| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Endocrine Care |
Department of Medicine, Guy’s, King’s, and St. Thomas’ School of Medicine, St. Thomas’ Hospital, London, United Kingdom SE1 7EH
Address all correspondence and requests for reprints to: Dr. Richard Holt, South Academic Block Level D, Southampton General Hospital, Tremona Road, Southampton, United Kingdom SO16 6YD. E-mail: righ{at}soton.ac.uk
Abstract
Exercise is a potent stimulus for GH secretion. Aging and obesity are associated with a diminution of GH secretion. We wanted to determine whether age or fat mass is more important in regulating the GH response to exercise. Four groups of healthy men were studied: seven lean young men [age, <40 yr; body mass index (BMI), <25 kg/m2], six overweight young men (age, <40 yr; BMI, >27.5), seven lean older men (age, >60 yr; BMI, <25), and 6 overweight older men (age, 60 yr; BMI, >27.5). The men performed a maximal exercise test.
GH secretion was higher in the younger men than in the older men. Peak GH was higher in the older lean men than in the older overweight men. There was no difference between the young groups. Fitness correlated negatively with age and positively with peak GH. In young men, there was no relation between BMI, bioimpedance, or leptin and GH secretion. In contrast, in older men there was an inverse correlation between measures of fat mass and GH secretion.
Age and physical fitness are more important than body fat in regulating exercise-induced GH secretion. These findings have important clinical implications if we are to prevent the frailty and morbidity associated with aging.
GH IS AN anabolic polypeptide secreted by the anterior pituitary gland in a pulsatile fashion (1). Apart from sleep, exercise is the most potent physiological stimulus to GH secretion (2, 3). GH increases within a few minutes of onset of both aerobic and resistance exercise and remains elevated for up to 30 min during recovery (4, 5). A maximal GH response is achieved at 70% of the maximal oxygen uptake (VO2 max) (6). The magnitude of this rise is dependent on a number of variables, including hypoxia, change in body temperature, and availability of energy substrates (7, 8). The underlying mechanisms that control GH release during exercise are poorly understood.
Both aging and obesity are associated with a reduction in GH secretion (4, 9). GH secretion increases gradually during childhood, peaks during puberty, and then declines during adult life (4). This progressive fall in GH secretion is associated with somatic changes that occur as part of the aging process (10). These include a loss of lean body mass, decreased muscle strength, and increased fat mass. The mechanisms leading to this fall in GH secretion are unclear. However, it is recognized that obesity is associated with a blunted response to all stimuli of GH and that significant weight loss is accompanied by restoration of normal GH secretion (9, 11). Furthermore, the severity of the secretory defect is proportional to the degree of obesity (9). It is unclear whether the fall in GH secretion with aging is secondary to the changes in body composition caused by aging or whether the fall in GH may precede these changes and may be causal for at least some components of the aging process, such as the accumulation of central obesity.
The aim of the current study was to examine the GH response to a maximal exercise test in young and old, and lean and overweight men to determine whether aging or adiposity was the more important determinant of GH production.
Subjects and Methods
Subjects
Four groups of healthy Caucasian men were studied: 1) men aged 20–40 yr with body mass index (BMI) less than 25 kg/m2 (n = 7), 2) men aged 20–40 yr with BMI more than 27.5 kg/m2 (n = 6), 3) men more than 60 yr old with BMI less than 25 kg/m2 (n = 7), and 4) men more than 60 yr with BMI more than 27.5 kg/m2 (n = 6). The two age groups were selected with the aim of obtaining subjects as far apart in age as possible while maintaining a fit elderly population capable of performing a maximal exercise test. To recruit volunteers who were able to exercise to a similar level, the overweight group was chosen to be representative of the normal population, and they were not clinically obese. The subjects were sedentary individuals who did not participate in any endurance or resistance training. The local ethics committee granted ethical permission, and all subjects gave written informed consent.
