Acidosis stimulates
,8-endorphin release during exercise
DEREK V. TAYLOR,
JAMES
ALEKSANDRA
CHICZ-DEMET,
G. BOYAJIAN,
NORMAN
JAMES, DANETTE
WOODS,
ARCHIE F. WILSON, AND CURT A. SANDMAN
Departments of Psychiatry and Medicine, University of California, Irvine 92668, and
State Developmental Research Institute, Fairview, Costa Mesa, California 92626
Taylor,
Derek V., James G. Boyajian,
Norman
James,
Danette Woods, Aleksandra
Chicz-Demet,
Archie F. Wilson, and Curt A. Sandman.
Acidosis stimulates P-endorphin
release during exercise. J. Appl. Physiol. 77(4): 1913-1918,
1994.-Elevated blood levels of /3-endorphin have been associated with high-intensity exertion, but the stimulus for P-endorphin release is unknown. Some studies of exercise have associated @-endorphin release with increased exertion levels, but
other evidence suggests that acidosis may stimulate the release
of /3-endorphin. This study examines acidosis as a possible stimulus for /3-endorphin release by examining the effects of arterial blood gases, whole blood lactate, and respiratory changes
on /?3-endorphin levels and by examining the effects of buffering
during exercise on these levels. Initially, seven healthy adult
males were evaluated during incremental exercise. During incremental exertion, indicators of acidosis correlated with endorphin levels: pH (r = -0.94), PCO~ (r = -0.83, HCO, (r =
-O&3), base excess (r = -0.94), and lactate (r = 0.89). A multivariate model showed that P-endorphin levels were predicted
best by the change in base excess. A time course analysis
showed that /3-endorphin responses peaked postexercise and
paralleled blood acid levels. Subsequently, subjects were compared after alkali loading and placebo during constant-intensity exercise at 85% of maximal exertion to determine whether
acidosis is necessary for endorphin release. Treatment with a
buffer, which effectively maintained pH above 7.40, significantly suppressed endorphin release (F = 3.07; P < 0.0001).
The results of this study indicate that acidosis rather than any
other physiological change associated with high-intensity exertion is the primary stimulus for /3-endorphin release.
endorphins; lactates; bicarbonates; opiates; respiration
PHYSICAL EXERTION leads to increase in the blood levels
of opioid peptide p-endorphin in humans (10,15,17), but
the physiological trigger is unknown. Exercise at an intensity >60% of maximum 0, uptake (VO,,,;
Refs. 10,
17) or at high muscle work load (15) significantly increases @-endorphin levels; lower-intensity exertion does
not increase endorphin levels (10, 27). Although these
studies concur that endorphin release occurs during
high-intensity exercise, they fail to identify the physiological trigger responsible for stimulating endorphin release.
Physiological changes occurring at higher-intensity
exertion may reveal the mediator of the endorphin response. Several studies suggested that ,B-endorphin is released when the anaerobic threshold is reached (3, 17,
23). Because the lactate turn point, characterized by production and release of large amounts of lactic acid (29)) is
a defining feature of the anaerobic threshold, it is possible that increased blood acidity rather than lactate ion is
responsible for the release of ,8-endorphin. Several exercise studies showed that P-endorphin release correlates
with both increased lactate production and decreased
0161-7567194
$3.00
Copyright
blood pH (3, 9, 16, 23). A remarkably similar result was
found in neonates that had experienced hypoxia (28). In
these studies, ,&endorphin levels at birth were closely
associated with decreased blood pH, suggesting that acidosis may be the primary stimulus for P-endorphin release.
If pH rather than lactate levels control /3-endorphin
release during exertion, then treatment with a blood
buffer should attenuate the endorphin response. The
present study examined the relationship between indicators of acidosis (i.e., arterial pH, HCO,, base excess, and
lactate) and endorphin release during incremental exertion. The influence of acidosis on endorphin release was
specifically examined by buffering pH changes during
high-intensity
exertion at a constant work load.
METHODS
Subjects
The study was approved by the Human Subjects Research
Committee of the University of California at Irvine (approval
no. 90-231), and each subject provided signed consent before
participating. Screening of subjects included medical history,
resting electrocardiogram, and spirometry test. Subjects were
excluded if they had any evidence of renal dysfunction, cardiac
abnormality, respiratory disease, or any other physical condition placing them at risk for maximal exercise testing. Seven
healthy adult males between 21 and 46 yr of age were tested.
