Physiologic Left Ventricular Cavity Dilatation in Elite Athletes

A. Pelliccia,F. Culasso,F. D. Di Paolo,Barry I. Maron

DOI: https://doi.org/10.7326/0003-4819-130-1-199901050-00005

IF: 39.2

1999-01-05

Annals of Internal Medicine

Abstract:Long-term athletic training is associated with cardiac morphologic changes, including increased left ventricular cavity dimension, wall thickness, and calculated mass, that are commonly described as athlete's heart (1-4). These changes seem to represent adaptations to the hemodynamic load produced by long-term, frequent, intensive exercise programs (5-7). The extent to which absolute left ventricular cavity dimension is increased by systematic training is modest in most athletes (3) but may be more substantial in others, raising the clinical dilemma of distinguishing athlete's heart from structural heart disease (8). This differential diagnosis has particularly important implications because the identification of certain cardiovascular diseases may constitute the basis for disqualifying an athlete from competition in an effort to minimize the risk for sudden cardiac death or disease progression (9). These considerations prompted us to use echocardiography to systematically evaluate the morphologic characteristics and physiologic limits of left ventricular cavity enlargement associated with intensive and long-term athletic conditioning in a large sample of highly trained athletes. Methods Study Sample The Institute of Sports Science, Rome, Italy, is a medical division of the Italian National Olympic Committee and is responsible for the physiologic and medical evaluation of competitive athletes who form the Italian national teams from which Olympic athletes are selected every 4 years. Since 1963, in accordance with the statutes of the Italian National Olympic Committee, it has been mandatory for all junior and senior national team members to undergo a preparticipation medical evaluation, usually on an annual basis (10). This program has routinely included a cardiovascular evaluation with history, physical examination, 12-lead and exercise electrocardiography, chest radiography, and (since 1985) echocardiography. Between January 1992 and December 1993, 1313 apparently healthy athletes were routinely evaluated as part of this program. Four of these athletes were subsequently excluded from the study sample because of cardiac abnormalities that could alter left ventricular cavity dimensions (bicuspid aortic valve with regurgitation in 3 athletes and mitral valve prolapse with regurgitation in 1 athlete). Thus, the final study sample comprised 1309 athletes who were judged to be free of structural cardiovascular or systemic disease. Blood pressure was consistently or predominantly less than 140/90 mm Hg in all participants. Athletes participated in 38 different sports (Table 1). Their ages ranged from 13 to 59 years (mean age, 24.3 6.0 years); 957 (73%) were male. Physical characteristics were as follows: body weight, 33 to 135 kg (mean, 71.6 14.9 kg); height, 139 to 214 cm (mean, 175.9 11.1 cm); and body surface area, 1.13 to 2.67 m2 (mean, 1.86 0.25 m2). Selected data from 267 of these athletes, all women, were presented as part of a previous analysis (11). Table 1. Demographic and Morphologic Characteristics of 1309 Athletes according to the Sports in Which They Participated All athletes had participated in vigorous training programs for substantial periods ranging from 2 to 34 years (mean, 7 years). Of the 1309 athletes, 340 (26%) had achieved international recognition in the Olympics or world championships; the remaining 969 competed at the national level. Echocardiography Two-dimensional and Doppler echocardiographic studies were performed by using a commercially available Hewlett-Packard (Andover, Massachusetts) instrument (Sonos 1500) with a 3.5-MHz transducer. Images of the heart were obtained in multiple cross-sectional planes by using standard transducer positions (12). M-mode echocardiograms were derived from two-dimensional images under direct anatomic visualization and were recorded at 100 mm/s. Measurements of end-diastolic and end-systolic left ventricular cavity dimensions, as well as anterior ventricular septal and posterior free-wall thicknesses, were obtained from the M-mode echocardiogram, in accordance with recommendations of the American Society of Echocardiography (13). To increase the accuracy of these measurements by reducing the possibility of erroneous inclusion of trabeculations or other overlying structures within the septal or free-wall thicknesses, measurements obtained from the M-mode echocardiogram were verified with the two-dimensional images. Posterior ventricular septal and anterolateral free-wall thicknesses were assessed from two-dimensional stop-frame images, as described elsewhere (14). Left ventricular mass was calculated from end-diastolic wall thicknesses and cavity dimension by using the formula proposed by Devereux (15). Left ventricular cavity dimension was normalized to body surface area and height, as suggested by de Simone and colleagues (16). Relative wall thickness was calculated as the ratio of the average of ventricular septal and posterior free-wall thicknesses to the radius of the internal ventricular cavity (17). Characteristics of left ventricular filling were obtained by using pulsed Doppler echocardiography, as described elsewhere (18, 19). Because the focus of this study was to determine the frequency with which athletes show left ventricular cavity enlargement in a range compatible with pathologic conditions, we relied primarily on absolute cavity dimensions (rather than values normalized to body surface area or other variables [16, 20, 21]) and arbitrarily chose a cut-off value of 60 mm for end-diastolic cavity dimension. This value exceeds the 95% prediction limits for normal left ventricular cavity dimension independent of the age, sex, height, or body surface area of most normal persons (21-23) and has been used clinically as a partition value for the detection of left ventricular cavity enlargement and hypertrophy (24, 25). Echocardiograms were obtained in each athlete during periods of intense training; however, because of unavoidable practical considerations, echocardiography could not always be performed when athletes were at the peak of their conditioning. When serial studies were available from the same athlete, the echocardiogram obtained at peak conditioning was used. Interobserver variability in measurements of left-ventricular end-diastolic cavity dimension was assessed in a subset of 35 athletes randomly selected from our study participants. Measurements were made by two of the authors independently and without knowledge of the identity of the participants. Concordance between the two observers was assessed by using the unpaired Student t-test and intraclass correlation. Statistical Analysis Data are expressed as the mean SD. Differences between means were assessed by using unpaired or paired Student t-tests where appropriate. A two-tailed P value less than 0.05 was considered statistically significant. We used multiple regression to analyze several continuous variables (height, weight, body surface area, age, systolic and diastolic blood pressure, and heart rate). Only continuous variables that significantly correlated with left ventricular cavity dimension were subsequently used in the covariance analysis, together with sex and type of sport (categorical variables). These two categorical variables were coded by using a series of (N 1) binary dummy variables, and their effect on left ventricular cavity dimension (that is, regression coefficients) was assessed after removing the effect of the continuous covariates. To evaluate the effect of type of sport and sex, we selected table tennis as the reference sport (because table tennis players had the smallest left ventricular cavity dimension) and female as the reference sex. We then calculated the regression coefficients for type of sport and both sexes. Subsequently, we assessed the relative effect of each type of sport on left ventricular cavity dimension by modifying the coefficients (impact factors) by the formula (I j I ref), where I j is the impact factor of the jth sport and I ref is the impact factor of the reference sport. These impact factors were presented in rank order from the lowest, set as zero (table tennis), to the highest (cycling). In addition, we clustered individual sports into three subgroups with high ( 4.4), medium (4.3 to 2.0), and low (<2.0) impact on the basis of the relative sport impact factors. Likewise, patients were divided into three subgroups for each sex according to body surface area: low ( 1.80 m2), medium (1.81 to 2.00 m2), and high (>2.00 m2) in men and low ( 1.50 m2), medium (1.51 to 1.70 m2), and high (>1.70 m2) in women. Mean values and SDs for left ventricular cavity dimension were stratified according to these subgroups of sport impact and body surface area in male and female athletes. To reduce the potential for false-positive results due to the large number of sport types, we assessed the significance of differences in mean left ventricular cavity dimensions in the different sports by using Bonferroni and Tukey multiple tests (26, 27). Finally, we assessed the significance of each variable in the covariance analysis by using the Wald test (26). Role of Study Sponsor This study was supported by the Italian National Olympic Committee and the Minneapolis Heart Institute Foundation. The funding sources had no impact on the gathering and interpretation of the results or on whether the report would be published. Results Cardiac Dimensions Left Ventricular Cavity In 725 of the 1309 participants (55%), the left ventricular cavity dimension was within generally accepted normal limits ( 54 mm) (21, 24). However, it exceeded the upper normal limits ( 55 mm) in the other 584 athletes (45%) and was substantially enlarged ( 60 mm) in 185 athletes (14% of the overall group). The end-diastolic left ventricular cavity dimension ranged from 38 to 66 mm (mean, 48.4 4.2 mm; 95th percentile value, 56 mm) in women and 43 to 70 mm (mean, 55.5 4.3 mm; 95th percentile val