Content uploaded by Richard A Feelders
Author content
All content in this area was uploaded by Richard A Feelders
Content may be subject to copyright.
Thyroid Hormone Concentrations, Disease, Physical
Function, and Mortality in Elderly Men
Annewieke W. van den Beld, Theo J. Visser, Richard A. Feelders, Diederick E. Grobbee, and
Steven W. J. Lamberts
Department of Internal Medicine (A.W.v.d.B., T.J.V., R.A.F., S.W.J.L.), Erasmus Medical Center, 3000 CA Rotterdam, The
Netherlands; and Julius Center for Patient Oriented Research (D.E.G.), University Medical Center Utrecht, 3508 GA
Utrecht, The Netherlands
Context: Physiological changes in thyroid hormone concentrations
might be related to changes in the overall physical function in the
elderly.
Objective: We determined to what extent thyroid hormone concen-
trations are related to physical function and mortality in elderly men.
Design: A longitudinal population study (the Zoetermeer study) was
conducted. Mortality was registered in the subsequent 4 yr.
Participants: Four hundred three independently and ambulatory
living men (aged 73–94 yr) participated.
Main Outcome Measures: The study examined the association be-
tween serum thyroid hormones and parameters of physical function
as well as the association with mortality.
Methods: TSH, free T
4
(FT4) total T
4
,T
3
,rT
3
, and T
4
-binding globulin
were measured. Physical function was estimated by the number of
problems in activities of daily living, a measure of physical perfor-
mance score (PPS), leg extensor strength and grip strength, bone
density, and body composition.
Results: Serum rT
3
increased significantly with age and the presence
of disease. Sixty-three men met the biochemical criteria for the low T
3
syndrome (decreased serum T
3
and increased serum rT
3
). This was
associated with a lower PPS, independent of disease. Furthermore,
higher serum FT4 (within the normal range of healthy adults) and rT
3
(above the normal range of healthy adults) were related with a lower
grip strength and PPS, independent of age and disease. Isolated low
T
3
was associated with a better PPS and a higher lean body mass. Low
FT4 was related to a decreased risk of 4-yr mortality.
Conclusions: In a population of independently living elderly men,
higher FT4 and rT
3
concentrations are associated with a lower phys-
ical function. High serum rT
3
may result from a decreased peripheral
metabolism of thyroid hormones due to the aging process itself and/or
disease and may reflect a catabolic state. Low serum FT4 is associated
with a better 4-yr survival; this may reflect an adaptive mechanism
to prevent excessive catabolism. (J Clin Endocrinol Metab 90:
6403– 6409, 2005)
F
EATURES OF AGING are in part similar to those of
hypothyroidism. In both conditions basal metabolic
rate decreases (1). Several changes in thyroid hormone con-
centrations occur during aging: serum TSH concentrations
decrease in healthy elderly humans, serum total and free T
3
levels demonstrate a clear, age-dependent decline, whereas
serum total and free T
4
(FT4) concentrations remain un-
changed (2). These changes are often associated with a poor
health status [reviewed by Mariotti et al. (2)]. Serum rT
3
,an
inactive metabolite of T
4
, seems to increase with age (3).
Together with the decrease in serum T
3
levels, this may
indicate a decreased peripheral hepatic metabolism of iodo-
thyronine during aging because liver type I deiodinase (D1)
is important for both serum T
3
production and rT
3
clearance.
However, evaluation of normal thyroid function in the el-
derly is complicated by an increased prevalence of nonthy-
roidal illness and by autoimmune subclinical hypothyroid-
ism (4).
Thyroid hormones are known to regulate the metabolic
thermostat by changing the basal metabolic rate. One may
hypothesize, therefore, that physiological changes in thyroid
hormone concentrations might be related to changes in the
overall physical function in the elderly.
We determined in a cross-sectional setting to what extent
thyroid hormone concentrations are related to age as well as
several physical characteristics of aging in independently
living, elderly men. In addition, we determined whether
potential associations between thyroid hormone concentra-
tions and age and physical status are due to the presence of
disease. Finally, we determined the relation between serum
thyroid hormones and 4-yr mortality.
