Growth Hormone & IGF Research 13 (2003) 225–238

www.elsevier.com/locate/ghir

Review

Growth hormone responses to sub-maximal and sprint exercise

Keith Stokes*
Department of Sport and Exercise Science, University of Bath, Bath BA2 7AY, UK

Received 9 July 2002; received in revised form 21 March 2003; accepted 21 March 2003

Abstract

Exercise is a potent stimulus for growth hormone (GH) release and a single bout of exercise can result in marked elevations in
circulating GH concentrations. The magnitude of the GH response to exercise will vary according to the type, intensity and duration
of exercise as well as factors such as the age, gender, body composition and fitness status of the individual performing the exercise.
However, the mechanisms regulating GH release in response to exercise are not fully understood. This review considers the GH
responses to sub-maximal and sprint exercise and discusses the factors that might affect GH release along with the mechanisms that
have been proposed to regulate exercise-induced GH release.
Ó 2003 Elsevier Science Ltd. All rights reserved.

Keywords: Prolonged exercise; High-intensity exercise; Repeated exercise; Metabolic responses; Endocrine; Human

1. Why is growth hormone of interest to an exercise is important to understand the relationship between
physiologist? exercise and GH release, and this review will attempt to
describe the natural response to acute bouts of exercise.
In recent years a great deal of attention has been paid However, the study of the relationship between ex-
to the abuse of growth hormone (GH) in sport, and ercise and GH release is complicated by a number of
whilst this alone might attract the interest of an exercise factors. First, historically and currently there have been
physiologist, there are other reasons why GH is studied a wide range of assays available for the measurement of
in sport and exercise physiology. Growth hormone is a GH. This makes the comparison of findings from dif-
potent anabolic agent, and the natural response to ex- ferent studies difficult. The Growth Hormone Research
ercise is, therefore, of great interest to sports performers Society has suggested that including a clear statement of
who are trying to maximise the anabolic effect of their methodology when reporting assay data might reduce
training regimen. Similarly, with the decline of resting this problem [1]. In addition, to improve standardisa-
circulating GH levels being a natural part of the ageing tion, it has been recommended that the GH reference
process, the use of exercise as an intervention to con- preparation should be a recombinant 22 kDa human
tribute to the maintenance of functional capacity could GH (IRP 88/624 [1] or IRP 98/574 [2]), each with a
have a major impact in the current climate of an ageing specific activity of 3 IU/mg rather than previously used
western society. Indeed, even a brief search on the In- pituitary derived reference preparations.
ternet provides evidence of the level of interest in what Second, although it is hoped that strict controls are in
are termed ‘‘GH supplements’’ as a so-called ‘‘fountain place in studies reporting on GH responses to exercise, it
of youth’’. Exercise might offer an alternative to these is possible that GH pulses as a result of the stress of
supplements as well as providing a multitude of other venous section, or stress related to anticipation might
health benefits. Therefore, there are many reasons why it affect the results. For example, some authors have re-
ported increased levels of GH following arterial and
venous puncture [3,4]. A further difficulty in comparing
*
Tel: +44-1225-384190; fax: +44-1225-383275. the findings of different studies relates to the fact that
E-mail address: [email protected]. some studies report GH release using measures such as

1096-6374/$ - see front matter Ó 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S1096-6374(03)00016-9
226 K. Stokes / Growth Hormone & IGF Research 13 (2003) 225–238

integrated GH concentrations and maximum GH con- literature reporting the effect of sub-maximal exercise
centrations, whilst others assess GH secretion using duration, when intensity is maintained, on the GH re-
cluster analysis or deconvolution analysis. These will sponse. However, in general, GH levels have been re-
give different insights into the relationship between ex- ported to peak at or near the end of sub-maximal
ercise and GH. In addition, whilst it is clear that exercise exercise. Table 1 gives some examples of the GH re-
alters circulating GH concentrations and that GH has sponse to sub-maximal treadmill running.
important actions in growth and metabolism, there is Sprint exercise is sometimes erroneously termed su-
little available evidence how GH alters cellular signalling pra-maximal since it is above V_ O2 max, or anaerobic
in tissues to mediate these actions. It is important that exercise since the energy for this form of exercise is
this issue is addressed in order that the mechanisms of predominantly from anaerobic metabolism, although
GH action are better understood. these terms are inaccurate for a number of reasons. In-
deed, the suggestion that exercise can be ‘‘supra’’-max-
imal is clearly paradoxical, and there is now evidence
2. The GH response to different types of exercise that there can be a large aerobic contribution to sprint
exercise. This review will use the term sprint exercise. As
One of the problems when studying the relationship with endurance exercise, running, cycling, swimming
between intense exercise and GH release is that there are and other modes of exercise can be included in the sprint
a number of different types of exercise that can be used. exercise category, as long as it is performed at an in-
For the purpose of this review, the terms sub-maximal tensity that is sustainable for only a short period of time
exercise and sprint exercise will be used, although even (up to about 60 s). Table 2 gives examples of the GH
within these categories there are a number of additional response to all-out sprinting on a cycle ergometer and
subdivisions. Sub-maximal exercise is any exercise that on a treadmill. It is clear from these examples that the
is below the maximal aerobic capacity, or maximal ox- GH response to sprint exercise is influenced by duration
ygen uptake ðV_ O2 max). As such, sub-maximal exercise of exercise [7]. In addition, mode of exercise may effect
includes any mode of exercise (e.g., walking, running, the GH response, for example, treadmill sprinting where
swimming, cycling) that is typically sustainable for more both arms and legs are active, has been suggested to
than about 5 min. It appears that low-intensity exercise elicit a greater GH response than sprint cycling, where
(50% of the lactate threshold) does not elicit GH se- movement is concentrated in the lower limbs, since a
cretion, whilst high-intensity exercise (50% of the dif- larger muscle mass is employed [8].
ference between the lactate threshold and V_ O2 max)
stimulates a GH response [5] suggesting that a threshold
intensity of exercise must be reached before a significant 3. Factors influencing exercise-induced GH release
increase in GH concentration can be detected. However,
there has been no systematic study of this threshold The mechanisms regulating GH secretion are yet to
concept, and recent evidence suggests that the GH re- be fully elucidated. However, it would appear that a
sponse to sub-maximal treadmill running displays a combination of hypothalamic release of GH releasing
linear dose–response relationship as GH secretion in- hormone (GHRH), withdrawal of somatostatin and
creases with rising intensity [6]. There is a similar lack of release of a GH releasing peptide (GHRP), such as the

Table 1
Examples of the mean peak GH response to sub-maximal exercise
Sub-maximal exercise Subjects Exercise Peak [GH] (lg/l)
Kanaley et al. [73] 8 F (non-obese) 30 min @ 70% V_ O2 peak 13.7
11 F (LBO) 30 min @ 70% V_ O2 peak 6.8
12 F (UBO) 30 min @ 70% V_ O2 peak 3.5
Pritzlaff et al. [6] 10 M (young active) 30 min @ 26% V_ O2 peak
2.5
30 min @ 47% V_ O2 peak
4.5
30 min @ 62% V_ O2 peak
5.5
30 min @ 76% V_ O2 peak
9.5
30 min @ 90% V_ O2 peak
14

Pritzlaff-Roy et al. [64] 8 F (young active) 30 min @ 33% V_ O2 peak
7

30 min @ 49% V_ O2 peak
12
30 min @ 62% V_ O2 peak
17
30 min @ 76% V_ O2 peak
22
30 min @ 86% V_ O2 peak
26
F, female; M, male; LBO, lower body obese; UBO, upper body obese.
 K. Stokes / Growth Hormone & IGF Research 13 (2003) 225–238 227

Table 2
Examples of the mean peak GH response to sprint exercise
Sprint exercise Subjects Exercise Peak [GH] (lg/l)
Nevill et al. [21] 5 F (ST) 30 s all-out treadmill sprint 26.9
6 F (ET) 30 s all-out treadmill sprint 10.4
6 M (ST) 30 s all-out treadmill sprint 44.0
6 M (ET) 30 s all-out treadmill sprint 15.9
Stokes et al. [7] 9 M (young active) 6 s all-out cycle sprint 4.0
30 s all-out treadmill sprint 18.5
Stokes et al. [8] 10 M (young active) 30 s all-out treadmill sprint 20.4
F, female; M, male; ST, sprint trained; ET, endurance trained.

