-enced by the level of preparation undertaken by these amateurathletes, such that the majority of the most marked abnormal-ities in cardiac structure or function as well as cardiacbiomarker changes were seen in those athletes training �35miles/wk before the marathon.

Similar effort-related changes have been documented in-vasively among other groups of athletes performing shorter-term higher-intensity exercise, with the magnitude of theincrease being related to the intensity of the exercise. Mod-erately strenuous cycling for 1 hour among trained competi-tive athletes was associated with an increase in pulmonary

TABLE 2. Baseline and Postmarathon Echocardiographic Indices (n�60)

Variable Baseline After the Marathon P

Left atrial dimensions, mm 34�4 33�3 0.17

Left atrial area, cm2 19�3 18�3 0.49

Right atrial area, cm2 17�2 16�3 0.45

LV end-diastolic volume, cm3 110�20 105�23 0.11

LV end-systolic volume, cm3 44�10 43�10 0.35

RV dimensions, mm 35�4 41�4 0.001

RV diastolic area, cm2 17�4 20�3 0.008

RV systolic area, cm2 10�2 13�2 0.004

LV ejection fraction, % 60�6 59�6 0.44

RV area change, % 41�7 33�7 �0.001

Mitral E-wave filling velocity, m � s�1 0.9�0.1 0.6�0.2 �0.001

Mitral A-wave filling velocity, m � s�1 0.5�0.1 0.7�0.12 �0.001

Ratio of mitral E/A 1.6�0.4 1.0�0.4 �0.001

TD-derived E� lateral, cm � s�1 12�2 8�2 �0.001

TD-derived A� lateral, cm � s�1 5�1 8�2 �0.001

TD-derived E� septal, cm � s�1 10�2 8�2 �0.001

TD-derived A� septal, cm � s�1 5�1 8�2 �0.001

RV base

VENDO, cm � s�1 11�2 9�2 0.007

�, % 18�7 14�4 �0.001

SR, s�1 1.2�0.4 1.2�0.3 0.25

RV mid

VENDO, cm � s�1 11�2 7�2 0.001

�, % 27�6 21�5 �0.001

SR, s�1 1.6�0.4 1.5�0.4 0.62

RV apex

VENDO, cm � s�1 8�2 6�1 0.007

�, % 38�8 29�7 �0.001

SR, s�1 2.4�0.6 2.3�0.6 0.86




TABLE 3. Stratification of Baseline and Postmarathon Variables According to Average Training Mileage in the 4 MonthsBefore the Marathon

P

Group A�35 miles/wk

(n�17)

Group B36 to 45 miles/wk

(n�20)

Group C�45 miles/wk

(n�23) Overall A vs B A vs C B vs C

Average training, miles/wk 29�6 43�3 51�2 �0.001 �0.001 �0.001 �0.001

Finishing time, min 268�39 245�41 225�29 0.03 0.16 0.19 0.009

Age, y 40�13 41�11 40�10 0.83 0.55 0.83 0.67

Male gender, % 59% 70% 74% 0.80 0.29 0.35 0.68

Heart rate, bpm

Before 63�7 62�9 59�8 0.17 0.55 0.07 0.21

After 101�16 99�20 103�10 0.66 0.53 0.87 0.39

Systolic blood pressure, mm Hg

Before 115�10 115�12 113�13 0.88 0.94 0.65 0.68

After 105�16 103�12 104�16 0.68 0.71 0.64 0.61

NT-proBNP, pg/mL, median (IQR)

Before 64 (25–77) 60 (30–81) 63 (19–85) 0.91 0.71 0.95 0.72

After 182 (112–219) 94 (56–211) 106 (64–175) 0.03 0.008 0.003 0.84

cTnT, ng/mL, median (IQR)

Before �0.01 �0.01 �0.01 0.99 0.99 0.99 0.99

After 0.09 (0.07–0.21) 0.02 (0.02–0.04) �0.01 (�0.01–0.01) �0.001 �0.001 �0.001 0.87