Subjects and Methods
Subjects
A cross-sectional, single-center study was conducted in 403 indepen-
dently living and ambulatory men, aged 73 yr and older. Only men were
investigated to obtain a rather large population. Names and addresses
of all male inhabitants 70 yr and older were obtained from the municipal
register of Zoetermeer, a medium-sized town in the midwestern part of
The Netherlands. A total of 1567 men were invited; 886 men did not
respond to the mailed invitation in which it was mentioned that only
subjects who lived independently and had no severe mobility problems
could participate. After exclusion of subjects who did not live indepen-
dently, subjects who were not physically or mentally able to visit the
First Published Online September 20, 2005
Abbreviations: ADL, Activities of daily living; BMD, bone mineral
density; CI, confidence interval; D1, type I deiodinase; FT4, free T
4
; kp,
kilopond; PPS, physical performance score; TBG, T
4
-binding globulin.
JCEM is published monthly by The Endocrine Society (http://www.
endo-society.org), the foremost professional society serving the en-
docrine community.
0021-972X/05/$15.00/0 The Journal of Clinical Endocrinology & Metabolism 90(12):6403–6409
Printed in U.S.A. Copyright © 2005 by The Endocrine Society
doi: 10.1210/jc.2005-0872
6403
on August 9, 2006 jcem.endojournals.orgDownloaded from
study center independently, and subjects with severe systemic illnesses,
403 men participated (25.7%). Participants signed informed consent. The
study was approved by the Medical Ethics Committee of Erasmus Med-
ical Centre Rotterdam.
The subjects were interviewed and examined by a single physician
using a standard and validated survey. We divided the cohort into four
groups according to the number of complaints or diseases present [0 (n ⫽
38), 1 (n ⫽ 87), 2 (n ⫽ 71), and ⱖ 3(n⫽ 207)]. Present diseases included
mainly hypertension; cerebral, coronary, and peripheral atherosclerosis;
mild congestive heart failure; chronic obstructive pulmonary disease;
diabetes; and arthrosis. However, none of the subjects were treated for
systemic, infectious, inflammatory, or malignant disorders at the time of
the investigation.
Eight subjects who used amiodarone or corticosteroids were excluded
from the analyses. Six subjects used thyroid hormone substitution and
were also excluded from the analyses. A number of participants were
taking medication for hypertension (n ⫽ 96), angina pectoris or a myo-
cardial infarction more than 6 months ago (n ⫽ 85), mild congestive heart
failure (n ⫽ 28), chronic obstructive pulmonary disease (n ⫽ 40), and
diabetes (n ⫽ 28).
Four years after the initial investigation, 75 men (19%) had died (eight
in the first year, 16 in the second year, 21 in the third year, and 30 men
in the fourth year). None of the subjects were lost to follow-up.
Hormone measurements
Blood samples were collected in the morning after an overnight fast.
Serum was separated by centrifugation and stored at ⫺40 C. TSH was
measured using an immunometric technique (Amerlite TSH-30; Ortho-
Clinical Diagnostics, Amersham, UK). FT4 was measured using the
Amerlite MAB FT4 assay (Ortho-Clinical Diagnostics). T
4
,T
3
, and rT
3
were all measured by in-house RIAs (5). T
4
-binding globulin (TBG) was
measured by Dynotest RIA (Brahms, Berlin, Germany). Intra- and in-
tervariability coefficients of all assays were less than 11%.
Subclinical hypo- and hyperthyroidism are defined as FT4 levels
within the normal range (between 11 and 25 pmol/liter) and TSH levels,
respectively, above (TSH ⬎ 4.3 mU/liter) and below (TSH ⬍ 0.4 mU/
liter) the 95% confidence limits as determined for this assay in 447
healthy blood donors aged 19 – 69 yr.
Physical function
Activities of daily living (ADL). Self-reported disability or satisfaction in
performing ADL was assessed by using a self-administered question-
naire modified from the Stanford Health Assessment Questionnaire as
described by Pincus et al. (6). A high score denotes high impairment in
ADL.
Physical performance
Lower extremity function, or physical performance, was assessed as
described by Guralnik et al. (7), including measurements of standing
balance, walking speed, and ability to rise from a chair. A summary
physical performance scale (PPS) was created by summing the category
scores for the walking, chair stand, and balance test. Mean scores of the
three tests as well as the summary performance scale were comparable
with subjects of the same age group investigated by Guralnik et al. (7).
Muscle strength
Isometric grip strength was tested using an adjustable handheld
dynamometer (JAMAR, Horsham, PA) in the nondominant hand (8).
Each test was repeated three times and the average, expressed in kilo-
ponds (kp), was used in the analysis.