putative endogenous GHRP-like ligand Ghrelin, medi- mechanism that stimulates anaerobic metabolism might
ate the effects of a number of candidate stimuli. During result in an increase in blood lactate concentrations
and after exercise there are a number of potential stimuli whilst also providing a signal for GH release.
for GH release, but the factors that determine the
magnitude of GH release remain unclear. Various 3.2. Increased hydrogen ion concentration in the blood
mechanisms regulating the GH response to exercise have
been proposed, as reviewed in the following sections. It has been suggested that any stimulus for GH re-
lease associated with lactate accumulation would be
3.1. Increased blood lactate concentrations more likely to act through hydrogen ion ðHþ Þ accu-
mulation [15]. Sutton et al. [14] studied the role of acid–
A promising candidate for regulating the secretion of base balance during exercise in the regulation of the
GH in exercise was identified when a correlation be- exercise-induced GH response. Acidotic, alkalotic and
tween blood lactate concentrations and GH concentra- control conditions were considered during a
45 min
tions was observed [9]. In a study employing arm graded exercise test to exhaustion. Although in the first
cranking exercise, leg bicycle exercise and treadmill 20 min of the exercise test acidosis resulted in a signifi-
running, at an intensity eliciting similar oxygen con- cantly higher circulating GH concentration than the
sumption in all trials, serum GH concentrations were alkalotic or control conditions, this did not extend to the
also found to be positively correlated with blood lactate second 20 min or the last 5 min stage. It was therefore
concentrations [10]. Furthermore, significant correla- concluded that, although exercise is a clear stimulus for
tions were found between plasma lactate and plasma GH release, the stimulatory mechanism acts indepen-
GH during 20 min of continuous ‘‘aerobic’’ exercise and dently of blood ½Hþ .
during 20 min of intermittent ‘‘anaerobic’’ exercise of However, it was not clear whether the same conclu-
nearly equal external work, with significantly higher sion could be drawn for short-term, high-intensity ex-
plasma lactate and plasma GH concentrations following ercise. Therefore, male subjects were administered
‘‘anaerobic’’ exercise [11]. NaHCO3 or NaCl placebo in a randomised double-
However, another study employing continuous and blind counterbalanced experiment with a crossover de-
intermittent exercise of equal external work identified a sign [15]. All-out high intensity exercise of 90 s duration
similar GH response in each trial, with a divergent resulted in a larger and faster rise in serum GH con-
pattern of blood lactate accumulation [12]. The differ- centrations in the placebo trial in all 10 subjects. Cor-
ence in the results of these two trials was attributed to relation analysis in the placebo trial alone demonstrated
the fact that ‘‘anaerobic’’ exercise was not employed by significant correlation between highest measured ‘‘peak’’
Karagiorgos et al. [12], reflected by higher measured GH and both peak ½Hþ  and peak lactate concentra-
blood lactate concentrations in the study of Van Helder tions. Overall correlations, combining data from both
et al. [11]. Further evidence that the GH response to trials, showed that highest measured ‘‘peak’’ GH con-
exercise is independent of lactate accumulation in the centration was correlated with peak ½Hþ  but not peak
blood is that artificial manipulation of blood lactate lactate concentration, suggesting that the GH response
levels using sodium lactate [13,14] have been shown to to high intensity exercise is more highly associated with
have no consistent effect on GH concentration. Indeed, peak venous ½Hþ  than with peak venous lactate con-
both Van Helder et al. [11] and Kozlowski [10] reflected centration. The use of combined oral and intravenous
on the findings of Sutton et al. [14] using sodium lactate, administration of either or NaCl placebo during an in-
and accepted that their results should not be considered cremental exercise test to exhaustion identified that
to demonstrate a causal link between blood lactate and NaHCO3 administration tended to suppress exercise-
serum GH concentrations. Instead, it is possible that a induced GH release at rest and during recovery from
228 K. Stokes / Growth Hormone & IGF Research 13 (2003) 225–238

exercise, except for the peak value at 60 min which was GH release [13,14] is evidence that systemic metabolic
similar in the two trials [16]. However, the mechanism disturbance might not regulate GH, and Karagiorgos et
by which an acute rise in blood ½Hþ  acts as a stimulus al. [12] found no correlation between any anaerobic
for GH release is not clear. metabolite or oxygen deficit and GH concentration.
Even if anaerobic metabolites in the blood do not have a
3.3. Oxygen demand/availability ratio role to play in the regulation of exercise-induced GH
release, this would not preclude the possibility that they
It is known that GH levels increase as oxygen de- are detected in the muscle [10,11,19,21–23].
mand increases and a significant correlation between It has been suggested that neural afferent signals from
initial O2 deficit and peak GH concentrations during 1 h muscle metabolic receptors, activated by local changes
of cycle ergometer exercise has been identified [17]. In in lactate concentration, oxygen concentration or pH,
addition, Raynaud et al. [18] demonstrated that indi- might participate in the activation of catecholamine re-
viduals dwelling at sea level who were non-adapted to lease [24,25]. Van Helder et al. [19] cited the close as-
hypoxia had a greater GH response to exercise under sociation between GH and the D/A ratio as support for
acute hypoxic conditions and when exercising at altitude the suggestion that similar metabolic receptors occur in
than when exercising under normal conditions. This the muscle with a role in GH regulation during exercise.
highlights a possible role for oxygen availability in reg- Ischaemic exercise, resulting in enhanced accumulation
ulating the GH response to exercise. However, under of lactate in the muscle, has been associated with a
hypoxic conditions, it would be expected that exercise twofold increase in GH concentrations compared to
would induce a greater metabolic response, and, there- normal exercise and this was attributed to the activation
fore, it might not be hypoxia per se that is regulating of muscle receptors [23]. However, Kjaer et al. [26]
GH release. In fact, Raynaud et al. [18] did report a blocked afferent nerve activity by epidural anaesthesia in
higher blood lactate response under hypoxic conditions order to test the hypothesis that afferent nervous activity
and when exercising at altitude than when exercising in from exercising muscle regulates GH release. Epidural
normoxic conditions. blockade had no effect on the GH response to exercise,
Therefore, it was suggested that the GH response to yet it was postulated that afferent nervous activity might
exercise might be proportional to the ratio of oxygen still have a role to play in the regulation of exercise-in-
demand to oxygen availability (D/A ratio) [19]. The D/A duced GH release, since elevated central motor activity
ratio was applied to the results of previously published might compensate for the reduction in afferent sensory
research, and a highly significant relationship between signals.
the GH response and the D/A ratio for continuous If afferent feedback from muscle metabolic receptors
‘‘aerobic’’ and intermittent ‘‘anaerobic’’ exercise [11] contributes to the regulation of GH release, then exer-
was identified [19]. Perhaps more surprisingly, there was cising under hypoxic conditions would be expected to
a highly significant relationship between GH and the exaggerate metabolic changes in contracting muscle and
D/A ratio in the data of Karagiorgos et al. [12]. This enhance the GH response to exercise. Epidural anaes-
finding was despite the fact that originally the results of thesia would blunt this response, yet Kjaer et al. [27]
that study identified no significant relationship between demonstrated that epidural anaesthesia during leg
blood lactate concentrations and the GH response to cycling exercise at
50% V_ O2 max under hypoxic con-
exercise or even between the oxygen deficit in continu- ditions enhanced, rather than blunted, the exercise-in-
ous exercise and serum GH concentrations. The data duced GH response. Epidural anaesthesia also reduced
from these studies [11,12], as well as those from other muscle strength and increased perceived exertion, sug-
previously published studies [18,20], were combined and gesting a role for ‘‘central command’’ in the exercise-
resulted in the demonstration of a close association be- induced GH release, rather than regulation by afferent
tween the D/A ratio and the GH response to exercise for feedback from receptors in exercising muscle.
a wide range of exercises [19]. The relationship between
the D/A ratio and GH release was proposed to suggest 3.5. Motor centre activity
that differences in oxygen demand and supply at the
muscle site might regulate GH release via muscle met- Activity in motor centres may directly stimulate pi-
abolic receptors [19]. tuitary hormone release, including GH, during exercise
[10,26,28,29]. Administration of tubocurarine is re-
3.4. Afferent signals from muscle metabolic receptors ported to induce a partial neuromuscular blockade,
which increases voluntary effort during exercise and
It might be that a combination of factors related to therefore necessitates higher activity in motor centres
anaerobic metabolism are involved in controlling GH [30]. Kjaer et al. [28] administered tubocurarine prior to
release [21]. However, the finding that there is no rela- exercise and demonstrated higher motor activity, mea-
tionship between externally administered lactate and sured through an increased rate of perceived exertion,
 K. Stokes / Growth Hormone & IGF Research 13 (2003) 225–238 229