LV diastolic dimensions, mm

Before 45�4 46�6 50�4 �0.001 0.31 �0.001 �0.001

After 44�4 45�4 47�4 0.42 0.18 0.36 0.60

LV ejection fraction, %

Before 64�6 58�4 55�5 �0.001 0.009 �0.001 0.01

After 63�5 59�7 57�7 0.006 0.01 0.002 0.50

TD-derived E� lateral, cm � s�1

Before 10�2 11�2 12�2 0.005 0.08 0.001 0.10

After 6�2 9�2 10�2 �0.001 �0.001 �0.001 0.05

% Decrease E� 44�18 23�10 19�12 �0.001 �0.001 �0.001 0.26

TD-derived A� lateral, cm � s�1

Before 5�1 6�1 6�1 0.02 0.04 0.007 0.32

After 9�2 7�2 7�2 0.02 0.03 0.009 0.69

% Increase A� 46�17 18�14 19�11 �0.001 �0.001 �0.001 0.68

RV diastolic dimensions, mm

Before 35�4 35�4 36�3 0.28 0.70 0.29 0.12

After 45�8 41�5 38�4 0.001 0.01 0.001 0.06

RV area change, %

Before 41�7 40�7 41�7 0.86 0.83 0.77 0.60

After 28�7 33�9 38�6 0.001 0.06 0.0002 0.04

Peak PASP (n�29), mm Hg

Before 20�4 21�3 20�2 0.81 0.53 0.77 0.80

After 45�5 37�5 32�4 �0.001 0.001 �0.001 0.05

Mean PASP (n�59), mm Hg

Before 13�3 12�3 12�3 0.82 0.54 0.70 0.78

After 35�7 23�5 18�4 �0.001 �0.001 �0.001 0.01

Mid RV �, %

Before 29�6 26�4 28�7 0.39 0.17 0.45 0.48

After 16�5 22�4 25�4 �0.001 0.003 �0.001 0.05




capillary wedge pressure from 8 to 15 mm Hg and mPAPfrom 14 to 26 mm Hg,16 whereas, among a group of ultra-marathon runners performing at high altitude, transientlyincreased PAP with RV dilatation and RV dysfunction and

symptoms were documented without an overall change in aless-sensitive troponin assay.17 Our data suggest that theincrease in pulmonary pressures is likely multifactorial,occurring in association with impaired LV relaxation but also

Figure 1. Baseline and postmarathon echocardiographic images from a participant who trained �35 miles per week and had anincrease in cTnT from �0.01 to 0.28 ng/mL and an increase in NT-proBNP from 68 to 220 pg/mL. A, Apical 4-chamber view at base-line and after the marathon demonstrating mild RV dilatation. B, Estimated RV systolic pressure at baseline (23 mm Hg) and after com-pletion of the marathon (48 mm Hg). C, Estimated mPAP calculated by using the pulmonary artery Doppler profile at baseline (accelera-tion time 137 ms, mPAP 8 mm Hg) and after completion of a marathon (acceleration time 110 ms, mPAP 22 mm Hg).




perhaps reflecting an intrinsic increase in pulmonary vascularresistance.

We excluded subjects with a history of cardiovasculardisease. In all participants, baseline echocardiographic indi-ces were normal, making the presence of occult hypertrophiccardiomyopathy, arrhythmogenic right ventricular dysplasia,myocarditis, and aortic stenosis unlikely.18 Because of thecombination of a low cardiac risk group and a lack of

evidence of LV wall motion abnormalities suggestive ofischemia in our study and in previous studies19, it is unlikelythat the cTnT elevation is due to epicardial coronary ische-mia. Furthermore, the average finish time for our slowestcohort was 4 hours and 28 minutes, the minimum periodrequired for myofibril degradation and release of this struc-tural protein.20 Also, the pattern of cTnT release21 is differentthan that expected with ischemic injury22; therefore, alterna-tive reasons for the elevation in troponin must be considered.In addition to bound troponin, there is also a free cytoplasmiccomponent comprising approximately 8% of cTnT. Releaseof this cytoplasmic cTnT may explain the early rise of thismarker after myocyte damage.20 Why reversible leakageoccurs is unclear, but this leakage may involve oxidativestress,23 hypoxia,24 or transient ischemia.25 In acute pulmo-nary embolism, similar changes occur with RV dilatation,dysfunction, mild elevations in pulmonary pressure, andincreased troponin without obvious ischemia.26,27 Impor-tantly, the parallel changes in both echocardiography and theenhanced specificity of the troponin assay employed28 sug-gest that the source of cTnT in our subjects is probably ofcardiac rather than of skeletal muscle origin.