The subjects varied in their levels of physical conditioning. All
subjects were in good health and physically active, and six had
participated in timed running events >5 km. The other subject’s activities included jogging regularly, recreational basketball, and skiing. Subject characteristics are displayed in Table 1.
Procedures
Each subject visited the exercise laboratory on four occasions: 1) screening and orientation to running on treadmill, 2)
incremental exercise with blood sampling, and 3) and 4) constant-intensity exercise with placebo or buffer, respectively. On
the first visit, after screening, each subject completed a graded
exercise test to exhaustion following the Bruce protocol (4) as
orientation to running on the treadmill. 0, uptake (VO,) and
electrocardiogram were measured during 3 min of standing
preexercise, exercise, and 10 min of recovery (7 min of walking
followed by 3 min of standing).
Respiratory,
blood gas, and endorphin
measurements during
incremental exercise. On the second visit, the subjects returned
to complete an incremental exertion test (as above) to study
the relationship among arterial blood gases, respiration, and
p-endorphin. One hour before the test a catheter (22-gauge arterial catheter, Arrow, Reading, PA) was placed in radial artery
of the right arm. The catheter was kept patent by continuous
flushing with normal saline at a rate of 20 ml/h, and the subjects rested for 30 min before an initial sample (7.5 ml) was
obtained. Additional samples were collected every 3 min corre-
0 1994 the American
Physiological
Society
1913
1914
ACIDOSIS
1. Subject characteristics
TABLE
Subj No.
Age, yr
1
27
46
23
21
2
3
4
5
6
7
Mean
STIMULATES
Height,
19
AI SD
27
24
26.7 k
i% 2 max, maximum
9.0
cm
180
178
176
175
182
179
180
179 2 6.1
Weight,
kg
82.5
73.2
65.9
68.2
78.2
67.6
74.5
72.9 + 6.1
VO, -,
ml
l
kg-’
l
min-’
59.7
54.7
66.2
63.9
53.6
54.4
63.1
59.4 -t 5.2
0, uptake.
sponding to the stages of exercise, at maximum exertion, and at
2, 5, and 10 min postexercise.
HCO; or placebo loading during constant-intensity
exertion.
During the third and fourth visits, subjects completed a 20-min
constant-intensity
exercise requiring 85% of Vozmax, as determined from the results of the incremental
exercise. The order
of treatment with buffer or placebo was randomly balanced and
double blind. Each experiment began with the placement of an
arterial catheter (see above) followed by placement of an intravenous catheter into the right forearm that was also kept patent with normal saline. After an initial arterial sample was collected, the subject ingested within a IO-min period either 0.3
g/kg of sodium bicarbonate
dissolved in 500 ml of water with
cherry-flavored
syrup or a placebo solution
of water with
cherry-flavored
syrup followed by 1 h of rest to allow for absorption. Buffering
was maintained
with intravenous
isotonic
sodium bicarbonate,
with the amount calculated
from (projected base excess at 85% of Vo, maxin meq/l) X (body weight in
kilograms)
X (0.2 l/kg body wt of extracellular
fluid). Base excess for running at 85% of %Jo, maxfor each subject was interpolated from the best-fit curve of base excess vs. VO,. For the
placebo experiment,
an equal volume of normal saline was infused. The rate of administration
was adjusted so that the entire volume would be infused during the preexercise and exercise periods.
Exercise commenced
with a 3-min warm-up period at 1.7
mph on a 5% grade followed by a 3-min transition
phase and
then 20 min at a constant work load, which required VO, of
285% of Vo, max.Blood samples were taken at the end of preexercise at the end of the warm-up period, every 2 min of exercise,
and at 2, 5, and 10 min postexercise.
For each subject, both
buffer loading and placebo experiments
were performed
at the
same time of day with exercise beginning between 9 A.M. and
12 P.M.
Physiological measurements. All exercises were performed on
a treadmill (Marathon,
California
Medical, Brea, CA). Ventilatory measurements
during exercise were made on a breath-bybreath basis, averaged, and reported for each 15-s interval. The
expired air of the subjects was sampled through a large “J”
valve (model 2700, Hans Rudolph,
Kansas City, MO; dead
space 100 ml) and routed via a capillary tube into a mass spectrometer (model 1100, Perkins-Elmer,
Pomona, CA) at a rate
of 60 ml/min. vo2, CO, production
(ho,),
respiratory
rate
(RR), and respiratory
exchange ratio (R) were calculated from
minute ventilation
(iTE) and exhaled gas measurements.
