The effects of biological maturation on thermoregulatory responses to exercise in hot ambient conditions have been well documented (5,17). The whole-body sweating rate of an adult male is approximately 40% greater than that of a prepubertal boy, reflecting a larger output per gland and greater gland sensitivity to thermal stimuli. Consequently, children rely to a greater extent on cutaneous blood flow for convective heat loss during exercise. At a given work level, heat production per kilogram of body mass is inversely related to body size-a disadvantage in the small child that is offset by a greater surface area:mass ratio. In addition, the rate of heat acclimatization is slower in children than adults.
The implications of these features of prepubescent subjects in respect to endurance performance and predisposition to heat injury, however, are not clear. Squire (23) concludes that "because children are inherently less efficient regulators than adults, they are at even greater risk for heat illness." Bar-Or (3) agrees, contending that "children might be expected to be less effective thermoregulators than adults and less tolerant of climatic heat, particularly when exposed to climatic extremes." This putative intolerance of prepubertal subjects to exercise in the heat has been explained not only by an inferior sweating response but also their "hypokinetic" cardiac response to exercise, as indicated by a lower cardiac output at a given level of metabolic demand (oxygen uptake) compared with adults (24).
This viewpoint of children as an at-risk group for heat injury has led to particular considerations for the safety of prepubertal athletes who train and compete in hot climatic conditions. Reflecting this concern, specific guidelines for children exercising in the heat have been published (1).
Others have argued that no clear evidence exists that the unique thermoregulatory responses to exercise in the heat in children can be translated into reduced exercise capacity or greater risk of heat injury compared with adults. In their review of six studies that directly compared physiologic responses of children and adults to exercise in the heat, Armstrong and Maresh (2) point out that rise in core temperature was consistently similar in the two groups. Recent studies have supported this viewpoint. Inbar et al. (12) found similar changes in rectal temperature among prepubertal boys and adult men cycling at 50% V˙O2max in 41°C, 21% humidity conditions. Rivera Brown et al. (15) have reported no significant differences in rectal temperature change, exercise tolerance, or cardiovascular responses between heat-acclimatized girls and adult women performing exercise in a hot outdoor environment (33°C, 55% humidity).
Rowland et al. (22) consider the finding of a low Q/V˙O2 in prepubertal subjects to be biologically spurious, because children and adults do not exercise with the same absolute V˙O2. When variables are appropriately adjusted for body size, no quantitative or qualitative maturational differences in cardiovascular responses to exercise in thermoneutral conditions have been observed (14,22).
The extent that thermal stress injury (heat stroke, exhaustion) occurs in prepubertal athletes is unclear. Reports of such incidences are not evident in the published literature. Indeed, in a 10-yr survey of medical records in a tropical region of Australia (Cairns), Brun and Mitchell (6) could not identify a single case of heat-related illness in a child athlete. Anecdotal information indicates, however, that such cases have occurred in the United States (D. Costa, personal communication 2007).
This study was designed to further address these issues, directly comparing endurance performance and physiologic variables in nonacclimatized prepubertal boys and young men during steady-work submaximal cycling at approximately 65% peak V˙O2 in both cool (~19°C) and hot (~31°C) ambient conditions with moderate humidity. Specifically, this investigation sought to assess maturational differences in 1) exercise tolerance in a hot climatic environment, 2) physiologic responses to upright cycling in the heat, particularly those that might limit endurance performance, and 3) the effect of change in ambient temperature on exercise endurance.
Eight adult males (mean age 31.8 ± 2.0 yr) and eight boys (mean age 11.7 ± 0.4 yr) were recruited for exercise testing. This study was performed in Massachusetts in the winter, spring, and early summer months, and all subjects were therefore considered to be nonacclimatized to exercise in the heat. None had any history of heat stress illness.
The men were all members of the hospital medical residency staff. Seven reported regular physical activity (running, walking, weight lifting) up to three times a week, but none were trained athletes or were involved in formal exercise training programs. All were healthy nonsmokers and were taking no medications that would affect cardiopulmonary function.
Data on the boys in this study have previously been published in an investigation of the factors that limit exercise endurance in the heat in children (19). Subjects were healthy and physically active, as all had participated on a recreational sports team (soccer, basketball) within the previous 3 months. However, none was considered to be a trained athlete. All were prepubertal (Tanner 1) by direct genital observation by a physician.
This study was reviewed and approved by the institutional review board of the Baystate Medical Center. Informed written consent was obtained from all adult participants. For the prepubertal subjects, informed permission and assent were provided by a parent and child, respectively.