compared to exercise without tubocurarine administra- vivo bioassay [33]. The assay is understood to measure
tion. At the same time exercise with tubocurarine elic- variants of GH other than 22 kDa GH, and various lines
ited a greater GH response than exercise alone, of evidence suggest that BGH is distinct from IGH [33].
suggesting that central motor activity might play a role However, the physiological significance of BGH, be-
in the regulation of the GH response to exercise. In a yond linear bone growth [32], is yet to be elucidated.
further study using epidural anaesthesia to block affer- It has recently been suggested that skeletal muscle
ent nerve activity, decreased muscle strength and higher proprioceptor afferents might contribute to the regula-
rates of perceived exertion during exercise with epidural tion of BGH release, since electrical stimulation has
blockade inferred increased motor centre activity [26]. been shown to result in increased release of BGH, but
Since epidural anaesthesia had no apparent effect on the not IGH, in rats [34]. Vibration-induced activation of
GH response to exercise it was suggested that this in- muscle spindle afferents in the tibialis anterior, but not
crease in motor centre activity compensated for the lack the soleus, has also been shown to result in elevated
of afferent nervous input. plasma BGH concentrations in human [35]. Immuno-
However, Kjaer et al. [29] did not observe any de- assayable GH was not altered by vibration of either
crease in the GH response to exercise with both afferent muscle. These studies suggest that there is a proprio-
sensory blockade by epidural anaesthesia combined with ceptive mechanism by which the release of BGH is
electrically induced cycling, to offset motor centre ac- regulated, whilst no such mechanism exists for IGH.
tivity. These results suggested that blood-borne humoral However, performing an all-out 30 s sprint on a cycle
feedback mechanisms and autonomic (i.e., spinal) re- ergometer at fast (
160 rpm) pedalling rates has been
flexes are capable of inducing the GH response to ex- shown to result in a greater IGH response when com-
ercise. It was postulated that a decrease in plasma pared with pedalling at slower ð
130 rpm) pedalling
glucose, as observed in this study, was of great impor- rates with 9 out of 10 subjects following this pattern of
tance in the control of the GH response to exercise. A response, although the difference was not significant
further study by Kjaer et al. [31] compared electrically ðP ¼ 0:05Þ [8]. This is despite the fact that there are no
induced leg cycling in tetraplegic humans with voluntary differences in the blood [8,36] or muscle [36] metabolic
arm cranking at a work rate similar to that achieved responses to sprinting at different pedalling rates. It is,
during the involuntary leg stimulation trials. It was therefore, possible that a proprioceptive mechanism
observed that GH concentrations increased as a result of might contribute to the regulation of IGH, as well as
voluntary arm exercise, but not with involuntary leg BGH, release.
stimulation. These results support the suggestion that an
intact central nervous system and activity in motor 3.7. Catecholamines
centres, as well as afferent nerves from exercising mus-
cles, are needed for the GH response to exercise. In During progressively incremental exercise, blood
addition, arm cranking exercise appeared to result in an catecholamines, adrenaline (A) and noradrenaline (NA)
exaggerated GH response, probably due to higher motor have been shown to rise with increasing exercise inten-
centre activity relative to work output, since the subjectsÕ sity [37]. In addition, catecholamines can directly stim-
arm muscles were weakened by partial paralysis. This ulate GH release from rat pituitary tissue in vitro [38]. In
finding provides further support for a role of motor exercising humans, a significant positive correlation
centre activity in the regulation of the GH response to between plasma [NA] and serum GH concentrations has
exercise. These studies do not consider a possible role of been identified [10]. These findings were supported by
psychological perception of exercise and the associated those of Chwalbinska-Moneta et al. [39] who also
stress signalling which might parallel motor centre ac- demonstrated significant correlations between catechol-
tivity. However, the potential importance of motor amine concentrations and serum GH concentrations
centre activity in regulating GH release in response to during and after an incremental exercise test. In addi-
exercise is evident. tion, the GH-threshold identified by Chwalbinska-
Moneta et al. [39] during incremental exercise was
3.6. A proprioceptive mechanism reported to occur at a similar work load as what was
identified as both the [A]-threshold and the [NA]-
Another potent pituitary growth factor has been threshold, as well as the lactate threshold.
identified that cannot be measured by standard GH Peripheral markers of heightened adrenergic outflow,
immunoassay and that has apparently distinct mecha- that is [A] and [NA], have been shown to precede and
nisms regulating its release [32]. This growth factor has correlate with exercise-induced GH concentrations [40].
been labelled as bioassayable GH (BGH) to differentiate A time delay between peak-[A] or peak-[NA] and peak-
between it and immunoassayable GH (IGH) since the [GH] of
20 min was identified and changes in exercise
method measures GH activity using changes in tibial intensity did not alter this interval. In addition,
epiphyseal width in hypophysectomised rats as an in increasing intensity resulted in a linear relationship
230 K. Stokes / Growth Hormone & IGF Research 13 (2003) 225–238

between the increment (change from baseline to peak) in lation of GH release as a result of exercise. It is possible
GH and the increment in A as well as the increment in that the exact mechanism regulating GH release is de-
NA. Multiple linear regression showed that the domi- pendent on the form of exercise being performed, and it
nant relationship was between incremental changes in certainly appears that environmental factors, such as
GH and NA. These results suggest that higher exercise temperature have an impact upon the GH response to
intensities might drive increased GH secretion, at least in exercise. A proposed model for the regulation of exer-
part, by central adrenergic activation [40]. However, It cise-induced GH release would be that, at the onset of
has been suggested that a decrease in pH in contracting exercise, impulses in motor centre elicit a workload de-
muscles due to accelerated lactate production and as- pendent increase in GH secretion [28]. Blood-borne
sociated metabolic changes may stimulate the sympa- metabolic error signals and/or neural afferent signals
thetic outflow by neural afferent signals from muscle from muscle metabolic receptors, proprioceptive feed-
metabolic receptors causing rapid release of catechol- back and, at higher exercise intensities, central adren-
amines [24,25]. Catecholamines released following af- ergic activation might then feed back to modulate
ferent signals from muscle metabolic receptors might, in further GH release.
turn, play a role in the regulation of GH release.

3.8. Change in core temperature 4. Roles of GHRH and somatostatin in the GH response
to exercise
Brenner et al. [41] identified the importance of motor
centre activity in the regulation of exercise-induced GH It has been proposed that relatively low intensity
release, but suggested that the role of the increase in core exercise induces moderate GH responses through acti-
temperature associated with exercise cannot be ruled vation of the central cholinergic system, resulting in a
out. An early study identified greater GH release in se- reduction in hypothalamic somatostatin release [45].
ven out of nine subjects when exercising for 20 min at a However, it appears that there is an upper limit to this
temperature of 21 °C compared with 4 °C [42]. Later process and at higher exercise intensities, once hypo-
studies support this finding, and it has been shown that thalamic somatostatinergic tone is completely sup-
the use of a thermal clamp technique during exercise pressed, further increases in GH release must be
halves the GH response, whilst passive heating results in mediated by an increase in GHRH secretion.
similar patterns of both temperature change and serum The importance of the inhibition of somatostatinergic
GH concentrations to those seen in exercise in cold tone as a result of exercise was demonstrated by Di
conditions [43]. Luigi et al. [46], who observed a suppression of the GH
Using a single subject, Buckler et al. [42] reported that response to treadmill exercise, at 60% V_ O2 max, fol-
there was a significant positive correlation between the lowing pretreatment with the somatostatin analogue
rate of change in core temperature and the rate of change octreocide, in humans. In contrast, administration of the
in GH concentrations. A similar finding was identified somatostatin inhibitor, pyridostigmine, has been shown
when 30 min of cycling was performed at room temper- to enhance exercise-induced GH release [47,48], sug-
ature (23 °C) compared with hot conditions (40 °C) [41]. gesting that pyridostigmine and exercise might act in-
Based on this evidence, the rate of rise in body tempera- dependently in eliciting the GH response to strenuous
ture during exercise might be more important than the exercise [48]. Administration of GHRH at the start of an
magnitude of change, and a sufficiently slow rise in core incremental exercise test lasting 25 min, with an addi-
temperature might not elevate GH concentrations [42]. tional stage at 100% V_ O2 max until exhaustion has also
However, it has also been suggested that a threshold exists been shown to have an additive effect on the GH re-
whereby increases in core temperature not exceeding sponse [45]. However, co-administration of GHRH and
0.5 °C do not elicit increases in systemic GH concentra- GH-releasing peptide-2 (GHRP-2) at the start of exer-
tions, whilst increases greater than 0.6 °C result in sig- cise further potentiated GH release [45], as did the ad-
nificant increases in circulating GH concentrations [44]. ministration of GHRP-2 before 30 min of exercise [49].
In addition, when revisiting earlier work [43] a core tem- It is possible that GHRP-2 acts via a mechanism po-
perature threshold for activation of GH release was re- tentiating the effect of endogenous GHRH and/or by
ported to exist at around 38–38.5 °C, whereafter a opposing the central actions of somatostatin [49].
significant exponential relationship between elevations in Growth hormone releasing peptides (e.g., GHRP-2)
core temperature and GH was identified [44]. act through specific receptors [50] for which a specific
endogenous ligand has been identified and named
3.9. Summary ghrelin [51]. Ghrelin has been shown to stimulate GH
release in a dose-dependent manner with a potency
From the discussion above it is clear that there are a similar to that of GHRH in culture, and intravenous
number of factors which might contribute to the regu- injection of ghrelin into male rats has been shown to
 K. Stokes / Growth Hormone & IGF Research 13 (2003) 225–238 231

