A

(ima4c©

K

tmsfhcbmwp

fsuet3mlmg

1d

Available online at www.sciencedirect.com

Journal of Science and Medicine in Sport 13 (2010) 475–478

Original paper

Influence of sex on the “Athlete’s Heart” in trained cyclists

Thomas Rowland a,∗, Melissa Roti b

a Department of Pediatrics, Baystate Medical Center, United Statesb Department of Movement Science, Westfield State College, United States

Received 2 June 2009; received in revised form 22 October 2009; accepted 26 October 2009

bstract

Compared to females, male endurance athletes have generally been considered to demonstrate greater values of cardiac mass and volumethe “athlete’s heart”). However, studies addressing this issue have frequently failed to match training volumes between groups or providednadequate adjustment of variables for body size and composition. This study compared echocardiographic anatomic features in 8 female and 8

ale competitive cyclists with a similar training history. Conforming to most previous reports, left ventricular mass and end diastolic dimension,djusted for fat free body mass and body surface area, respectively, were greater in the males (3.56 ± 0.83 g kg−1 versus 2.50 ± 0.38 g kg−1;1.7 ± 2.2 mm and 37.4 ± 2.5 mm per BSA0.5). This study indicated that when training volume as well as body size and composition areonsidered, male endurance athletes exhibit greater cardiac dimensions and mass compared to their female counterparts.

ights r

mb

(eeVtauivmibmsB

2009 Sports Medicine Australia. Published by Elsevier Ltd. All r

eywords: Cardiac hypertrophy; Athletes Echocardiography

The metabolic demands of endurance exercise trainingrigger parallel adaptations of enhanced skeletal muscle

etabolic capacity (mitochondrial biogenesis) and expan-ion of the cardiovascular system. Cardiac anatomicaleatures of this response—collectively termed the “athlete’seart”—include augmented ventricular volume and increasedardiac mass.1,2 In conjunction with increases in circulatinglood volume, these changes are reflected in greater maxi-al stroke volume, cardiac output and oxygen uptake (VO2),hich are in turn translated into improvements in enduranceerformance.

Certain observations have indicated that cardiac structuraleatures in highly endurance trained women are less con-picuous than those of their male counterparts.3–6 Among 10niversity female endurance athletes, for instance, Georget al. found mean left ventricular mass (expressed relativeo fat free body mass, FFM) to be 2.80 g kg−1 compared to.50 g kg−1 in male athletes.3 Among 600 female and 738ale athletes in 27 sports, Pellicia et al. reported ventricu-

ar mass indexed to body surface area to be 31% greater inales.4 However, Morales et al. could find no significant

ender-related difference in ventricular mass in training-

∗ Corresponding author.E-mail address: thomas.rowland@bhs.org (T. Rowland).

bctsvp

440-2440/$ – see front matter © 2009 Sports Medicine Australia. Published by Eloi:10.1016/j.jsams.2009.10.488

eserved.

atched distance runners when expressed either to FFM orody surface area (BSA).7

In making group comparisons by sex it is necessary thata) training volume be equal, and (b) cardiac measures bexpressed relative to appropriate variables that effectivelyliminate the influences of both body size and composition.8,9

entricular mass and volume are presumed to relate closely tohe volume of metabolically active tissue (i.e., for endurancethletes, the mass of contracting skeletal muscle), whichnfortunately is impossible to measure and varies accord-ng to sport activity. Consequently, various anthropometricariables have been utilised as surrogate measures. Fat freeass (lean body mass), in absolute terms or expressed by

ts allometric power function, has been considered to be theest proxy for metabolically active muscle and provides theost accurate means of adjusting variables such as cardiac

ize and mass in athletic males and females. Height2.7 andSA1.5 have also been utilised.10,11

Comparisons in which cardiac variables are related toody mass (the ratio standard) clearly will lead to spuriousonclusions if groups differ significantly in body composi-

ion. The value of indexing cardiac measurements to BSAurface area, a traditional means of “normalising” theseariables to body size, depends on the nature and body com-osition homogeneity of the subject population. For instance,

sevier Ltd. All rights reserved.

4 ce and

twf

toSla

1

yttvppctits

tditaiTUo2

cctsFB

rm5siTtcQvat

wwt(

aaeFfatwLttF

tt8t(emSghcpb>

vpa

piwM

2

wa1fs6

76 T. Rowland, M. Roti / Journal of Scien

he muscular male Olympic power lifter and the obese youngoman might have equal body surface areas but clearly dif-

ering amounts of metabolically active muscular tissue.This study matched two groups of competitive cyclists by

raining and competition history to further delineate effectsf sex on the anatomic characteristics of the “athlete’s heart.”tandard echocardiographic determinations of left ventricu-

ar structural features in the supine position were related toppropriate measures of body size and composition.