The initial testing consisted of a progressive cycle test for determination of peak oxygen uptake (V˙O2). On two separate subsequent sessions, constant-work tests to exhaustion were performed at an intensity of approximately 65% peak V˙O2 in a cool and moderately hot indoor thermostat-controlled room at moderate levels of relative humidity. The ambient conditions on the first of these submaximal tests were randomly alternated to prevent any possible order effect. The ambient testing environmental conditions, equipment, and procedures were identical for the boys and men.
Progressive testing for determination of peak V˙O2.
Body weight and height were measured with a balanced beam scale and stadiometer, respectively, and, for descriptive purposes, body mass index was calculated as mass (kg) divided by height (m) squared. Testing was performed at a room temperature of 20-21°C with moderate levels of relative humidity (range 48-72%).
Subjects pedaled at a constant cadence of 50 rpm on an upright, mechanically braked Monark cycle ergometer (Model 818, Varberg, Sweden) with 3-min work stages. Initial and incremental loads were 25 and 50 W for the boys and men, respectively. All subjects met the criteria for exhaustive work effort, defined by inability to maintain the pedaling cadence in association with subjective evidence of fatigue (sweating, hyperpnea) and a heart rate > 185 bpm (boys) or > 170 bpm (men) and/or respiratory exchange ratio (RER, V˙CO2/V˙O2) > 1.00 (boys) or > 1.10 (men).
Standard open-circuit techniques for measurement of gas-exchange variables were used with a Q-Plex Cardiopulmonary Exercise System (Quinton Instrument Company, Seattle, WA). Minute ventilation was assessed with a pneumotachometer in the expiratory line. Expired air was collected in a 6-L mixing chamber, with aliquots analyzed for oxygen and carbon dioxide by zirconia oxide and infrared analyzers, respectively. Physiologic variables were averaged for 15-s time periods, and peak V˙O2 was defined as the mean of the two highest values during the final minute of exercise. The system was calibrated before and after each test with standard gases of known oxygen and carbon dioxide concentration.
Heart rate was measured for a 10-beat duration on an electrocardiogram. Standard Doppler echocardiographic techniques were used to estimate stroke volume (20). A 2.0-MHz transducer was directed inferiorly from the suprasternal notch to assess velocity of blood flow in the ascending aorta. The velocity-time integral (VTI), the integration of the area under the velocity curve for individual beats over time, was averaged from the 5-10 highest and most well-defined curves. Stroke volume was then calculated as the product of average VTI and the cross-sectional area of the aortic root, obtained from the diameter of the sinotubular junction measured by two-dimensional echocardiography (long-axis view), with the subject seated at rest on the cycle. Reliability and concurrent validity of this method have been previously reported from this laboratory (18,21).
Cardiac output (Q) was calculated as the product of stroke volume and heart rate, and both cardiac output and stroke volume were expressed relative to body surface area as cardiac index and stroke index, respectively. Arterial venous oxygen difference was calculated from the Fick equation as the absolute value of V˙O2/Q.
The two constant-load tests, one in hot and the other in cool ambient conditions, were performed at a work load of approximately 65% peak V˙O2, as calculated by a regression equation between submaximal work loads and V˙O2 during the initial progressive test. For descriptive purposes, temperature and humidity were averaged from measurements at the beginning and end of the test. No fan was used. Ambient conditions for the hot environment were 31.1 ± 0.3°C, 50 ± 5% relative humidity for the men and 31.0 ± 0.3°C, relative humidity 57 ± 2% for the boys. In the cool environment, the mean ambient temperature was 19.8 ± 1.0°C for the men and 19.6 ± 0.6°C for the boys, with relative humidities of 58 ± 10 and 66 ± 11%, respectively. For ambient temperature, mean values were significantly different for condition but not for group, whereas differences in humidity reached statistical significance for group but not condition.
Subjects urinated completely immediately before exercise testing and were then weighed nude on an electronic scale with a precision of ± 0.010 kg (Model WWS-250, AmCells Corporation, Carlsbad, CA). A rectal temperature thermistor (model 491B, Cincinnati Sub-Zero Products, Inc., Cincinnati, OH) was inserted 8 cm (boys) and 12 cm (men) beyond the anal sphincter and connected to a microprocessor-based thermometer with a resolution of 0.1°C and accuracy of ± 0.05% (Model 08402-20, Cole-Parmer Instrument Company, Vernon Hills, IL).
During exercise, subjects had free access to a container of cool water and were permitted to drink ad libitum but without encouragement. Volume of water consumed was measured at the end of each test.