induce GH release [51]. These findings suggest that overall mean 6 h GH levels. In contrast, an acute dose of
ghrelin circulating in the blood might elicit GH release, anti-somatostatin serum to male rats increased only GH
however, in healthy human adults and in GH-deficient nadir levels. In addition, administration of an acute dose
adults, sub-maximal exercise did not result in significant of anti-GHRH serum raised GH nadir levels in females
changes in plasma ghrelin concentration despite eliciting but had no effect in males. These findings suggest that
GH release [52]. Therefore, it would appear that there is the secretory pattern of somatostatin plays an important
a role for an endogenous GHRP-like ligand (ghrelin) in role in the sexually dimorphic GH secretion patterns in
regulating the GH response to exercise, but that sys- rats.
temic ghrelin is not involved [52]. Women have been identified as having greater day-
time GH serum concentrations [58] and greater 24 h GH
secretion [59,60] than men although higher integrated
5. Large inter-individual variation in the GH response to serum GH concentrations in young women have been
exercise shown to be strongly influenced by serum oestradiol
concentrations [59]. Jaffe et al. [61] compared GH se-
It has often been reported that there is a great degree cretion in women in the early follicular phase of the
of inter-individual variation in the GH response to ex- menstrual cycle, when oestrogen levels are comparable
ercise [7,18,53], and when individuals work at the same between sexes, with GH secretion in men and found it to
sub-maximal exercise intensity relative to their maximal be similar in both groups. In addition, a positive cor-
exertion, this variation remains, suggesting that inter- relation between plasma oestradiol concentrations and
individual variation is not due to exercise intensity [53]. GH secretion was identified [61]. These findings suggest
One approach that has been taken is to divide subjects that higher oestradiol concentrations in women, rather
into responders and non-responders [54], but this is than sex per se, result in greater GH secretion in women
sometimes difficult as, for example, there appears to be a than in men. However, the level(s) at which oestradiol
continuum of GH responses to sprint exercise [7]. exerts its regulatory control is not clear [62]. The ap-
Variations in the GH response to an exercise stimulus parent importance of oestradiol in determining average
are mainly due to differences in the time to peak re- daily GH secretion means that women using the con-
sponse, and the maximal value [4,7]. However, despite traceptive pill will have markedly different daily GH
large inter-individual variation, within an individual the secretion from women not using the contraceptive pill,
GH response to sub-maximal exercise has been shown to and that different formulations of contraceptive pill may
be similar over several months at a range of intensities mediate different effects depending on their oestradiol
[4]. In addition, integrated (area under the curve) GH content.
concentrations following a single 30 s sprint on a cycle Despite the fact that total secretion rates are similar
ergometer had a standard error of measurement of for males and females matched for age, relative adi-
2:9  54:3 lg l1 [55]. There is, therefore a certain in- posity and oestradiol concentrations, there do appear to
trasubject reliability in the GH response to sub-maximal be differences in patterns of GH secretion between men
and sprint exercise. and women [60,61]. Specific patterns of GH secretion
are regulators of GH bioactivity in rats and therefore
the importance of the GH secretory pattern in the reg-
6. Roles of gender, age and body composition in the GH ulation of GH action in humans has been considered
response to exercise [63]. Women have been shown to have more GH pulses
with interpulse concentrations twice as high as those of
6.1. Gender men. These reported higher interpulse GH concentra-
tions in women than in men bears a similarity to the
A number of studies have considered the sex-related differences in GH secretory patterns in male and female
differences in GH secretion in rats. Jansson et al. [56] rats [56]. In addition, GH secretion in men has been
described high amplitude GH pulses with low GH found to be dominated by large nocturnal pulses with
concentrations between pulses in male rats compared relatively low GH secretion throughout the rest of the
with less regular pulses with higher interpulse concen- day. In contrast, women had a much more uniform
trations in females. The role of GHRH and somato- pulsatile pattern of secretion throughout the day,
statin in these different secretory patterns of male and spending nearly twice as much time in active GH se-
female rats was studied by Painson and Tannenbaum cretion than men [61]. It is possible that the differences in
[57] using passive immunisation with specific antisera. In the pattern of GH secretion between men and women
female rats, a single acute dose of anti-somatostatin are attributable to a lesser role of somatostatin in wo-
serum resulted in increased plasma GH concentrations men [61]. In any case, it appears that the pattern of GH
at all time points for 6 h after administration, as well as secretion might impact upon important aspects of
an increase in GH peak amplitude, GH nadir levels, and growth and metabolism in humans [63].
232 K. Stokes / Growth Hormone & IGF Research 13 (2003) 225–238

Wideman et al. [53] reported that women had greater found age to be the most important predictor of GH
serum GH concentrations at rest and during 30 min of area under the curve in all subjects with lower GH area
high-intensity sub-maximal exercise, despite the women under the curve in older individuals, along with a greater
performing their trials in the early-follicular phase, when metabolic clearance rate (MCR). However, more re-
oestradiol levels are similar between men and women. cently it was shown that the administration of a body
The higher serum GH concentrations were attributed to weight adjusted GH bolus resulted in there being no
increased mass per secretory burst in women than men, correlation between age and MCR [68].
and women also reached peak GH concentrations ear- There is a paucity of literature regarding the effect of
lier in exercise. Despite these differences, the relative ageing on the GH response to sub-maximal and sprint
responses to a single bout of sub-maximal exercise were exercise, although there is some evidence that the growth
similar between men and women [53]. With increases in hormone response to resistance exercise is blunted with
exercise intensity, GH secretion has been shown to in- age [69]. The major limitation of work in this area is the
crease in a linear dose–response relationship in both fact that younger individuals tend to be able to do more
men [6] and women [64] (see Table 1). However, women work in a bout of exercise than older individuals.
were found to have a greater GH response to exercise at
all intensities, and it was also found that women reached 6.3. Body composition
peak GH concentrations earlier in exercise than men [6].
Mean (24 h) serum GH concentration has been
6.2. Age demonstrated to be negatively correlated with percent-
age body fat allied with a progressive increase in entropy
There is evidence that the pituitary GH pool is pre- of 24 h GH profiles with increasing percentage body fat
served with increased age yet daily GH secretion rate [66]. Abdominal adiposity has been associated with de-
has been negatively correlated with age [65,66]. This age- creased spontaneous 24 h GH secretion [70], and it has
related fall in GH secretion appears to be more pro- been estimated that each unit increase in BMI, at a given
nounced in men than in women [37]. However, which age, reduces the daily secretion rate by 6% [65]. Re-
aspect, or aspects, of secretion rate contribute to the cently, it has been shown that abdominal visceral fat
observed changes in secretion rate with age (mass of (AVF), measured using single-slice computed tomogra-
secretory burst, frequency of secretory burst or basal phy scans, is a strong predictor of 24-h GH secretion,
secretion rate) is not entirely clear. Iranmanesh et al. [65] with an inverse, curvilinear relationship between AVF
observed age to be a major negative statistical determi- and 24-h GH secretion independent of age and gender
nant of GH burst frequency and also endogenous GH [71]. In fact, the results of this study suggest that factors
half-life. However later research found that GH secre- affecting GH secretion such as age and gender might
tory burst amplitude varied inversely with age, without exert their influence in part via effects on AVF [71].
identifying any significant correlation between age and Intra-abdominal fat mass has been shown to be the
burst frequency or endogenous GH half-life [66]. From major determinant of stimulated GH secretion in heal-
the results of that study Veldhuis et al. [66] suggested thy non-obese adults [72], suggesting that abdominal
that the primary impact of age, acting with altered body adiposity might also alter the GH response to exercise.
composition, is to diminish the amount or mass of GH In fact, obese women have been shown to have a re-
secreted per burst, possibly mediated by an increase in duced GH response to sub-maximal exercise when
somatostatinergic inhibitory tone and/or decreased ac- compared with non-obese women [73] (see Table 1). Six
tivity of hypothalamic GHRH. In addition, Veldhuis et hour integrated GH concentrations were half as great in
al. [66] used an approximate entropy statistic to evaluate those with upper body obesity compared with those with
the relative degree of serial orderliness or regularity of lower body obesity, although no statistically significant
24 h serum GH concentration profiles and observed a differences were identified when subjects were separated
reduced regularity of GH secretion with age. This sug- by body fat distribution. In the same study, 16 weeks of
gests that with increasing age there is disruption in the aerobic training did not alter the GH response to exer-
pathways directing GH secretion, possibly as a result of cise despite improvements in aerobic capacity, although
a reduction in the co-ordination of the secretion of no changes in fat mass were identified [73]. The evidence
GHRH and somatostatin. seems to suggest, therefore, that high body fat percent-
It has been estimated that for men with a normal age and, in particular, high levels of AVF, will result in
body mass index (BMI), an indirect measure of body reduced GH responses to exercise.
fatness, each decade of increasing age reduces the GH In addition to gender, age and body composition
production rate by 14% and the GH half-life by 6% [65]. factors such as sex-steroid hormones, nutritional status,
Vahl et al. [67] studied the significance of age on the physical fitness/exercise training, quality and quantity
pharmacokinetics of a single fixed dose exogenous pulse, of sleep and medication use might also affect exercise-
mimicking endogenous conditions, in normal adults and induced GH secretion. All of these factors must be
 K. Stokes / Growth Hormone & IGF Research 13 (2003) 225–238 233

