Download Impairment in Cognitive and Exercise Performance during

0021-972X/01/$03.00/0
The Journal of Clinical Endocrinology & Metabolism
Copyright © 2001 by The Endocrine Society
Vol. 86, No. 1
Printed in U.S.A.
Impairment in Cognitive and Exercise Performance
during Prolonged Antarctic Residence: Effect of
Thyroxine Supplementation in the Polar
Triiodothyronine Syndrome*
H. LESTER REED, KATHLEEN R. REEDY, LAWRENCE A. PALINKAS,
NHAN VAN DO, NANCY S. FINNEY, H. SAMUEL CASE, HOMER J. LEMAR,
JAMES WRIGHT, AND JOHN THOMAS
Endocrine Service (H.L.R., N.V.D., N.S.F., H.J.L.), Departments of Medicine and Clinical Investigation
(J.W.), Madigan Army Medical Center, Tacoma, Washington 98431; U.S. Food and Drug
Administration (K.R.R.), Rockville, Maryland 20857; Department of Family and Preventive Medicine
(L.A.P.), University of California at San Diego, La Jolla, California 92093; Department of Exercise
Science and Physical Education (H.S.C.), Western Maryland College, Westminster, Maryland 21157;
and Office of Naval Research (J.T.), Arlington, Virginia 22217
placebo group, in contrast, showed a reduced M-t-S score (11.2 ⫾ 1.3%;
P ⬍ 0.0003) and serum free T4 (5.9 ⫾ 2.4%; P ⬍ 0.02), compared with
baseline. The change in M-t-S score was correlated with the change
in free T4 (P ⬍ 0.0003) during both periods, and increases in serum
TSH preceded worsening scores in depression, tension, anger, lack of
vigor, and total mood disturbance (P ⬍ 0.001) during period 2. Additionally, the submaximal work rate for a fixed O2 use decreased
22.5 ⫾ 4.9% in period 1 and remained below baseline in period 2
(25.2 ⫾ 2.3%; P ⬍ 0.005) for both groups. After 4 months of AR, the
L-thyroxine supplement was associated with improved cognition,
which seems related to circulating T4. Submaximal exercise performance decrements, observed during AR, were not changed with this
L-thyroxine dose. (J Clin Endocrinol Metab 86: 110 –116, 2001)
ABSTRACT
Humans who work in Antarctica display deficits in cognition, disturbances in mood, increased energy requirements, a decline of thyroid hormone products, and an increase of serum TSH. We compared
measurements in 12 subjects, before deployment (baseline), with 11
monthly studies during Antarctic residence (AR). After 4 months of
AR (period 1), half of the subjects (T4 group) received L-thyroxine [64
nmol䡠day⫺1 (0.05 mg䡠day⫺1)]; and the other half, a placebo (placebo
group) for the next 7 months of AR (period 2). During period 1, there
was a 12.3 ⫾ 5.1% (P ⬍ 0.03) decline on the matching-to-sample
(M-t-S) cognitive task and an increase in depressive symptoms, compared with baseline. During the intervention in period 2, M-t-S scores
for the T4-treated group returned to baseline values; whereas the
H
UMANS WHO LIVE at high latitudes are exposed to
environmental extremes of photoperiod length, low
temperatures, low relative humidity, seasonal changes in
activity, increased electromagnetic radiation, and both social
and geographic isolation. More than 280 million people live
in circumpolar regions; however, very little is known about
the cumulative effects of this environment on human physiology. Antarctica provides an ideal natural laboratory for
the study of human responses to extended severe winter
conditions (1). Some of these responses may be applicable to
residents in more temperate climates.
Military and civilian members of the United States Antarctic Program, who live in a residence in Antarctica (AR)
above the 70o S latitude, for extended periods of time, experience deficits in cognition and alterations in mood (2, 3).
In 1989, the incidence of self-reported depression (62.1%),
irritability (47.6%), and concentration or memory deficit
(51.5%) was significant (P ⬍ 0.001) (4).
Antarctic residents have also developed a constellation of
physiological and hormonal changes called the polar T3 syndrome (5–7). This syndrome is characterized by an elevation
in TRH-stimulated TSH (6) and nonstimulated TSH (7, 8) in
the absence of pituitary resistance to thyroid hormones (6).
Additionally, a small decline in serum free T3 (FT3) and free
T4 (FT4), a doubling in both T3 distribution volume and
plasma appearance and clearance rate, as well as a small
decrease in T4 distribution volume further help define this
condition (8, 9). Physiologically, and presumably as part of
hypothermic cold adaptation (10), this group of residents has
a fall in body temperature (11) and an apparent 40% increase
in daily energy requirements (6, 9).
The circulating thyroid hormone values observed with AR
suggest, at least in part, a cerebral and pituitary hypothyroxinemia, which seems associated with cognitive and mood
symptoms consistent with this hypothesis. Hypothyroidism
Received February 1, 2000. Revision received May 31, 2000. Rerevision received September 11, 2000. Accepted September 19, 2000.
Address all correspondence and requests for reprints to: H. Lester
Reed, Division of Medicine, Middlemore Hospital, Private Bag 93311,
Otahuhu, Auckland 6, New Zealand. E-mail: [email protected].
* Presented in part at the 80th Annual Meeting of The Endocrine
Society, New Orleans, Louisiana, June 24 –27, 1998, and the 71st Annual
Meeting of The American Thyroid Association, Portland, Oregon, September 16 –20, 1998. This work is supported in part by National Science
Foundation Grant OPP-9418466 and Madigan Army Medical Center
Department of Clinical Investigation Grant 96083. The opinions expressed herein are those of the authors and are not to be construed as
reflecting the views of the Department of the Army, the Department of
Defense, National Science Foundation, or U.S. Food and Drug
Administration.