In addition to changes in cTnT, we noted significantpost-race changes in NT-proBNP in our subjects, which webelieve reflect changes in diastolic filling after strenuousexercise. Although one suggestion may be that the elevationin natriuretic peptide concentrations in athletes might reflecta physiological response to increased natriuresis, the lack ofchange in serum sodium in our study subjects makes thisunlikely. Notably, post-race values of NT-proBNP weretypically higher than those used for the exclusion of heartfailure, and like cTnT levels were strongly influenced byprerace preparation. Although elevation of natriuretic peptideconcentrations after strenuous exercise was demonstratedpreviously,21 to our knowledge this is the first study tocorrelate changes in NT-proBNP to the development ofabnormalities on echocardiography and to demonstrate dif-ferences in release magnitude as a function of an athlete’straining.

Our study has several limitations. Although pulsed-Dopp-ler and TD are purported to be less sensitive to alterations inloading conditions and more sensitive to changes in systolicand diastolic ventricular function, they are not load insensi-tive.29 A contribution of the increased running time (mean 43minutes) and more prolonged exercise to the observed in-crease in biomarkers and echocardiographic abnormalitiesamong the less-prepared cohort cannot be excluded. How-ever, similar to work by others,21 we found that the associa-tion between marathon finish time and the increase introponin was poor (r�0.16). We did not perform follow-upstudies to determine the chronology of the changes in cardiacstructure and function. However, our group previously re-ported follow-up data on a similar cohort of subjects partic-ipating in the 2003 Boston Marathon.30 In that smaller study,some similar, albeit smaller and statistically nonsignificant,changes involving the RV and the LV were detected imme-diately after completion of the marathon. Follow-up echocar-diographs in this cohort, obtained within 3 to 4 weeks ofcompletion of the marathon, demonstrated that indices of

Figure 2. Baseline and postmarathon cTnT in groups stratifiedaccording to training mileage. At baseline, cTnT was undetect-able in any of the groups. After completion of a marathon, agraded cTnT release effect was observed as a function of extentof training, with athletes who trained less (�35 miles/wk beforethe race) demonstrating significantly greater increases in post-race cTnT, compared with athletes who trained with a higherweekly mileage.

Figure 3. Baseline and postmarathon NT-proBNP values ingroups stratified according to training mileage. There was nodifference among groups with different levels of training at base-line. Levels of NT-proBNP were increased in all subgroups withprolonged exercise; however, these increases were exaggeratedin athletes who trained less (�35 miles/wk before the race).




systolic function had normalized, whereas changes in indicesof diastolic function persisted. Early recovery and the tran-sient nature of the echocardiographic changes involving boththe RV and LV have also been reported among other cohortsundergoing more strenuous endurance pursuits.17,31 We alsodid not perform serial measurement of serum biomarkersafter the marathon; however, among marathon runners with asimilar mean finish time to our cohort, Scharhag et almeasured cTnT and cardiac troponin I before the marathonand at 15 minutes, 3 hours, and 24 hours after participation.21

They found that cTnT, measured using an identical assay tothat used in our study, unlike cardiac troponin I, was highestat 15 minutes after completion of a marathon run anddeclined thereafter to undetectable levels at 24 hours. Indeed,numerous studies, like this one, have previously shown thattroponin T concentrations normalize by 24 hours after endur-ance sports.32–34 We did not individually estimate the RApressure, but instead we uniformly added 10 mm Hg. Al-though this may have overestimated the RV systolic pressurein some subjects, an overall increase would still have beenobserved. Also, the increase in the estimated mPAP lendsfurther support to the RV systolic pressure data.

To our knowledge, our study is the first to successfullycorrelate participation in endurance sports with biochemicaland echocardiographic evidence of cardiac injury and dys-function and to demonstrate a strong relationship betweenextent of training and the presence and magnitude of suchcardiovascular abnormalities after marathon running. How-ever, there are no data to suggest that there are long-termsequelae to the increase in biomarkers and echocardiographicevidence of injury in this setting. In contrast, many studiessuggest that endurance exercise is associated with a reductionin cardiovascular risk and an increased life expectancy.7,8 Ourstudy does suggest that, to protect against elevations incardiac biomarkers and echocardiographic evidence of car-diac dysfunction associated with endurance exercise, appro-priate preparation is important.

