Medicine & Science in Sports & Exercise:
APPLIED SCIENCES: Physical Fitness and Performance
Growth in Peak Aerobic Power during Adolescence
GEITHNER, CHRISTINA A.1; THOMIS, MARTINE A.2; EYNDE, BAVO VANDEN2; MAES, HERMINE H. M.3; LOOS, RUTH J. F.2,4; PEETERS, MAARTEN2; CLAESSENS, ALBRECHT L. M.2; VLIETINCK, ROBERT5; MALINA, ROBERT M.6; BEUNEN, GASTON P.2
1 Department of Exercise Science, Gonzaga University, Spokane, WA; 2 Center For Physical Development Research; Faculty of Kinesiology and Rehabilitation Sciences; K. U. Leuven, BELGIUM; 3 Virginia Institute for Psychiatric and Behavioral Genetics, Department of Human Genetics, Virginia Commonwealth University, Richmond, VA; 4 Pennington Biomedical Research Center, Human Genomics Laboratory, Baton Rouge, LA; 5 Center for Human Genetics; Faculty of Medicine, K. U. Leuven, BELGIUM; and 6 Tarleton State University, Stephenville, TX
Address for correspondence: Christina A. Geithner, Ph.D., Department of Exercise Science, Gonzaga University, 502 E. Boone Avenue, Spokane, WA 99258; E-mail: email@example.com.
Submitted for publication October 2003.
Accepted for publication May 2004.
GEITHNER, C. A., M. A. THOMIS, B. VANDEN EYNDE, H. H. M. MAES, R. J. F. LOOS, M. PEETERS, A. L. M. CLAESSENS, R. VLIETINCK, R. M. MALINA, and G. P. BEUNEN. Growth in Peak Aerobic Power during Adolescence. Med. Sci. Sports Exerc., Vol. 36, No. 9, pp. 1616–1624, 2004.
Purpose: To model the growth of peak aerobic power during adolescence in both sexes followed longitudinally from 10 to 18 yr.
Methods: Peak aerobic power (peak V̇O2) was measured annually during a maximal treadmill test with the Bruce protocol. Height and weight were measured semiannually. The Preece-Baines Model I growth function was used to fit curves to data for individuals with ≥ six observations for peak aerobic power to estimate age at peak velocity (PV) for peak V̇O2 (age at PVPV̇O2), PVPV̇O2 (L·min−1·yr−1), and value at PVPV̇O2 (L·min−1) for each individual. Curves were successfully fitted for 83 individuals (48 males, 35 females). The model was also fitted to individual data for height and weight to estimate ages at peak height velocity (PHV) and peak weight velocity (PWV). Age at PVPV̇O2 was compared with ages at PHV and PWV. Pearson correlation coefficients were calculated between ages at PV and PV for peak V̇O2, height, and weight.
Results: Mean ages at PVPV̇O2 are 12.3 ± 1.2 yr for females and 14.1 ± 1.2 yr for males. Peak V̇O2 increases in both sexes throughout adolescence, with males having higher values than females at all ages. Age at PVPV̇O2 occurs nearly coincident with PHV and before PWV in both sexes. Correlation coefficients among ages at PHV, PWV, and PVPV̇O2 suggest a general maturity factor for body size and aerobic power.
Conclusion: Growth in peak V̇O2 exhibits a clear growth spurt in both sexes during adolescence. The growth spurt occurs earlier in females but is of greater magnitude in males.
Aerobic power is an important component of health-related physical fitness and is viewed as the primary indicator of cardiorespiratory fitness. Peak V̇O2 is the generally accepted term used when measuring or discussing maximal aerobic power in children and adolescents (2,3).
Longitudinal studies of maximal aerobic power in European adolescents are limited largely to boys and ordinarily include rather small samples (7,24,26,28). Most of the studies are short term and do not ordinarily span the entire period of the adolescent growth spurt, i.e., acceleration, maximum growth or peak velocity, and deceleration. This limitation may not permit an adequate description of adolescent growth of peak V̇O2and, more specifically, the growth curve for this indicator of cardiorespiratory fitness. Available longitudinal studies of peak V̇O2 vary in methodology, specifically, cycle ergometer or treadmill, and method of estimating peak velocity of growth in aerobic power. A related factor is habitual activity and/or systematic aerobic training during adolescence, which results in an increase in peak V̇O2 (14,20). There is considerable interest in the improvement of aerobic power with training, but it is difficult to partition the changes associated with training from those that accompany normal adolescent growth and maturation, or interactions among training, growth, and maturation (17,19). Longitudinal rather than cross-sectional data and analyses are needed to address these issues (2,4,20).