110
COGNITION AND EXERCISE IN ANTARCTICA
is known to be associated with cognitive deficits, mood alterations (12, 13), and changes in visual evoked potentials
(14). Cognition and mood are also affected in subclinical
hypothyroidism, where serum TSH is minimally elevated
and the peripheral products are normal (15). A recent report
suggests that memory may be affected in some individuals,
even when the serum TSH is in the upper half of the normal
range (16). Administration of T4 improves mood and cognitive performance in individuals with subclinical hypothyroidism (15).
We consequently hypothesized that normalizing these circulating thyroid hormone parameters with T4 supplementation may improve cognitive performance and mood state
(15). In this paper, we report the effects of T4 [64 nmol䡠day⫺1
(0.05 mg䡠day⫺1)]) and placebo, during AR, on cognition,
mood, resting and exercise O2 use, and serum thyroid
hormones.
Materials and Methods
Subjects
Twelve (11 male and 1 female) healthy euthyroid subjects, all members of the annual military party that wintered over at McMurdo Sound,
Antarctica, participated in this study. The protocol was approved by the
Madigan Army Medical Center Institutional Review Board, and all
subjects gave written informed consent. Subjects were similar to one
another with regard to age, body mass index (BMI), body surface area
(BSA), exercise capacity, and resting O2 consumption [resting metabolic
rate (RMR)] at the beginning of the protocol (Table 1). Our original study
group included 14 subjects, 1 of whom had subclinical hypothyroidism
and another of whom was noncompliant with the protocol, causing both
to be excluded from further study. No subject had a history of depressive
or thyroid disease, and all were screened by a military physical and
psychological assessment. Available diet contained a minimum of 1,182
nmol/day (150 ␮g) iodine (7, 9), and no chronic medications were taken.
These subjects were studied in September, 1996, while in Port Hueneme,
CA (34° 09⬘ N; 119° 12⬘ W) before departing for Antarctica (baseline),
then again between 10 and 18 days after arrival at McMurdo Sound,
Antarctica (77° 51⬘ S, 166° 37⬘ E), in October, 1996, and monthly thereafter
through August, 1997. The environmental conditions of temperature
TABLE 1. Subject demographics with metabolic measures
PG
T4G
Total
n
Gender (M/F)
Age (Yr)
BMI (kg䡠m⫺2)
BSA (m2)
VO2max (mL䡠min⫺1䡠kg⫺1)
RMR (mL䡠min⫺1䡠m⫺2)
Tty Body temp (C)
6
(5/1)
32.0 ⫾ 2.7
27.3 ⫾ 2.1
1.99 ⫾ 0.11
40.9 ⫾ 4.4
124 ⫾ 2
36.4 ⫾ 0.2
6
(6/0)
31.2 ⫾ 2.5
26.4 ⫾ 0.7
1.99 ⫾ 0.03
39.8 ⫾ 3.1
127 ⫾ 9
36.5 ⫾ 0.1
12
(11/1)
31.6 ⫾ 1.8
27.1 ⫾ 1.1
1.99 ⫾ 0.05
39.6 ⫾ 2.4
126 ⫾ 4
36.4 ⫾ 0.1
At the end of period-2
BMI (kg䡠m⫺2)
BSA (m2)
VO2max (mL䡠min⫺1䡠kg⫺1)
RMR (mL䡠min⫺1䡠m⫺2)a
Tty Body Temp (C)b
25.4 ⫾ 1.5
1.94 ⫾ 0.12
43.9 ⫾ 2.9
140 ⫾ 12
35.6 ⫾ 0.4
26.3 ⫾ 0.6
1.99 ⫾ 0.03
41.4 ⫾ 1.5
152 ⫾ 8
35.2 ⫾ 0.2
25.7 ⫾ 0.8
1.96 ⫾ 0.06
43.5 ⫾ 1.5
147 ⫾ 7
35.4 ⫾ 0.2
Baseline
The mean and ⫾ SEM obtained in the baseline and the final month
in period-2 for BMI, BSA, V̇O2max, RMR, and body temperature at the
Tty are shown in the table. This presentation allows for consistent
comparison with values that only have two measures over the study
such as V̇O2max. Statistical analyses which used all available data
were carried out according to Materials and Methods.
a
P ⬍ 0.05 over the study without group effect.
b
P ⬍ 0.0001 over the study without group effect.
111
and photoperiod during the study are shown in Fig. 1, with period 1
lasting the first 4 months of AR and ending with near-total sunlight, and
period 2 extending from February to August during the austral autumn
and winter.
During outdoor activity, each subject wore standard polar coldweather clothing, with which the face and hands are commonly exposed.
The minimum outside exposure was approximately 0.5 h䡠day⫺1 (6, 9),
indoor fluorescent lighting of normal intensity was used, and all subjects
maintained routine 8-h䡠day⫺1 sleep cycles. Indoor living compartment
temperature was between 18 and 25 C, although there is a substantial
vertical thermal gradient in Antarctica (17). The laboratory air temperature was between 22.6 and 23.7 C, and relative humidity was between
63.7 and 81.5% during blood sampling and metabolic testing.
Study protocol
After baseline measurements, the subjects were rank-ordered, then
randomly assigned to either a placebo group (PG) or T4-administration
group (T4G). Both groups consumed (daily) a pharmaceutical-grade,
opaque, white, gelatin capsule beginning in September, 1996, and ending after the last measurement in August, 1997. Placebo capsules were
consumed by both groups for the first 4 months of AR (period 1) in a
single-blind fashion (Fig. 1). After 4 months (period 2), in a double-blind
protocol, the placebo capsules were replaced with capsules containing
64 nmol (0.05 mg) l-thyroxine (Levoxyl; Daniels Pharmaceuticals, Inc.,
St. Petersburg, FL) for the T4G, while the PG continued with placebo.