AcknowledgmentsDr Neilan is supported by an Irish Board for Training in Cardiovas-cular Medicine and Department of Health and Children Cardiovas-cular Health Strategy Travelling Fellowship, an American Society ofEchocardiography Research Fellowship award, and an AmericanHeart Association Post-Doctoral Fellowship Grant. We acknowledgethe assistance of Dr Marvin Adner, medical director of the BostonAthletic Association, for permission to perform the study; Drs AnnaC. Johansson, John G. Morgan, Cynthia Taub, Mordehay Vaturi, andDali Fan for their contributions in data collection; and Susan Philipand Jeff Sirek from GE Healthcare for their equipment loan andtechnical support.

DisclosuresDr Januzzi has received honoraria for speaker’s presentations fromDade, Roche, and Ortho/Johnson&Johnson and an educational grantfrom Roche. Dr Douglas has received honoraria for speaker’spresentations from GE Healthcare. Dr Picard has received researchsupport from GE. Drs Lewandrowski and Lee-Lewandrowski havereceived research support from Roche, and Dr Lewandrowski hasreceived honoraria for speaker’s presentations from Roche. No otherauthors report disclosures.

References1. Leon AS, Connett J, Jacobs DR Jr, Rauramaa R. Leisure-time physical

activity levels and risk of coronary heart disease and death. The MultipleRisk Factor Intervention Trial. JAMA. 1987;258:2388–2395.

2. Paffenbarger RS Jr, Hyde RT, Wing AL, Lee IM, Jung DL, Kampert JB.The association of changes in physical-activity level and other lifestylecharacteristics with mortality among men. N Engl J Med. 1993;328:538–545.

3. Kratz A, Wood MJ, Siegel AJ, Hiers JR, Van Cott EM. Effects ofmarathon running on platelet activation markers: direct evidence for invivo platelet activation. Am J Clin Pathol. 2006;125:296–300.

4. Douglas PS, O’Toole ML, Woolard J. Regional wall motion abnor-malities after prolonged exercise in the normal left ventricle. Circulation.1990;82:2108–2114.

5. Roberts WO, Maron BJ. Evidence for decreasing occurrence of suddencardiac death associated with the marathon. J Am Coll Cardiol. 2005;46:1373–1374.

6. Maron BJ, Poliac LC, Roberts WO. Risk for sudden cardiac death asso-ciated with marathon running. J Am Coll Cardiol. 1996;28:428–431.

7. Williams PT. Relationship of distance run per week to coronary heartdisease risk factors in 8283 male runners. The National Runners’ HealthStudy. Arch Intern Med. 1997;157:191–198.

8. Sarna S, Sahi T, Koskenvuo M, Kaprio J. Increased life expectancy ofworld class male athletes. Med Sci Sports Exerc. 1993;25:237–244.

9. Adams JE 3rd, Bodor GS, Davila-Roman VG, Delmez JA, Apple FS,Ladenson JH, Jaffe AS. Cardiac troponin I: a marker with high specificityfor cardiac injury. Circulation. 1993;88:101–106.

10. Sahn DJ, DeMaria A, Kisslo J, Weyman A. Recommendations regardingquantitation in M-mode echocardiography: results of a survey of echo-cardiographic measurements. Circulation. 1978;58:1072–1083.

11. Yock PG, Popp RL. Noninvasive estimation of right ventricular systolicpressure by Doppler ultrasound in patients with tricuspid regurgitation.Circulation. 1984;70:657–662.

12. Kitabatake A, Inoue M, Asao M, Masuyama T, Tanouchi J, Morita T,Mishima M, Uematsu M, Shimazu T, Hori M, Abe H. Noninvasiveevaluation of pulmonary hypertension by a pulsed Doppler technique.Circulation. 1983;68:302–309.

13. Sohn DW, Chai IH, Lee DJ, Kim HC, Kim HS, Oh BH, Lee MM, ParkYB, Choi YS, Seo JD, Lee YW. Assessment of mitral annulus velocity byDoppler tissue imaging in the evaluation of left ventricular diastolicfunction. J Am Coll Cardiol. 1997;30:474–480.