considered in the design and interpretation of experi- creased GHRH release, increased circulating FFA and
mental research. modulation by IGF-I have all been considered.
Lanzi and Tannenbaum [78] found spontaneous GH
release to be inhibited within 1 to 2 h after a single
7. Effect of time of day on exercise-induced GH release subcutaneous injection of rGH in rats, and it remained
completely suppressed for up to 4 h after the rGH in-
A 20 min incremental treadmill running test, with the jection. They also demonstrated no difference in the
last 5 min at 90% V_ O2 max, has been shown to induce an duration or magnitude of attenuation of the GH re-
increase in GH regardless of whether the exercise was sponse according to an acute (single subcutaneous in-
performed in the morning (between 07:00 and 08:00 h) jection) or chronic (5 day) injection regimen. Passive
or the afternoon (between 15:00 and 16:00 h) in women immunisation with specific somatostatin antiserum re-
[74]. Peak GH concentrations were measured at the end versed the rGH-induced blunting of the spontaneous
of each exercise bout and returned to resting levels GH response by restoring the amplitude of the GH se-
20 min after exercise and no changes were identified in cretory bursts. The fact that immunoneutralisation of
either the magnitude or the pattern of the exercise-in- somatostatin prevented the attenuation of spontaneous
duced GH response in the morning or afternoon. These GH release after GH pre-treatment provides strong
results agree with a study showing no diurnal variation support for a role for somatostatin in GH autoregula-
in GH release in response to insulin-induced hypo- tion. However, since the normal pattern of pulsatile GH
glycaemia [75]. Scheen et al. [22] compared continuous secretion was not restored by passive immunisation with
bed-rest with 3 h of high (60% V_ O2 max) and low (40% somatostatin antiserum, the possibility of a GH-induced
V_ O2 max) intensity, arm cranking and leg cycling exer- inhibitory effect on hypothalamic GHRH cannot be
cise at three different times of day, where exercise was discounted.
initiated at approximately 05:00, 14:30 and 23:30 h. The In a further study Lanzi and Tannenbaum [79] also
results of this study demonstrated that exercise elicits a demonstrated a role for somatostatin in the attenuation
clear GH response regardless of time of day with 5- to of exogenous GHRH-induced GH release in rats. Serial
6-fold increases in plasma GH concentrations. In addi- injections of GHRH at 2 h intervals elicited 4- to 6-fold
tion, there was no difference in the magnitude of the increases in GH release when GHRH was administered
exercise-induced GH response at three different times of at times of peak spontaneous GH release, but only a
day suggesting that there is no diurnal rhythm in the GH minimal GH response was observed during trough pe-
response to prolonged sub-maximal exercise. riods. There was no evidence of desensitisation of so-
Similar findings were also reported for moderately matotropes since high GH responsiveness to exogenous
trained young men, who performed 30 min of treadmill GHRH was maintained at a time of spontaneous se-
exercise on three separate occasions at 07:00, 19:00 and cretory episode following a previous exogenous GHRH
24:00 h. No significant difference in the magnitude of challenge during a trough period. These results dem-
the GH response was identified for these three trials, onstrate the importance of the cyclical increase in en-
although the increase in serum GH concentrations was dogenous hypothalamic somatostatin secretion in
followed by a period of suppression of GH secretion in preventing desensitisation of the pituitary to GHRH. In
the 07:00 and the 19:00, but not the 24:00 h trial [76]. It the same study a single subcutaneous rGH injection 3 h
would appear, therefore, that exercise is a robust stim- prior to GHRH administration severely attenuated the
ulus that can overcome underlying diurnal rhythms. GHRH-induced GH response. Passive immunisation
However, both low- and moderate- intensity sub-max- with specific somatostatin antiserum reversed the blun-
imal exercise have been shown to suppress GH release ted GH response and completely restored GH respon-
in the first part of nocturnal sleep, when GH concen- siveness to GHRH. This was consistent with the
trations are usually at their peak, and increases GH contention that GH feedback is exerted, at least in part,
release in the second part of sleep, when GH concen- by somatostatin. In addition, the understanding that
trations are normally lower, without altering total GH GH receptor mRNA is colocalised in somatostatin-
release [77]. positive neurons in the periventricular nucleus of the rat
hypothalamus [80] further supports these findings.
7.1. Effect of repeated bouts of exercise on GH release In normal adults repeated GHRH administration has
been shown to result in an attenuated GH response to
There is a great deal of evidence suggesting that GH, the second stimulus [81]. However, the administration of
like a number of other hormones, regulates its own se- arginine, which acts to suppress somatostatin release,
cretion via a negative feedback mechanism, although the with the second bolus of GHRH restored the respon-
nature of this autoregulation is not entirely clear. A siveness of the somatotroph and, in fact, even potenti-
number of possibilities exist and the role of GH auto- ated the GH response. This suggests that the attenuation
feedback, increased somatostatinergic tone and/or de- of the GH observed using repeated boluses of GHRH
234 K. Stokes / Growth Hormone & IGF Research 13 (2003) 225–238

alone was not due to a GHRH-induced reduction in the sponse in exercise bouts separated by more than 30 min.
size of the pool of GH available for release. In addition, Lanzi and Tannenbaum [78] also measured IGF-I and
the fact that arginine administration reinstated the did not observe an increase in plasma IGF-I concen-
GHRH-induced GH response following the second tration, but reported GH release to be suppressed for 4 h
stimulation implies an important role for somatostatin after rGH administration in rats. This implies that the
in GH autoregulation. GH negative feedback loop can function independently
The GH response to repeated bouts of exercise has of IGF-I, although there does remain the possibility that
also been studied, and Kanaley et al. [82] demonstrated locally synthesised IGF-I in the pituitary gland might
an augmented GH response to repeated bouts of 30 min play a role.
exercise at 70% V_ O2 max separated by either 60 min or Cappon et al. [85] also considered GH autoinhibition,
210 min of recovery. Each exercise bout resulted in a whereby GH feeds back on itself directly, however, this
distinct GH pulse and the apparent progressive increase seemed unlikely as GH was only slightly elevated at the
in GH response tended to be greater with a longer (210 end of each recovery period. Alternatively, an increase
min) recovery period. The augmented response with in FFA as a result of the first exercise bout might have
repeated bouts of exercise provided evidence that the blocked GH release directly at the pituitary level [88].
depletion of pituitary stores with repeated stimuli to GH The potential role of FFA in the GH feedback loop was
release is unlikely. Jaffe et al. [83] supported this view- studied by Pontiroli et al. [89]. Infusion of methionyl-
point, suggesting that pituitary GH content far exceeded GH (met-GH) blocked the response to exogenous
the amount of GH released in their study and yet they GHRH and administration of acipimox, an antilipolytic
demonstrated a suppression of the GH response to re- agent, and pyridostigmine, to block hypothalamic so-
peated GHRH administration. The GH response to the matostatin release, did not restore the GH response to
second of two 65 min bouts of exercise at 70% V_ O2 max GHRH. This indicates that inhibition of the GH re-
separated by 3 h of recovery has also been shown to sponse to GHRH can occur independently of circulating
result in a greater GH response than a single 65 min plasma FFA levels and hypothalamic somatostatin re-
bout of exercise at 70% V_ O2 max [84]. From these results lease and was probably mediated by GH autofeedback
it would appear that exercise provides sufficient stimulus at the pituitary gland.
to overcome the autonegative feedback demonstrated Repeated sprint exercise has also been shown to re-
using pharmacological interventions [82]. sult in an attenuated GH response when GH is still el-
However, Cappon et al. [85] demonstrated that the evated immediately prior to a second sprint [8]. In this
GH response to 10 min of constant power cycling ex- study, the GH response to the second sprint was atten-
ercise, at an intensity corresponding to 50% of the dif- uated in 6 out of 10 subjects who displayed moderate to
ference between the lactate threshold and V_ O2 max, was large increases in serum GH concentrations following
dramatically attenuated as a result of previous exercise the first sprint, and whose GH levels remained elevated
bouts. In addition, they demonstrated an acute, GH before the second sprint. A GH response to the second
independent, exercise-induced increase in IGF-I. A pu- sprint was identified in two of the subjects and, in both
rified IGF-I preparation has been shown to inhibit GH cases, the GH response to the first sprint was moderate
release from pituitary cells in culture [86], and the in- and GH had returned to pre-exercise levels before the
fusion of rhIGF-I has also been demonstrated to sup- second sprint. This suggests that GH autoinhibition can
press pulsatile and GHRH-stimulated GH secretion in be responsible for the suppression of GH release in re-
male subjects [61]. There is an apparent role for so- peated sprint bouts with 60 min between sprints. In-
matostatin in this long-loop feedback, since Berelowitz creasing the recovery period between sprints to 240 min,
et al. [86] also observed IGF-I to stimulate a dose- by which time serum GH had returned to pre-exercise
related release of somatostatin from hypothalamic ex- levels, resulted in a partial recovery of the GH response
plants. However intraventricular IGF infusion in ewes to a second sprint [90].
had no effect on GH secretion, whereas intrapituitary
infusion resulted in the inhibition of GH release [87]
providing strong evidence for a direct effect of IGF-I at 8. Effect of training on acute GH release
the level of the pituitary.
However, Cappon et al. [85] found that exercise-in- A number of studies have concluded that exercise
duced IGF-I levels were not significantly higher than training has no effect on resting GH concentrations
baseline within 30 min of recovery, whilst the recovery when comparing sedentary individuals and athletes [91–
between exercise bouts was 50 min, suggesting that IGF- 94], sprint-trained and endurance-trained athletes [21]
I did not play a role in the attenuation of the GH re- and following endurance training [95]. However, the
sponse to exercise in their study. In addition, it would endogenous pulsatility of GH concentrations makes the
appear that exercise-induced IGF-I would not have a measurement of ‘‘resting’’ GH concentrations almost
role in the regulation of the exercise-induced GH re- meaningless. In addition, it is likely that in some of the
 K. Stokes / Growth Hormone & IGF Research 13 (2003) 225–238 235