. Methods

Eight highly trained female cyclists (mean age 30.4 ± 1.6ears) were recruited for cardiac assessment and exerciseesting. All were in good health and taking no medicationshat would influence cardiac performance. Anthropometricalues, cardiac dimensions and mass, and maximal aerobicower during a progressive upright cycle exercise were com-ared to those of a group of eight highly competitive maleyclists (mean age 33.7 ± 2.9 years). Some of the data fromhe men have been included in an assessment of differencesn cardiovascular responses to exercise from those of non-rained adults males.12 No athletes in either group reportedignificant participation in other sports or resistance training.

Training and competition histories as obtained by ques-ionnaire were comparable in the two groups. Femalesescribed an average of 8.6 years (range 3–18) of compet-tive cycle training, compared to 9.9 years (range 6–15) inhe males. Number of months of the year engaged in trainingveraged 10 (range 7–12) in the females and 11 (range 8–12)n the men, with average five training rides a wk for both.he distribution of racing categories, as designated by the.S. Cycling Federation, was identical for the two groups:ne cyclist in each group was category 1, two were category, three category 3, and two category 4.

Height and weight were measured by stadiometer andalibrated balance-beam scale, respectively. Scapular and tri-eps skinfolds were averaged from three measurements onhe right side. Percent body fat was derived from the sex-pecific skinfold equations of McArdle, Katch, and Katch.13

at free mass (FFM) was then calculated as body mass − (%F × body mass).

Left ventricular dimensions and wall thickness wereecorded by echocardiographic two-dimensional-guided M-ode techniques with subjects in the supine position (Sonos

500, Philips Healthcare, Andover, MA). Triplicate mea-urements were averaged in the parasternal long axis viewmmediately distal to the tips of the mitral valve leaflets.he left ventricular dimension (LVED) was measured from

he trailing edge of the ventricular septum to the endo-ardial surface of the posterior wall coincident with the

wave of the electrocardiogram (end diastole). The leftentricular end-systolic dimension (LVES) was recordeds the shortest vertical distance from the ventricular sep-um to the endocardial surface of the ventricular posterior

((f(

Medicine in Sport 13 (2010) 475–478

all during systole. Ventricular septal (LVS) and posteriorall (LVPW) thicknesses were measured at end dias-

ole. Left ventricular shortening fraction was calculated asLVED − LVES)LVED × 100.

The single-dimension measures of LVED, LVES, LVS,nd LVPW were indexed to the square root of body surfacerea (BSA).14 Left ventricular end diastolic size was alsoxamined as estimated LV volume (LVED3) normalised toFM. Left ventricular mass (LVM) was calculated by the cubeormula 1.04 [(LVED + LVPW + LVS)3 − LVED3] − 14 withll values in centimeters.15 LVM was expressed relativeo BSA, BSA1.5, height2.7, and FFM, as well as FFMb,here b is the scaling exponent in the allometric equationVM = aFFMb. The latter was determined by log transforma-ion of both LBM and FFM, with least squares regression usedo solve for values of b in the equation log LVM = log a + b logFM.

For determination of aerobic fitness, the athletes cycled inhe upright position with a progressive protocol to exhaus-ion on a mechanically braked Monark ergometer (model18; Varberg, Sweden). Stage duration was 3 min with ini-ial and incremental loads of 50 W with a cadence of 50 rpmin the last stages cadence was increased to 70–80 rpm). Gasxchange variables were measured with a Q-Plex Cardiopul-onary Exercise System (Quinton Instrument Company,eattle) using standard open circuit techniques. Peak oxy-en uptake (peak VO2) was defined as the mean of the twoighest 15-s average values recorded in the final min of exer-ise. A peak effort was established as the point when theedaling cadence could no longer be maintained, supportedy subjective evidence of fatigue as well as a peak heart rate170 bpm or maximal respiratory exchange ratio >1.10.

Differences in anthropometric and echocardiographicariables between the men and women were analysed by inde-endent Student’s t-test. Statistical significance was defineds p < 0.05.

This study was performed in accordance with ethicalrinciples set forth in the Declaration of Helsinki. Writtennformed consent was obtained from all subjects. This studyas approved by the institutional review board of the Baystateedical Center.