Anthropometry
Standing height was measured to the nearest centimeter using a Holtain (Crymych, Dyfed, UK) stadiometer. Weight was measured to the nearest 25 g using an electronic weighing scale (Tanita mode 1 TBF-305; Tanita UK Ltd., Yiewsley, Middlesex, UK). BMI was calculated from height and weight using the formula weight in kilograms/height in meters2. The percent body fat was measured by bioimpedance using the Tanita mode 1 TBF-305. This method correlates well with values obtained from dual energy x-ray absorptiometry (12). Waist circumference was measured using a metal tape measure.
Exercise protocol
All subjects had fasted for at least 6 h and had restrained from strenuous activity for 24 h before testing. VO2 max was determined during an incremental workload test to exhaustion with an electromagnetically brake bicycle ergometer (Lode Excalibur Sport version 2.0, Groningen, The Netherlands). The fitness test was combined with continuous analysis of expired oxygen, carbon dioxide content, and minute ventilation using the CDX/D measurement charts (Medical Graphics, London, UK). An open circuit, indirect calorimetry system was used to collect and analyze expired gases every 2 sec during exercise. Known gases were used to provide a three-point calibration regression for oxygen and carbon dioxide fractions. The ventilatory volumes were measured from inspired volumes with a Hans Rudolph pulmonary tachometer (Hans Rudolph Inc., Kansas City, MO) (13). Calibration against standard gases, operating temperature, and barometric pressure was performed before each test. All subjects began pedaling at a work rate of 50 watts. Every 2 min the workload was increased by 40 watts until the subject had reached a respiratory exchange ratio of 1 or more, the point at which the production of carbon dioxide equals the consumption of oxygen. After this, the workload was increased by 20 watts every 2 min until maximal exertion. This protocol was adopted to ensure that the length of the test lasted approximately 12–15 min in total. Subjects were encouraged vocally during the test. To ensure that a valid VO2 max was attained, subjects were encouraged to continue until there was no further rise in VO2, and the respiratory exchange ratio was 1.1 or more. At the beginning of the recovery, subjects were advised to continue cycling at low intensity (70 watts) for a further 3 min. The within subject coefficient of variation (CV) for VO2 max with this test is less than 10%. The within-subject CVs for peak GH and the integrated measurement of GH response, expressed as the area under the curve (AUC), were 29.4% and 26.3%, respectively (13). These values are at least as reproducible as other GH provocation tests (14). The between-subject CVs for peak GH and AUC are 74% and 68%, respectively.
Blood samples were taken from an indwelling plastic cannula, which was inserted 30 min before exercise. Blood samples were obtained 5 min before exercise, at the start of exercise, every 4 min during exercise, at peak exercise, and then at 5, 10, 20, 30, and 45 min during recovery. Blood for lactate was collected into chilled glass tubes containing 10 mg sodium fluoride and 5 g potassium oxalate and stored on ice until analysis. Blood for GH, IGF-I, leptin, and insulin determinations was collected into plain glass tubes. The specimens were allowed to coagulate at room temperature and were then centrifuged for 10 min at 2500 rpm at 4 C. The serum was separated and stored at -20 C until analysis.
Assays
Serum GH was measured by a two-site immunoradiometric assay, which has been previously described (15). The intraassay CVs at analyte values of 1, 10, and 50 mU/liter were 3%, 1%, and 1.5%, respectively. The interassay CVs at analyte values of 1.68, 12.1, and 27.2 mU/liter were 10%, 4.1%, and 5.4%, respectively. The level of sensitivity of the assay is 0.2 mU/liter.
Serum leptin was measured by a commercial immunoradiometric assay (DSL-23100 human leptin immunoradiometric assay kit, Diagnostic Systems Laboratories, Inc., Webster, TX). The intraassay CVs at analyte values of 2.75, 13.5, and 73.6 ng/ml were 3.7%, 4.9%, and 2.6%, respectively. The interassay CVs at analyte values of 2.83, 14.35, and 73.9 ng/ml were 6.6%, 5.3%, and 3.7%, respectively. The level of sensitivity of the assay is 0.1 ng/ml.