studies considering the effect of training on resting GH smaller exercise-induced GH response to a 30 s sprint
concentrations the assays employed would have been on a cycle ergometer compared with pre-training de-
unable to satisfactorily measure the low levels of GH spite a 5% increase in mean power output during the
associated with human subjects at rest [66]. sprint [97]. Weltman et al. [96] suggested that reduced
Some studies have considered the effect of training exercise-induced GH concentrations following training
state on the GH response to an acute bout of exercise. may be a result of a combination of reduced GH se-
There is little agreement between these studies as to cretion and enhanced GH clearance. In support of this
whether exercise training increases [93,95] or decreases contention is the suggestion that the half-life of en-
[91,96,97] the GH response to a single exercise bout. dogenous GH is shorter in exercising than resting in-
The GH response to a 30 min run at 60% of V_ O2 max dividuals [98].
was found to be significantly greater in runners (min- However, the effect that exercise training has on the
imum mileage of 40 miles per week) compared with GH response to exercise is not entirely clear and might
moderately active controls [93]. In agreement with depend on a number of factors including the type and
these findings, the GH response to a standard swim- frequency of training and the duration of the training
ming training session (15  200 m with 20 s rest be- period. In addition, only three of the studies mentioned
tween sets) was enhanced in nine top-level male [91,93,97] compared the response of trained individuals
endurance swimmers (national team members) follow- with untrained control subjects and of these only one
ing 18 weeks of training despite no change in swim- [97] completed longitudinal study incorporating a con-
ming velocities during these sets [95]. In addition, it has trol group. It would appear, therefore, that further re-
been shown that endurance-trained athletes have a search is required in this area before a meaningful
greater GH response to insulin-induced hypoglycaemia conclusion can be drawn.
than untrained individuals [92].
However, an incremental sub-maximal cycling test
has been shown to elicit a smaller GH response in well- 9. Summary
trained compared with untrained cyclists [91]. In ad-
dition, six weeks of sub-maximal exercise training was This review has focussed on the natural GH response
associated with a 45% reduction in the GH response to to intense exercise, and has identified a number of
a 20 min constant load cycle ergometer test [96]. In this the factors that influence exercise-induced GH release.
study, absolute power output was the same in pre- and Fig. 1 summarises these factors, although it is accepted
post-training tests and over the training period power that this list might not be exhaustive, and that the
output expressed as a percentage of maximum power mechanisms of action and relationships between many
output decreased from 82% to 70%. These findings of these factors are not fully understood. It is clear,
suggest that, in sub-maximal exercise, there might be a therefore, that further research is needed in order to
critical relative exercise intensity required to elicit GH determine the stimulus for exercise-induced GH secre-
release [96]. Six weeks of combined speed and speed- tion in response to a single bout of exercise, as well as
endurance training has also been found to result in a to understand the effect of multiple bouts of exercise on

Fig. 1. Summary of the factors affecting the GH response to intense exercise.

236 K. Stokes / Growth Hormone & IGF Research 13 (2003) 225–238

the GH response to exercise, both in a single day, and as [15] S.E. Gordon, W.J. Kraemer, N.H. Vos, J.M. Lynch, H.G.
part of a programme of training. A better understanding Knuttgen, Effect of acid–base balance on the growth hormone
of these factors could be of benefit to sports performers, response to acute high-intensity cycle exercise, J. Appl. Physiol. 76
(1994) 821–829.
but also to normal individuals attempting to slow the [16] A.N. Elias, A.F. Wilson, S. Naqvi, M.R. Pandian, Effects of blood
process of ageing. pH and blood lactate on growth hormone, prolactin, and
gonadotropin release after acute exercise in male volunteers, Proc.
Soc. Exp. Biol. Med. 214 (1997) 156–160.
Acknowledgements [17] C. Lassarre, F. Girard, J. Durand, J. Raynaud, Kinetics of human
growth hormone during submaximal exercise, J. Appl. Physiol. 37
(1974) 826–830.
The author thanks Dr. Mary Nevill, Dr. Henryk
[18] J. Raynaud, L. Drouet, J.P. Martineaud, J. Bordachar,
Lakomy, and Prof. George Hall for the assistance given J. Coudert, J. Durand, Time course of plasma growth hormone
in the preparation of this review. during exercise in humans at altitude, J. Appl. Physiol. 50 (1981)
229–233.
[19] W.P. VanHelder, K. Casey, M.W. Radomski, Regulation of
References growth hormone during exercise by oxygen demand and avail-
ability, Eur. J. Appl. Physiol. 56 (1987) 628–632.
[1] Growth Hormone Research Society, Consensus guidelines for the [20] J.R. Sutton, Hormonal and metabolic responses to exercise in
diagnosis and treatment of growth hormone (GH) deficiency in subjects of high and low work capacities, Med. Sci. Sports 10
childhood and adolescence: summary statement of the GH (1978) 1–6.
Research Society, J. Clin. Endocrinol. Metab. 85 (2000) 3990– [21] M.E. Nevill, D.J. Holmyard, G.M. Hall, et al., Growth hormone
3993. responses to treadmill sprinting in sprint- and endurance-trained
[2] J. Seth, A. Bristow, Standardisation of human growth hormone athletes, Eur. J. Appl. Physiol. 72 (1996) 460–467.
(hGH) assays, Available from http://www.ukneqas.org.uk/nib- [22] A.J. Scheen, O.M. Buxton, M. Jison, et al., Effects of exercise on
scgh.pdf (accessed 1 December 2002) 2000. neuroendocrine secretions and glucose regulation at different
[3] G. Copinschi, M. Hartog, J.M. Earll, R.J. Havel, Effect of various times of day, Am. J. Physiol. 274 (1998) E1040–E1049.
blood sampling procedures on serum levels of immunoreactive [23] M. Viru, E. Jansson, A. Viru, C.J. Sundberg, Effect of restricted
human growth hormone, Metab. Clin. Exp. 16 (1967) 402–409. blood flow on exercise-induced hormone changes in healthy men,
[4] J. Raynaud, A. Capderou, J.P. Martineaud, J. Bordachar, Eur. J. Appl. Physiol. 77 (1998) 517–522.
J. Durand, Intersubject variability in growth hormone time course [24] D.L. McCloskey, J.H. Mitchell, Reflex cardiovascular and respi-
during different types of work, J. Appl. Physiol. 55 (1983) 1682– ratory responses originating in exercising muscle, J. Physiol. 224
1687. (1972) 173–186.
[5] N.E. Felsing, J.A. Brasel, D.M. Cooper, Effect of low and high [25] M. Kjaer, Epinephrine and some other hormonal responses to
intensity exercise on circulating growth hormone in men, J. Clin. exercise in man, J. Sports Med. 10 (1989) 2–15.
Endocrinol. Metab. 75 (1992) 157–162. [26] M. Kjaer, N.H. Secher, F.W. Bach, S. Sheikh, H. Galbo,
[6] C.J. Pritzlaff, L. Wideman, J.Y. Weltman, et al., Impact of acute Hormonal and metabolic responses to exercise in humans: effect
exercise intensity on pulsatile growth hormone release in men, J. of sensory nervous blockade, Am. J. Physiol. 257 (1989) E95–
Appl. Physiol. 87 (1999) 498–504. E101.
[7] K.A. Stokes, M.E. Nevill, G.M. Hall, H.K.A. Lakomy, The time [27] M. Kjaer, B. Hanel, L. Worm, et al., Cardiovascular and
course of the human growth hormone response to a 6 s and a 30 s neuroendocrine responses to exercise in hypoxia during impaired
cycle ergometer sprint, J. Sports Sci. 20 (2002) 487–494. neural feedback from muscle, Am. J. Physiol. 277 (1999) R76–
[8] K.A. Stokes, M.E. Nevill, G.M. Hall, H.K.A. Lakomy, Growth R85.
hormone response to repeated maximal cycle ergometer exercise at [28] M. Kjaer, N.H. Secher, F.W. Bach, H. Galbo, Role of motor
different pedaling rates, J. Appl. Physiol. 92 (2002) 602–608. center activity for hormonal changes and substrate mobilization in
[9] J.R. Sutton, J.D. Young, L. Lazarus, J.B. Hickie, J. Maksvytis, humans, Am. J. Physiol. 253 (1987) R687–R695.
The hormonal response to physical exercise, Austr. Ann. Med. 18 [29] M. Kjaer, N.H. Secher, J. Bangsbo, et al., Hormonal and
(1969) 84–90. metabolic responses to electrically induced cycling during
[10] S. Kozlowski, J. Chwalbinska-Moneta, M. Vigas, H. Kaciuba- epidural anesthesia in humans, J. Appl. Physiol. 80 (1996)
Uscilko, K. Nazar, Greater serum GH response to arm than leg 2156–2162.
exercise performed at equivalent oxygen uptake, Eur. J. Appl. [30] E. Asmussen, S.H. Johansen, M. Jorgensen, M. Nielsen, On the
Physiol. 52 (1983) 131–134. nervous factors controlling respiration and circulation during
[11] W.P. VanHelder, R.C. Goode, M.W. Radomski, Effect of exercise-experiments with curarization, Acta. Physiol. Scand. 63
anaerobic and aerobic exercise of equal duration and work (1965) 343–350.
expenditure on plasma growth hormone levels, Eur. J. Appl. [31] M. Kjaer, S.F. Pollack, T. Mohr, et al., Regulation of glucose
Physiol. 52 (1984) 255–257. turnover and hormonal responses during electrical cycling in
[12] A. Karagiorgos, J.F. Garcia, G.A. Brooks, Growth hormone tetraplegic humans, Am. J. Physiol. 271 (1996) R191–R199.
response to continuous and intermittent exercise, Med. Sci. Sports [32] S. Ellis, M.A. Vodian, R.E. Grindeland, Studies on the bioassay-
11 (1979) 302–307. able growth hormone-like activity of plasma, Recent Prog. Horm
[13] M.S. Vigas, S. Nemeth, L. Jurcoricora, J. Mikula, L. Komadel, Res. 34 (1978) 213–238.
The importance of lactate in exercise-induced growth hormone [33] G.E. McCall, K.L. Gosselink, A.J. Bigbee, R.R. Roy, R.E.
release in man. Radioimmunoassays: methodology and applica- Grindeland, V.R. Edgerton, Muscle afferent-pituitary axis: a novel
tions in physiology and clinical studies, Horm. Metab. Res. 5 pathway for modulating the secretion of a pituitary growth factor,
(Suppl) (1974) 166–169. Ex. Sport Sci. Rev. 29 (2001) 164–169.
[14] J.R. Sutton, N.L. Jones, C.J. Toews, Growth hormone secretion [34] K.L. Gosselink, R.E. Grindeland, R.R. Roy, et al., Skeletal
and acid–base alteration at rest and during exercise, Clin. Sci. muscle afferent regulation of bioassayable growth hormone in the
Mol. Med. 50 (1976) 241–247. rat pituitary, J. Appl. Physiol. 84 (1998) 1425–1430.
 K. Stokes / Growth Hormone & IGF Research 13 (2003) 225–238 237