. Results

The male cyclists were larger than the females, with meaneight 69.2 ± 5.7 kg versus 58.9 ± 5.8 kg, height 177 ± 6

nd 168 ± 6 cm, and body surface area 1.85 ± 0.10 and.67 ± 0.10 m2 (for all comparisons p < 0.05). Percent bodyat was significantly greater in the females (16.1 ± 2.2% ver-us 10.0 ± 2.5%, respectively). Average calculated FFM was1.7 ± 5.6 kg in the males and 49.4 ± 4.5 kg in the females

p < 0.05). Maximal aerobic power was greater in the males5.10 ± 0.59 L min−1, 73.7 ± 7.0 mL kg−1 min−1) than theemales (3.51 ± 0.31 L min−1, 59.8 ± 4.5 mL kg−1 min−1)p < 0.05).

T. Rowland, M. Roti / Journal of Science and

Table 1Left ventricular resting supine echocardiographic measures in male andfemale cyclists. Values are mean (standard deviation).

Men Women

LVED (mm) 56.6 (2.6) 48.4 (3.4)*

LVED (mm per BSA0.5) 41.7 (2.2) 37.4 (2.5)*

LVES (mm) 34.7 (2.8) 30.6 (3.5)LVES (mm per BSA0.5) 25.7 (2.7) 23.7 (2.7)VSd (mm) 9.0 (0.7) 7.7 (1.1)VSd (mm per BSA0.5) 6.8 (0.7) 6.0 (0.9)LVPWd (mm) 8.1 (1.6) 6.5 (0.9)LVPWd (mm per BSA0.5) 6.0 (1.2) 5.1 (0.7)Shortening fraction (%) 37.8 (4.8) 36.1 (5.1)LVED3 (cm3 FFM kg−1) 2.95 (0.45) 2.33 (0.45)LVM (g) 220 (51) 124 (16)*

LVM (g per BSA1.0) 120 (28) 74 (10)*

LVM (g per BSA1.5) 65 (15) 39 (5)*

LVM (g ht−2.7) 49.1 (11.2) 28.4 (3.7)*

LVM (g FFM kg−1) 3.56 (0.83) 2.50 (0.38)*

LVM (g FFM kg−0.29) 12.8 (2.7) 9.0 (1.3)*

LVED = left ventricular end diastolic dimension; LVES = left ventricularend-systolic dimension; VSd = ventricular septum, diastole; LVPWd = leftvL

saawtgsru(ilwtw

rar(Fw

3

ocmgm

tamasfec

asLtmio1rr

esvsmam

devetetpase

vCfavlfinm

p

entricular posterior wall, diastole; BSA = body surface area (m2);VM = left ventricular mass (g); ht = height (m); FFM = fat free mass.* p < 0.05 males versus females.

Echocardiographic measures of left ventricular dimen-ions relative to various variables adjusting for body sizend composition are outlined in Table 1. Both in absolutend size-normalised terms, values of LVED and LV massere significantly greater in males. Adjusted values of ven-

ricular septal thickness and posterior wall thickness werereater in the males but gender differences barely escapedtatistical significance (p = 0.06 and 0.10, respectively). Theatios of left ventricular end diastolic dimension to ventric-lar septal thickness (6.1, 6.2) and posterior wall thickness7.0, 7.3) in the males and females, respectively, were sim-lar. No significant sex differences were observed in restingeft ventricular shortening fraction. Group average relativeall thicknesses (ratio of average of septal and posterior wall

hickness to LVED) were 0.132 and 0.130 for the men andomen, respectively (p > 0.05).The allometric scaling exponent b for FFM was 0.29 in

espect to LVM. Correlation coefficients for normalising vari-bles for LVM for the combined groups were FFM (kg)= 0.69, FFM (kg0.29) r = 0.62, BSA (m1.0) r = 0.58, BSAm1.5) r = 0.61, and height (m2.7) r = 0.62. Thus, FFM1.0 andFM0.29 were observed to provide the greatest relationshipith LVM.

. Discussion

Valid comparisons of structural and functional featuresf the athlete’s heart between men and women are highly

ontingent upon several considerations: (a) the type of sportust be similar2, (b) since females can be expected possess

reater percent body fat and relatively less muscle mass thanales, values of variables which relate to active metabolic

amuo

Medicine in Sport 13 (2010) 475–478 477

issue (ventricular mass, volume) must be adjusted by anppropriate normalising factor,8,9 and (c) the groups must beatched for training history (duration, frequency, intensity)

nd competitive level. In conforming to these requisites, thistudy found that size- and composition-adjusted morphologiceatures of the athlete’s heart, measured by supine restingchocardiogram, were larger in male compared to femaleompetitive cyclists.