Cross-sectional studies and cross-sectional analyses of longitudinal data indicate that peak V̇O2 increases gradually and progressively from about 8 to 16 yr in boys and 8 to 13 yr in girls. Peak V̇O2 tends to show a plateau after 13 yr in girls and increases at a slower rate after 16 yr in boys (2,14). Longitudinal studies show a growth spurt, i.e., a nonlinear increase, in peak V̇O2 during adolescence in both sexes (13,19,23). The shape of the adolescent growth curve in peak V̇O2 is similar to that for height. However, peak V̇O2 continues to increase after maximal rate of growth in height is attained (peak height velocity, PHV). The trend is less consistent in girls, although some longitudinal data indicate that peak V̇O2 continues to increase for several years after PHV (2,3,13,19).
The estimated maximum rate of growth (peak velocity, PV) in peak V̇O2 (PVPV̇O2, L·min−1·yr−1) during the adolescent growth spurt occurs near the time of PHV and peak weight velocity (PWV) (7,10,13,19,23,26). Although most studies are limited to boys, PVPV̇O2 occurs earlier in girls, but its magnitude is greater in boys. Studies vary in the robustness of estimated ages at PVPV̇O2, PHV, and PWV, and with one exception (19) do not employ mathematical curve-fitting to estimate parameters of the adolescent growth spurt in peakV̇O2, height, and weight.
A variety of methods have been used to fit curves to individual data for height and weight to derive parameters of the adolescent growth spurt (12,18). The growth spurt in peak V̇O2(age at PVPV̇O2 and PVPV̇O2, L·min−1·yr−1) has been approached with curve-fitting procedures for individual longitudinal data in only the Saskatchewan Growth Study (19). The Preece-Baines Model I growth function (PB Model I, a modified logistic model; 21) was fitted to individual data for height, weight, and peak V̇O2to derive the parameters of the adolescent growth spurt. PVPV̇O2 occurred in the same year as PHV and PWV in both sexes (19), but the sample size for girls was relatively small in contrast to that for boys, which may limit generalizations.
The purpose of this study is to mathematically model adolescent growth of peak V̇O2in a sample of boys and girls followed longitudinally from 10 to 18 yr of age. The parameters of the growth spurt are compared between boys and girls, and relationships among parameters of the growth spurt in peakV̇O2, height, and weight are evaluated.
The sample consisted of 105 twin pairs (210 subjects), all of whom were selected from the East Flanders Prospective Twin Registry (15) for the Leuven Longitudinal Twin Study (16). Determination of zygosity (11) and design of the study (6,15,16) have been described previously. The study was approved by the Medical Ethics Committee of the Faculty of Physical Education and Physiotherapy at the Katholieke Universiteit Leuven. Parents gave informed consent for their children’s participation, and permission was given by the subjects as well.
The twins were followed longitudinally from 10 to 18 yr of age (6). Height and weight were measured semiannually from 10 through 16 yr, and then measured again at 18 yr. Peak V̇O2 was measured annually over the same age interval. Skeletal maturity status was assessed annually with the Tanner-Whitehouse 2 method (27), which involves assessment of left hand-wrist x-rays and comparison to sex-specific British standards. First observations of the subjects occurred at a prepubertal stage (stage I of breast development in girls and genital development in boys) at a mean age of 10.3 ± 0.3 yr.
Standing height was measured with a Harpenden stadiometer with the subjects in bare feet and the head in the Frankfort horizontal plane. Measurements were made by three consecutive experienced anthropometrists during the course of the study (1985–1999). Interobserver reliability was verified each time before a different anthropometrist started measuring. The intraclass correlation coefficient for stature (R) was 0.997, the mean difference between repeated measurements was less than 0.1 cm, the technical error was 0.18 cm, and the coefficient of variation was 0.10 (6). Weight was measured to the nearest 0.1 kg using a beam balance scale with subjects in swimsuits.