[35] G.E. McCall, R.E. Grindeland, R.R. Roy, V.R. Edgerton, [54] T. Raastad, T. Bjoro, J. Hallen, Hormonal responses to high- and
Muscle afferent activity modulates bioassayable growth hor- moderate-intensity strength exercise, Eur. J. Appl. Physiol. 82
mone in human plasma, J. Appl. Physiol. 89 (2000) 1137– (2000) 121–128.
1141. [55] K. Stokes, M. Nevill, G. Hall, H. Lakomy, Reliability of the
[36] P.W. Cherry, H.K.A. Lakomy, L.H. Boobis, M.E. Nevill, Rapid growth hormone response to a single 30-s cycle ergometer sprint,
recovery of power output in females, Acta. Physiol. Scand. 164 in: Proceedings of the 12th Commonwealth International Sport
(1998) 79–87. Conference, July 19–23, 2002, Manchester, UK (Abstract).
[37] A. Weltman, C.M. Wood, C.J. Womack, et al., Catecholamine [56] J.O. Jansson, S. Eden, O. Isaksson, Sexual dimorphism in
and blood lactate responses to incremental rowing and running the control of growth hormone secretion, Endocr. Rev. 6 (1985)
exercise, J. Appl. Physiol. 76 (1994) 1144–1149. 128–150.
[38] A. Giustina, J.D. Veldhuis, Pathophysiological basis of the [57] J.-.C. Painson, G.S. Tannenbaum, Sexual dimorphism of somato-
neuroregulation of the somatotropic (GH) axis in experimental statin and growth hormone-releasing factor signalling in the
animals and the human, Endocr. Rev. 19 (1998) 717–797. control of pulsatile growth hormone secretion in the rat, Endo-
[39] J. Chwalbinska-Moneta, H. Krysztofiak, A. Ziemba, K. Nazar, crinology 128 (1991) 2858–2866.
Kaciuba-Uscilko H. Threshold increases in plasma growth hor- [58] B.E. Engstrom, F.A. Karlsson, L. Wide, Gender differences in
mone in relation to plasma catecholamine and blood lactate diurnal growth hormone and epinephrine values in young adults
concentrations during progressive exercise in endurance-trained during ambulation, Clin. Chem. 45 (1999) 1235–1239.
athletes, Eur. J. Appl. Phyisol. 73 (1996) 117–120. [59] K.Y. Ho, W.S. Evans, R.M. Blizzard, et al., Effects of sex and age
[40] A. Weltman, C.J. Pritzlaff, L. Wideman, et al., Exercise-dependent on the 24-h profile of growth hormone secretion in man:
growth hormone release is linked to markers of heightened central importance of endogenous estradiol concentrations, J. Clin.
adrenergic outflow, J. Appl. Physiol. 89 (2000) 629–635. Endocrinol. Metab. 64 (1987) 51–58.
[41] I.K.M. Brenner, J. Zamecnik, P.N. Shek, R.J. Shephard, The [60] S.M. Pincus, E.F. Gevers, I.C.A.F. Robinson, et al., Females
impact of heat exposure and repeated exercise on circulating stress secrete growth hormone with more process irregularity than males
hormones, Eur. J. Appl. Physiol. 76 (1997) 445–454. in both humans and rats, Am. J. Physiol. 270 (1996) E107–E115.
[42] J.M.H. Buckler, The relationship between changes in plasma [61] C.A. Jaffe, B. Ocampo-Lim, W. Guo, et al., Regulatory mecha-
growth hormone levels and body temperature occurring with nisms of growth hormone are sexually dimorphic, J. Clin. Invest.
exercise in man, Biomedicine 19 (1973) 193–197. 102 (1998) 153–164.
[43] M.C. Cross, M.W. Radomski, W.P. VanHelder, S.G. Rhind, R.J. [62] E.E. Muller, V. Locatelli, D. Cocchi, Neuroendocrine control of
Shephard, Endurance exercise with and without thermal clamp: growth hormone secretion, Physiol. Rev. 79 (1999) 511–607.
effects on leukocytes and leukocyte subsets, J. Appl. Physiol. 81 [63] C.A. Jaffe, D.K. Turgeon, K. Lown, R. Demott-Friberg, P.B.
(1996) 822–829. Watkins, Growth hormone secretion pattern is an independent
[44] M.W. Radomski, M. Cross, A. Buguet, Exercise-induced hyper- regulator of growth hormone actions in humans, Am. J. Physiol.
thermia and hormonal responses to exercise, Can. J. Physiol. 283 (2002) E1008–E1015.
Pharmacol. 76 (1998) 547–552. [64] C.J. Pritzlaff-Roy, L. Wideman, J.Y. Weltman, et al., Gender
[45] H.C.M. Maas, W.R. de Vries, I. Maitimu, E. Bol, C.Y. Bowers, governs the relationship between exercise intensity and growth
H.P.F. Koppenschaar, Growth hormone responses during stren- hormone release in young adults, J. Appl. Physiol. 92 (2002) 2053–
uous exercise: the role of GH-releasing hormone and GH- 2060.
releasing peptide-2, Med. Sci. Sports Exerc. 32 (2000) 1226–1236. [65] A. Iranmanesh, G. Lizarralde, J.D. Veldhuis, Age and relative
[46] L. Di Luigi, F.G. Conti, A. Casini, et al., Growth hormone and adiposity are specific negative determinants of the frequency and
insulin-like growth factor I responses to moderate submaximal magnitude of growth hormone (GH) sectretory bursts and the
acute physical exercise in man: effects of octreocide, a somato- half-life of endogenous GH in healthy men, J. Clin. Endocrinol.
statin analogue, administration, Int. J. Sports Med. 18 (1997) Metab. 73 (1991) 1081–1088.
257–263. [66] J.D. Veldhuis, A.Y. Liem, S. South, et al., Differential impact of
[47] T.J. Marcell, R.A. Wiswell, S.A. Hawkins, K.M. Tarpenning, age, sex steroid hormones, and obesity on basal versus pulsatile
Age-related blunting of growth hormone secretion during exercise growth hormone secretion in men as assessed in an ultrasensitive
may not be solely due to increased somatostatin tone, Metabolism chemiluminescence assay, J. Clin. Endocrinol. Metab. 80 (1995)
48 (1999) 665–670. 3209–3222.
[48] W.R. De Vries, S.A. Abdesselam, T.J. Schers, et al., Complete [67] N. Vahl, N. Moller, T. Lauritzen, J.S. Christiansen, J.O.L.
inhibition of hypothalamic somatostatin activity is only partially Jorgensen, Metabolic effects and pharmacokinetics of a growth
responsible for the growth hormone response to strenuous hormone pulse in healthy adults: relation to age, sex, and body
activity, Metabolism 51 (2002) 1093–1096. composition, J. Clin. Endocrinol. Metab. 82 (1997) 3612–
[49] L. Wideman, J.Y. Weltman, J.T. Patire, et al., Synergy of 3618.
L -arginine and GHRP-2 stimulation of growth hormone in men [68] T.K. Hansen, Pharmacokinetics and acute lipolytic actions of
and women: modulation by exercise, Am. J. Physiol. 279 (2000) growth hormone: impact of age, body composition, binding
R1467–R1477. proteins and other hormones, Growth Horm. IGF Res. 12 (2002)
[50] F. Camanni, E. Ghigo, E. Arvat, Growth hormone-releasing 342–358.
peptides and their analogs, Front. Neuroendocrinol. 19 (1998) [69] G. Pyka, R.A. Wiswell, R. Marcus, Age-dependent effect of
47–72. resistance exercise on growth hormone secretion in people, J. Clin.
[51] M. Kojima, H. Hosoda, Y. Date, M. Nakazato, H. Matsuo, K. Endocrinol. Metab. 75 (1992) 404–407.
Kangawa, Ghrelin is a growth-hormone-releasing acylated pep- [70] N. Vahl, J.O.L. Jorgensen, C. Skjaerbaek, J.D. Veldhuis, H.
tide from stomach, Nature 402 (1999) 656–660. Orskov, J.S. Christiansen, Abdominal adiposity rather than age
[52] R. Dall, J. Kanaley, T.K. Hansen, et al., Plasma ghrelin levels and sex predicts mass and regularity of GH secretion in healthy
during exercise in healthy subjects and in growth hormone- adults, Am. J. Physiol. 272 (1997) E1108–E1116.
deficient patients, Eur. J. Endocrinol. 147 (2002) 65–70. [71] J.L. Clasey, A. Weltman, J. Patrie, et al., Abdominal visceral fat
[53] L. Wideman, J.Y. Weltman, N. Shah, S. Story, J.D. Veldhuis, A. and fasting insulin are important predictors of 24-hour GH release
Weltman, Effects of gender on exercise-induced growth hormone independent of age, gender, and other physiological factors,
release, J. Appl. Physiol. 87 (1999) 1154–1162. J. Clin. Endocrinol. Metab. 86 (2001) 3845–3852.
238 K. Stokes / Growth Hormone & IGF Research 13 (2003) 225–238