Serum IGF-I was measured by the Nichols advantage chemiluminescence immunoassay kit. The intraassay CVs at analyte values of 63, 208, and 766 nmol/liter were 4.8%, 5.2%, and 4.4%, respectively. The interassay CVs at analyte values of 62, 215, and 811 nmol/liter were 7.1%, 5.7%, and 7.4%, respectively. The level of sensitivity of the assay is 6 nmol/liter.
Serum insulin was measured by the Immulite insulin chemiluminescent enzyme immunometric assay kit (Diagnostic Products, Los Angeles, CA). The intraassay CVs at analyte values of 10.7, 41, and 439 pmol/liter were 4.8%, 5.4%, and 3.8%, respectively. The interassay CVs at analyte values of 10.7, 41, and 439 pmol/liter were 5.8%, 7.6%, and 4.8%, respectively. The level of sensitivity of the assay is 2 pmol/liter.
Plasma lactate was measured by the fully automated YSI, Inc. 2000 industrial lactate analyzer (Yellow Springs, OH).
Statistics
Results are given as the mean ± SEM. The data were analyzed in two ways. First, the differences in serum levels of GH, IGF-I, and insulin and measures of adiposity within the four groups were analyzed by ANOVA, and when P < 0.05, the calculation was completed with Fisher’s least significant difference test. Leptin was log transformed because of its log-normal distribution. Secondly, age, fat mass, and fitness were further analyzed as a continuous whole. The relationships among age, anthropometrical characteristics, fitness, and GH were assessed by simple regression analysis. As the variables are not strictly independent of each other, multivariate analysis was also performed. The computer system and program used for analyzing the results were Apple Macintosh (Apple Computer Inc., Cupertino, CA) and StatView (Abacus Concepts, Inc., Berkeley, CA). Values below the limit of detection were assigned values at the lower limit of detection of the assay.
Results
Subject characteristics
The baseline characteristics are summarized in Table 1
. There was no significant difference in
baseline fasting GH levels among the four groups. Fasting serum insulin
was significant lower in the two lean groups compared with the
overweight groups (P = 0.0012). There was no
significant difference in the peak lactate levels reached during
exercise, indicating that each group exercised to a similar intensity.
Serum IGF-I levels were significantly higher in the two young groups
compared with the older groups (P = 0.02). Baseline
serum leptin concentrations were significantly higher in the overweight
groups compared with the lean groups (P = 0.006). There
was a strong correlation between BMI and measures of adiposity (data
not shown).
|
In all subjects GH increased progressively throughout the exercise
period, peaked 10 min (range, 0–30 min) after exercise, and thereafter
decreased during recovery (Fig. 1). The
peak GH and GH AUC responses were significantly greater in the young
groups compared with the older groups (P < 0.05; Fig. 2). When age was considered as a
continuous variable for the whole group, the GH response correlated
inversely with age (peak GH: r = -0.6; P =
0.0009; AUC: r = -0.6; P = 0.0005). Even within
the young group, the GH response correlated inversely with age (peak
GH: r = -0.7; P = 0.009; AUC: r = -0.7;
P = 0.005).
|
|
Peak GH levels were significantly higher in the old lean group
compared with the old overweight group (P = 0.03), but
there was no significant difference in AUC (Fig. 2
). In the old groups,
GH secretion (peak and AUC) was correlated with both physical and
biochemical markers of obesity (Table 2).