Peak V̇O2 (L·min−1) was measured during a maximal exercise test on a Woodway ELG2 treadmill using the Bruce protocol (9). An open circuit system with a calibrated flow sensor adapted to BTPS (body temperature and pressure, saturated with water vapor) was used. Oxygen concentration was determined with an O2 analyzer (Oxymat M, Siemens) according to STPD (standard temperature and pressure, dry; 0°C, 760 mm Hg, no water vapor). An infrared CO2 analyzer (Ultramat M, Siemens) was used to measure CO2. Data for ECG, O2, and CO2 were recorded every 30 s during the test. The test began at 2.7 km·h−1 and a 10% incline, progressing with increases of 1.3 km·h−1 in speed and 2% incline every 3 min. Subjects were encouraged to run to exhaustion and subjective ratings of exertion were assessed using the Borg scale for rating of perceived exertion (8). Criteria for a maximal test and test termination were a respiratory quotient greater than 1.0, a heart rate of 180 beats·min−1 or more, a ventilatory equivalent for oxygen > 30, and/or physical exhaustion. If the subject did not meet at least three of the four test criteria, the data were not kept as part of the individual’s data record and were recorded as missing data. Test-retest reliability of peak V̇O2 was estimated using the procedure of Van’t Hof et al. (29). Correlation coefficients between mean values for peak V̇O2 for the sample were plotted by measurement intervals to find the y-intercept, a proxy for in-field test-retest reproducibility. Test-retest reliability of peak V̇O2 was between 0.79 and 0.82.
The Preece-Baines Model I growth function (Simplex, PB Model I, 21) was fitted to individual records of 146 subjects with six or more observations for peak V̇O2 between 10 and 18 yr of age to derive estimates of age at PVPV̇O2 (yr), PVPV̇O2 (L·min−1·yr−1), and value at PVPV̇O2 (L·min−1). Growth curves were successfully fitted for 83 individuals (48 males, 35 females; labeled as “fitters”). Fitted curves for peak V̇O2 were checked visually to verify the correspondence between the raw increments and the fitted velocities and to count the number of data points above, below, and on the fitted curve as well as the number of runs. Subsequently, a runs test (25) was performed to ascertain any bias in the fitting with the PB Model I. The raw data for peak V̇O2 for fitters were also evaluated individually to determine whether peak V̇O2 continued to increase throughout adolescence with the highest value being the last value measured, or whether peak V̇O2 reached a peak value and then leveled off (i.e., reached a plateau) or declined.
The 83 fitters included 44 individuals from 22 twin pairs and 39 single members of twin pairs. Curves were fitted for 32 individuals (22 males, 10 females) on the first attempt without adjusting the raw data. Data for peak V̇O2 were adjusted for the other 51 individuals (25 males, 26 females) following a protocol analogous to that used in deriving a better fit for height (6). Estimated values for adult stature and length at birth were used to obtain better fits to the longitudinal data for height with the PB Model I (6). There are no equations or models for estimating adult values for peak V̇O2; thus, a repeat datum was used for some individuals to obtain a better fit to the data. The data were further adjusted using several additional methods: 1) a datum was dropped if it was lower than those at previous ages; 2) a datum of 0.10 was added at 0.00 yr if the distance curve generated from the raw, unadjusted data began at a point higher than the first measurement value for peak V̇O2 or began at a point less than 0.00 L·min−1; and/or 3) a repeat datum of maximal value for peak V̇O2 was added at subsequent age(s) of observation if the data did not level off or declined after the maximal value in peak V̇O2 was observed (PB Model I requires a leveling off of the data, as occurs for stature in late adolescence or early adulthood).
The peak V̇O2 data for the remaining 63 individuals (36 males and 27 females) were not successfully fitted with the PB Model I (labeled as “nonfitters”). The nonfitters included 22 individuals from 11 twin pairs and 41 single members of twin pairs.
Differences in means for age (yr), measurement intervals (time between successive measurements, yr), peak V̇O2 (L·min−1), and velocity increments in peak V̇O2 (L·min−1·yr−1) between fitters and nonfitters were compared with sex-specific, independent, two-tailed t-tests. Means for peak V̇O2 were also plotted by chronological age for fitters and nonfitters within sex.
Descriptive statistics for the PB Model I estimates of age at PVPV̇O2, PVPV̇O2, and value at PVPV̇O2 were calculated by sex for all subjects for whom curves were fitted. Corresponding parameters for height and weight based on the PB Model I were already available (6; unpublished). Sex-specific means and standard deviations for PVPV̇O2 were also calculated by half-year intervals for 3 yr before and after PVPV̇O2 (−3.0 to +3.0 yr), and then plotted relative to age at PVPV̇O2 to generate mean constant velocity curves for peak V̇O2 by sex. Ages at PVPV̇O2 were compared with ages at PHV and PWV to examine the timing of the growth spurt in peak V̇O2 relative to corresponding values for body size. Correlation coefficients were also calculated between ages at PVPV̇O2, PHV, and PWV, and PV̇O2, PHV, and PWV to examine relationships among these indicators of function and body size during the adolescent growth spurt.