Isometric leg extensor strength was measured as described by Hsieh
and Philips (9) and van den Beld et al. (10) using the Hoggan MicroFET
handheld dynamometer. To obtain one measure of leg muscle strength,
maximum leg extensor strength was defined as the maximum strength
for the right or left leg in a position of 120 degrees. Statistical analyses
were based on the physical unit measurement, moments, obtained by
multiplying the maximum strength (newtons) and the distance of the
dynamometer to the knee joint (meters).
Bone mineral density (BMD) and body composition
Total-body BMD was measured using dual-energy x-ray absorpti-
ometry (Lunar, Madison, WI), as were hip BMDs at the femoral neck,
trochanter, and Ward’s triangle. In addition, total and trunk lean body
mass and fat mass were measured (11, 12). Quality assurance including
calibration was performed routinely every morning for dual-energy
x-ray absorptiometry, using the standard provided by the manufacturer.
Body mass index was calculated as the weight in kilograms divided
by the square of the height in meters.
Data analyses
Results are expressed, unless otherwise stated, as mean and sd. Vari-
ables, which were not normally distributed, were logarithmically trans-
formed. Comparisons between groups were made by using ANOVA.
Differences are given with corresponding 95% confidence intervals
(CIs). Relations between variables were assessed using linear regression
stated as linear regression coefficient (beta) and 95% CI. Multiple re-
gression analysis was used to adjust for age and body mass index and
determine the contribution of different independent variables to the
dependent variable. Univariate general linear model was used to de-
termine the significance between groups and adjust for covariates. Un-
less otherwise mentioned, all analyses are done after adjustment for age.
Results
Serum hormones and their relation with age
Mean age of this population was 77.8 yr (range 73–94 yr).
Mean age did not differ between the groups with no to three
or more diseases.
As shown in Table 1, within the population of elderly
men, serum rT
3
concentrations were significantly posi-
tively related with age, whereas TSH and FT4 concentra-
tions were not. There was a tendency in this population for
TBG levels to increase and T
3
levels and T
3
to TBG ratio to
decrease with age. Although T
4
levels were positively
related with age in this population (Table 1), none of the
subjects had T
4
levels above the range of healthy adults
(⬎138 nmol/liter). However, 53 subjects had T
4
levels
below the range of healthy adults (⬍64 nmol/liter). All the
relations described in Table 1 were independent of the
presence of disease.
Subclinical hyper- and hypothyroidism
One subject with newly discovered overt hypothyroidism
and two subjects with newly discovered overt hyperthyroid-
ism were excluded from further analyses.
Six subjects met the biochemical criteria for subclinical
hypothyroidism (TSH ⬎ 4.3 mU/liter and FT4 between 11
and 25 pmol/liter), and 44 subjects met the biochemical
criteria for subclinical hyperthyroidism (TSH ⬍ 0.4 mU/liter
TABLE 1. Relations of the thyroid hormones with age in a
population of elderly men
All subjects

SE P value
TSH (mIU/liter) ⫺0.01 0.01 0.24
FT4 (pmol/liter) 0.05 0.04 0.23
T
4
(nmol/liter) 0.68 0.23 0.003
T
3
(nmol/liter) ⫺0.006 0.003 0.07
rT
3
(nmol/liter) 0.005 0.001 ⬍0.001
TBG (mg/liter) 0.11 0.06 0.06
T
4
to TBG ratio 0.06 0.01 0.71
T
3
to TBG ratio ⫺0.01 0.001 0.07
6404 J Clin Endocrinol Metab, December 2005, 90(12):6403– 6409 van den Beld et al. • Thyroid Hormone Concentrations in Elderly Men
on August 9, 2006 jcem.endojournals.orgDownloaded from
and FT4 between 11 and 25 pmol/liter). Age did not differ
between the euthyroid and subclinically hypo- or hyperthy-
roid groups.
Subjects with subclinical hyperthyroidism had by defini-
tion normal FT4 levels. However, within this normal range,
subjects with TSH levels less than 0.4 mU/liter had slightly
higher FT4 levels than subjects with normal TSH levels
(17.5 ⫾ 0.46 vs. 16.5 ⫾ 0.17 pmol/liter, P ⫽ 0.05). T
3
levels did
not differ between these groups.
It should be stressed that, although we classified subjects
with TSH less than 0.4 mU/liter and normal FT4 levels as
having subclinical hyperthyroidism, this does not necessar-
ily mean that they all have subclinically increased thyroid
function. Serum TSH may also be decreased by high age,
illness, or drugs (see below). If we used the more conserva-
tive cut-off serum TSH level of 0.1 mU/liter, only six patients
met the criteria for subclinical hyperthyroidism.
