ArticlePDF Available

Fasting enhances growth hormone secretion and amplifies the complex rhythms of growth hormone secretion in man

Authors:

Abstract

Studies in man have shown that the episodic release of growth hormone (GH) is infrequent and erratic, and unlike that in the rat does not appear to have discernible ultradian periodicities. However, these observations in nonfasted subjects may be invalid since mixed nutrients have unpredictable effects on GH release. Moreover, in the fed state basal GH levels are frequently undetectable, thus rendering the identification of low amplitude pulses unreliable. Accordingly, the 24-h pulsatile pattern of GH secretion obtained from repetitive venous sampling in six normal adult male subjects was examined during a control fed day and during the first and fifth days of a 5-d fast. The GH data were analyzed using two distinct methods: a discrete pulse detection algorithm (Cluster analysis) and Fourier expansion time-series, which allows fixed periodicities of secretory activity to be resolved. The 5-d fast resulted in a significant increase in discrete GH pulse frequency (5.8 +/- 0.7 vs. 9.9 +/- 0.7 pulses/24 h, P = 0.028), 24 h integrated GH concentration (2.82 +/- 0.50 vs. 8.75 +/- 0.82 micrograms.min/ml; P = 0.0002), and maximal pulse amplitude (5.9 +/- 1.1 vs. 12.3 +/- 1.6 ng/ml, P less than 0.005). While multiple low-amplitude sinusoidal periodicities were present on the control fed day, time-series analysis revealed enhancement of circadian and ultradian cycles on the first and fifth days of fasting. Concomitantly, fasting resulted in a decline (day 1 vs. day 5) in serum concentrations of somatomedin C (1.31 +/- 0.22 vs. 0.77 +/- 0.18 U/ml) and glucose (4.9 +/- 0.2 vs. 3.2 +/- 0.2 mmol/liter), and a marked rise in free fatty acid (0.43 +/- 0.12 vs. 1.55 +/- 0.35 mmol/liter) and acetoacetate (35 +/- 6 vs. 507 +/- 80 nmol/liter). We conclude that the acute nutritional status is an important determinant of spontaneous pulsatile GH secretion in man. Fast-induced enhancement of GH release is achieved through combined frequency (discrete pulses) and amplitude (sinusoidal periodicities) modulation. Such alterations in somatotropic hormone release may play an important role in substrate homeostasis during starvation.
Fasting
Enhances
Growth
Hormone
Secretion
and
Amplifies
the
Complex
Rhythms
of
Growth
Hormone
Secretion
in
Man
Klan
Y.
Ho,*
Johannes
D.
Veldhuls,*
Michael
L.
Johnson,*
Richard
Furlanetto,9
William
S.
Evans,*
K.
G.
M. M.
Alberti,11
and
Michael
0.
Thomer*
Departments
of
*Internal
Medicine
and
tPharmacology,
University
of
Virginia
Medical
School,
Charlottesville,
Virginia
22908;
§The
Childrens
Hospital
of
Philadelphia,
University
of
Pennsylvania,
Philadelphia,
Pennsylvania
19104;
and
"lthe
Royal
Victoria
Infirmary,
Newcastle
Upon
Tyne,
England
Abstract
Studies
in
man
have
shown
that
the
episodic
release
of
growth
hormone
(GH)
is
infrequent
and
erratic,
and
unlike
that
in
the
rat
does
not
appear
to
have
discernible
ultradian
periodicities.
However,
these
observations
in
nonfasted
subjects
may
be
in-
valid
since
mixed
nutrients
have
unpredictable
effects
on
GH
release.
Moreover,
in
the
fed
state
basal
GH
levels
are
fre-
quently
undetectable,
thus
rendering
the
identification
of
low
amplitude
pulses
unreliable.
Accordingly,
the
24-h
pulsatile
pattern
of
GH
secretion
obtained
from
repetitive
venous
sam-
pling
in
six
normal
adult
male
subjects
was
examined
during
a
control
fed
day
and
during
the
first
and
fifth
days
of
a
5-d
fast.
The
GH
data
were
analyzed
using
two
distinct
methods:
a
discrete
pulse
detection
algorithm
(Cluster
analysis)
and
Four-
ier
expansion
time-series,
which
allows
fixed
periodicities
of
secretory
activity
to
be
resolved.
The
5-d
fast
resulted
in
a
significant
increase
in
discrete
GH
pulse
frequency
(5.8±0.7
vs.