Cuff blood-pressure measurements were made in the left arm, using the standard auscultatory technique. Diastolic pressure was defined by muffling of sounds. Values of mean arterial pressure (MAP) were calculated as 1/3(systolic − diastolic) + diastolic.
Subjects pedaled at a constant cadence of 50 rpm and watched a video movie during the cycling. The test was terminated and endurance time recorded when the subject declared he was too fatigued to continue. To prevent possible staff bias, and to limit external factors influencing perception of fatigue, no verbal encouragement was provided by the testing staff. Heart rate, stroke volume, blood pressure, and rectal temperature were measured at 5-min intervals, and V˙O2 was determined every 10 min (with a 2-min collection of expired air). Safety criteria for terminating the test were a rectal temperature > 39°C, heart rate > 200 bpm, or symptoms of nausea, dizziness, chills, exhaustion, change in sensorium, or headache.
On completion of exercise, subjects removed the rectal temperature probe, dried with a towel, urinated, and then were reweighed nude. Percent dehydration was determined as [(initial weight − final weight)/initial weight] × 100.
All statistical analyses were conducted using SAS version 9.1 for Windows. Group comparisons of peak V˙O2 and maximal cardiac output were performed with independent t-tests. Time series data (e.g., change in a variable from beginning to end-exercise, or behavior of a variable throughout the course of the exercise sessions) were analyzed using linear mixed models in SAS PROC MIXED, allowing for explicit modeling of the correlation structure induced in the data because of repeated measurements. In each case, unstructured covariance matrices were successfully fit to the data, minimizing the assumptions required for the analysis.
This is a modern analogue of the more traditional repeated-measures ANOVA (10).
Statistical significance was set a priori for all analyses at P < 0.05. When no significant differences were found in variables that were germane to physiological mechanisms of interest in this study, post hoc power analyses were performed to support any conclusions that these variables did not behave differently among the age groups.
Mean weight of the boys was 47.1 ± 8.4 kg, height 155 ± 8 cm, body surface area 1.44 ± 0.16 m2, and BMI 19.7 ± 2.6 kg·m−2. Respective values for the men were 81.8 ± 6.5 kg, 174 ± 4 cm, 1.97 ± 0.06 m2, and 26.8 ± 2.4 kg·m−2. Mean peak V˙O2 for the boys was 44.2 ± 4.7 mL·kg−1·min−1 and 40 ± 7.1 mL·kg−1·min−1 for the men (P > 0.05). Average peak heart rate and respiratory exchange ratio values were 187 ± 13 bpm and 1.03 ± 0.05 for the boys, and 180 ± 7 and 1.14 ± 0.07 for the men. Maximal stroke index and Q index values on the progressive test were 70 ± 13 mL·m−2 and 12.81 ± 1.90 L·min−1·m−2 and 62 ± 8 mL·m−2 and 11.14± 1.33 L·min−1·m−2 for the boys and men, respectively (P > 0.05).
There were no significant differences in relative exercise intensity (percent peak V˙O2) between ambient temperature condition or subject group. In the cool environment, boys cycled at an average of 62.9 ± 3.9% peak V˙O2 and the men 66.8 ± 6.1% peak V˙O2. In the hot condition, exercise intensities were 65.4 ± 6.6 and 65.9 ± 2.9% peak V˙O2 in the boys and men, respectively.
Exercise endurance time was significantly shorter in the hot compared with the cool environment for the boys and men (by an average of 29.2 and 29.0%, respectively), but there were no significant differences between the groups in either ambient condition (hot: boys 29.30 ± 6.19 min, men 30.46 ± 8.84 min; cool: boys 41.38 ± 6.30 min, men 42.88 ± 11.79 min) (Fig. 1). No subjects demonstrated evidence of abnormal heat stress during testing or reached safety criteria for test termination.
Volume of water consumed during the cool and hot conditions was similar for the boys and men. The boys drank 240 ± 95 and 229 ± 104 mL in the hot and cool tests, whereas respective values for the men were 430 ± 160 and 424 ± 285 mL. When expressed relative to body mass, no significant differences in average subject voluntary fluid intake were observed between groups in either climatic condition (hot: boys 5.03 ± 2.01 mL·kg−1, men 5.19 ± 1.92 mL·kg−1; cool: boys 4.77 ± 2.19 mL·kg−1, men 5.17 ± 3.64 mL·kg−1).