Serum hormones and their relation with the presence
of disease
Subjects who met the biochemical criteria for subclinical
hyperthyroidism had significantly more diseases, compared
with euthyroid or subclinical hypothyroid subjects (2.34 ⫾
0.16 vs. 2.08 ⫾ 0.06 vs. 1.26 ⫾ 0.53, P ⬍ 0.05)
Summarized values of the hormone measurements as well
as the values in the groups divided according to the number
of diseases present (for etiology, see Subjects and Methods) are
presented in Table 2. Only serum rT
3
concentrations differed
among the four groups, with the highest levels in the group
with three or more diseases.
According to the normal ranges in healthy adults, 137
subjects had normal T
3
and rT
3
levels (Fig. 1, group A); 123
subjects had T
3
levels within the normal range and rT
3
levels
above the normal range (⬎0.32 nmol/liter, group B); 63 sub-
jects had T
3
levels below (⬍1.35 nmol/liter) and rT
3
levels
above the normal range, which is characteristic for the low
T
3
syndrome or nonthyroidal illness (group C); and 66 sub-
jects had T
3
concentrations below and rT
3
concentrations
within the normal range (group D).
Subjects with the biochemical criteria for the low T
3
syn-
drome were significantly older than those without low T
3
and high rT
3
(Fig. 1). Furthermore, these subjects had sig-
nificantly more diseases than the other subjects [2.35 (95% CI
2.09; 2.61) vs. 2.06 (95% CI 1.94; 2.17), P ⫽ 0.04]. After ad-
justment for age, there remained a trend toward significance
(P ⫽ 0.07).
Serum hormones and their relation with physical
characteristics
The 44 subjects with subclinical hyperthyroidism had a sig-
nificantly lower lean body mass than euthyroid and subclini-
cally hypothyroid subjects [50.6 kg (95% CI 49.1; 52.3) vs. 51.6
(95% CI 51.1; 52.1) and 58.4 kg (95% CI 52.4; 64.4), P ⬍ 0.05] and
slightly lower bone density values. This latter relation seemed
be explained through the first relation because after adjustment
for lean body mass, the trend with bone density was no longer
present. No other significant differences in physical character-
istics were observed between these groups.
Among the four groups illustrated in Fig. 1, there was a
significant difference in PPS, independent of age. Subjects
with low T
3
and high rT
3
concentrations had the lowest PPS,
whereas subjects with low T
3
and normal rT
3
had the highest
scores. Furthermore, lean body mass was significantly higher
in this latter group. The other parameters did not signifi-
cantly differ between the groups.
However, considering the values of the physical charac-
teristics in Fig. 1, there seemed to be a trend that subjects with
high rT
3
levels have lower scores of these physical charac-
teristics. Therefore, we repeated the analyses dividing sub-
jects into two groups: one group with elevated rT
3
concen-
trations (⬎0.32 nmol/liter) and one group with normal rT
3
levels. After adjusting for age, it appeared that PPS, muscle
strength (leg extensor strength and isometric grip strength),
and lean body mass were significantly lower in subjects with
high rT
3
concentrations. Because the number of diseases was
higher in this group, we also adjusted for disease. Although
the strength of the relations became slightly less, the direction
of the relations remained similar (Table 3).
Independent of age and disease, increasing serum FT4
concentrations, within the normal range, were related to
lower PPS [beta ⫽⫺0.11 point/(nanomoles per liter) (95% CI
⫺0.18; ⫺0.03), P ⫽ 0.006] and isometric grip strength [beta ⫽
⫺0.24 kp/(nanomoles per liter) (95% CI ⫺ 0.53; ⫺0.11), P ⫽
0.004]. Figure 2 shows the correlation between rT
3
and FT4
concentrations. This relation is independent of T
3
levels and
has a correlation coefficient of 0.62 (P ⬍ 0.001).