9.9±0.7
pulses/24
h,
P
=
0.028),
24
h
integrated
GH
con-
centration
(2.82±0.50
vs.
8.75±0.82
g
min/ml;
P
=
0.0002),
and
maximal
pulse
amplitude
(5.9±1.1
vs.
12.3±1.6
ng/ml,
P
<
0.005).
While
multiple
low-amplitude
sinusoidal
periodic-
ities
were
present
on
the
control
fed
day,
time-series
analysis
revealed
enhancement
of
circadian
and
ultradian
cycles
on
the
first
and
fifth
days
of
fasting.
Concomitantly,
fasting
resulted
in
a
decline
(day
1
vs.
day
5)
in
serum
concentrations
of
soma-
tomedin
C
(1.31±0.22
vs.
0.77±0.18
U/ml)
and
glucose
(4.9±0.2
vs.
3.2±0.2
mmol/liter),
and
a
marked
rise
in
free
fatty
acid
(0.43±0.12
vs.
1.55±0.35
mmol/liter)
and
acetoace-
tate
(35±6
vs.
507±80
nmol/liter).
We
conclude
that
the
acute
nutritional
status
is
an
impor-
tant
determinant
of
spontaneous
pulsatile
GH
secretion
in
man.
Fast-induced
enhancement
of
GH
release
is
achieved
through
combined
frequency
(discrete
pulses)
and
amplitude
(sinusoidal
periodicities)
modulation.
Such
alterations
in
so-
matotropic
hormone
release
may
play
an
important
role
in
substrate
homeostasis
during
starvation.
Introduction
Growth
hormone
(GH)'
is
secreted
episodically
in
man.
Al-
though
it
is
well
established
that
GH
release
is
entrained
to
the
This
work
was
presented
in
part
at
the
68th
Annual
Meeting
of
the
Endocrine
Society,
Anaheim,
CA,
25-27
June
1986.
Receivedfor
publication
2
February
1987
and
in
revisedform
16
November
1987.
1.
Abbreviations
used
in
this
paper:
GH,
growth
hormone;
GHRH,
onset
of
slow-wave
sleep
(1,
2),
at
other
times
during
the
day
release
is
intermittent,
erratic,
and
unpredictable
with
no
rec-
ognizable
pattern
or
periodicity
(3-8).
This
stands
in
contrast
to
the
rat
that
possesses
an
ultradian
pattern
characterized
by
rhythmic
release
of
GH
approximately
every
3-4
h
(9).
Thus,
in
man,
GH
release
is
nyctohemeral
and
does
not
appear
to
display
distinct
cyclic
ultradian
periodicity
throughout
the
24-h
period.
Different
nutrients
have
different
effects
on
GH
release.
For
example,
glucose
and
fatty
acids
suppress
GH
release
while
certain
amino
acids
stimulate
secretion
(10).
Most
of
the
24-h
studies
of
GH
secretion
to
date
have
been
carried
out
in
the
fed
state
(3-6,
8).
Accordingly,
inferences
regarding
the
absence of
demonstrable
cyclic
periodicity
in
man
may
not
be
valid
be-
cause
of
the
unpredictable
effects
of
mixed
nutrients.
In
addi-
tion,
basal
GH
concentrations
in
the
fed
state
are
frequently
undetectable
by
conventional
radioimmunoassay,
thus
ren-
dering
the
identification
of
small
pulses
impossible
or
unreli-
able.
Fasting
would
be
expected
to
eliminate
the
unpredictable
effects
of
mixed
nutrients
on
GH
secretion.
Moreover,
fasting
lowers
somatomedin
C
levels
(1
1)
and
in
so
doing
removes
feedback
inhibition
of
GH
release
(12).
This
open-loop
system
would
be
predicted
to
elevate
basal
GH
concentrations
and
allow
identification
of
spontaneous
pulsatility
with
greater
ac-
curacy.
In
view
of
the
preceding
considerations,
our
aim
was
to
determine
the
serial
effects
of
fasting
on
the
24-h
pulsatile
pattern
of
GH
secretion
in
man.
We
hypothesized
that
the
otherwise
concealed
characteristics
of
periodic
GH
secretion
might
be
uncovered
by
fasting.