14. Urheim S, Edvardsen T, Torp H, Angelsen B, Smiseth OA. Myocardialstrain by Doppler echocardiography: validation of a new method toquantify regional myocardial function. Circulation. 2000;102:1158–1164.

15. Apple FS, Quist HE, Doyle PJ, Otto AP, Murakami MM. Plasma 99th

percentile reference limits for cardiac troponin and creatine kinase MBmass for use with European Society of Cardiology/American College ofCardiology consensus recommendations. Clin Chem. 2003;49:1331–1336.

16. Hopkins SR, Gavin TP, Siafakas NM, Haseler LJ, Olfert IM, Wagner H,Wagner PD. Effect of prolonged, heavy exercise on pulmonary gasexchange in athletes. J Appl Physiol. 1998;85:1523–1532.

17. Davila-Roman VG, Guest TM, Tuteur PG, Rowe WJ, Ladenson JH, JaffeAS. Transient right but not left ventricular dysfunction after strenuousexercise at high altitude. J Am Coll Cardiol. 1997;30:468–473.

18. Maron BJ, Shirani J, Poliac LC, Mathenge R, Roberts WC, Mueller FO.Sudden death in young competitive athletes: clinical, demographic, andpathological profiles. JAMA. 1996;276:199–204.

19. Kalliokoski KK, Laaksonen MS, Luotolahti M, Laine H, Takala TO,Nuutila P, Knuuti J. Myocardial perfusion after marathon running. ScandJ Med Sci Sports. 2004;14:208–214.

20. Katus HA, Remppis A, Scheffold T, Diederich KW, Kuebler W. Intra-cellular compartmentation of cardiac troponin T and its release kinetics inpatients with reperfused and nonreperfused myocardial infarction.Am J Cardiol. 1991;67:1360–1367.

21. Scharhag J, Herrmann M, Urhausen A, Haschke M, Herrmann W, KindermannW. Independent elevations of N-terminal pro-brain natriuretic peptide and cardiactroponins in endurance athletes after prolonged strenuous exercise. Am Heart J.2005;150:1128–1134.

22. Neumayr G, Gaenzer H, Pfister R, Sturm W, Schwarzacher SP, Eibl G,Mitterbauer G, Hoertnagl H. Plasma levels of cardiac troponin I afterprolonged strenuous endurance exercise. Am J Cardiol 2001;87:369–371,A10.




23. Usatyuk PV, Natarajan V. Role of mitogen-activated protein kinases in4-hydroxy-2-nonenal-induced actin remodeling and barrier function inendothelial cells. J Biol Chem. 2004;279:11789–1197.

24. Ogawa S, Gerlach H, Esposito C, Pasagian-Macaulay A, Brett J, Stern D.Hypoxia modulates the barrier and coagulant function of cultured bovineendothelium: increased monolayer permeability and induction of proco-agulant properties. J Clin Invest. 1990;85:1090–1098.

25. Wu AH, Ford L. Release of cardiac troponin in acute coronary syn-dromes: ischemia or necrosis? Clin Chim Acta. 1999;284:161–174.

26. Goldhaber SZ, Visani L, De Rosa M. Acute pulmonary embolism:clinical outcomes in the International Cooperative Pulmonary EmbolismRegistry (ICOPER). Lancet. 1999;353:1386–1389.

27. Giannitsis E, Muller-Bardorff M, Kurowski V, Weidtmann B, WiegandU, Kampmann M, Katus HA. Independent prognostic value of cardiactroponin T in patients with confirmed pulmonary embolism. Circulation.2000;102:211–217.

28. Wu AH, Apple FS, Gibler WB, Jesse RL, Warshaw MM, Valdes R Jr.National Academy of Clinical Biochemistry Standards of LaboratoryPractice: recommendations for the use of cardiac markers in coronaryartery diseases. Clin Chem. 1999;45:1104–1121.

29. Jacques DC, Pinsky MR, Severyn D, Gorcsan J 3rd. Influence of alter-ations in loading on mitral annular velocity by tissue Doppler echocar-diography and its associated ability to predict filling pressures. Chest2004;126:1910–1918.