The
flow signal from a pneumotachograph
(model 50 MC, Meriam,
Cleveland, OH) connected to the J valve via 3 m of flexible
tubing (3.5 cm ID) was integrated
to determine
expired VE.
Continuous
cardiac monitoring
was obtained with a three-lead
electrocardiogram
(II, aVF, and V5, Brentwood
Instruments,
Torrance, CA).
Blood gases. Blood gases were measured using blood gas analyzers (Corning 178 and Corning CO-oximeter
2500, Norwood,
MA) that measured pH, Pco,, and PO, by separate electrodes
ENDORPHIN
RELEASE
and 0, saturation
by spectrophotometry.
The first machine
also provided mathematically
derived HCO, and base excess
measurements.
Lactate. Whole blood lactate was analyzed using a lactate
analyzer (model YSI 23L, Yellow Springs Instruments,
Yellow
Springs, OH).
P-Endorphin.
Blood samples were collected in chilled EDTA
venipuncture
tubes containing 500 KIU/ml of aprotinin
(Sigma
Chemical, St. Louis, MO). Plasma was separated by centrifugation (2,000 g; 15 min) and stored in polypropylene
tubes at
-70°C until assayed in duplicate. Plasma ,&endorphin levels
were assayed by a solid-phase two-site immunoradiometric
assay (Nichols Institute Diagnostics,
San Juan Capistrano, CA).
The minimum
detectable dose of this method is 14 pg/ml(95%
confidence limit) with coefficient of variance of 4.1% (intra-assay) and 9.0% (interassay) at the highest concentration measured in the present study. Cross-reactivity with related opiates
(and also oxytocin) is ~0.01% at 5 pg/ml and that with adreno-
corticotropic hormone is 0.028% at 900 rig/ml.
Statistics
All statistics were performed using BMDP386 statistical software for the IBM personal computer (BMDP Statistical Soft-
ware, Los Angeles, CA). A trend analysis of incremental exercise data was performed for each subject to assessthe effects of
the independent variables (IV) (base excess, lactate, pH, Pco~,
PO,, HCO,, R, RR, VCO,, VO,, and VE) on the dependent variable (DV) (P-endorphin) and to obtain a line or curve of best fit
for each subject. To more accurately compare the correlation
coefficients in light of the small sample size, McNemar’s r to z
transformation
(19) was used to reduce the heterogeneity
of
variance among the subjects. Multiple regression, implementing BMDP program 2R, was used to obtain a weighted linear
combination of the IVs (base excess, lactate, pH, Pco~, PO,,
HCO,, R, RR, ko2, VO,, and VE) that best predicted the
amount of variance accounted for by the DV (@-endorphin).
Experiment-wise
type I error was controlled
at cy = 0.05 using a
Bonferroni
adjustment
for all planned comparisons
and post
hoc test of correlation coefficients.
A repeated
measures multivariate
analysis of variance,
BMDP 4V, was performed to assess the effects of the IV “treatment” (placebo vs. buffer) on DV (pH, base excess, HCO,, lactate, and P-endorphin)
over 14 treatments
(time points during
exercise). Analysis of covariance was used to test the influence
of the covariates (blood lactate and HCO,) on the DV (P-endorphin).
RESULTS
Incremental
Exercise
high aerobic capacity
. The subjects demonstrated
WO 2max59.4 t 5.2 ml kg-’ min-‘) as shown in Table 1.
Figure 1 shows time courses for base excess, p-endorphin, and VO,. Because exercise time varied for each subject, the time scale is expressed as percentage of maximal
time so that 100% = end of test (15.8 t 4.0 min). ,&Endorphin levels peaked at 23.0 t 2.7 min. This peak was
immediately preceded by nadirs of base excess (21.3 t 4.1
min) and pH (21.7 t 4.05 min). In contrast, the 7j02 peak
(15.5 t 3.23 min) coincided with the end of test.