Therefore, the mean dose of 32.3 ⫾ 0.5 nmol䡠m⫺2䡠day⫺1 (25.1 ⫾ 0.4
␮g䡠m⫺2䡠day⫺1) represented an attempt to normalize the 65-nmol䡠m⫺2 T4
deficit previously suggested to be present during period 2 (9). This
conversion occurred in the month of February and preceded the March
testing period by a mean of 26.5 days. Pill counts were maintained
during the monthly capsule distribution, and a representative rate of
medication compliance of 92.5% was noted in March.
Biochemical measurements
Blood for thyroid hormone analysis was collected just before the
metabolic measures at baseline and then monthly after arrival in Antarctica (Fig. 1). Sampling was completed between 0530 –1100 h, just
before metabolic testing and after a 12-h fast. Blood was allowed to clot
at room temperature, then was separated and stored at ⫺70 C. From
Antarctica, all samples were transported in October, 1997, at ⫺70 C, to
Tacoma, WA, where they remained at this temperature until they were
coassayed, in duplicate, using a batch method for subject and assay. FT4
and FT3 were analyzed using commercially available kits (AxSYM; Abbott Laboratories, Abbot Park, IL) with an intraassay coefficient of variance (CV) of 6% and 7%, respectively, and an assay detection limit of 5.15
pmol/L (0.4 ng/dL) and 1.69 pmol/L (1.1 pg/mL), respectively. TSH
was measured by a commercial kit (Diagnostic Systems Laboratories,
Inc., Webster, TX) with an intraassay CV of 4% and effective lower
detection limit of 0.03 mU/L. The reference ranges and conversion to SI
units used for these assays are: FT4, 2.19 –23.81 pmol/L (ng/dL ⫻
12.87 ⫽ pmol/L); FT3, 2.22–5.35 pmol/L (pg/mL ⫻ 1.536 ⫽ pmol/L);
and TSH, 0.5–5.1 mU/L.
Cognitive testing
After orientation and training, subjects achieved a mean proficiency
of 74% accuracy with the matching-to-sample task (M-t-S) (18 –21). In
this task, a grid pattern of red and green squares is presented on a
computer screen and, after a delay, the grid is replaced with 2 similar
patterns, 1 of which is the original. This test of attention, spatial and
short-term memory, and pattern recognition uses 20 trials and has been
reported to show 10% differences in similar paired groups (20). Subjects
used individual identical computers, located in quiet areas, to carry out
this monthly testing. Our within-subject CV for this test was 10.5%
during repeated testing at baseline.
Each month, subjects also completed the profile of mood states and
the Center for Epidemiological Studies depression (CES-D) scale (22, 23).
The profile of mood states is a 65-item, self-report mood questionnaire
that obtains data on 6 factors: tension-anxiety, depression-dejection,
anger-hostility, vigor-activity, fatigue-inertia, and confusion-bewilderment. A total mood disturbance score was calculated by summing the
112
REED ET AL.
JCE & M • 2001
Vol. 86 • No. 1
FIG. 1. The protocol design indicated
for the two groups shows that both
groups consumed a placebo (hatched
bar) for the first 4 months (period 1) of
the study. During period 2, the T4G
(black bar) received 64 nmol䡠day⫺1 (0.05
mg䡠day⫺1) T4, and the PG (hatched bar)
continued the placebo . The arrow indicates arrival in McMurdo, and the
mean temperature (solid line) and photoperiod (dashed line) are indicated.
scores of the individual factors after weighting the vigor-activity score
negatively, thereby providing a global estimate of affective state. This
test of mood has been used in previous polar studies in Antarctica (23)
and to test changes in mood with therapy of hypothyroidism (12). The
CES-D scale (22) was used to measure depressive symptoms, where
respondents described their mood, over the preceding week, by rating
each of the 20 items, on a scale from 0 –3.
Metabolic measurements and exercise protocols
While subjects were dressed in shorts, socks, and an undershirt, body
weight and a tympanic temperature (Tty) (Model HH-300, Exergen
Corp., Watertown, MA) were measured. Our one female subject also had
urine measured for pregnancy determination. The subjects then rested
for 20 min, in the supine position, covered with a light blanket. Subsequently, standard O2 utilization measures were obtained for 10 min
(SensorMedics Metabolic Cart, Model 2900z; SensorMedics Corp., Yorba
Linda, CA) (24, 25). The two identical metabolic carts were calibrated
using barometric pressure and temperature corrections with standard
concentrations of O2 and CO2 and were reassessed before each new
subject. The O2 and CO2 analyzers have an accuracy of 0.03% and 0.05%,
respectively, and the ventilation volume is accurate to ⫾3%.
After measurement of the resting O2 uptake, the subjects began the
submaximal testing with the cycle ergometer (Monark model 818E;
Monark, Vansbro, Sweden). Each subject pedaled at 50 rpm for 3 min
at stair-step work rates of 0, 25, 50, and 75 W. Data were collected from
the steady-state final 2 min of each work rate, hereafter referred to in
watts (24). During all submaximal tests, peak values of O2 consumption
(V̇O2) were below 50% maximum O2 use (V̇O2max) (26). Submaximal
studies were carried out at baseline and then twice at each monthly
period, always following the RMR measure, and separated by a 20-min
rest period. V̇O2max was measured at baseline and at the end of period
2 using standard criteria (27).
Calculations from V̇O2max and exercise performance. Each submaximal test
was used to fit a linear regression model of V̇O2 vs. work rate in watts
(P ⬍ 0.01) (26); and from that an intercept, the regressed RMR (RMRr),
and slope (⌬V̇O2/⌬Watt) were derived. A standardized work rate (WRs)
was calculated using a midrange V̇O2 of 400 mL䡠min⫺1䡠m⫺2 for each
individual (26). Because of the submaximal nature of the work rate and
an unchanged respiratory exchange ratio between the resting RMR
(0.814 ⫾ 0.027) and the completion of the submaximal period (0.844 ⫾
0.022), we report only the V̇O2, as is customary (24).