[72] N. Vahl, J.O.L. Jorgensen, A. Jurik, J.S. Christiansen, Abdominal [86] M. Berelowitz, M. Szabo, L.A. Frohman, S. Firestone, L. Chu,
adiposity and physical fitness are major determinants of the age- Somatomedin-C mediates growth hormone negative feedback by
associated decline in stimulated GH secretion in healthy adults, effects on both the hypothalamus and the pituitary, Science 212
J. Clin. Endocrinol. Metab. 81 (1996) 2209–2215. (1981) 1279–1281.
[73] J.A. Kanaley, M.M. Weatherdrupp-Dentes, E.B. Jaynes, M.L. [87] T.P. Fletcher, G.B. Thomas, F.R. Dunshea, L.G. Moore, I.J.
Hartman, Obesity attenuates the growth hormone response to Clarke, IGF feedback effects on growth hormone secretion in
exercise, J. Clin. Endocrinol. Metab. 84 (1999) 3156–3161. ewes: evidence for action at the pituitary but not the hypothalamic
[74] E.A. Galliven, A. Singh, D. Michelson, S. Bina, P.W. Gold, P.A. level, J. Endocrinol. 144 (1995) 323–331.
Deuster, Hormonal and metabolic responses to exercise across [88] F.F. Casanueva, L. Villanueva, C. Dieguez, et al., Free fatty acids
time of day and menstrual cycle phase, J. Appl. Physiol. 83 (1997) block growth hormone (GH) releasing hormone-stimulated GH
1822–1831. secretion in man directly at the pituitary, J. Clin. Endocrinol.
[75] R.S. Nathan, E.J. Sachar, G. Langer, M.A. Tabrizi, F.S. Halpern, Metab. 65 (1987) 634–642.
Diurnal variations in the response of plasma prolactin, cortisol [89] A.E. Pontiroli, R. Lanzi, L.D. Monti, E. Sandoli, G. Pozza,
and growth hormone to insulin-induced hypoglycemia in normal Growth hormone (GH) autofeedback on GH response to GH-
men, J. Clin. Endocrinol. Metab. 49 (1979) 231–235. releasing hormone. Role of free fatty acids and somatostatin, J.
[76] J.A. Kanaley, J.Y. Weltman, K.S. Pieper, A. Weltman, M.L. Clin. Endocrinol. Metab. 72 (1991) 492–495.
Hartman, Cortisol and growth hormone responses to exercise at [90] K.A. Stokes, M.E. Nevill, G.M. Hall, H.K.A. Lakomy, Effect of
different times of day, J. Clin. Endocrinol. Metab. 86 (2001) 2881– recovery period on human growth hormone responses to repeated
2889. sprint cycling, Med. Sci. Sports Exerc. 34 (2002) S72.
[77] W. Kern, B. Perras, R. Wodick, H.L. Fehm, J. Born, Hormonal [91] S.R. Bloom, R.H. Johnson, D.M. Park, M.J. Rennie, W.R.
secretion during nighttime sleep indicating stress of daytime Sulaiman, Differences in the metabolic and hormonal response to
exercise, J. Appl. Physiol. 79 (1995) 1461–1468. exercise between racing cyclists and untrained individuals, J.
[78] R. Lanzi, G.S. Tannenbaum, Time course and mechanism of Physiol. 258 (1976) 1–18.
growth hormoneÕs negative feedback effect on its own spontane- [92] K.J. Mikines, M. Kjaer, C. Hagen, B. Sonne, E.A. Richter, H.
ous release, Endocrinology 130 (1992) 780–788. Galbo, The effect of training on responses of b-endorphin and
[79] R. Lanzi, G.S. Tannenbaum, Time-dependent reduction and other pituitary hormones to insulin-induced hypoglycemia, Eur. J.
potentiation of growth hormone (GH) responsiveness to GH- Appl. Physiol. 54 (1985) 476–479.
releasing factor induced by exogenous GH: a role for somato- [93] J.C. Bunt, R.A. Boileau, J.M. Bahr, R.A. Nelson, Sex and
statin, Endocrinology 130 (1992) 1822–1828. training differences in human growth hormone levels during
[80] K.A. Burton, D.K. Clifton, R.A. Steiner, Growth hormone prolonged exercise, J. Appl. Physiol. 61 (1986) 1796–1801.
receptor messenger RNA is colocalized in somatostatin neurons [94] T. Barreca, E. Reggiani, R. Franceschini, et al., Serum prolactin,
in the male rat hypothalamus, in: 21st Annual Meeting of the growth hormone and cortisol in athletes an sedentary subjects
Society for Neuroscience (Abstract), New Orleans LA, 1991, p. after submaximal and exhaustive exercises, J. Sports Med. 28
432. (1988) 89–92.
[81] E. Ghigo, E. Arvat, F. Valente, et al., Arginine reinstates the [95] M. Bonifazi, E. Bela, C. Lupo, G. Martelli, B. Zhu, G. Carli,
somatotrope responsiveness to intermittent growth hormone- Influence of training on the to exercise of adrenocorticotropin and
releasing hormone administration in normal adults, Neuroendo- growth hormone plasma concentrations in human swimmers, Eur.
crinology 54 (1991) 291–294. J. Appl. Physiol. 78 (1998) 394–397.
[82] J.A. Kanaley, J.Y. Weltman, J.D. Veldhuis, A.D. Rogol, M.L. [96] A. Weltman, J.Y. Weltman, C.J. Womack, et al., Exercise
Hartman, A. Weltman, Human growth hormone response to training decreases the growth hormone (GH) response to acute
repeated bouts of aerobic exercise, J. Appl. Physiol. 83 (1997) constant-load exercise, Med. Sci. Sports Exerc. 29 (1997)
1756–1761. 669–676.
[83] C.A. Jaffe, R. DeMott-Friberg, A.L. Barkan, Suppression of [97] K.A. Stokes, M.E. Nevill, P.W. Cherry, G.M. Hall, H.K.A.
growth hormone (GH) secretion by a selective GH-releasing Lakomy, The effect of 6 weeks of sprint training on the growth
hormone (GHRH) antagonist, J. Clin. Invest. 92 (1993) 695–701. hormone response to repeated maximal cycle ergometer exercise,
[84] O. Ronson, E. Haug, B.K. Pedersen, R. Bahr, Increased neuro- J. Physiol. 528 (2000) 51P.
endocrine response to a repeated bout of endurance exercise. [98] D.L. Thompson, J.Y. Weltmen, A.D. Rogol, D.L. Metzger, J.D.
[85] J. Cappon, J.A. Brasel, S. Mohan, D.M. Cooper, Effect of brief Veldhuis, A. Weltman, Cholinergic and opioid involvement in
exercise on insulin-like growth factor I, J. Appl. Physiol. 76 (1994) release of growth hormone during exercise and recovery, J. Appl.
2490–2496. Physiol. 75 (1993) 870–878.