Content uploaded by David Christopher Nieman
Author content
All content in this area was uploaded by David Christopher Nieman on May 28, 2015
Content may be subject to copyright.
467
--
The Effects of Moderate Exercise Training on Natural Killer
Cells and Acute Upper Respiratory Tract Infections
D.
C. Nieman,
S.
L. Nehlscn-Cunnarella,
P.
A. Markoff;
A.
J.
Balk-Lurnberfon,
H.
Yang,
D.
B.
W.
Chrirrorz,
J.
W.
Lee, and
K.
Arabafzis
Department of Health Science, School of Public Health Lorna Linda University. lmmunology Center, Lorna Linda
University Medical Center.
I
'
Abstract
I
D.
C. Nieman,
S.
L.
Nehlsen-Cannarella,
P.
A.
1
Markoff;
A.
J.
Balk-Lamberron,
H.
Yang,
D.
B.
W.
Chriaon.
J.
W.
Lee,
K.
Arabafzis, The Effects of Moderate Exercise
Training on Natural Killer Cells and Acute Upper Respira-
tory Tract Infections. Int
J
Sports Med, Vol 11, No 6, pp
467-473,1990.
I
I
Accepted after revision: February 12, 1990
A randomly controlled 15-wk exercise train-
ing
(ET)
study (five 45-min sessionsjwk, brisk walking at
60% heart rate reserve) with a group of 36 mildly obese,
sedentary women was conducted to investigate the relation-
ship between improvement in cardiorespiratory fitness,
changes in natural killer (NK) cell number and activity, and
acute upper respiratory tract infection (URI) symptoma-
tology. The study was conducted using a 2 (exercise and
nonexercise groups)
x
3 (baseline, 6-, and 15-wk testing
sessions) factorial design, with data analyzed using re-
peated measures ANOVA. No significant change in NK
cell number occurred as a result of ET as measured by the
CD16 and Leu-19 monoclonal antibodies. ET did have a
significant effect on NK cell activity
(E:T
50:l) especially
during the initial 6-wk period [F(2,68)
=
12.34, p
<
0.001
1.
Using data from daily logs kept by each subject, the exercise
group was found to have significantly fewer URI symptom
daysjincident than the nonexercise group (3.6k0.7 vs
7.0
k
1.4 days, respectively, p
=
0.049). Improvement in
cardiorespiratory fitness was correlated significantly with a
reduction in URI symptom daysjincident (r
=
0.37,
p
-
0.025) and a change in
NK
cell activity from baseline to
six but not 15 wks (r
=
0.35, p= 0.036). In summary, mod-
erate ET is associated with elevated NK cell activity after six
but not 15 weeks, and reduced URI symptomatology in
comparison to a randomized, sedentary control group.
Key
words
~
I
exercise, upper respiratory tract infections.
I
natural killer cell activity. immune system
i
Introduction
Acute upper respiratory tract infections (URI)
impose a significant burden each year in terms of days of dis-
ability, lost school or work days, and medical costs (23). The
Centers for Disease Control has estimated that 429 million
URI occur annually in the US resulting in $2.5 billion in direct
costs (37). The 1987 US incidence rates per 100 persons per
year for all acute respiratory conditions was 80.1, led by in-
fluenza (38.2) and the common cold (25.9) (23).
There is a growing interest in the effects of both
acute and chronic exercise on immunosurveillance and host
protection (14, 18-22,26,30,33,35,40). Although it is gener-
ally believed that exercise training increases resistance to URI,
little scientific evidence exists to support this contention. Most
studies have examined the effects of acute submaximal and
maximal exercise on changes in immune system parameters
and function, providing little direct evidence as to how these
alterations affect resistance to URI. Few prospective studies
have been conducted in this area, and none have utilized ran-
domized sedentary control groups (35,40).
A better understanding of the relationship be-
tween exercise and URI may be gained by investigating the ef-
fects of exercise training on natural killer (NK) cells and their
activity. NK cells, first discovered in the early 1970s, are
unique
in
that they express spontaneous cytolytic activity
against
a
variety of tumor and virus-infected cells (12,41). Un-
like T lymphocytes, NK cells do not require the involvement of
major histocompatibility antigens to initiate cytotoxicity. NK
cells account for 10- 15
%
of mononuclear cells in the periph-
eral blood, and have been observed to respond rapidly to for-
eign materials and initially control them until the antigen spe-
cific immune system begins to respond. Thus interest in NK
cells and their activity has grown because they represent a
major first line defense system against viral infection (1 7,4 1).