Sex-specific mean errors of estimate for the curve-fitting of peak V̇O2were 0.097 L·min−1 for males and 0.068 L·min−1 for females. These represent 2.5% and 2.9% of mean values at 18.4 yr (young adult) of 3.88 L·min−1 and 2.33 L·min−1, for males and females, respectively, which compared favorably with the residual standard deviation for peak V̇O2 pooled over all subjects in the Saskatchewan Growth Study (3.0%, 20). The numbers of data points above and below the fitted curves, on average, did not differ (2.29 vs 2.45, respectively), and the runs test was not significant (z = 0.31, P = 0.38). Hence, there was no tendency for systematic bias toward under- or overestimation of peak V̇O2 with the PB Model I.
Comparison of fitters and nonfitters.
Male fitters had slightly but significantly lower mean ages at measurement than nonfitters (14.05 ± 2.70 yr vs 14.06 ± 2.71 yr; P < 0.05) and had significantly higher mean values for peak V̇O2 (2.79 ± 0.85 L·min−1 vs 2.68 ± 0.78 L·min−1, P < 0.05). Male fitters and nonfitters did not differ in measurement intervals and increments in peak V̇O2. Female fitters and nonfitters did not differ significantly in age, measurement intervals, peak V̇O2, and increments in peak V̇O2 Peak V̇O2 did not differ between fitters and nonfitters within sex, except in late adolescence (Fig. 1). Review of raw data for each individual indicated that peak V̇O2 increased throughout adolescence in less than one half of the fitters (N = 32, 23 males, 9 females), and leveled off (reached a plateau) or declined after reaching a maximal value in the majority (N = 51, 25 males, 26 females).
FIGURE 1Means for pe...Image Tools
Cross-sectional treatment of the data.
Peak V̇O2 increased, on average, gradually and progressively with age in both sexes from 10 through about 12.5 yr of age. Subsequently, peak V̇O2 leveled off in females and continued to increase with age in males (Fig. 1). The increase in peak V̇O2 through 18 yr was characteristic of male fitters, whereas it leveled off in male nonfitters between 16 and 18 yr of age (Fig. 1). Peak V̇O2 was, on average, greater in males than in females at all ages.
Longitudinal treatment of the data.
The longitudinal analysis of peak V̇O2 showed a similar age-related pattern as the cross-sectional analysis (Figs. 1 and 2), but the fitting of curves to individual data for peak V̇O2 presented a somewhat different pattern in girls (Fig. 2). Peak V̇O2 continued to increase in girls throughout adolescence but at a lower rate after about 13 yr of age. There was a clear growth spurt in peak V̇O2 in both sexes (Figs. 3 and 4), but the timing and magnitude of the growth spurt in peak V̇O2 differed significantly between boys and girls (Fig. 3, Table 1). Peak rate of growth in peak V̇O2 (PVPV̇O2 L·min−1·yr−1) occurred significantly earlier (P < 0.05) in females (12.3 ± 1.2 yr) than in males (14.1 ± 1.2 yr), whereas the magnitude of peak V̇O2 was significantly greater (P < 0.05) in males (1.01 ± 0.40 L·min−1) than in females (0.58 ± 0.30 L·min−1).
Growth spurts in peak V̇O2, height and weight.
Mean age at PVPV̇O2 was coincident with mean age at PHV in both sexes (Table 1). Peak V̇O2 velocities are shown aligned on age at PVPV̇O2 in Figure 4. Because ages at PVPV̇O2 were coincident with ages at PHV for both sexes (Table 1), the data shown in Figure 4 also illustrate estimated peak V̇O2 velocities relative to PHV. In contrast, mean ages at PVPV̇O2 occurred approximately 0.3 yr and 0.7 yr earlier, on average, than mean ages at PWV for males and females, respectively (Table 1).
The timing of individual ages at PVPV̇O2 is shown relative to the growth spurts in height and weight in Table 2. Timing was considered coincident if PVPV̇O2 occurred within ± 0.125 yr of PHV and PWV (5). Although mean age at PVPV̇O2 (14.1 ± 1.2 yr) occurred close in time to age at PHV (14.0 ± 1.0 yr) in boys, more individual ages at PVPV̇O2 occurred after age at PHV. Mean ages at PVPV̇O2 and PHV were identical in girls, 12.3 ± 1.2 yr and 12.3 ± 1.0 yr, respectively, and ages at PVPV̇O2 in individual girls were equally distributed before and after PHV. On the other hand, PVPV̇O2 occurred before PWV in the majority of boys and girls. The pattern of the timing of PVPV̇O2 relative to PHV and PWV did not differ significantly by sex (chi-square ≤ 0.996, P = 0.608, df = 2 for PHV; chi-square ≤ 1.981, P = 0.371, df = 2 for PWV). However, a significant difference in the pattern of the timing of PVPV̇O2 relative to PHV compared with PWV was evident within sex (chi-square ≤ 23.372, P = 0.001, df = 4 for males; chi-square ≤ 10.821, P = 0.029, df = 4 for females). PHV and PWV were not coincident in either sex. The small number of cases in which PVPV̇O2 occurred coincidentally with PHV and PWV was not surprising because the interval used to define coincidence was small (±0.125 yr). Broader intervals (±0.25 and ±0.50 yr) were also evaluated, and the distributions of individual ages at PVPV̇O2 relative to PHV and PWV were similar.