Mortality
After adjustment for age, serum FT4 concentrations,
within the normal range, were significantly related with an
increased risk of 4-yr mortality [relative risk 1.27 (95% CI
1.01–1.60)]. Serum TSH and T
3
were not related to mortality
TABLE 2. Descriptive values of the thyroid hormones in a population of elderly men in the total group as well as divided by the number
of diseases
All subjects No diseases One disease Two diseases Three or more diseases
Mean 95% CI Mean 95% CI Mean 95% CI Mean 95% CI Mean 95% CI
TSH (mU/liter) 1.16 1.06–1.27 1.42 0.84–2.00 1.17 0.95–1.38 1.21 0.92–1.49 1.10 0.99–1.21
FT4 (pmol/liter) 16.6 16.3–16.9 16.6 15.6–17.5 15.9 15.3–16.5 17.1 16.2–17.9 16.7 16.3–17.1
T
4
(nmol/liter) 80.6 79.0–82.0 75.6 70.7– 80.6 78.7 75.0–82.4 81.4 77.5– 85.2 82.0 79.8–84.1
T
3
(nmol/liter) 1.43 1.41–1.45 1.41 1.33–1.48 1.47 1.42–1.52 1.43 1.38–1.48 1.42 1.38–1.45
rT
3
(nmol/liter) 0.33 0.32–0.34 0.31 0.29–0.34 0.31 0.29–0.32 0.33 0.31–0.35 0.35
a
0.33–0.36
TBG (mg/liter) 18.0 17.6 –18.4 17.2 15.8–18.5 18.1 17.2–19.0 18.1 17.1–19.0 18.1 17.5–18.7
T
4
to TBG ratio 4.65 4.55–4.76 4.62 4.34–4.91 4.46 4.26–4.67 4.66 4.39–4.92 4.74 4.59–4.90
T
3
to TBG ratio 0.084 0.08– 0.09 0.085 0.079– 0.091 0.083 0.080– 0.087 0.084 0.078– 0.089 0.084 0.077–0.092
a
Differences between groups, P ⬍ 0.01.
van den Beld et al. • Thyroid Hormone Concentrations in Elderly Men J Clin Endocrinol Metab, December 2005, 90(12):6403– 6409 6405
on August 9, 2006 jcem.endojournals.orgDownloaded from
nor was the T
3
to TBG ratio. Subjects with the low T
3
syn-
drome did not have a higher 4-yr mortality risk than subjects
without this syndrome. Also, subjects with subclinically hy-
po- (n ⫽ 6) or hyperthyroidism (n ⫽ 44) did not show a higher
4-yr mortality than euthyroid subjects. The relation between
FT4 concentrations and mortality was independent of the
presence of disease and parameters of physical function.
Discussion
We found in a population of independently living, elderly
men that a substantial number of subjects (one third of the
population) had T
3
levels below the normal range of healthy
adults. Half of these subjects also had elevated rT
3
levels,
which is characteristic for the low T
3
syndrome or nonthy-
roidal illness. This group had a higher age, more diseases,
TABLE 3. Relations between normal and elevated rT
3
concentrations with physical performance in elderly males
Age adjusted Age and disease adjusted
Normal rT
3
High rT
3
Normal rT
3
High rT
3
Physical performance (pts) 8.69 (8.37–9.01) 8.21 (7.87– 8.54)
a
8.65 (8.33–8.97) 8.25 (7.92–8.58)
Activities of daily living (pts) 10.6 (10.0–11.2) 10.8 (10.2–11.4)
Max. LES (Nm) 105.1 (102.3–107.9) 101.3 (98.5–104.2)
b
104.9 (102.1–107.6) 101.6 (98.7–104.5)
Isometric grip strength (kp) 35.1 (34.2–36.0) 33.7 (32.8 –34.7)
a
35.0 (34.1–35.9) 33.8 (32.9 –34.8)
Lean mass (kg) 52.4 (51.7–53.1) 50.9 (50.2–51.7)
c
52.4 (51.7–53.1) 50.9 (50.1–51.7)
Fat mass (kg) 21.4 (20.6–22.2) 20.8 (20.0–21.6)
Neck BMD (g/cm) 0.89 (0.87–0.91) 0.87 (0.85–0.89)
Max. LES, Maximum leg extensor strength; pts, points; Nm, physical unit measure (maximum strength in newtons multiplied by the distance
of the dynamometer of the knee in meters).
a
P ⬍ 0.05.
b
P ⬍ 0.10.
c
P ⬍ 0.001.
FIG. 1. Overview of the values of T
3
and rT
3
within a population of 403 elderly men. The spotted lines indicate the normal values of T
3
and
rT
3
. LES, Maximum leg extensor strength; IGS, isometric grip strength.
6406 J Clin Endocrinol Metab, December 2005, 90(12):6403– 6409 van den Beld et al. • Thyroid Hormone Concentrations in Elderly Men
on August 9, 2006 jcem.endojournals.orgDownloaded from
and a lower physical performance. Subjects with isolated low
T
3
levels (and normal rT
3
concentrations) had the best phys-
ical performance and the highest lean body mass. Further-
more, subjects with high rT
3
concentrations (independent of
the T
3
level) had worse physical performance scores and
lower grip strength. These high rT
3
levels (above the normal
range of healthy adults) were accompanied by high FT4
levels (within the normal range). Low FT4 concentrations
were related to a decreased risk of 4-yr mortality.