Nearly all studies to date have measured the
acute effect of exercise, both maximal (2, 6,
9,
16) and sub-
maximal (1, 7, 29, 36), on NK cell activity and numbers. We
are aware of only two prospective studies that have measured
the effects of exercise training on resting NK cell activity in
human subjects (5,40).The studies arrived at different conclu-
sions, with Crist et al. (5) reporting a significant increase in NK
cell activity following three months of exercise training and
Watson et al. (30) reporting
a
decrease. In randomized con-
trolled studies utilizing mice. daily exercise training has been
468
In(.
J.
Sports
Med.
1
I
(1
990)
--
D.
C.
Nietvuriri
ct
al.
associated with a considerable increase in NK cell activity (8)
and improved resistance to infection (4). Pcdersen et al. (28)
have reported increased levels of NK cell activity at rest in
trained vs untrained males, and proposed that the constantly
elevated NK cell activity may promote better resistance
against URI.
If a relationship exists between exercise train-
ing and decreased risk of URI, a concomitant increase in NK
cell activity would seem to be an important correlate. To inves-
tigate the effect of moderate exercise training on URI sympto-
matology and NK cell number and activity, we conducted a
randomly controlled 15-week brisk walking study with a
group of mildly obese, sedentary, premenopausal women.
This type of subject was selected because women of this age
have been reported to have the highest rates of URI among
adults (23), and to experience considerable cardiorespiratory
benefits from regimens of brisk walking (1 3).
Methods
Subjects
The methods for this study have been de-
scribed elsewhere (24). A brief description of our methods will
be summarized here. Fifty women were selected for this study
who met the following criteria: 25-45 years of age, mildly
obese (10-40% overweight), premenopausal, 155- 170 cm in
height, not presently on an exercise program or a reducing
diet, a nonsmoker without a history of alcohol or drug abuse,
no current use of medications (except oral contraceptives), ab-
sence of hypertension and diabetes, and no family history of
heart disease. Those who qualified for the study were in-
structed that they would be randomly assigned to an exercise
(EX) or nonexercise (NEX) group. All subjects agreed to ac-
cept their randomized placement preceding the study, and
were blinded as to the aims of the study. Each subject volun-
tarily signed an informed consent statement approved by the
Loma Linda University Institutional Review Board for
Human Studies.
Experimental Design
The 15-week research project extended from
the last weekend of January to mid-May 1989. Testing was
conducted at 3 time periods: baseline, 6 weeks, and 15 weeks.
Since exercise training was conducted on weekdays, testing
was peformed on Sundays to allow EX subjects a minimum of
36 h of recovery from the last exercise session.
All subjects reported to the Loma Linda Uni-
versity Human performance Lab for testing at 0700 h follow-
ing 12 h of fasting. After resting for at least 5 min, blood
samples were collected. Subjects returned throughout the day
for assessment of the following: height and weight, body com-
position (hydrostatic weighing and 7-site skinfold tests),
resting 12-lead EKG, and 12-lead EKG graded exercise test-
ing with metabolic measurements. If a subject exhibited overt
symptoms of URI, the appointment was rcschcduled.
Bociy Cornposition
crr
zd Trecrcirniii Testing
Body density was determined by hydrostatic
weighing, as outlined by Pollock et al. (31 ),with the formula of
Siri (34) used to determine the percent body fat. Residual
volume was measured with the Jaeger constant volume, vari-
able pressure plethysmograph (Erich Jaeger, GmbH
&
Co.,
Wiirzburg, FRD). Body mass index (BMI) was calculated
using the formula kglm'.
Maximal graded exercise testing was con-
ducted using the Bruce treadmill protocol on the Quinton
44000 stress test system and Q55 treadmill (Quinton Instru-
ment Co., Seattle, WA). Metabolic measurements were taken
with the Sensor Medics MMC Horizon System 4400 meta-
bolic cart (Sensor Medics, Anaheim, CA).