Analysis
The data were subjected to an ANOVA for within-period, betweenperiod, and group differences. If differences within a period existed,
then individual step-wise repeated model fitting or iterative regression
analyses were carried out to determine the structure of the change
(Systat; Systat Inc., Evanston, IL), and the parameters were compared in
a paired fashion when appropriate (28). Relationships between serum
measures and cognitive function were carried out with linear regression,
ANOVA, and analysis for covariance. Unless otherwise stated, significance was determined at the P ⬍ 0.05 level, and ⫾ sem are listed.
Results
Physiological parameters of subjects (Table 1)
Tty declined from baseline by 0.87 ⫾ 0.10 C for the pooled
value representing all of period 1 and by 1.21 ⫾ 0.10 C for the
pooled value representing all of period 2 (P ⬍ 0.0001). No
group difference was noted. V̇O2max, the time to reach
V̇O2max, the maximum work achieved with V̇O2max, resting
heart rate, and body weight, BMI, and BSA were not different
between groups, nor was there a significant change in these
measures over the study.
Thyroid hormone measurements (Table 2 and Fig. 2)
Serum FT4 in the PG declined from baseline (13.6 ⫾ 0.3
pmol/L) over the entire study, by a mean of 4.72 ⫾ 1.89%
(P ⬍ 0.017) and, specifically, by 5.93 ⫾ 2.37% as a pooled
measure for all of period 2. FT4 in the T4G was not different
from the PG in baseline or in period 1. However, in period
2, the pooled monthly value increased 8.10 ⫾ 2.63% over
baseline (P ⬍ 0.03), and this change was different from the
PG (P ⬍ 0.006) represented by the final measurement in
period 2, of 14.3 ⫾ 0.6 pmol/L, compared with 12.9 ⫾ 0.3
pmol/L for the PG.
Serum FT3 declined slightly from baseline by a mean
3.67 ⫾ 1.30% in both groups (P ⬍ 0.04) over the entire study.
COGNITION AND EXERCISE IN ANTARCTICA
113
TABLE 2. Thyroid hormone and mood scores with group comparisons
Baseline
Value
AR (months)
At the end of period-1
T4G
Total
PG
T4G Placebo
Given
Total
PG
T4G T4 Given
Total
0
0
0
4
4
4
11
11
11
1.92 ⫾ 0.34
12.6 ⫾ 0.3
3.82 ⫾ 0.12
2.43 ⫾ 0.29
12.9 ⫾ 0.3
3.82 ⫾ 0.17
1.45 ⫾ 0.35
14.3 ⫾ 0.6
3.61 ⫾ 0.23
1.94 ⫾ 0.26
13.5 ⫾ 0.4
3.72 ⫾ 0.14
Thyroid values
TSH (mU/L)a,b 2.11 ⫾ 0.43 2.01 ⫾ 0.43 2.06 ⫾ 0.28
FT4 (pmol/L)a,b 13.6 ⫾ 0.3 13.1 ⫾ 0.5 13.4 ⫾ 0.3
FT3 (pmol/L)c
3.90 ⫾ 0.17 3.92 ⫾ 0.15 3.90 ⫾ 0.11
Mood measure
Fatigued
Confusiond
CESDepression
Scoree
At the end of period-2
PG
1.93 ⫾ 0.27 1.91 ⫾ 0.67
12.4 ⫾ 0.4 13.0 ⫾ 0.4
3.92 ⫾ 0.23 3.73 ⫾ 0.17
4.33 ⫾ 1.23 5.67 ⫾ 1.76 5.00 ⫾ 1.04 6.00 ⫾ 1.13 5.67 ⫾ 1.31 5.83 ⫾ 0.82 7.00 ⫾ 1.75 4.83 ⫾ 2.04 5.92 ⫾ 1.32
4.50 ⫾ 0.56 4.17 ⫾ 1.17 4.33 ⫾ 0.62 4.17 ⫾ 0.40 4.83 ⫾ 1.17 4.50 ⫾ 0.60 4.50 ⫾ 0.62 2.67 ⫾ 1.05 3.58 ⫾ 0.65
5.67 ⫾ 1.28 5.83 ⫾ 0.79 5.75 ⫾ 0.72 12.00 ⫾ 2.41 8.00 ⫾ 1.64 10.00 ⫾ 1.52 10.00 ⫾ 2.25 10.50 ⫾ 2.14 10.25 ⫾ 1.48
The mean and ⫾ SEM obtained in the baseline and the final month in period-1 and period-2 during AR for serum TSH, FT4, FT3 and mood
measures of fatigue, confusion and depression by the CESD-score are shown. As in Table 1, representative data are shown for the final month
of the time period indicated and analyses carried out using all 12 monthly measures.
a
P ⬍ 0.006 group effect period-2.
b
P ⬍ 0.02 difference over study with respect to time for PG.
c
P ⬍ 0.05 time effect without group effect.
d
P ⬍ 0.05 group effect in period-2.
e
P ⬍ 0.05 for time effect in period-1 and P ⫽ 0.07 for time effect over study.
FIG. 2. The six PG subjects were used
to determine a 12-month change in serum TSH over the study (P ⬍ 0.001).
The mean (solid line) and ⫾SEM (䡠 䡠 䡠) of
the individually fitted sine function parameters are shown (P ⬍ 0.01).
The pooled values for period 2 showed no group difference
and are represented by the final monthly measurement in
period 2, of 3.82 ⫾ 0.17 pmol/L for the PG and 3.61 ⫾ 0.23
pmol/L for the T4G.