Weight loss and degree of dehydration were minimal in both groups in each ambient condition. The boys lost 0.050 ± 0.080 kg while cycling in the heat and 0.130 ± 0.070 kg in the cool room. These represent dehydration rates of 0.11 ± 0.17 and 0.28 ± 0.15%, respectively. Weight in the men decreased by 0.68 ± 0.07 kg in the heat and 0.19 ± 0.08 kg in the cool exercise conditions, indicating dehydration rates of 0.23 ± 0.28 and 0.20 ± 0.32%, respectively.
Rectal temperatures (Tre) at rest and end of exercise are outlined in Table 1. No significant difference was observed in Tre between groups or conditions, and values rose progressively in all tests. Peak Tre and ΔTre were similar in all ambient conditions and between boys and men.
Cardiovascular variables at 5 min and end-exercise for the two groups are presented in Table 2. Average heart rate was greater in the heat than in the cool condition and higher in boys than in the men at all exercise times. However, no group-time interaction was observed in either testing environment. There was no significant difference in change in heart rate between men and boys. Mean arterial pressure and values of arterial venous oxygen difference were greater in the men, but no interaction effect was observed.
Stroke index remained stable in all tests, and values were similar by condition and group. Cardiac index rose slowly during exercise in both groups and was significantly higher at end-exercise than at 5 min, both in men and boys. Values were not significantly different between groups and conditions, and no group-time interaction was seen in either ambient condition. No effects of condition or time were observed in arterial oxygen difference or blood pressure in men or boys.
Power analysis revealed that if a new study were conducted with similar differences and variability in the outcome variables among groups and conditions as in the present study, an N of 45 per group would be required to find a significant difference of the size observed in the present study (at P < 0.05 and a power of 0.80) in cardiac index. Similar analyses suggest that sample sizes of 64, 140, and 17 would be required to find significant differences in stroke index, heart rate, or arterial venous oxygen difference, respectively. The required number for endurance time was 681 and 579 for the hot and cool conditions, respectively.
This study, which directly compared adult males and prepubertal boys, failed to reveal evidence of maturational differences in thermoregulatory responses to exercise or endurance exercise performance in a moderately hot (~31°C) environment. In a testing model in which significant dehydration was prevented, the two groups were similar in (a) endurance capacity at the same relative work load (~65% peak V˙O2), (b) the effect of difference in ambient temperature conditions on performance (~29% decline in the heat), and (c) cardiovascular responses to both environments.
Despite concern regarding a unique vulnerability of children with exercise in the heat, few other investigations have directly compared cardiovascular, thermoregulatory, and performance responses of pre- and postpubertal subjects to exercise in hot ambient conditions. Most frequently cited is the study of Drinkwater et al. (8), in which five nonacclimatized prepubertal girls and five college-aged women walked at 30% V˙O2max for two 50-min bouts in three ambient temperatures (28, 35, and 48°C). No fluid replacement was given. Whereas all subjects completed the first walk in the lower two temperatures, four of the five girls had to be withdrawn from exercise at 48°C (118°F) because of tachycardia and fatigue. Only two of the girls could finish the second walk at 35°C (compared with all the women), and only one girl and three women could begin the second exercise at 48°C.
The authors conclude that compared with adults, prepubertal girls have a low tolerance for exercise in the heat, reflecting inferior cardiovascular responses (based on their signs of facial flushing, dizziness, and marked fatigue). However, no group differences in rise of rectal temperature were observed, and cardiac index and change in cardiac index were similar in the two groups.
Rivera-Brown et al. (15) compared responses of athletic, heat-acclimatized girls and women to submaximal cycling to exhaustion in sunny outdoor conditions with a mean ambient temperature of 33°C. Fluids were consumed during exercise in volumes relative to body size. No significant differences between the two groups were seen in endurance time, sweating rate, rise in rectal temperature, or heat storage. Compared with the women, girls demonstrated similar end-exercise values for rectal temperature, stroke index, cardiac index, and forearm skin blood flow.
Inbar et al. (12) describe thermoregulatory responses of eight prepubertal and eight young men cycling at 50% peak V˙O2 for three 20-min bouts with 7-min rest periods in an environment of 41 ± 0.67°C, 21% relative humidity. Subjects were encouraged to drink water throughout the test. Maximal values and increases in rectal and skin temperature were similar in the two groups. Although sweating rate was less in the prepubertal boys, this group demonstrated the greatest mass-relative evaporative cooling and sweating efficiency (ratio between evaporative heat loss and total sweating rate). The authors conclude that compared with adults, prepubertal boys are more effective thermoregulators during exercise in the heat.