We
employed
an
immunora-
diometric
assay
with
enhanced
sensitivity
for
the
measurement
of
GH
and
characterized
the
episodic
pattern
of
GH
secretion
using
two
distinct
methods:
a
discrete,
statistically
based,
pulse
detection
algorithm,
and
the
Fourier
expansion
time-series.
These
two
approaches
are
able
to
resolve
episodic
pulsations
and
regularly
recurring
periodicities,
respectively.
Since
pitu-
itary
GH
release
is
in
part
dependent
on
hypothalamic
GH-re-
leasing
hormone
(GHRH)
secretion
(13,
14),
we
administered
GHRH-40
at
the
end
of
each
24-h
study
period
to
ascertain
whether
altered
responsiveness
to
GHRH
accompanied
changes
in
GH
release
during
fasting.
Methods
Subjects
Six
normal
healthy
male
volunteers
(ages
21-36
yr)
of
normal
body
weight
(BMI
21.8-28.0
kg/m2)
were
studied.
Each
subject
gave
written
informed
consent,
approved
by
the
Human
Investigation
Committee
growth
hormone-releasing
hormone;
IGHC,
integrated
growth
hor-
mone
concentration;
SmC,
somatomedin
C.
968
Ho,
Veldhuis,
Johnson,
Furlanetto,
Evans,
Alberti,
and
Thorner
J.
Clin.
Invest.
©
The
American
Society
for
Clinical
Investigation,
Inc.
0021-9738/88/04/0968/08
$2.00
Volume
81,
April
1988,
968-975
of
the
University
of
Virginia
School
of
Medicine.
All
subjects
had
normal
sleep
habits,
had
not
taken
any
transmeridian
ffights
within
the
previous
4
wk,
and
were
not
taking
any
medication.
Study
design
Each
subject
was
studied
on
three
occasions
during
which
blood
sam-
ples
were
obtained
every
20
min
over
a
24-h
period.
These
studies
were
undertaken
during
(a)
a
control
fed
day,
in
which
three
meals
were
served
at
0900,
1300,
and
1800
hours,
and
on
(b)
the
first
and
(c)
fifth
days
of
a
5-d
fast
in
which
only
water,
potassium
chloride
(20
meq/d),
and
vitamin
supplements
(I
multivitamin
capsule/d)
were
given.
In
three
subjects,
the
control
fed
study
was
performed
before
the
two
fasted
studies,
while
in
the
remaining
three,
the
control
study
was
undertaken
last.
A
period
of
2
wk
separated
the
fed
and
the
fasted
studies.
The
subjects
were
admitted
to
the
Clinical
Research
Center
at
0700
hours
and
a
cannula
with
a
heparin
lock
was
inserted
into
a
forearm
vein.
A
1-h
period
was
allowed
before
commencement
of
the
studies
at
0800
hours.
Subjects
were
encouraged
to
ambulate
on
study
days
and
daytime
naps
were
prohibited.
Sleeping
was
permitted
after
2200
hours
and
the
period
of
sleep
recorded
by
the
ward
staff.
An
intravenous
bolus
dose
of
I
jAg/kg
GHRH-40
was
administered
at
the
end
of
each
of
the
three
24-h
studies,
and
blood
sampling
continued
at
the
same
frequency
for
a
further
3
h.
Blood
samples
for
glucose,
free
fatty
acids,
,B-hydroxybutyrate,
acetoacetate,
somatomedin
C
(SmC),
and
blood
biochemistries
were
obtained
during
the
5-d
fast.
Subjects
were
also
weighed
daily
and
a
urine
analysis
performed
daily
for
uri-
nary
ketones.
Assays
Samples
from
each
individual's
three
studies
(n
=
72
+
9
for
each
study)
were
run
in
triplicate
in
the
same
assay.
GH
was
measured
using
the
Nichols
Institute
(San
Juan
Capistrano,
CA)
hGH
immunoradio-
metric
assay
which
was
modified
to
enhance
the
limit
of
sensitivity
to
0.2
ng/ml.
The
intraassay
coefficients
of
variation
were
6.3, 6.4,
and
3.9%
at
4,
7.3,
and
19.8
ng/ml,
respectively.
The
interassay
coefficients
of
variation
were
<
9.2%
at
the
above
levels.
Blood
glucose
was
mea-
sured
using
the
glucose
oxidase
method
and
a
glucose
analyzer
(Beck-
man
Instruments,
Fullerton,
CA)
(15).