Serum TSH in the PG showed an effect of time (P ⬍ 0.001),
over the 12-month study, with a sine distribution, where the
mean amplitude and period are shown in Fig. 2 [1.14䡠sine
(0.812䡠months in Antarctica), (P ⬍ 0.01)]. The model predicted a period of 7.74 months, with peak values (46.1 ⫾
6.8%) in November and (46.0 ⫾ 6.7%) in July above the
mesor. The minimum was predicted during March as 46.1 ⫾
6.8% below the mesor (Fig. 3). Serum TSH was not different
between the PG and the T4G in baseline or over period 1.
However, the pooled period 2 mean serum TSH in the T4G
was reduced to1.64 ⫾ 0.33 mU/L or 24.1 ⫾ 0.2% below the
mesor for the PG (P ⬍ 0.00001) and, although somewhat
lower, was not different than baseline. The same PG seasonal
pattern for TSH was not observed for the T4G.
Cognitive and mood assessment (Table 2 and Fig. 3)
Cognitive assessment. M-t-S scores, which were similar between groups at baseline (PG, 76.8 ⫾ 2.4%; T4G, 71.0 ⫾ 2.2%),
with a mean score for the whole study group of 73.9 ⫾ 2.2%
correct, remained unchanged from baseline for the first 3
months of period 1 (PG, 73.6 ⫾ 3.1%; T4G, 72.4 ⫾ 2.8%). By
the final month of period 1 (January), the M-t-S score decreased to 64.4 ⫾ 3.3% (P ⬍ 0.03) (PG, 60.7 ⫾ 2.3%; T4G, 68.2 ⫾
114
JCE & M • 2001
Vol. 86 • No. 1
REED ET AL.
FIG. 3. The mean changes, compared
with baseline, for the paired M-t-S raw
test score percent-correct (⫾SEM) are
shown with respect to duration of Antarctic Residence. The whole study
group (n ⫽ 12) is compared in the first
and last month of period 1 (P ⬍ 0.02) (F;
solid line). Changes in performance are
represented midway through period 2
for both the T4G (f; dashed line), who
were administered 64 nmol䡠day⫺1 (0.05
mg䡠day⫺1) T4 (P ⫽ not significant) during all of period 2 and the PG (F; solid
line) (P ⬍ 0.0003).
6.4%) or 12.3 ⫾ 5.1% below baseline measures. In contrast,
during period 2, the T4G increased their score to 73.6 ⫾ 5.0%
or ⫹4.0 ⫾ 6.7% (not significant) above baseline, while the PG
remained 11.2 ⫾ 1.3% below baseline (P ⬍ 0.0003) with a
mean score of 68.3 ⫾ 3.0%.
Correlation of cognitive and thyroid hormone assessment. The
percent change from baseline in individual subjects’ M-t-S
scores was related to the % change in serum FT4 from baseline, by both the end of period 1 (January) (P ⬍ 0.04) and
throughout all of period 2 (P ⬍ 0.01). Over the entire study,
between January (period 1) and through period 2, for each
1.0% change in FT4, there was a 1.13 ⫾ 0.26% change in the
M-t-S score (P ⬍ 0.0003). This regression has an intercept of
⫺5.5 ⫾ 2.6% change in M-t-S score, with no change in FT4.
Mood assessment. Self-reported symptoms of depression, as
measured by the CES-D scale, increased from baseline during period 1 in both groups (P ⬍ 0.05) (Table 2). Depressive
symptoms remained higher, compared with baseline,
throughout period 2 in both groups, but the difference did
not reach significance (P ⫽ 0.07). The T4G reported less
fatigue-inertia (P ⬍ 0.01) and confusion-bewilderment (P ⬍
0.05), during period 2, than the PG.
Correlation of mood and thyroid hormone assessment. In both
groups, increases in serum TSH, during period 2, preceded
high scores for depression-dejection, tension-anxiety, angerhostility, lack of vigor-activity, and total mood disturbance
(P ⬍ 0.001). Declines in serum FT3, during period 2, preceded
high scores for worsening fatigue-inertia and confusionbewilderment in both groups (P ⬍ 0.05).
Metabolic and exercise assessment (Table 1)
The RMR increased from baseline in both groups by 11.0 ⫾
3.6% for the pooled value of all of period 1 (P ⬍ 0.05) and by
19.3 ⫾ 4.7% for the pooled value in period 2 (P ⬍ 0.005). The
individual group values, pooled for all of period 2, increased
over baseline by 19.8 ⫾ 2.8% (PG) and 18.9 ⫾ 9.0% (T4G), but
they did not differ from one another. Representative values
for the final month of each period are listed in Table 1. There
was no within-period change.
The WRs decreased from baseline (18.2 ⫾ 0.9 W) over the
study (P ⬍ 0.0001), to a mean for all of period 1 of 13.3 ⫾ 0.4
W or a 22.5 ⫾ 4.9% decline. For all of period 2, the mean of
13.6 ⫾ 0.4 W represented a 25.2 ⫾ 2.3% decrease from baseline. No difference was found between the groups or within
a period.
The RMRr increased over the study (P ⬍ 0.02). From baseline (231 ⫾ 8 mL䡠min⫺1䡠m⫺2), it increased for all subjects by
16.5 ⫾ 3.2% as a pooled value for all of period 1 (P ⬍ 0.01),
and the pooled value for period 2 remained elevated above
baseline by 15.8 ⫾ 3.1% (P ⬍ 0.04). The pooled value in period
2 for each group (PG, 264 ⫾ 6 mL䡠min⫺1䡠m⫺2; T4G, 262 ⫾ 7
mL䡠min⫺1䡠m⫺2) remained above baseline, without a difference between groups. There was no within-period change
detected.