,&Endorphin
levels during rest, incremental exercise,
and recovery significantly correlated with pH (r = -0.94;
P < O.Ol), PCO, (r = -0.85; P < O.Ol), HCO, (r = -0.88; P
<O.Ol), base excess (r = -0.94; P < O.Ol), and lactate (r =
0.89; P < 0.01). Effects during exertion only (recovery
l
l
ACIDOSIS
EXERCISE
=
E
s-
400
350
STIMULATES
ENDORPHIN
RECOVERY
400
h
END OF TEST
j
1:-‘\
i /’ \.&
\
i I.
1915
RELEASE
r
I
.
f 300
g 250
PC: 200
0
2 150
Y 100
4
b
50
m
0
5
-100
-20
<
po
-25-30
I
-20
’
-25
I
-15
!
-10
I
e-Le--.--i
-5
0
5
BASE EXCESS (mM1
FIG. 2. Values of @-endorphin corresponding to base excess values
during rest, incremental exercise, and recovery are shown with corresponding regression line.
f O
-E -5
g
L-30
changes in base excess and RR: ,&endorphin
9.43(base excess) - 1.20RR.
Constant-Intensity
-
1
k
1
1
80 -
= 83.65-
Exertion
The individually selected treadmill speeds and inclinations, on the basis of the second incremental exercise
test, allowed subjects to maintain a constant level of exertion, requiring a VO, of 50.9 t 11.0 ml kg-‘. mine1 or
85.6 $- 15.9% of vogrnax throughout constant-intensity
testing. VO, did not increase significantly with time of
exercise nor was there a significant difference in intensity between placebo (50.6 t 12.7 ml. kg-‘. min-‘) and
buffer (51.3t 9.0 ml kg-’ min-‘) conditions.
Buffering treatments significantly elevated pH values
throughout the entire test compared with during the plaas shown in Fig. 3.
cebo condition (F, 1o = 28;P < O.OOOl),
Mean pH was 7.43't 0.04 in the buffer condition and 7.32
t 0.05 during the placebo condition. In both conditions,
pH decreased significantly with time (F,,,, = 26; P <
0.001) during constant-intensity
exercise. During the
placebo condition, pH dropped from 7.42 & 0.02 to 7.31 t
0.05. Buffering elevated pH to 7.50 t 0.01 at the start of
exercise and prevented values at the end of exercise (7.41
t 0.05) from dropping below the normal resting range.
Elevation of pH at the start of the test Pras due to an
l
l
,
0
50
PERCENT
SUBJECT
100
OF TOTAL
200
150
EXERCISE
TIME
_._.a #I _ _ _ w#2 ______.#3 .......... #4
.ee.-....e
#5
#6 w..-..-.#7
FIG. 1. P-Endorphin, arterial base excess,and 0, uptake #o& of
each subject during incremental exercise test are graphed on standard
time line. Time of incremental exercise test was standardized by dividing each time point by total exercise time and expressing it as percent.
Separate line type is used for each subject. Data for each subject are
represented by best-fit curves. Plots for VO, are biphasic linear,
whereas those for base excess and fi-endorphin are fourth-order quadratic.
PREIWARM-UP
i
:
l
CONSTANT
:‘
INTENSITY
RECOVERY
= BUFFER
= PLACEBO
and rest values excluded) yielded similar values for pH (r
= -0.93; P < O.Ol), HCO, (r = -0.87; P > O.Ol),base
excessjr = -0.93; P < O.Ol),and lactate (r = 0.94;P <
0.01). VE (r = 0.79; P < 0.05) and RR (r = 0.81; P < 0.05)
were also significantly
correlated with P-endorphin
values.