Correlation coefficients among ages at and PV of peak V̇O2, PHV, and PWV are shown in Table 3. Correlation coefficients among ages at PV were moderate and significant in males (0.51, 0.59, 0.63; P < 0.001), but low and with one exception not significant in females (0.07, 0.27, 0.38; P < 0.05). In contrast, correlation coefficients between ages at PV and the respective PV for the three variables were low. Only the correlation coefficients between age at PHV and PHV were significant (P < 0.05), −0.36 in females and −0.37 in males.
The correlation coefficients between PHV and PWV and age at PVPV̇O2, though low, suggested opposite trends in males and females. The correlation coefficient between PHV and age at PVPV̇O2 was negative and significant in males (−0.35, P < 0.05) but approached zero in females (0.06). In contrast, the correlation coefficient between PWV and age at PVPV̇O2 was negative and not significant in females (−0.22) but approached zero in males (0.03). Corresponding correlation coefficients between PV̇O2 and ages at PHV and PWV were low and negative in both sexes (−0.01 and −0.13 in males, −0.07 and −0.10 in females).
Male fitters had greater peak V̇O2 than male nonfitters in late adolescence (Fig. 1). The two groups, however, did not differ in skeletal maturity. The skeletal age-chronological age difference (SA − CA) was 0.06 ± 0.73 for male fitters and −0.06 ± 0.88 yr for male nonfitters. Thus, both groups would be classified as average or “on time” in maturity status. This suggests that the differences in peak V̇O2 between male fitters and nonfitters are not attributable to individual differences in maturity status (17). The greater peak V̇O2 of fitters is difficult to explain based on the data available but may be affected by other factors, for example, intensity of activity, performance efficiency or skill, motivation during the test, etc. Female fitters and nonfitters did not differ in peak V̇O2 (Fig. 1) and in the SA − CA difference, with values of −0.74 ± 0.67 for fitters and −0.96 ± 0.75 yr for nonfitters. Both groups of females were thus later maturing.
Some of the unsuccessful attempts at fitting with the PB Model I may be explained by individual differences in the pattern of growth in peak V̇O2 after peak values were attained. It was reasonably common in this sample for peak V̇O2 to decline after PVPV̇O2 was attained, which indicates a more variable pattern of adolescent growth than those for height and weight. The PB Model I was designed to model growth in height. It assumes a sigmoid curve with five parameters (21). As such, it may not be appropriate for peak V̇O2, particularly when the number of data points do not far exceed the minimum number required for curve fitting with the model. The fit of the PB Model I to individual height data can be enhanced by adding length at birth and an estimate of adult stature (6). This is not possible with peak V̇O2 data and may contribute to some of the unsuccessful attempts at fitting individual data.
Growth characteristics of peak V̇O2 during adolescence.
Peak leveled off (reached a plateau) or declined after reaching a maximal value in the majority of subjects. It increased throughout adolescence in less than one half of the subjects. Peak V̇O2 is a measurement of both function and performance. It varies with motivation, tolerance of fatigue, training history, and level of habitual physical activity. It is also highly correlated with muscle mass, which has its own growth spurt during adolescence (17).
The cross-sectional growth pattern in peak V̇O2 in the present sample (Fig. 1) was consistent with cross-sectional studies and cross-sectional analyses of longitudinal data in both sexes (2,14,17). The longitudinal analysis, however, suggested a different pattern in adolescent females. Peak V̇O2 continued to increase in girls throughout the growth spurt, although the rate slowed after about 12.5 yr of age. This age approximated mean ages at PVPV̇O2 and height in this sample of girls, 12.3 yr. Data for Canadian girls, which were analyzed in a similar manner, also exhibited a progressive increase in peak V̇O2 from 8 to 13 yr, but the limited age range of the sample did not permit evaluation of changes in peak V̇O2 in later adolescence (20). This trend is consistent with longitudinal data for Norwegian (1), German (24), Dutch (13), and British (3) girls.