The ⌬V̇O2/⌬Watt increased from baseline (9.17 ⫾ 0.34
mL䡠min⫺1䡠watt⫺1) to 9.91 ⫾ 0.22 mL䡠min⫺1䡠watt⫺1 for a
pooled value for all of period 1, and it remained elevated at
10.14 ⫾ 0.19 mL䡠min⫺1䡠watt⫺1 for the pooled value of period
2 (P ⬍ 0.04). This change represents a 9.2 ⫾ 3.8% increase in
period 1 and a 12.0 ⫾ 4.1% increase in period 2, over baseline.
No group difference over the study and no change within a
period were detected.
Discussion
A T4 supplement of 64 nmol (0.05 mg) per day will improve both cognition and some self-reported mood scores
during extended residence in Antarctica. Reductions in serum FT4 are correlated with poor responses on cognitive
tests, and an elevated serum TSH precedes declines in mood.
COGNITION AND EXERCISE IN ANTARCTICA
This supplement does not change the reduced submaximal
exercise performance, body temperature, FT3, or elevated
RMR observed during the same period.
Cognition and mood
A seasonal component for affective illness has been described; and, although the specific mechanisms are unknown, thyroid hormones have been suggested as a contributor (29, 30). Human cognitive performance declines
under conditions of experimental cold air exposure and during military operations in cold geographical locations (18 –
21). A 29% decline in the M-t-S score is noted while humans
are exposed to cold air at 4 C (20). Tyrosine administered before
the cold exposure reverses this cognitive defect which has been
termed: cold induced amnesia (20). Living in heated housing
conditions does not correct thyroid hormone changes observed
during residence in cold environments (5, 8, 21).
Cognition and mood with T4 intervention. The relationship between the change in serum FT4 and the change in cognition
is present both in period 1 and period 2, supporting the early
association between these two variables within 4 months of
AR. Changes in FT4 can account for approximately 56% of the
decline in cognition in Antarctica, whereas other hormonal,
environmental, and psychological features are also possible
contributors. A fall in body temperature, as we report, is a
hallmark of human hypothermic cold adaptation (10, 31).
The effect of a 1.0-C reduction in Tty on mood and cognition
during AR is unknown, although it may depress both. Based
on the lack of a difference in RMR, RMRr, WRs, or pulse rate
between groups, and the absence of a statistical decline in
TSH below baseline for the T4G, it is unlikely that the treatment group was overreplaced with T4 in period 2. Additionally, it would be unusual for overreplacement to improve
cognition in this protocol (32).
The specific role of T3 in the genesis of hypothyroidassociated psychological decrements was recently reported
by Bunevièius, et al. (12). Our study does not address this
issue directly, except that FT3, which is slightly decreased in
both groups, over the study, is associated with more fatigue,
confusion, and depression (CESD) independent of T4 administration. Although not significant, the serum FT3 tended
to be 5.5% lower in the T4-treated group, compared with
placebo, by the last month of the study. The reduced FT3,
which may be statistically significant with an increased study
population, suggests a thyroidal contribution to FT3. The doubling of T3 clearance, with cold exposure that is independent of
TSH and T4 (33, 34), may help explain why compensation could
be marginal. Because only 56% of the changes in cognition are
related to changes in FT4, it is possible that declines in central
nervous system (CNS) T3 may contribute to some of the remaining cognitive decrement during AR (12).
The majority of CNS T3 is generated locally by type-II
deiodinase (5⬘DI-II). A tissue-specific increase in 5⬘DI-II activity during cold exposure, as described for rodent brown
adipose tissue, could occur in human brain during AR and
mild reductions in body temperature. Therefore, the serum
TSH would be maintained at suboptimal levels in the setting
of small decreases in serum FT4, as we observe. In this cold
exposure model, we speculate that selective brain tissues
115
with previously low local T3 contributions by 5⬘DI-II or
Type-I deiodinase, such as the hypothalamus, may become
increasingly dependent on circulating T4. With T4 administration, a specific CNS carrier protein, transthyretin (TTR),
which carries T4 preferentially to T3, could ensure a homogeneous distribution of T4 to these CNS sites (35). Binding to
TTR with its low-affinity sites would be augmented with
increased serum FT4, and disassociation from TTR to brain
tissue could be facilitated because of the relative reduction of
T4 in the CNS.
Serum TSH
A semiannual pattern of serum TSH is noted in Belgium,
where differences of 29% occur between a bimodal peak in
December and July and a trough in May (36). Although our
findings agree with this report, the photoperiod in Antarctica
is opposite; therefore, either the outdoor temperature exposure or the rate of change of photoperiod may play a significant role in this observation (30). Sawhney, et al. (37)
report peaks of serum TSH between 3– 4 months and again
after 11–12 months of AR. Our subjects arrived in October
and displayed the peaks after 1–3 and 9 –11 months of AR.
This seasonal TSH pattern, which is common to both hemispheres, suggests the possibility of reductions during the
rapid change in photoperiod near an equinox in March and
September (Figs. 1 and 2) or stimulation with 1–3 months of
winter weather conditions. Factors such as body temperature, photoperiod, circulating FT4, dietary iodine, decreased
androgen status (38), or cytokine alterations (38) may facilitate chronic TSH stimulation or a phase shift in the circadian
pattern while in Antarctica. These possibilities could contribute to the augmented peaks of our seasonal curve, compared with those observed in Belgium (36). It is unlikely that
the changes in thyroid hormones that we report are attributable to depression alone (13). Because this seasonal curve
was not observed with the T4G, who had a 24% reduction in
TSH and an 11% increase in FT4 in period 2, compared with
the PG, we would suggest that reduced circulating FT4 could
contribute significantly to its development.
Metabolic measures
Increased energy requirements associated with AR and
other polar sites (37–39) have been inferred from dietary
records (5, 9, 39). Without a change in resting heart rate and
V̇O2max, we would favor (as suggested by some, for hyperthyroidism) a skeletal muscle or peripheral vascular tone
etiology to account for increased O2 use with AR (27, 40).