30. Neilan TG, Yoerger DM, Douglas PS, Marshall JE, Halpern EF, LawlorD, Picard MH, Wood MJ. Persistent and reversible cardiac dysfunctionamong amateur marathon runners. Eur Heart J. 2006;27:1079–1084.

31. Douglas PS, O’Toole ML, Hiller WD, Hackney K, Reichek N. Cardiacfatigue after prolonged exercise. Circulation. 1987;76:1206–1213.

32. Apple FS, Quist HE, Otto AP, Mathews WE, Murakami MM. Releasecharacteristics of cardiac biomarkers and ischemia-modified albumin asmeasured by the albumin cobalt-binding test after a marathon race. ClinChem. 2002;48:1097–1100.

33. Herrmann M, Scharhag J, Miclea M, Urhausen A, Herrmann W, KindermannW. Postrace kinetics of cardiac troponin T and I and N-terminal pro-brainnatriuretic peptide in marathon runners. Clin Chem. 2003;49:831–834.

34. Siegel AJ, Lewandrowski EL, Chun KY, Sholar MB, Fischman AJ,Lewandrowski KB. Changes in cardiac markers including B-natriureticpeptide in runners after the Boston marathon. Am J Cardiol. 2001;88:920–923.

CLINICAL PERSPECTIVEIncreased cardiac biomarkers and alterations in cardiac function are well-described sequelae to participation in endurancesports. However, although multiple studies have independently demonstrated these alterations, none have linked theseabnormalities. In fact, to date, convincing associations between the two have not been demonstrated, and risk factors forthe development of signs of myocardial damage or dysfunction are unknown. Therefore, we designed this study todetermine the associations between the observed transient cardiac dysfunction and injury among amateur participants inendurance sports and to elicit the factors associated with the development of this dysfunction and injury. We found thatthe extent and the degree of the transient cardiac injury and dysfunction were significantly influenced by the degree ofpreparation and training. Temporary changes indicating heart stress occurred in marathon runners who trained less than 35miles per week in the months before the event. However, these changes were milder or absent in those who ran more than45 miles per week. The protection afforded by training was independent of age and gender. Notably, however, no datasuggest that there are long-term sequelae to the biomarker and echocardiographic changes in this setting, and indeed, incontrast, many studies suggest that endurance exercise is associated with a reduction in cardiovascular risk and an increasedlife expectancy. Our study does suggest that, to protect against elevations in cardiac biomarkers and echocardiographicevidence of cardiac dysfunction associated with endurance exercise, appropriate preparation is important.




Pamela S. Douglas, David Lawlor, Michael H. Picard and Malissa J. WoodM. Yoerger, Davinder S. Jassal, Kent B. Lewandrowski, Arthur J. Siegel, Jane E. Marshall,

Tomas G. Neilan, James L. Januzzi, Elizabeth Lee-Lewandrowski, Thanh-Thao Ton-Nu, DanitaNonelite Participants in the Boston Marathon

Myocardial Injury and Ventricular Dysfunction Related to Training Levels Among

Print ISSN: 0009-7322. Online ISSN: 1524-4539 Copyright © 2006 American Heart Association, Inc. All rights reserved.

is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Circulation doi: 10.1161/CIRCULATIONAHA.106.647461

2006;114:2325-2333; originally published online November 13, 2006;Circulation.

http://circ.ahajournals.org/content/114/22/2325World Wide Web at:

The online version of this article, along with updated information and services, is located on the

http://circ.ahajournals.org//subscriptions/

is online at: Circulation Information about subscribing to Subscriptions:

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

document. Permissions and Rights Question and Answer this process is available in the

click Request Permissions in the middle column of the Web page under Services. Further information aboutOffice. Once the online version of the published article for which permission is being requested is located,

can be obtained via RightsLink, a service of the Copyright Clearance Center, not the EditorialCirculationin Requests for permissions to reproduce figures, tables, or portions of articles originally publishedPermissions:


http://circ.ahajournals.org/content/114/22/2325

http://www.ahajournals.org/site/rights/

http://www.lww.com/reprints

http://circ.ahajournals.org//subscriptions/


Myocardial Injury and Ventricular Dysfunction Related to Training Levels Among Nonelite Participants...

Documents

Transcript of Myocardial Injury and Ventricular Dysfunction Related to Training Levels Among Nonelite Participants...