In agreement with previous studies, serum rT
3
concentra-
tions increased with age in our population, whereas approx-
imately one third of this population had T
3
levels below the
normal range of healthy adults (2). Serum total T
4
levels
increased with age in our population. This relation could not
be explained through an increase in TBG levels. It has to be
mentioned, however, that T
4
levels were never above the
normal range of healthy adults in our population. In contrast,
T
4
levels were below this normal range in 53 subjects.
These changes in thyroid hormone concentrations may be
explained by a decrease in peripheral (hepatic) thyroid hor-
mone metabolism with aging. First, aging may be accompa-
nied by a decreased activity of D1, which in turn leads to a
decrease in serum T
3
, due to a reduced peripheral conversion
of T
4
to T
3
, and an increase in serum rT
3
levels due to a
reduced rT
3
degradation in the liver (13, 14). In addition, a
reduced selenium intake may contribute to a decreased D1
activity in the elderly because selenium deficiency is known
to reduce the expression of the D1 selenoprotein (15). Second,
the observed increase in rT
3
levels with aging may in part be
explained by a reduced hepatic uptake of rT
3
. However, both
an impaired D1 activity and a decreased hepatic uptake of
thyroid hormones may also be due to disease or a poor
nutritional state rather than aging itself. The extent to which
the changes in thyroid hormone concentrations and their
relations with physical characteristics in this elderly popu-
lation were due to the aging process per se or the presence of
(nonthyroidal) illness was investigated by examining these
relations before and after adjustment for the presence of
disease (see below).
We determined whether changes in serum thyroid hor-
mone concentrations were related to characteristics of the
aging process, like physical functional status. Thyroid hor-
mones are known to play an essential role in many biological
processes in essentially every tissue. This is illustrated by the
clinical symptoms in hypothyroidism and thyrotoxicosis. We
hypothesized, therefore, that changes in peripheral thyroid
hormone metabolism in the elderly might be related to
changes in physical functional status.
As mentioned above, nonthyroidal illness is associated
with an increase in serum rT
3
concentrations (16, 17). Inter-
estingly, in this population a relatively large proportion of
the subjects met the biochemical criteria for the low T
3
syn-
drome, which is decreased serum T
3
and increased serum
rT
3
. It needs to be emphasized that the investigated popu-
lation was relatively healthy and that subjects with systemic
infectious, inflammatory, and malignant disorders were ex-
cluded. Known morbidity included mainly hypertension,
atherosclerotic disease, congestive heart failure, chronic ob-
FIG. 2. Correlation between rT
3
and FT4 concentrations in 403 elderly men. r, Correlation coefficient.
van den Beld et al. • Thyroid Hormone Concentrations in Elderly Men J Clin Endocrinol Metab, December 2005, 90(12):6403– 6409 6407
on August 9, 2006 jcem.endojournals.orgDownloaded from
structive pulmonary disease, diabetes, and arthrosis. Inde-
pendent of the presence of disease, subjects who met the
biochemical criteria for the low T
3
syndrome had a lower
physical function. It appeared, however, that not only sub-
jects with the low T
3
syndrome but all subjects with isolated
high rT
3
(and FT4) concentrations (groups B and C in Fig. 1)
had a lower physical function. To examine the potential
influence of disease on the associations found, the following
analyses were made. First, the increase in rT
3
levels with age
was independent of the presence of disease. Second, the
decrease in physical functional status with age was inde-
pendent of disease (data not shown). Third, when we ad-
justed for the presence of disease, similar linear regression
coefficients were obtained for the relations between rT
3
(and
FT4) and physical characteristics, this despite the observation
that rT
3
levels were slightly higher in subjects with the pres-
ence of several diseases or complaints. This may indicate that
rT
3
levels may reflect an individual’s physical functional
status, partially independent of disease. Increasing rT
3
levels
could then represent a catabolic state, eventually preceding
an overt low T
3
syndrome. In this respect, nutritional status
may also be a determinant of rT
3
concentrations because
caloric deprivation is also accompanied by an increase in rT
3
levels (3). A substantial number of subjects had low T3 levels,
remarkably without accompanying high rT
3
levels. Because
T
3
and rT
3
changes in nonthyroidal illness are usually con-
cordant, it seems unlikely that isolated low T
3
levels are due
to nonthyroidal illness. The higher physical performance
scores and lean body mass in subjects with isolated low T
3
levels support this. Although Mariotti et al. (18) carried out
a different study protocol, they support the finding of an
age-dependent reduction of peripheral thyroid hormone me-
tabolism at least partially independent of associated non-
thyroidal illness.