The direction and magnitude of the metabolic changes we
see are in agreement with observations during hyperthyroidism (24, 25, 27). However, because an 8 –15% increase in
resting energy expenditure observed during thyroid overreplacement should be detectable in our study, it is unlikely
that our T4 treatment group received excessive replacement
when measured by either metabolic parameters or serum
TSH (27, 41). Our study is limited by a small subject population, which is typical for this unique environment, and
small group differences may have been obscured by the large
between-period increases of 11–19% for metabolic measures
with AR.
116
REED ET AL.
The increased T3 in skeletal muscles or other metabolically
active tissues (9) may be dependent on a tissue-specific thyroid
receptor or uptake increase associated with cold sensitive T3
tissue binding (8) and increased T3 clearance. Thus, skeletal
muscle could extract T3 from the serum preferentially and show
little metabolic effect on T4 therapy, as long as the serum FT3
concentrations remain similar between the treatment and placebo groups. Tissue-specific uptake (42), action (43, 44), and
receptor isoform distribution (45) are well known.
We conclude that T4 supplementation can improve declines
in cognition and mood, but it does not normalize exercise performance, body temperature, or serum FT3 observed during
AR. Both the metabolic and cognitive features of this syndrome
may well exist at other high-latitude extremes where screening
for mood and thyroid disorders would be prudent. Further
study is needed to determine the relevance of this work to lower
latitude, temperate climate winters.
Acknowledgments
This study could not have been completed without the dedicated
support of the subjects from the Naval Support Force Antarctica, Dr.
Mark Staudacher, and the helpful suggestions of Dr. K. M. M. Shakir
during protocol development and manuscript review. The study group
appreciates review of this manuscript by Drs. T. Dillard and F. Flynn and
the editorial assistance of Ms. Christine Reed. We are also grateful for
the support from the Antarctic Support Associates; in particular, Ms.
Roberta Score, Mr. Russel Bixby, and Mr. Robert Robbins.
References
1. Weller G, Bently CR, Elliot DH, Lanzerotti LJ, Webber P. 1987 Laboratory
Antarctica: research contributions to global problems. Science. 238:1361–1368.
2. Strange RE, Klein WJ. 1973 Emotional and social adjustment of recent U.S.
winter-over parties in isolated Antarctic stations. In: Edholm OG, Gunderson
EKE, eds. Polar human biology: proceedings of the SCAR/IUPS/IUBS symposium on human biology and medicine in the Antarctic. London: Heineman
Medical; 410-416.
3. Palinkas LA, Cravalho M, Browner D. 1995 Seasonal variation of depressive
symptoms in Antarctica. Acta Psychiatr Scand. 91:423– 429.
4. Palinkas LA. 1992 Going to extremes: the cultural context of stress, illness and
coping in Antarctica. Soc Sci Med. 35:651– 664.
5. Do NV, LeMar H, Reed HL. 1996 Thyroid hormone responses to environmental cold exposure and seasonal change: a proposed model. Endocrinol
Metab. 3:7–16.
6. Reed HL, Ferreiro JA, Shakir KM, Burman KD, O’Brian JT. 1988 Pituitary and
peripheral hormone responses to T3 administration during Antarctic residence. Am J Physiol. 254:E733–E739.
7. Harford RR, Reed HL, Morris MT, Sapien IE, Warden R, D’Alesandro MM.
1993 Relationship between changes in serum thyrotropin and total and lipoprotein cholesterol with prolonged Antarctic residence. Metabolism.
42:1159 –1163.
8. Reed HL. 2000 Environmental influences upon thyroid hormone regulation.
In: Braverman LE, Utiger RD, eds. Werner and Ingbar’s the thyroid: a fundamental and clinical text, ed 8. Philadelphia: Lippincott-Raven; 257–265.
9. Reed HL, Silverman ED, Shakir KM, Dons R, Burman KD, O’Brian JT. 1990
Changes in serum triiodothyronine (T3) kinetics after prolonged Antarctic
residence: the polar T3 syndrome. J Clin Endocrinol Metab. 70:965–974.
10. Savourey G, Barnavol B, Caravel JP, Feuerstein C, Bittel JH. 1996 Hypothermic general cold adaptation induced by local cold acclimation. Eur
J Physiol. 73:237–244.
11. Reed HL, Brice D, Shakir KM, Burman KD, D’Alesandro MM, O’Brian JT.
1990 Decreased free fraction of thyroid hormones after prolonged Antarctic
residence. J Appl Physiol. 69:1467–1472.
12. Bunevic̆ius R, Kaz̆anavic̆ius G, Z̆alinkevic̆ius R, Prange A. 1999 Effects of
thyroxine as compared with thyroxine plus triiodothyronine in patients with
hypothyroidism. N Engl J Med. 340:424 – 429.
13. Jackson I. 1998 The thyroid axis and depression. Thyroid. 8:951–956.
14. Tamburini G, Tacconi P, Ferrigno P, et al. 1998 Visual evoked potentials in
hypothyroidism: a long-term evaluation. Electromyogr Clin Neurophysiol.
38:201–205.
15. Monzani F, Del Gerra P, Caraccio N, et al. 1993 Subclinical hypothyroidism:
neurobehavioural features and beneficial effect of l-thyroxine treatment. Clin
Investig. 71:367–371.
JCE & M • 2001
Vol. 86 • No. 1
16. Kraus RP, Phoenix E, Edmonds MW, Nicholson IR, Chandarana PC, Tokmakejian S. 1997 Exaggerated TSH responses to TRH in depressed patients
with “normal” baseline TSH. J Clin Psychiatry. 58:266 –270.