A stepwise multiple regression was used to determine
the best predictor of ,&endorphin change. Base excess
accounted for all the variance due to acidosis eliminating
:
:
further need for pH, HCO,, or lactate in the equation
[
i
:: , -1-1
P-endorphin = 52.4 - 8.9(base excess) (R2 = 0.88).The
10
20
30
0
strong relationship between base excess and ,@-endorphin
is demonstrated in Fig. 2. The only respiratory measure
TIME (MINUTES)
that accounted for further variance was RR, and the final
FIG. 3. Means k SD of pH at each time point during constant-inmodel R2 = 0.93 related changes in ,&endorphin to tensity exercise for buffer and placebo conditions.
i
::: i
1
?
i
ACIDOSIS
1916
2o
PRE/WARM-UP
;
:
:
:
:
:
:
i
15
z
g
T
10
-
v-y&
T
:
-20
T
\-
CONSTANT
v
Cl
INTENSITY
RECOVERY
;
:
:
’
= BUFFER
= PLACEBO
I
10
’
0
STIMULATES
:
:
*
I
20
PREIWARM-UP
;
:1
15.0
t
CONSTANT
v
0
g 12.5
E
w 10.0
INTENSITY
= BUFFER
= PLACEBO
T
I
1
#-
a
)li
7.5
cl3
4
5.0
~
2.5 t
&+zg0.0
0
/
l!l
:
i
!
I
10
20
r PREMIARM-UP
180
-
e160
-
s
-
140
i
CONSTANT
:
:
:
:
:
:
:
:
:9
v
0
0
!
I
RECOVERY
= BUFFER
= PLACEBO
T
I
T
I
lo
20 l
lNTENSITY
,:
I
-a
,
-a
30
TIME (MINiiTES)
TIME (MINUTES)
t
RELEASE
2oo
30
4.
17.5
‘;;
3
,
Means + SD of base excess at each time point during constant-intensity exercise for buffer and placebo conditions.
FIG.
ENDORPHIN
T
::
:
:
2oo
RECOVE sRY
:,
u
‘+-p-v
:
--Iaa
:::-.
:
:
:
FIG. 6. Means + SD of &endorphin
at each time point during constant-intensity exercise for buffer and placebo conditions.
PREIWARM-UP
E
180
-
zi
.4
2
160
-
140
-
120
-
100
-
E
Q,
a
0
0
2
Y
a
Ig
TIME (MINUTES)
FIG. 5. Means + SD of blood lactate at each time point during constant-intensity exercise for buffer and placebo conditions.
almost eightfold increase in the base excess (from 1.3 t
1.6 mM in the placebo condition to 10.3 t 2.1 mM in the
buffer condition). Similar to changes in pH (Fig. 4), base
excess values were higher in the buffer condition than
those in the placebo condition (& 1o = 52; P < 0.001) and
decreased equally with time in both conditions (& 130=
43; P < 0.001). However, buffering did not prevent mean
base excess levels from dropping slightly below zero at
the end of exercise for some subjects. Base excess values
at the end of exercise were -9.0 t 3.3 meq/l during the
placebo condition compared with -1.6 t 3.6 meq/l during the buffer condition.
Although lactate values generally increased with time
(F 13,130 = 51; P < O.OOl), the effect of exercise time on
lactate values was clearly greater in the buffer condition
than that in the placebo condition (F13 130= 1.8; P < 0.05),
since lactate increased from 1.0 t 0.3 to 9.8 t 3.2 mM
during buffering but increased from 0.8 k 0.1 to 6.4 t 2.1
mM during the placebo condition (Fig. 5).
P-Endorphin increased linearly with time across both
conditions (F,, 130= 9.0; P < O.O), and there was a strong
tendency for P’-endorphin values to be lower across the
buffer condition (Fl 1o = 3.87; P = 0.08, Fig. 6). Using
blood HCO, and lactate as covariants revealed that
:
:
:
:
:
80
60
40
,
30
i
CONSTANT
INTENSITY
:
:
:
:
; v = BUFFER
i 0 = PLACE80
:
:
:
:
:.
:
:
:
:
20 ’
0
.t
I
20
10
:
TIME (MINUTES)
FIG. 7. Adjusted
mean P-endorphin
and lactate as covariates.
values with effects of HCO;
treatment with buffer depressed the rate of fl-endorphin
increase compared with treatment with placebo (F,, 128=
3; P < 0.0001). Figure 7 shows the adjusted means from
the analysis of /3-endorphin across time when the effects
of lactate and blood HCO, are controlled.