In contrast, there was no significant difference in peak GH or AUC
between the young lean and young overweight groups. Furthermore, in the
young groups there was no correlation between BMI, bioimpedance,
leptin, or fat mass and GH secretion. The peak GH was reached earlier
in the recovery stage in the young overweight group (mode, 0 min;
range, 0–10 min) than in the young lean group (mode, 10 min; range,
5–30 min). There was no correlation between lean body mass and either
peak GH or AUC in the young or old men or in the group as a whole.
|
VO2 max was significantly lower in the old
overweight group compared with the other groups (P <
0.05; Table 1). When considered as a whole, age, waist circumference,
and serum leptin were negatively correlated with
VO2 max. In contrast, VO2
max did not correlate with bioimpedance and was only weakly correlated
with BMI (Table 3). When the young and
old groups were analyzed separately, the associations between
VO2 max and measures of adiposity were only found
in the old group (Table 3), suggesting the age is a more important
determinant of VO2 max than adiposity. After
multivariate analysis of the whole group, age and individual measures
of adiposity were independently associated with
VO2 max.
|
VO2 max demonstrated a curvilinear
relationship with peak GH secretion (r = 0.72; P =
0.0008) and GH AUC (r = 0.7; P = 0.0012; Fig. 3). The association was
stronger in the old men (peak GH: r = 0.8; P =
0.0007; GH AUC: r = 0.7; P = 0.005) than in the
young men (peak GH: r = 0.5; P = 0.05; GH AUC:
r = 0.5; P = 0.05). After multivariate analysis of
the whole group using exercise, adiposity, and age, peak GH secretion
was correlated with VO2 max only
(P = 0.04), and GH AUC was inversely correlated with
age only (P = 0.03). When multivariate analysis using
VO2 max and measures of adiposity was performed
in the old group, only VO2 max was independently
associated with GH secretion.
|
Baseline IGF-I correlated positively with peak GH secretion
(r = 0.7; P = 0.0002) and GH AUC (r = 0.7;
P = 0.0001; Fig. 4).
|
This is the first study to compare simultaneously the effects of adiposity and aging on the exercise-induced generation of GH. We have shown that GH production declines significantly with age, and this variable is more important in determining the response than adiposity. In young subjects, adiposity had no effect on GH secretion, and an association between fat and GH secretion was only seen in the older men. VO2 max declined with age and was predictive of GH secretion. On multivariate analysis, only VO2 max and age were associated with GH secretion, with no independent effect of obesity. The serum IGF-I concentration was predictive of the GH response to exercise. This study has several important clinical implications. The decline in GH appears to be independent of changes in body composition, and therefore the age-related fall in GH secretion, or somatopause, may be causal in determining the changes in body composition with age. VO2 max, as a marker of fitness, is an independent predictor of peak GH secretion, suggesting that the maintenance of physical fitness is important in preserving GH secretion. Serum IGF-I is a good marker of integrated GH secretion.
Numerous studies have shown associations between increasing age and declining GH secretion (5, 9, 16, 17, 18, 19). Our data confirm the decline in GH production and serum IGF-I with age. Increasing age was associated with a decline in VO2 max. This is a surrogate marker for fitness and therefore supports the hypothesis that the decline in GH production with age is associated with a decline in physical fitness. On multivariate analysis, we found that peak GH secretion was associated with VO2 max, whereas the AUC was associated with age. Although age has previously been found to be independently associated with an increase in the volume of distribution and the MCR of GH (20), the current study does not define the relative contributions of secretion and clearance of GH to the reduced GH concentrations in the older group. Furthermore, it remains unclear whether the decline in GH reflects the effects of aging itself or the influences of age-associated changes in lifestyle, such as a decline in maximum aerobic capacity. It is also unclear from this study whether GH production regulates fitness or vice versa.