The findings in these reports, then, are supported by those in the present study, which indicate that among euhydrated subjects, thermoregulatory and cardiovascular responses to exercise in the heat are similar in prepubertal male children and young adult men. Moreover, these results suggest that, at least in conditions in which subjects remain well hydrated, children carry no greater risk for heat stress injury than their adult counterparts.
Observations in this study indicate no evidence of an inferior cardiovascular response to exercise in the heat, confirming the findings of Rivera-Brown (15) and those of studies comparing children and adults in normothermic conditions (14,22). No significant differences between the two groups were seen in cardiac index or stroke index, or in changes in heart rate, mean arterial pressure, or arterial venous oxygen difference during exercise.
Earlier studies have suggested that children are no more prone to lower voluntary fluid intake (and are, therefore, at no greater risk for dehydration) during exercise than adults (4,11). In this study directly comparing the two groups, no differences were seen in either ad libitum fluid intake or level of dehydration. Voluntary water intake largely compensated for fluid losses in the approximately 30 min of exercise in the hot ambient condition in boys and men. From weight-loss and fluid-intake data, it can be calculated that in the hot condition, children replaced an average of 83% and adults 76% of their sweat, urine, and respiratory fluid losses by water intake. These high replacement percentages may relate to the moderately hot exercise condition. Involuntary hydration in boys cycling intermittently for 3.5 h at 39°C, 45% humidity, replaced 66% of fluid losses in the study of Bar-Or et al. (4), and Greenleaf et al. describe a 70% replacement rate in young men exercising in 39.8°C (11).
Common experience indicates that exercise is tolerated less well in hot compared with cool climatic conditions. The magnitude of this effect of ambient temperature on performance is illustrated by Galloway and Maughan, who found that adult cyclists pedaling at 70% V˙O2max could persist for an average of 93 min in an 11°C environment, 80 min in 21°, and 50 min in 31°C (9). Similarly, in the present study, an increase in ambient temperature by approximately 11°C (19-31°C) resulted in averages decline in endurance performance of 29% in boys and men. At least within this temperature range, then, biological maturation does not seem to affect the influence of change in ambient temperature on endurance performance.
The physiologic mechanism by which endurance performance capacity is limited in the heat has not been well clarified. By the traditional explanation, diminished perfusion of skeletal muscle as a consequence of sweating-induced dehydration (fall in blood volume) and a "steal" of blood flow to the cutaneous circulation for heat dissipation have been held responsible (16). In this study, however, dramatic differences in endurance performance relative to ambient temperature were observed in both men and boys in the absence of significant dehydration, and there was no evidence of falls in skeletal muscle blood perfusion (as indicated by stable arterial venous oxygen difference and mean arterial pressure). These findings are consistent with a model in which a rise in core temperature serves as the critical factor limiting endurance performance, mediated by the role of the central nervous system as a "protective governor" (7,13). By this explanation, sensations of fatigue that limit exercise in the heat (dizziness, hyperpnea, weakness) reflect brain perception rather circulatory insufficiency. This study suggests that if this model is correct, central factors responding to critical levels of core temperature are qualitatively and quantitatively similar in boys and adult men.
Limitations of this study need to be recognized. The small number of subjects reflects the difficulty of recruiting subjects to have rectal temperatures performed during exercise testing. The post hoc power analysis suggests, however, that this number of subjects provided accurate group comparisons. That is, this analysis indicated that a large number of subjects would be required to find a significant difference in the relevant comparisons, a consequence of an observed effect size so small that even if it were found to be statistically significant, it would not be likely be considered physiologically relevant. That being noted, the number of subjects used in this investigation suggests that the results should be viewed with caution until they are confirmed in a larger sample. The effect of a 7-8% difference in relative humidity between the groups is uncertain.
In assessing the implications of this study, several caveats are important. First, adult-child physiologic responses to exercise were compared in moderate, but not severely hot ambient conditions. Bar-Or (3) suggests that prepubertal children might be at particular risk for insufficient thermoregulatory responses to exercise when ambient temperature exceeds skin temperature by 10°C (generally an air temperature > 45°C). Second, because subjects were given free access to fluids, no patterns of physiologic responses to exercise could be related to lack of hydration. The extent that dehydration might influence the cardiovascular and thermoregulatory responses to exercise in the heat observed in this study is not known. Third, the findings were restricted to male subjects who were healthy and not athletically trained. Additional research will need to clarify maturational differences in physiologic and performance responses to exercise in the heat relative to gender, level of physical fitness, and high-risk individuals (such as those with obesity, diabetes mellitus, or cystic fibrosis).