DISCUSSION
The findings in this study during incremental exercise
indicate that acidosis stimulates the release of @endorphin and that prevention of acidosis by buffering during
high-intensity
exercise largely prevents the increase in
P-endorphin levels. Changes in measurements of acidbase status, including blood lactate concentration,
decreased arterial pH, base excess, HCO,‘, and CO, levels,
significantly corresponded with higher levels of P-endorphin. These data are consistent with prior studies of exercise (9, 16, 23) as well as hypoxic stress (28) in which
acidosis correlated with /3-endorphin. A multivariate
model showed that base excess accounted for 88% of the
variance and is the most important variable in predicting
/3-endorphin response. The small amount of variance
that RR accounts for may be the result of reciprocal interaction between the opioid system and respiration.
ACIDOSIS
STIMULATES
Manipulation
of acid-base status during constant-intensity exertion confirmed that acidosis is the primary
stimulus affecting the rate of P-en.dorphin release. During 20 min of exertion at 85% of Vozmax, buffering successfully maintained pH above 7.41. Buffering was accompanied by an increase in the rate of blood lactate
elevation and, at the same time, by a suppression of the
endorphin response. These effects are opposite those expected if lactate ion were responsible for stimulating the
increase in @-endorphin. Other research established that
during exercise, when lactate levels exceed the buffering
capacity of the cell (0.4 meq/l), lactate diffuses out of the
cell in its conjugate acid form and is buffered in equimolar quantities by HCO, (2). Therefore, the effect of buffering on blood lactate confounds attempts to maintain a
constant acid-base status. Our use of both lactate and
HCO, as covariates statistically corrected for these effects in a manner consistent with modeling done by
Beaver et al. (2). It is evident that HCO, loading strongly
suppresses the endorphin response even though it increases lactic acidosis.
Although during exercise @-endorphin showed a concomitant increase with respiration, the overall pattern of
response suggests an acid-stimulated
response. Respiratory measurements (VE, VO,, and RR) correlated with
@-endorphin when only exertion was considered. As seen
with other studies (9, ll), ,&endorphin
values peaked
during the lO-min recovery period after exercise, paralleling lactic acidosis, whereas VO,, VE, and RR decreased
immediately at the end of exercise. In this study, considering the recovery period in analysis removed correlations with respiratory measurements,
suggesting that
these measurements increase coincidentally with P-endorphin and are not causally related. The time lag between acidosis and endorphin response (-1.5 min) is
more consistent with the known response latency of ,&
endorphin to other stresses (14). Therefore,
acidosis
seems to be the primary stimulus for ,8-endorphin release. The correlation between respiration and p-endorphin during incremental exercise may be due to the effect
of acidosis on these two measurements by separate mechanisms. It has been established that acidosis stimulates
respiration through action on chemosensitive neurons in
the medullary respiratory control center (8,13). The hypothalamus is a likely site for pH-sensitive neurons controlling endorphin response because it controls pituitary
releases of P-endorphin (1). Therefore, acidosis may effect both respiration and endorphin release but at separate sites.
The response of P-endorphin to acidosis may be consistent with indirect evidence that ,&endorphin inhibits
respiration during exertion (12, 18) and hypoxia (25).
More specifically, blocking receptors for ,&endorphin
has been shown to have a direct effect on the neurons
responsible for respiratory control in the medulla (26).
The existence of dual mechanisms for acid-stimulated
respiratory changes and ,&endorphin
release would not
exclude acidosis from having a specific role during respiratory challenge. In fact, if the role of ,&endorphin is to
prevent hyperventilation
during respiratory challenge,
then acidosis would be an ideal stimulus. Acidosis would
ENDORPHIN
1917
RELEASE
stimulate respiration and at the same time stimulate ,&
endorphin as a feedback inhibitor.
The results of this study have implications for perinatology. Perinatal stress is characterized by unusually
high neonatal endorphin levels (6,7) resulting in respiratory suppression that is treatable with opiate blockers
(5). Also, there is evidence that exposure to high endorphin levels during perinatal development may contribute
to developmental pathological conditions such as autism
(22, 24) and sudden infant death syndrome (21). If buffering of the blood decreases the release of fl-endorphin
in utero, deleterious effects of endorphin release might
be mitigated by fetal alkalinization. Fetuses may be more
responsive to the effects of buffering because both acidosis (30) and P-endorphin (20) have been shown to more
strongly affect respiratory control centers during early
development. Further study in animal models would be
useful in determining the therapeutic benefits of buffering treatment.