Previous studies have shown that increasing adiposity is associated with a decline in GH production (9, 11, 21) and increased clearance of GH (20). In contrast to these studies, we only found a relationship between biochemical and physical indicators of fat mass and GH secretion in the older age groups. In contrast to this, there was no association between fat mass and GH secretion in the young groups, implying that age is more important than fat mass in determining the exercise-induced generation of GH in young men. It is intriguing to note that the GH peak occurred earlier in the young overweight group than in the lean group. It is not clear why this happened, but we speculate that this may reflect a quicker rise in body temperature in the overweight group. In contrast to the elderly groups, there was no difference in VO2 max between the young lean and overweight groups. As the GH response correlated strongly with VO2 max, the lack of an effect of obesity in the young groups may reflect similar levels of fitness. Indeed, physical fitness may be the underlying reason behind the disparity in the GH response in the older groups. On multivariate analysis the effect of adiposity was lost when age and fitness were taken into account. Therefore, these data support the hypothesis that GH production regulates body composition, and the age-related changes in body composition are a consequence of a fall in the GH concentration with age.
The bulk of circulating IGF-I is derived from the liver, where its secretion is regulated by GH and nutritional status (22). Unlike GH, circulating IGF-I concentrations remain stable throughout the day. In normally nourished individuals, it has been suggested that serum IGF-I can be used as a biochemical marker of GH secretion (23, 24, 25). However, in clinical practice, serum IGF-I does not always correlate well with either arginine or insulin-stimulated GH secretion in normal and GH-deficient adults (26). In contrast, our findings that serum IGF-I concentrations correlated well with GH production during maximum exercise suggest that exercise-stimulated GH release is qualitatively different from pharmacologically stimulated release and may be physiologically relevant (27).
Although the study has produced interesting results, it has some weaknesses. The effects of aging on GH secretion were larger than we had expected. This was particularly marked in the younger age groups, where there were significant differences within the group attributable to age. In retrospect, it may have been better not to have adopted a factorial design for age and studied a wider range of ages as a continuous variable. We had chosen to adopt a factorial design to maximize the study power in the time available. However, despite this, the sample sizes may have been too small. Finally, we accept that although bioimpedance correlates with dual energy x-ray absortiometry measures of fat mass, it is not the gold standard. Nevertheless, we were reassured by the close correlation between all measures of adiposity.
In conclusion, our study confirmed the decline in GH secretion seen with age. We have shown that adiposity is not an independent regulator of GH secretion during exercise. In contrast, the variation in GH secretion was more closely related to changes in physical fitness as assessed by VO2 max. This has important clinical implications if we are to prevent the frailty and morbidity associated with old age.
Acknowledgments
We thank Jenny Jones, Abu Fofanah, Sue Bowes, Bill Johnson, and Robert Jupp for the assistance with the laboratory assays.
Footnotes
This work was supported by Department of Medicine, Guy’s, King’s, and St. Thomas’ School of Medicine, and formed the basis of a B.Sc. project for E.W.
Abbreviations: AUC, Area under the curve; BMI, body mass index; CV, coefficient of variation; VO2 max, maximal oxygen uptake.
Received January 26, 2001.
Accepted August 29, 2001.
References
This article has been cited by other articles:
 |
|
 |
|
  J. Gibney, M.-L. Healy, and P. H. Sonksen The Growth Hormone/Insulin-Like Growth Factor-I Axis in Exercise and Sport Endocr. Rev., October 1, 2007; 28(6): 603 - 624. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Weltman, J. Y. Weltman, C. P. Roy, L. Wideman, J. Patrie, W. S. Evans, and J. D. Veldhuis Growth hormone response to graded exercise intensities is attenuated and the gender difference abolished in older adults J Appl Physiol, May 1, 2006; 100(5): 1623 - 1629. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. R Christ, M. Zehnder, C. Boesch, R. Trepp, P. E Mullis, P. Diem, and J. Decombaz The effect of increased lipid intake on hormonal responses during aerobic exercise in endurance-trained men. Eur. J. Endocrinol., March 1, 2006; 154(3): 397 - 403. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Wong and V. Harber Lower Excess Postexercise Oxygen Consumption and Altered Growth Hormone and Cortisol Responses to Exercise in Obese Men J. Clin. Endocrinol. Metab., February 1, 2006; 91(2): 678 - 686. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