A small number of subjects met the biochemical criteria for
subclinical hypothyroidism, compared with previous find-
ings in older populations (19). However, it should be realized
that our population consisted entirely of males, and subclin-
ical hypothyroidism is less prevalent in males than females.
The relatively low number of subjects with subclinical hy-
pothyroidism may also be due to selection of relatively
healthy elderly subjects. This would suggest that subclinical
hypothyroidism is associated with significant morbidity and
mortality. Although subclinical hypothyroidism is indeed
associated with increased cardiovascular morbidity (20), in
very old subjects, it is associated with reduced mortality (19).
Lean body mass was significantly lower in subjects with
subclinical hyperthyroidism, compared with subjects with
subclinical hypothyroidism. Due to the small number of sub-
jects with subclinical hypothyroidism, the power of the anal-
yses involving this group is very small.
A relatively large number of subjects (44, 11%) were iden-
tified with subclinical hyperthyroidism using a TSH cut-off
level of 0.4 mU/liter. In a substantial number of these sub-
jects, the low TSH concentrations may be explained by the
age-related decline in serum TSH as well as by nonthyroidal
illness. If the more stringent and generally accepted TSH
cut-off level of 0.1 mU/liter was used, only six subjects (1.5%)
were identified with subclinical hyperthyroidism.
Recently it has been described that in a population-based
study, subclinical hyperthyroidism predicts mortality (21).
We could not confirm these results. However, higher FT4
levels, within the normal range (independent of TSH levels),
were associated with a higher risk of 4-yr mortality. This
relation appeared to be very strong and independent of, for
example, disease bone density and specific medication
known to influence thyroid function. Although no definitive
conclusions can be made, it should be mentioned that we
were not informed about the possible subsequent develop-
ment of overt hyperthyroidism during 4-yr follow-up in the
population studied. Also, no information about autoimmune
thyroid antibodies was available. Our findings are in agree-
ment with a recent study by Gussekloo et al. (19), who were
informed about the development of overt thyroid dysfunc-
tion. They also found that low FT4 levels were associated
with a longer life span in a population of men and women
aged older than 85 yr. Unlike their findings, we did not find
an association between elevated TSH levels and a lower
mortality. This might be due to the low prevalence of ele-
vated TSH levels in our population. Although serum rT
3
concentrations were also inversely related to parameters of
physical ability, they did not predict mortality. Remarkably
the presence of nonthyroidal illness, associated with a num-
ber of diseases, was not predictive of mortality.
In conclusion, in a population of independently living
elderly men, serum rT
3
concentrations increase with age and
the presence of disease. In this relatively healthy population,
a large proportion met the biochemical criteria for the low T
3
syndrome. Higher FT4 and rT
3
concentrations are associated
with a lower physical functional status. Higher serum rT
3
concentrations may result from a decreased peripheral me-
tabolism of thyroid hormones due to the aging process itself
and/or disease and may reflect a catabolic state. The inverse
relations between T
3
and physical performance and lean
body mass and between FT4 and mortality may indicate that
a lower activity of the thyroid hormone axis is beneficial
during the aging process. Possibly it serves as an adaptive
mechanism to prevent excessive catabolism.
Acknowledgments
Received April 21, 2005. Accepted September 8, 2005.
Address all correspondence and requests for reprints to: A. W. van
den Beld, M.D., Ph.D., Department of Internal Medicine, Room Bd230,
Erasmus Medical Center, P.O. Box 2040, 3000 CA Rotterdam, The Neth-
erlands. E-mail: a.vandenbeld@erasmusmc.nl.