The finding of greater echocardiographic measures ofdjusted heart mass and volume among male cyclists in thistudy is consistent with most previous reports.3–6 AbsoluteVM was 220 ± 28 g in males compared to 124 ± 16 g inhe females, and 30–37% greater values were observed in the

ales when LVM was related to the most appropriate normal-sing factor (FFM or FFM0.29). Similarly, adjusted LVED (ane-dimensional indicator of LV end diastolic volume) was2% larger in the males. These values are similar to thoseeported by George et al. in sex comparisons of distanceunners and cross-country skiers.3

These sex differences are presumably related to the differ-ntial effect of estrogen and testosterone on myocyte proteinynthesis, as triggered by myocardial stretch, increasedentricular pressure, and tachycardia which accompaniesports training.1,16 While testosterone serves to enhanceyocardial hypertrophy,17,18 animal studies clearly indicate

n anti-hypertrophic effect of estrogen by several differentechanisms.19–21

Other determinants might play central roles in genderifferences in structural cardiac modifications observed inndurance athletes. Perrault and Turcotte argued that theentricular enlargement observed in these athletes is bestxplained by increased blood volume in combination withraining-induced resting bradycardia.22 By this argument,ccentric hypertrophy during training is a secondary responseo minimise wall stress by LaPlace’s Law as increases inlasma volume cause ventricular dilatation. Others havegreed that training-induced increases in blood volume mighterve as the primary stimulus for the athlete’s ventricularnlargement and hypertrophy.23,24

Potential sex-related differences in responses of bloodolume expansion with sports training are thus of interest.onvertino reviewed studies addressing this issue,23 with the

ollowing conclusions: (a) highly trained male and femalethletes both typically demonstrate a 20–25% greater bloodolume (when expressed per kg body weight) than nonath-etes, (b) male athletes have higher blood volumes thanemale athletes, (c) the percent increase in blood volumen training studies is similar in both sexes, and (d) amongonathletes, weight-normalised blood volume is greater inales.This information regarding blood volume raises the

ossibility that quantitative differences in the anatomic char-

cteristics of the athlete’s heart between men and womenight simply reflect similar differences observed in the

ntrained state.5 The latter, in turn, may be manifestationsf differential effects of specific sex hormones on blood vol-

4 ce and

uvu

ceisaoutmf

4

itstwwth

P

•

•

C

i

R

testosterone propionate on cardiac output, renal blood flow, and bloodvolume in mice. Proc Soc Exp Biol Med 1973;144:671–3.

78 T. Rowland, M. Roti / Journal of Scien

me. Both estrogen and testosterone act to increase bloodolume,25,26 but the comparative magnitude of this effect isnknown.

This investigation examined structural features of adultyclists from a cross-sectional perspective, and thus thextent that these findings represent effects of athletic train-ng versus genetic pre-selection is problematic. Detrainingtudies in competitive athletes indicate a decline in cardiacnatomic and performance variables, but even with cessationf training these values remain at the upper extremes of thentrained population.24 Thus, both influences can be assumedo contribute. But whether sex might influence the relative

agnitude of training and genetic remains a challenge foruture investigators.

. Conclusion

In summary, this study confirmed findings in previousnvestigations that, when adjusted for body size and composi-ion, the cardiac dimensions and mass of the athlete’s heart inupine highly trained female endurance athletes are less thanhose of comparably trained males. Gaining insights in factorshich define sex-related differences in cardiac size and massith exercise training may provide a better understanding of

he basic mechanisms that surround the process of myocardialypertrophy, both in athletic and clinical populations.

ractical implications

Male endurance athletes are characterised by greater car-diac volume and mass than comparably trained females,even when sex differences in body composition and sizeare considered.Physicians assessing athletes for possible heart diseaseneed to be cognisant that significant myocardial thick-ening in females is more likely to present a true cardiacabnormality than in males.

onflict of interest

The authors have no conflict of interest in equipment usedn this study.

eferences

1. George KP, Wolfe LA, Burggraf GW. The ‘athletic heart syndrome’. Acritical review. Sports Med 1991;11:300–31.

2. Naylor LH, George K, O’Driscoll G, Green DJ, et al. The athlete’s heart:

a contemporary appraisal of the ‘Morganroth hypothesis’. Sports Med2008;38:69–90.