Although this study examines the effects of buffering
in a limited number of conditioned males, it strongly suggests that endorphin release during exertion is stimulated primarily by metabolic acidosis. During incremental exertion, P-endorphin release corresponded with increasing acidosis as measured by arterial blood gases and
whole blood lactate. Base excess was the best predictor of
P-endorphin release. A time course analysis revealed that
,8-endorphin mirrored the production
of acid with the
nadir in base excess closely preceding the peak of ,&endorphin during recovery. Most importantly,
buffering of
blood successfully attenuated the rate of ,8-endorphin release, confirming that release of endorphin during exercise is dependent on pH changes.
The
assistance
of Scott
Iteen,
David
Crook,
Sheila
Tai, Paul
Nguyen,
and Laura Tai is greatly
appreciated.
Comments
from Penny
Fidler were valuable
in revising
the text.
This study was supported
in part by National
Institute
of Child
Health
and Human
Development
Grant
ROl-HD-28413-01
to C. A.
Sandman.
Address
for reprint
requests:
D. V. Taylor,
Research
Box 5A, 2501
Harbor
Blvd., Costa Mesa, CA 92626.
Received
19 November
1993;
accepted
in final
form
6 June
1994.
REFERENCES
1. Barna,
I., and J. I. Koenig.
Effects
of mediobasal
hypothalamic
lesion on immunoreactive
ACTH/beta-endorphin
levels in cerebrospinal fluid, in discrete
brain regions,
in plasma, and in pituitary
of
the rat. Brain Res. 593: 69-76, 1992.
2. Beaver,
W. L., K. Wasserman,
and B. J. Whipp.
Bicarbonate
buffering
of lactic acid generated
during exercise.
J. Appl. Physiol.
60: 472-478,1986.
3. Brooks,
S., J. Burrin,
M. E. Cheetham,
G. M. Hall,
T. Yeo,
and C. Williams.
The responses
of the catecholamines
and betaendorphin
to brief maximal
exercise
in man. Eur. J. Appl. Physiol.
Occup. Physiol.
57: 230-234,
1988.
4. Bruce,
R. A. Exercise
testing
of patients
with coronary
heart disease: principles
and normal
standards
for evaluation.
Ann. Clin.
Res. 3: 323-332,
1971.
5. Chernick,
V., and R. J. Craig.
Naloxone
reverses
neonatal
depression
caused
by fetal asphyxia.
Science
Wash. DC 216: 12521253,1982.
6. Davidson,
S. I., H. Gil-Ad,
2. Rogovin,
S. Laron,
and H.
Reisner.
Cardiorespiratory
depression
and plasma
P-endorphin
levels in low-birth
weight infants
during the first day of life. Am. J.
Dis. Child. 141: 145-148,
1987.
7. Divers,
W. J., R. D. Stewart,
M. M. Wilkes,
and S. S. Yen.
1918
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
ACIDOSIS
STIMULATES
Amniotic fluid beta-endorphin and alpha-melanocyte: stimulating
hormone immunoreactivity
in normal and complicated pregnancies. Am. J. Obstet. Gynecol. 144: 539-542, 1982.
Eldridge, F. L., J. P. Kiley, and D. E. Millhorn.
Respiratory
responses to medullary hydrogen ion changes in cats: different effects of respiratory and metabolic acidoses. J. Physiol. Land. 358:
285-297,1985.
Goldfarb, A. H., B. D. Hatfield, D. Armstrong,
and J. Potts.
Plasma beta-endorphin
concentration: response to intensity and
duration of exercise. Med. Sci. Sports Exercise 22: 241-244, 1990.
Goldfarb, A. H., B. D. Hatfield, J. Potts, and D. Armstrong.
Beta-endorphin time course response to intensity of exercise: effect of training status. ht. J. Sports Med. 12: 264-268, 1991.
Goldfarb,
A. H., B. D. Hatfield, G. A. Sforzo, and M. G.
Flynn. Serum beta-endorphin levels during a graded exercise test
to exhaustion. Med. Sci. Sports Exercise 19: 78-82, 1987.
Grossman, A., P. Bouloux, P. Price, P. L. Drury, K. S. Lan,
T. Turner, J. Thomas, G. M. Besser, and 3. Sutton. The role
of opioid peptides in the hormonal responses to acute exercise in
man. Clin. Sci. Land. 67: 483-491, 1984.