Exercise Progranz
For 1'5 weeks the EX group followed a closely
supervised walking program on a measured course. This con-
sisted of five 45-min sessions each week at an intensity of 60%
of heart rate reserve. To ensure that the subjects exercised at a
proper intensity, heart rates were monitored by checking pulse
rates every 0.8 km. At the completion of 45 min the supervisor
recorded their walking distance to the nearest 0.16 km. During
the 15-week study, the NEX group was instructed not to parti-
cipate in any exercise outside of normal daily activity.
Subject Recording of UR
I
Symptoms
Log books for daily recording of health prob-
lems and exercise patterns were given to each subject at base-
line. Careful verbal and written instructions were given. Sub-
jects recorded health problems each day of the 15-week study,
using 10 codes utilized in a previous study by the Centers for
Disease Control (I 1). The coded health problems included:
cold (runny nose, cough, sore throat), allergy (itchy eyes,
stuffy nose), headache, fever,
nausea/vomiting/diarrhea,
fatigue/tiredness, muscle/joint/bone problem or injury,
menstrual cramps, other (describe in blank), or none. URI was
defined when subjects coded for a cold with or without
supporting symptoms of headache, fever, fatigueltiredness,
or nausea/vomiting/diarrhea. An episode of URI was
deemed to have commenced if symptoms were coded for a
minimum of 48 h and separated by at least one week from a
previous episode. Mode of exercise was recorded using 25
coded activities with duration and/or distance reported with
each.
Blood Analvsis
Heparinized whole blood was used for NK cell
number and activity assays and EDTA whole blood for
complete blood counts (CBC). CBC were performed on Coul-
ter S-Plus IV instrumentation with visual cell differentials in
our clinical hematology laboratory.
Peripheral blood lymphocytes were isolated
by Ficoll-Hypaque centrifugation. Using direct immuno-
fluorescence staining of cell surfaces in lysed whole blood with
mouse anti-human monoclonal antibodies (mAb) from Bec-
ton Dickinson lmmunocytometry Systems (Mountain View,
CA),
we analyzed the blood samples for NK cells (CD
16
and
Leu-19
[CD
undesignatcd]). Blood samples wcre also ana-
lyzed for other lymphocyte subpopulations, data which are
being presented clsewhcrc (24). Dual-stained samplcs were
The
Eflecls
of
hfoderule E.~ercise Truir~ing or? Nuturul
Killer
Int.
J.
Sports
itled.
11
(1
990)
469
-
Table
1
Metabolic Parameters at Baseline, 6 weeks, and 15 weeks. (Meansf SE)
Exercise Group Nonexercise Group Effect
(N
=
18) (N= 18) Group
x
Time
Variable Baseline
6
weeks
15
weeks Baseline
6
weeks
15
weeks
p
value
HRstage
2
(bpm) 162
f
3
152
f
3t 146
f
4t 153
f
2 152
f
2 150
f
3 0.003
HRm, (bpm) 183
f
2 184
f
2 183 f2 179
f
2 182
f
2 180
f
2 0.320
VEsIage
2
(Imin-
')
53.7
f
1.9 44.6f1.7t 44.1f1.4t 48.4f1.8 47.8
f
1.9 47.4
f
1.8
<
0.001
VE~,
(~rnin-') 81.8
f
2.7 85.3
f
2.4 84.8
f
3.2 81.1
f
3.2 85.1
f
2.9 88.2f 2.8 0.408
~02
stage
2
(ml.kglmin-') 21.1
f
0.4' 19.2
f
0.3t 18.9f 0.2t 19.3f 0.3 19.0
f
0.2 18.8
f
0.2
<
0.001
vOzmax (ml.kg1 min-') 25.7
f
0.9 26.4
f
0.7 26.3f 0.7 25.0f 0.9 25.1
f
0.9 24.7
f
0.9 0.482
HR,heartrate; Stage2, Bruceprotocol2.5mph, 12%elevation; Max,max~maleffort;
~~,ventilation;~02,oxygenuptake.