17. D’Alesandro MM, Lopez A, Reed HL, Harford R. 1991 Indoor temperature
variations in McMurdo, Antarctica. Antarctic J US. 26:237–238.
18. Thomas J, Schrot J. 1988 Naval Medical Research Institute performance assessment battery documentation. NMRIR.88 –7. Bethesda, MD: Naval Medical
Research Institute.
19. Thomas JR, Ahlers ST, House JF, Schrot J. 1989 Repeated exposure to moderate cold impairs matching-to-sample performance. Aviat Space Environ
Med. 60:1063–1067.
20. Shurtleff D, Thomas JR, Schrot J, Kowalski K, Harford R. 1994 Tyrosine
reverses a cold-induced working memory deficit in humans. Pharmacol Biochem Behav. 47:935–941.
21. Hodgdon JA, Hesslink RL, Hackney AC, Vickers RR, Hilbert RP. 1991
Norwegian military field exercises in the arctic: cognitive and physical performance. Arctic Med Res. 50(Suppl 6):132–136.
22. Radloff LS. 1977 The CES-D scale: a new self-report depression scale for
research in the general population. Appl Psychological Measurement.
1:385– 401.
23. Palinkas LA, Johnson JC, Boster JS, Houseal M. 1998 Longitudinal studies of
behavior and performance during a winter at the south pole. Aviat Space
Environ Med. 69:73–77.
24. Kimura H, Kawagoe Y, Kaneko N, Fessler HE, Hosoda S. 1996 Low efficiency
of oxygen utilization during exercise in hyperthyroidism. Chest. 110:1264 –
1270.
25. Lim VS, Zavala DC, Flanigan MJ, Freeman RM. 1986 Basal oxygen uptake:
a new technique for an old test. J Clin Endocrinol Metab. 62:863– 868.
26. Gaesser GA, Brooks GA. 1975 Muscular efficiency during steady-rate exercise:
effects of speed and work rate. J Appl Physiol. 38:1132–1139.
27. Martin 3rd WH, Spina RJ, Korte E, et al. 1991 Mechanisms of impaired
exercise capacity in short duration experimental hyperthyroidism. J Clin Invest. 88:2047–2053.
28. Feldman HA. 1988 Families of lines: random effects in linear regression analysis. J Appl Physiol. 64:1721–1732.
29. Sher L, Rosenthal NE, Wehr TA. 1999 Free thyroxine and thyroid-stimulating
hormone levels in patients with seasonal affective disorder and matched controls. J Affect Disord. 56:195–199.
30. Wehr TA. 1998 Effect of seasonal changes in day length on human neuroendocrine function. Horm Res. 49:118 –124.
31. Savourey G, Caravel J, Barnavol B, Bittel JH. 1994 Thyroid hormones changes
in a cold air environment after local cold acclimation. J Appl Physiol.
76:1963–1967.
32. Kathmann N, Kuisle U, Bommer M, Naber D, Nuller OA, Engel RR. 1994
Effects of elevated triiodothyronine on cognitive performance and mood in
healthy subjects. Neuropsychobiology. 29:136 –142.
33. Hesslink Jr RL, D’Alesandro MM, Armstrong 3rd DW, Reed HL. 1992 Human cold air habituation is independent of thyroxine and thyrotropin. J Appl
Physiol. 72:2134 –2139.
34. Reed HL, D’Alesandro MM, Kowalski KR, Homer LD. 1992 Multiple cold air
exposures change oral triiodothyronine kinetics in normal men. Am J Physiol.
263:E85–E93.
35. Schreiber G, Southwell BR, Richardson SJ. 1995 Hormone delivery systems
to the brain-transthyretin. Exp Clin Endocrinol. 103:75– 80.
36. Maes M, Mommen K, Hendrickx D, et al. 1997 Components of biological
variation, including seasonality, in blood concentrations of TSH, TT3, FT4,
PRL, cortisol and testosterone in healthy volunteers. Clin Endocrinol (Oxf).
46:587–598.
37. Sawhney RC, Malhotra AS, Nair CS, et al. 1995 Thyroid function during a
prolonged stay in Antarctica. Eur J Appl Physiol. 72:127–133.
38. Lugg DJ. 2000 Antarctic medicine. J Am Med Assoc. 283:2082–2084.
39. Johnson RE, Kark RM. 1947 Environment and food intake in man. Science.
105:378 –379.
40. Olson BR, Klein I, Benner R, Burdett R, Trzepacz P, Levey GS. 1991 Hyperthyroid myopathy and the response to treatment. Thyroid. 1:137–141.
41. Al-Adsani H, Hoffer LJ, Silva JE. 1997 Resting energy expenditure is sensitive
to small dose changes in patients on chronic thyroid hormone replacement.
J Clin Endocrinol Metab. 82:1118 –1125.
42. Everts ME, de Jong M, Lim CF, et al. 1996 Differential regulation of thyroid
hormone transport in liver and pituitary: its possible role in the maintenance
of low T3 production during nonthyroidal illness and fasting in man. Thyroid.
6:359 –368.
43. Sherman SI, Ringel MD, Smith MJ, Kopelen HA, Zoghbi WA, Ladenson PW.
1997 Augmented hepatic and skeletal thyromimetic effects of tiratricol in
comparison with levothyroxine. J Clin Endocrinol Metab. 82:2153–2158.
44. Lin B, Coughlin S, Pilch PF. 1998 Bidirectional regulation of uncoupling
protein-3 and GLUT-4 mRNA in skeletal muscle by cold. Am J Physiol.
275:E386 –E391.
45. Anderson GW, Mariash CN, Oppenheimer JH. 2000 The molecular actions of
thyroid hormone. In: Braverman LE, Utiger RD, eds. Werner and Ingbar’s the
thyroid: a fundamental and clinical text, ed 8. Philadelphia: Lippincott-Raven;
174 –195.