The nonlinear increase in peak V̇O2 in females evident in the longitudinal analysis might be explained, in part, by the later maturity status (SA − CA difference) of the sample. Only one early-maturing girl (SA − CA > 1.0 yr) was included in the sample of fitters. It is thus likely that many subjects were still growing near the termination of the study, including growth in aerobic power. In a cross-sectional analysis of data for Dutch girls, mean peak V̇O2 was virtually flat from 12 through 16 yr in early maturing girls but increased gradually from 12 through 17 yr in later maturing girls (13). An additional factor might be maturity-associated variation in size and body composition; early maturing girls tend to be, on average, heavier and fatter than late maturing girls (17). The observations underscore the need for longitudinal analysis of longitudinal data in order to better understand changes in peak V̇O2 during adolescence. Cross-sectional data or cross-sectional analysis of longitudinal data may mask individual differences in the growth curve. The observations also highlight the need to follow growth in peak V̇O2 into the later ages of adolescence and/or into young adulthood.
The greater absolute values and increments in peak V̇O2 in males compared with females during adolescence in the present study are consistent with the literature. Sex differences in peak V̇O2 have been attributed to greater hemoglobin concentration, fat-free mass, stroke volume, stroke volume at peak V̇O2, and higher intensity of activity or sports play in males (2,3,4,14,22). The effect of body mass on peak V̇O2 has been examined via dimensional analysis (2,22); however, there is wide variability in individual longitudinal mass scaling exponents for maximal aerobic power and there are sex differences in mass scaling exponents even when body size and composition are similar (22). Biological differences between males and females that influence aerobic fitness, other than body size and as yet unidentified, have thus been postulated as sources of the sex differences in peak V̇O2 (22).
Timing and magnitude of the growth spurt in peakV̇O2.
Mean ages at PVPV̇O2 in the present study are quite similar to those reported for Canadian males and females (Table 4). Estimated values at PVPV̇O2 in the present sample of Belgian adolescents are also quite similar to estimates for Canadian adolescents (Table 4). Mean peak V̇O2 at PVPV̇O2 for males in the present study, 2.99 L·min−1, is virtually identical to estimates for Norwegian and Dutch adolescents, 2.9 L·min−1 (1,13). In contrast, mean peak V̇O2 at PVPV̇O2 for the females in the present study, 1.86 L·min−1, is somewhat lower than those reported for other European samples, 2.2–2.4 L·min−1 (1,13,23). On the other hand, maximum velocities of growth in peak V̇O2 (PVPV̇O2, L·min−1·yr−1) in the present study are greater than estimates for Canadian males and females (Table 4). Variation in estimated parameters of the adolescent growth spurt in peak V̇O2 among samples of European adolescents may be explained by different analytical protocols, subtle differences in testing protocols, variable intervals between observations, age at first observation, duration of the respective studies, and, of course, sampling and population variability, for example, level of habitual physical activity, maturity status, and so on.
Relationships to growth spurts in height and weight.
Ages at PHV and PWV and PV for height and weight for the present sample are reasonably similar to and well within the range of corresponding estimates for longitudinal samples of European adolescents (17). The estimated age at PHV for girls falls at the older end of the range of reported means, which reflects the later maturation of the sample. Mean age at PVPV̇O2 occurred, on average, coincident with mean age PHV and before mean age at PWV in both sexes. This is also consistent with other longitudinal studies (7,13,19,26). The growth spurt in peak V̇O2 is associated with rapid growth in height and weight, muscle mass, and organs of the cardiorespiratory system at this time. However, the present analysis indicates considerable individual differences in the distribution of ages at PVPV̇O2 relative to ages at PHV and PWV (Table 2).
Correlation coefficients among ages at PVPV̇O2, PHV, and PWV in males (0.51 to 0.63, P < 0.01) suggested simultaneous regulation of adolescent growth in body size and physiological function. In contrast, age at PVPV̇O2 was poorly related to ages at PHV and PWV in females (0.06 and 0.27, respectively), although ages at PHV and PWV were significantly related (0.38, P < 0.05). The correlation coefficients are consistent with the generally simultaneous growth in body size and fat-free mass—specifically muscle mass, heart, and lungs, during the male adolescent growth spurt. PWV occurs, on average, closer to PHV in boys than in girls (Table 1), so that sex differences in correlation coefficients may be related to the sex differences in changes in body composition during adolescence. The adolescent spurt in boys includes principally gains in height and muscle mass, whereas fat mass changes slightly (4,17). In contrast, the spurts in height and fat-free mass are less intense in girls than in boys, whereas there is a continuous rise in fat mass (3,4,17).