References
1. Piers LS, Soares MJ, McCormack LM, O’Dea K 1998 Is there evidence for an
age-related reduction in metabolic rate? J Appl Physiol 85:2196–2204
2. Mariotti S, Franceschi C, Cossarizza A, Pinchera A 1995 The aging thyroid.
Endocr Rev 16:686 –715
3. Magri F, Cravello L, Fioravanti M, Busconi L, Camozzi D, Vignati G, Ferrari
E 2002 Thyroid function in old and very old healthy subjects. J Endocrinol
Invest 25:60 –63
4. Chiovato L, Mariotti S, Pinchera A 1997 Thyroid diseases in the elderly.
Baillieres Clin Endocrinol Metab 11:251–270
5. Eelkman Rooda SJ, Kaptein E, Visser TJ 1989 Serum triiodothyronine sulfate
in man measured by radioimmunoassay. J Clin Endocrinol Metab 69:552–556
6. Pincus T, Summey JA, Soraci Jr SA, Wallston KA, Hummon NP 1983 As-
sessment of patient satisfaction in activities of daily living using a modified
Stanford Health Assessment Questionnaire. Arthritis Rheum 26:1346–1353
7. Guralnik JM, Seeman TE, Tinetti ME, Nevitt MC, Berkman LF 1994 Vali-
dation and use of performance measures of functioning in a non-disabled older
population: MacArthur studies of successful aging. Aging (Milano) 6:410 – 419
6408 J Clin Endocrinol Metab, December 2005, 90(12):6403– 6409 van den Beld et al. • Thyroid Hormone Concentrations in Elderly Men
on August 9, 2006 jcem.endojournals.orgDownloaded from
8. Hamilton A, Balnave R, Adams R 1994 Grip strength testing reliability. J Hand
Ther 7:163–170
9. Hsieh CY, Phillips RB 1990 Reliability of manual muscle testing with a
computerized dynamometer. J Manipulative Physiol Ther 13:72–82
10. van den Beld AW, de Jong FH, Grobbee DE, Pols HA, Lamberts SW 2000
Measures of bioavailable serum testosterone and estradiol and their relation-
ships with muscle strength, bone density, and body composition in elderly
men. J Clin Endocrinol Metab 85:3276 –3282
11. Gotfredsen A, Jensen J, Borg J, Christiansen C 1986 Measurement of lean
body mass and total body fat using dual photon absorptiometry. Metabolism
35:88–93
12. Mazess RB, Barden HS, Bisek JP, Hanson J 1990 Dual-energy x-ray absorp-
tiometry for total-body and regional bone-mineral and soft-tissue composition.
Am J Clin Nutr 51:1106–1112
13. Pearce CJ 1991 The euthyroid sick syndrome. Age Ageing 20:157–159
14. Donda A, Lemarchand-Beraud T 1989 Aging alters the activity of 5⬘-deiodi-
nase in the adenohypophysis, thyroid gland, and liver of the male rat. Endo-
crinology 124:1305–1309
15. Olivieri O, Girelli D, Stanzial AM, Rossi L, Bassi A, Corrocher R 1996
Selenium, zinc, and thyroid hormones in healthy subjects: low T
3
/T
4
ratio in
the elderly is related to impaired selenium status. Biol Trace Elem Res 51:31–41
16. Goichot B, Schlienger JL, Grunenberger F, Pradignac A, Sapin R 1994 Thy-
roid hormone status and nutrient intake in the free-living elderly. Interest of
reverse triiodothyronine assessment. Eur J Endocrinol 130:244–252
17. Visser TJ 1994 Role of sulfation in thyroid hormone metabolism. Chem Biol
Interact 92:293–303
18. Mariotti S, Barbesino G, Caturegli P, Bartalena L, Sansoni P, Fagnoni F, Monti
D, Fagiolo U, Franchesci C, Pinchera A 1993 Complex alteration of thyroid
function in healthy centenarians. J Clin Endocrinol Metab 77:1130 –1134
19. Gussekloo J, van Exel E, de Craen AJ, Meinders AE, Frolich M, Westendorp
RG 2004 Thyroid status, disability and cognitive function, and survival in old
age. JAMA 292:2591–2599
20. Hak AE, Pols HA, Visser TJ, Drexhage HA, Hofman A, Witteman JC 2000
Subclinical hypothyroidism is an independent risk factor for atherosclerosis
and myocardial infarction in elderly women: the Rotterdam Study. Ann Intern
Med 132:270 –278
21. Parle JV, Maisonneuve P, Sheppard MC, Boyle P, Franklyn JA 2001 Predic-
tion of all-cause and cardiovascular mortality in elderly people from one low
serum thyrotropin result: a 10-year cohort study. Lancet 358:861–865
JCEM is published monthly by The Endocrine Society (http://www.endo-society.org), the foremost professional society serving the
endocrine community.
van den Beld et al. • Thyroid Hormone Concentrations in Elderly Men J Clin Endocrinol Metab, December 2005, 90(12):6403– 6409 6409
on August 9, 2006 jcem.endojournals.orgDownloaded from