SmC,
unesterified
fatty
acids,
acetoacetate,
and
,#-hydroxbutyrate
concentrations
were
measured
using
methods
as
previously
described
(16,
17)
and
as
adapted
from
reference
18.
Data
analysis
Two
distinct
mathematical
methods
were
used
to
characterize
the
ef-
fects
of
fasting
on
episodic
GH
secretion.
Pulse
analysis.
Discrete
GH
pulses
were
identified
using
Cluster
analysis,
an
objective
multipoint,
statistically
based
pulse
detection
algorithm
that
defines
significant
hormone
excursions
in
relation
to
the
actual
experimental variance
(19).
A
pulse
is
defined
as
a
statisti-
cally
significant
increase
in
a
cluster
of
hormone
values
followed
by
a
statistically
significant
decrease
in
a
second
cluster
of
values.
Based
upon
prior
objective
optimization
studies,
the
cluster
size
for
nadirs
and
peaks
(the
number
of
points
used
in
testing
nadirs
against
peaks)
was
set
at
1
and
2,
respectively
(20).
Pulse
criteria
were
set
to
minimize
the
occurrence
of
false
positive
pulses
on
signal-free
noise
to
<
5%.
Identified
properties
of
pulsatile
GH
release
included:
pulse
frequency,
maximal
peak
amplitude,
mean
interpulse
GH
concentrations,
inte-
grated
GH
concentration
(IGHC
or
area
under
the
curve),
and
the
fraction
of
GH
secreted
in
pulses.
Time-series
analysis
of
periodic
hormone
release.
Serial
GH
data
were
also
analyzed
for
the
occurrence
of
significant
sinusoidal
period-
icities
using
transformation,
as
described
previously
(21).
Since
the
Fourier
series
is
a
linear
combination
of
terms,
the
standard
deviations
of
the
individual
coefficients
can
be
determined
by
linear
least-squares
estimation
models
(22).
Using
within-sample
variance
from
all
experi-
mental
replicates,
this
approach
permits
one
to
define
all
individual
significant
periodicities
(as
sine
or
cosine
functions)
and
estimate
their
corresponding
amplitudes
with
statistical
confidence
limits.
An
un-
derlying
periodicity
in
the
data
was
considered
significant
whenever
its
amplitude
exceeded
zero
by
at
least
1.96
SD
(P
<
0.025
at
each
tail)
compared
with
signal-free
noise.
Only
such
periodicities
were
consid-
ered
further.
Statistical
analysis.
Paired
two-tailed
t-statistics
were
used
to
test
for
significant
changes
in
the
means
of
parameters
measured
during
control
and
fasting
sessions.
Results
are
expressed
as
mean±SEM.
Results
The
six
subjects
lost
a
mean
of
4.8±0.3
kg
during
the
5
d
of
fasting.
Fig.
1
shows
changes
in
blood
glucose,
free
fatty
acids,
,8-hydroxybutyrate,
and
acetoacetate
concentrations
during
the
5-d
fast.
There
was
a
progressive
fall
(P
<
0.0002)
in
blood
glucose
from
4.9±0.2
mmol/liter
(88±3
mg/dl)
to
3.2±0.2
mmol/liter
(57±3
mg/dl).
Fasting
resulted
in
a
fivefold
in-
crease
in
free
fatty
acid
levels
(0.43±0.12
mmol/liter
to
1.55±0.35
mmol/liter),
a
14-fold
rise
in
acetoacetate
concen-
trations
(35±6
to
507±80
nmol/liter)
and
a
greater
than
sixty-
fold
increase
in
,B-hydroxybutyrate
concentrations
(57±10
to
3,945±659
nmol/liter).
There
was
a
slow
but
progressive
decline
in
circulating
SmC
concentrations
with
increasing
duration
of
fast.
The
changes
in
SmC
were
significant
by
the
fifth
fasted
day
(0.77±0.18
vs.
1.31±0.22
U/ml
on
day
1;
P
<
0.02).
The
24-h
IGHC
(micrograms
per
minute
per
milliliter)
in-
creased
after
1
d
of
fasting
(2.82±0.50
vs.
7.82±1.12,
control
vs.
day
1;
P
=
0.0009)
and
remained
elevated
on
day
5
(8.75±0.82).
The
pulsatile
component
(Fig.
2)
of
GH
release
followed
a
similar
pattern
with
an
increase
on
day
1
(2.07±0.47
vs.