Kiley, J. P., F. L. Eldridge, and D. E. Millhorn.
The roles of
medullary extracellular and cerebrospinal fluid pH in control of
respiration. Respir. Physiol. 59: 117-130, 1985.
Kjaer, A., U. Knigge, F. W. Bach, and J. Warberg. Histamineand stress-induced secretion of ACTH and beta-endorphin:
involvement of corticotropin-releasing
hormone and vasopressin.
Neuroendocrinology 56: 419-428,1992.
Kraemer, W. J., J. E. Dziados, L. J. Marchitelli,
S. E. Gordon, E. A. Harman, R. Mello, S. J. Fleck, P. N. Frykman,
and N. T. Triplett. Effects of different heavy-resistance exercise
protocols on plasma /3-endorphin concentrations. J. Appl. Physiol.
74: 450-459,1993.
Kraemer, W. J., S. J. Fleck, R. Callister, M. Shealy, G. A.
Dudley, C. M. Maresh, L. Marchitelli,
C. Cruthirds,
T.
Murray, and J. E. Falkel. Training responses of plasma betaendorphin, adrenocorticotropin,
and cortisol. Med. Sci. Sports Exercise 21: 146-153,1989.
McMurray,
R. G., W. A. Forsythe, M. H. Mar, and C. J.
Hardy. Exercise intensity related responses of beta-endorphin
and catecholamines. Med. Sci. Sports Exercise 19: 570-574, 1987.
ENDORPHIN
RELEASE
18. McMurray,
R. G., D. S. Sheps, and D. M. Guinan. Effects of
naloxone on maximal stress testing in females. J. Appl. Physiol. 56:
436-440,1984.
19. McNemar, Q. Psychological Statistics. New York: Wiley, 1962, p.
451.
20. Moss, I. R., L. M. Sugarman, and D. L. Goode. Endogenous
opioid effect on breathing during normoxia and hypoxia in developing swine. Biol. Neonate 52: 337-346,1987.
21. Myer, E. C., D. L. Morris, M. L. Adams, D. A. Brase, and
W. L. Dewey. Increased cerebrospinal fluid beta-endorphin immunoreactivity in infants with apnea and in siblings of victims of
sudden infant death syndrome. J. Pediatr. 111: 660~666,1987.
22. Panksepp, J., T. Vilberg, and N. J. Bean. Reduction of distress
vocalization in chicks by opiate-like peptides. Brain Res. Bull. 3:
663-667,1978.
23. Rahkila, P., E. Hakala, M. Alen, K. Salminen, and T. Laatikainen. Beta-endorphin and corticotropin release is dependent on
a threshold intensity of running exercise in male endurance athletes. tife Sci. 43: 551-558, 1988.
24. Sandman, C. A. The opiate hypothesis in autism and self-injury.
J. Child Adolesc. Psychopharmacol. 1: 235-246,199O.
25. Schaeffer, J. I., and G. G. Haddad. Ventilatory response to
moderate and severe hypoxia in adult dogs: role of endorphins. J.
Appl. Physiol. 65: 1383-1388,1988.
26. Trouth, C. O., F. J. Bada, Y. Pan, J. A. Holloway,
R. M.
Millis, and D. G. Bernard. Naloxone application to the ventrolatera1 medulla enhances the respiratory response to inspired carbon
dioxide. Life Sci. 49: 193-200, 1991.
27. Viru, A., 2. Tendzegolskis,
and T. Smirnova. Changes of betaendorphin level in blood during prolonged exercise. Endocrinol.
Exp. 24: 63-68,199O.
28. Wardlaw, S. L., R. I. Stark, and L. Boci. Plasma fl-endorphin
and lipotropin in the human fetus at delivery: correlation with arterial pH and PO,. J. Clin. Endrocrinol. Met& 70: 888-801,1979.
29. Wasserman, K. The anaerobic threshold measurement to evaluate exercise performance. Am. Rev. Respir. Dis. 129: S35-S40,1984.
30. Whittaker,
J. A., C. 0. Trouth, Y. Pan, R. M. Millis, and D.
G. Bernard. Age differences in responsiveness of brainstem chemosensitive neurons to extracellular pH changes. Life Sci. 46:
1699-1705,199O.