Cross-sectional and longitudinal analyses indicated a well-defined growth spurt in peak V̇O2 in males, but only the longitudinal treatment of the data indicated a growth spurt in peak V̇O2 in females. Growth spurts in height, weight, and peak V̇O2 occurred earlier in females than in males, but the magnitude of the growth spurts in the three variables was greater in males. Peak V̇O2 was greater in males than in females at all ages from late childhood through adolescence. Mean age at PVPV̇O2 occurred, on average, coincident with PHV and before PWV in both sexes, although there was considerable interindividual variation. Moderate, positive correlation coefficients among ages at PVPV̇O2, PHV, and PWV suggested simultaneous regulation of the timing of maximum growth in body size and aerobic power during male adolescence. Corresponding correlation coefficients suggested some dissociation of maximum growth in peak V̇O2 from maximum growth in height and weight during female adolescence. Longitudinal studies that span adolescence (about 9 –18 yr) and with sufficient numbers of data points and subjects are needed to further understand the growth of peak V̇O2, particularly in girls, during this major transitional period in the life cycle. The PB Model I has been applied to peak V̇O2 data in two longitudinal series to date. However, it is not applicable for peak V̇O2 data for all subjects, which emphasizes the need for further development of models that may be more appropriate for the longitudinal analysis of peak V̇O2 during adolescence.
This research was supported by Research Fund K. U. Leuven (OT/86/80), Nationale Bank van België, Fund for Medical Research (Belgium) (3.0038.82, 3.0008.90, 3.0098.91), and NATO (860823).
Martine Thomis was supported by the Fund for Scientific Research-Flanders (Belgium) as a Postdoctoral Fellow. Ruth J. F. Loos was supported by a grant of the Belgian American Educational Foundation (BAEF).
Justification for more than six authors: M. A. Thomis, B. Vanden Eynde, H. H. M. Maes, R. J. F. Loos, M. Peeters, A. Claessens, R. Vlietinck, and G. Beunen were involved with the organization of and/or data collection for the twin study from which the data came for this manuscript. C. A. Geithner and M. A. Thomis contributed equally to the writing of the manuscript, and G. Beunen and R. M. Malina provided significant input.
1. Anderson, K. L., V. Seliger, J. Rutenfranz, and J. Skrobak-Kaczynski. Physical performance capacity of children in Norway. Part IV: the rate of growth in maximal aerobic power and the influence of improved physical education of children in a rural community. Population parameters in a rural community. Eur. J. Appl. Physiol.
2. Armstrong, N., and J. R. Welsman. Assessment and interpretation of aerobic fitness in children and adolescents. Exerc. Sport Sci. Rev.
3. Armstrong, N., J. Williams, J. Balding, P. Gentle, and B. Kirby. The peak oxygen uptake of British children with reference to age, sex and sexual maturity. Eur. J. Appl. Physiol.
4. Beunen, G., and R. M. Malina. Growth and physical performance relative to the timing of the adolescent spurt. Exerc. Sports Sci. Rev.
5. Beunen, G. P., R. M. Malina, M. A. Van’t Hof, et al. Adolescent Growth and Motor Performance: A Longitudinal Study of Belgian Boys.
HKP Sport Science Monograph Series, Vol.1. Champaign, IL: Human Kinetics, 1988, pp. 1–102.
6. Beunen, G., M. Thomis, H. H. Maes, et al. Genetic variance of adolescent growth in stature. Ann. Hum. Biol.
7. Binkhorst, R. A., M. C. De Jong-Van De Kar, and Q. C. Q. Vissers. Growth and aerobic power of boys aged 11–19 years. In:Children and Sport: Pediatric Work Physiology
, J. Ilmarinen and I. Valimaki (Eds.). New York: Springer-Verlag, 1984, pp. 99–105.
8. Borg, G. A. V. Physical Performance and Perceived Exertion.
Lund, Sweden: Gleerup, 1962, pp. 1–62.
9. Bruce, R. A. Multistage treadmill test of submaximal and maximal exercise. In:Exercise Testing and Training of Apparently Healthy Individuals: A Handbook for Physicians
, American Heart Association (Ed.). New York: American Heart Association, 1972, pp. 32–34.
10. Cunningham, D. A., D. H. Paterson, C. J. R. Blimkie, and P. Donner. Development of cardiorespiratory function in circumpubertal boys: a longitudinal study. J. Appl. Physiol.
11. Derom, C., E. Bakker, R. Vlietinck, et al. Zygosity determination in newborn twins using DNA variants. J. Med. Genet.
12. Hauspie, R., and H. Chrastek-Spruch. Growth models: possibilities and limitations. In:Human Growth in Context
, F. E. Johnston, B. Zemel, and P. B. Eveleth (Eds.). London: Smith-Gordon, 1999, pp. 15–24.