5.74±0.78,
control
vs.
day
1;
P
=
0.0013)
but
no
further
increase
on
day
5
(4.25±0.67).
In
contrast,
although
there
was
a
strong
trend
in
the
nonpulsatile
component
of
GH
release
to
increase
on
day
1,
no
significant
change
was
noted
(0.75±0.11
vs.
2.08±0.53,
control
vs.
day
1;
P
=
0.10).
By
day
5,
however,
nonpulsatile
GH
release
was
markedly
enhanced
(4.50±0.75;
P
=
0.0002).
Fasting
also
resulted
in
a
progressive
increase
in
mean
GH
pulse
frequency
and
maximal
pulse
amplitude.
The
mean
number
of
peaks
per
24
h
increased
from
5.8±0.7
to
7.3±0.6
(P
<
0.004)
on
the
first
fasted
day
and
to
9.9±0.7
(P
<
0.03)
on
the
fifth
fasted
day.
There
was
also
a
significant
increase
in
the
mean
maximal
GH
peak
amplitude
between
the
control
day
(5.9±1.1
ng/ml)
and
the
first
fasted
day
(13.1±1.2
ng/ml,
P
=
0.003)
and
the
fifth
fasted
day
(12.3±1.6
ng/ml,
P
=
0.005).
The
mean
interpeak
GH
concentrations
were
mostly
below
4-
-:7
o
2-
E
0-
Day
of
fast
Figure
1.
Changes
in
the
mean±SEM
blood
glucose,
free
fatty
acid
(FFA),
#-hydroxybutyr-
ate
(#-OH),
and
aceto-
acetate
(AA)
concentra-
tions
during
5
d
of
fast-
ing
(n
=
6).
Sample
on
day
I
is
drawn
after
overnight
fast,
i.e.,
under
identical
condi-
tions
for
control
fed
day.
Growth
Hormone
Secretion
during
Fasting
969
r-i
pulsatile
non-pulsatile
D<O
OO
T
III
'
;
C
Dl
D5
Fait
Fast
ng/ml).
The
time
courses
of
the
GH
responses
to
GHRH
under
these
three
conditions
are
given
graphically
in
Fig.
3.
Further
analyses
using
integrated
GHRH-stimulated
GH
data
by
parametric
and
nonparametric
testing
also
indicated
the
absence
of
significant
differences
in
GHRH
responses
with
fasting.
Figure
2.
Mean±SEM
IGHC
during
the
control
fed
day
(C),
the
first
day
(Dl),
and
the
fifth
(D5)
day
of
a
5-d
fast.
The
IGHC
comprised
a
pulsatile
(o)
and
a
nonpulsatile
(m)
com-
ponent
(n
=
6).
the
limit
of
assay
detection
on
the
control
fed
day
(0.3±0.2
ng/ml
vs.
1.3±0.5
ng/ml
on
the
first
fasted
day,
P
=
0.04)
and
increased
to
3.4±0.8
ng/ml
on
the
fifth
fasted
day
(P
=
0.007
vs.
control;
P
=
0.03
vs.
day
1).
With
the
increase
in
interpulse
GH
concentrations
incremental
GH
peak
amplitude
(differ-
ence
between
peak
nadir
and
maximum)
increased
from
5.7±1.0
ng/ml
to
11.9±1.1
(P
=
0.006),
then
fell
back
to
8.6±1.3
ng/ml
(P
=
NS,
day
1
vs.
day
5).
In
Fig.
3
the
ampli-
fied
patterns
of
GH
secretion
induced
by
fasting
and
responses
to
GHRH-40
are
shown
in
all
six
subjects.
Table
I
summarizes
the
specific
changes
in
individual
peak
properties
quantitated
in
each
of
the
six
subjects
on
the
3
d
of
study.
To
look
for
regular
underlying
sinusoidal
periodicities
of
GH
secretion,
we
subjected
the
data
to
time-series
analysis
for
each
of
the
study
days.
Fig.
4
shows
such
an
analysis
in
a
representative
subject
for
whom
amplitudes
of
the
Fourier
transformations
are
plotted
against
periodicity
in
minutes.
No
single
dominant
periodicity
is
present
during
the
control
fed
day
as
reflected
in
the
flat
profile
of
the
periodogram;
however,
many
significant
ultradian
rhythms
existed
on
the
control
day.
On
the
first
fasted
day,
this
subject's
periodogram
shows
the