'p
<
0.05 between subjects baseline. 6 weeks or 15 weeks
tP
<
0.05 w~thin subjects 6-week or 15-week values versus baseline.
run and analyzed on a Becton Dickinson FACScan flow cy-
tometer using appropriate negative, isotype and lymphocyte
controls. Using the procedure of Pross and Maroun
i32), the activity of NK cells was assessed using a standard
'~r release assay at baseline, 6 weeks, and 15 weeks. Blood
lymphocytes (effectors) were isolated within 2 h of venipunc-
ture. The target cells were K562 cells, a human ery-
throleukemic cell line (American Tissue Culture). Target cells
were used 24 h after subculturing during log phase of the
growth cycle. Effector and target cells were tested at the effec-
tor: target (E:T) ratio of 50:1 in six replicates. Controls in-
cluded cultures of untreated cells (spontaneous release), and
cells treated with 3
'%
SDS (total release). NK cell activity was
calculated using the following formula:
VoNK cell
=
experiment mean
-
spontaneous mean
x
100
activity total mean
-
spontaneous mean
.
.
In all NK cell activity testing, percent spon-
taneous release was less than 5 %of total release.
Statistical Analysis
Results are expressed as mean
f
SE.
A
2
x
3
repeated measures ANOVA with one between-subjects factor
(EX vs NEX) and one within-subject factor (time of testing)
was used to analyze the data. When Box's M suggested that the
assumptions necessary for the univariate approach were not
tenable, the multivariate approach to repeated measures
ANOVA was used (27). In the latter case, Pillais trace statistic
was used as the test statistic. With regard to comparison among
specific means, only seven comparisons were of interest to us.
These were the contrast of the baseline measures with the 6th
and 15th week measurements within the EX and NEX groups
and the contrast between the EX and NEX groups at each of
the three measurement points. The Dunn-Sidak procedure
(15) was used to test these comparisons. Pearson correlations
were used to determine the association between change in car-
diorespiratory fitness, NK cell activity, and URI symptoma-
tology. Comparison between groups for age, BMI. and URI
were evaluated by simple univariate t-tests.
Results
Subjects
Of the 50 women originally selected for the
study, 36 complied with all phases of testing and intervention.
Dropouts numbered 14, with 7 from each group. Eight sub-
jects (3 EX, 5 NEX) dropped out before baseline testing for
personal reasons (job change, marriage, family problems,
change in desire to participate), and 6 others (4 EX and 2 NEX
subjects) started but failed to complete the study (2 preg-
nancies,
1
surgery,
1
injury,
1
failure to avoid dieting, and
1
job
conflict). Dropouts did not differ from study participants in
BMI, but were significantly younger (29.9
+
1.0 and
34.4+ 1.1 years, respectively, p
=
0.005).
The average age for EX and NEX groups was
36.0
k
1.6 and 32.8
+
1.4 years, respectively (NS). Average
BMI was 28.3
k
0.7 and 27.8+ 0.9 for EX and NEX groups,
respectively, indicating mild obesity (31). Average weight at
baseline was 76.7
f
1.9 and 76.4
f
2.5 kg and, at 15 weeks,
76.7
+
1.7 and 78.0f 2.6 kg for EX and NEX groups, respec-
tively [F(2,67)
=
6.65, p
=
0.0021. No significant change in
body fat percent occurred in either group.
Training Program
All subjects in the EX group complied strictly
with the training regimen, exercising 5 days/week, 45
min/session. All but 10.5% of total exercise sessions for the
entire group were supervised. During the first 6 weeks, sub-
jects averaged 5.0f 0.1 km/session, and during the re-
mainder of the study, 5.1
f
0.1 km/session. The average
measured exercise heart rate was 138
f
1 bpm, which corre-
sponded to 60% heart rate reserve or 62
f
2% VOzmax. Using
data from the personal logs kept by each subject, all NEX sub-
jects were found to have retained their pre-study sedentary
habits.
Graded Esercise Testing
The results of the graded exercise tests are sum-
marized in Table 1. In comparison to NEX, EX subjects ex-
perienced significant improvements in submaximal car-
diorespiratory performance ability as demonstrated by the re-
duction in heart rate, ventilation, and oxygen uptake during
stage 2 of the Bruce's treadmill test. Most of the improvcment
in submaximal fitness occured during the initial
6
weeks.