3. George KP, Wolfe LA, Burggraf GW, Norman R, et al. Electrocardio-graphic and echocardiographic characteristics of female athletes. MedSci Sports Exerc 1995;27:1362–70.

Medicine in Sport 13 (2010) 475–478

4. Pellicia A, Maron BJ, Culasso F, Spataro A, Caselli G, et al. Athlete’sheart in women. JAMA 1996;276:211–5.

5. Mumford M, Prakash R. Electrocardiographic and echocardiographiccharacteristics of long distance runners. Am J Sports Med 1981;9:23–8.

6. Whyte GP, George K, Sharma S, Firozzi S, Stephens A, Senior R,et al. The upper limit of physiological cardiac hypertrophy in elitemale and female athletes: the British experience. Eur J Appl Physiol2004;92:592–7.

7. Morales MC, Gleim GW, Marino ND, Glace BW, Coplan NL, etal. The role of gender in echocardiographically determined left ven-tricular mass in equally trained populations of runners. Am Heart J1992;124:1104–6.

8. George KP, Batterham AM, Jones B. The impact of scalar variable andprocess on athlete–control comparisons of cardiac dimensions. Med SciSports Exerc 1998;30:824–30.

9. Batterham AM, George KP, Whyte G, Sharma S, McKenna W, et al.Scaling cardiac structural data by body dimensions: a review of theory,practice, and problems. Int J Sports Med 1999;20:495–502.

10. Daniels SR, Kimball TR, Morrison JA, Khoury P, Witt S, Meyer RA, etal. Effect of lean body mass, fat mass, blood pressure, and sexual matu-ration on left ventricular mass in children and adolescents. Circulation1995;92:3249–54.

11. De Simone G, Daniels SR, Devereux RB, Meyer RA, Roman MJ, deDivitiis O, et al. Left ventricular mass and body size in normotensivechildren and adults: assessment of allometric relations and impact ofoverweight. J Am Coll Cardiol 1992;20:1251–60.

12. Rowland T, Roti M. Cardiac responses to progressive upright exercisein adult male cyclists. J Sports Med Phys Fitness 2004;44:179–85.

13. McArdle WD, Katch KI, Katch VL. Exercise physiology. Philadelphia:Lea & Febiger; 1981. p. 379.

14. Gutgesell HP, Rembold CM. Growth of the human heart relative tobody surface area. Am J Cardiol 1990;65:662–8.

15. Devereux RB, Reichek N. Echocardiographic determination of left ven-tricular mass in man: anatomic validation of the method. Circulation1977;55:613–8.

16. Lips DJ, de Windt LJ, van Kraaij DJW, Doevendans PA, et al. Molec-ular determinants of myocardial hypertrophy and failure: alternativepathways for beneficial and maladaptive hypertrophy. Eur Heart J2003;24:883–96.

17. Van Eickels M, Grohe C, Cleutjens JPM, Janssen BJ, Wellens HJJ,Doevendans PA, et al. 17�-Estradiol attenuates the development ofpressure-overload hypertrophy. Circulation 2001;104:1419–23.

18. Pelzer T, Shamim A, Wolfges S, Schumann M, Neyes L, et al.Modulation of cardiac hypertrophy by estrogens. Adv Exp Med Biol1997;432:83–9.

19. Babiker FA, De Windt LJ, van Eickels M, Grohe C, Meyer R, Doeven-dans PA, et al. Estrogenic hormone action in the heart: regulatorynetwork and function. Cardiovasc Res 2002;53:709–19.

20. Schaible TF, Malhotra A, Ciambrione G, Scheuer J, et al. The effectsof gonadectomy on left ventricular function and cardiac contractileproteins in male and female rats. Circ Res 1984;54:38–49.

21. Behrendt H. Effect of anabolic steroids on rat heart muscle cells. CellTissue Res 1977;180:303–15.

22. Perrault HM, Turcotte RA. Do athletes have “the athlete heart?”. ProgPediatr Cardiol 1993;2:40–50.

23. Convertino VA. Blood volume: its adaptation to endurance training.Med Sci Sports Exerc 1991;23:1338–48.

24. Ehsani AA, Spina RJ. Loss of cardiovascular adaptations after physicalinactivity. Cardiol Clin 1997;15:431–8.

25. Broulik PD, Kochakain CD, Dubovsky J. Influence of castration and

26. Stachenfield NS, Taylor HS, Keefe DL. Mechanisms for estrogenand progesterone effects on plasma volume. Med Sci Sports Exerc2003;35:S198 [abstract].