13. Kemper, H. C. G., R. Verschuur, and J. W. Ritmeester. Maximal aerobic power in early- and late-maturing teenagers. In:Children and Exercise XII
, J. Rutenfranz, R. Mocellin, and F. Kliimt (Eds.). Champaign, IL: Human Kinetics, 1986, pp. 213–225.
14. Krahenbuhl, G. S., J. S. Skinner, and W. M. Kohrt. Developmental aspects of maximal aerobic power in children. Exerc. Sports Sci. Rev.
15. Loos, R. J. F., R. Derom, C. Derom, et al. The East-Flanders Prospective Twin Survey (Belgium): a population-based register. Twin Res.
16. Maes, H. H. M., G. P. Beunen, R. F. Vlietinck, et al. Inheritance of physical fitness in 10-yr-old twins and their parents. Med. Sci. Sports Exerc.
17. Malina, R. M., C. Bouchard, and O. Bar-Or. Growth, Maturation, and Physical Activity
, 2nd Ed. Champaign, IL: Human Kinetics, 2004, pp. 235–250, 307–365.
18. Marubini, E., and S. Milani. Approaches to the analysis of longitudinal data. In:Human Growth: A Comprehensive Treatise
, Vol. 3, 2nd Ed., F. Faulkner and J. M. Tanner (Eds.). New York: Plenum, 1986, pp. 79–94.
19. Mirwald, R. L., and D. A. Bailey. Maximal Aerobic Power, a Longitudinal Analysis.
London, Ontario: Sports Dynamics, 1986, pp. 1–80.
20. Mirwald, R. L., D. A. Bailey, N. Cameron, and R. L. Rasmussen. Longitudinal comparison of aerobic power in active and inactive boys aged 7.0 to 17.0 years. Ann. Hum. Biol.
21. Preece, M. A., and M. J. Baines. A new family of mathematical models describing the human growth curve. Ann. Hum. Biol.
22. Rowland, T., P. Vanderburgh, and L. Cunningham. Body size and the growth of maximal aerobic power in children: A longitudinal analysis. Pediatr. Exerc. Sci.
23. Rutenfranz, J., K. L. Andersen, V. Seliger, et al. Maximal aerobic power affected by maturation and body growth during childhood and adolescence. Eur. J. Pediatr.
139:106 –112, 1982.
24. Rutenfranz, J., K. L. Andersen, V. Seliger, F. Klimmer, I. Berndt, and M. Ruppel. Maximum aerobic power and body composition during the puberty growth period: similarities and differences between children of two European countries. Eur. J. Pediatr.
25. Siegel, S. Nonparametric Statistics for the Behavioral Sciences.
New York: McGraw-Hill Book Company, Inc., 1956, pp. 52–60, 247, 252–253.
26. Sprynarova, S., J. Parizkova, and V. Bunc. Relationships between body dimensions and resting and working oxygen consumption in boys aged 11 to 18 years. Eur. J. Appl. Physiol.
27. Tanner, J. M., R. H. Whitehouse, N. Cameron, W. A. Marshall, M. J. R. Healy, and H. Goldstein. Assessment of Skeletal Maturity and Prediction of Adult Height (TW2 Method)
, 2nd Ed. London: Academic Press, 1983, pp. 1–99.
28. Vanden Eynde, B., D. Vienne, M. Vuylsteke-Wauters and D. Van Gerven. Aerobic power and pubertal peak height velocity in Belgian boys. Eur. J. Appl. Physiol.
29. Van’t Hof, M. A., J. Simons, and G. BEUNEN. Data quality evaluation and data processing. In:Somatic and Motor Development of Secondary Schoolboys: Norms and Standards
, M. Ostyn, J. Simons, G. Beunen, R. Renson, and D. Van Gerven (Eds.). Leuven, Belgium: Leuven University Press, 1980, pp. 45–48.
MAXIMAL AEROBIC POWER, PEAK V̇O2; ADOLESCENT GROWTH SPURT; AGE AT PEAK VELOCITY; PREECE-BAINES; CURVE-FITTING
©2004The American College of Sports Medicine
What does "Remember me" mean?
By checking this box, you'll stay logged in until you logout. You'll get easier access to your articles, collections,
media, and all your other content, even if you close your browser or shut down your
To protect your most sensitive data and activities (like changing your password),
we'll ask you to re-enter your password when you access these services.
What if I'm on a computer that I share with others?
If you're using a public computer or you share this computer with others, we recommend
that you uncheck the "Remember me" box.
Highlight selected keywords in the article text.
Data is temporarily unavailable. Please try again soon.