Risk for Ischemic Heart Disease and All-Cause Mortality in Subclinical Hypothyroidism

Article (PDF Available)inJournal of Clinical Endocrinology & Metabolism 89(7):3365-70 · August 2004with222 Reads
DOI: 10.1210/jc.2003-031089 · Source: PubMed
Abstract
We investigated possible associations between subclinical hypothyroidism and atherosclerotic diseases (ischemic heart disease and cerebrovascular disease) and mortality. Of 2856 participants (mean age 58.5 yr) in a thyroid disease screening between 1984 and 1987, 257 subjects with subclinical hypothyroidism (TSH > 5.0 mU/liter) and 2293 control subjects (TSH range 0.6-5.0 mU/liter) were analyzed. In the baseline cross-sectional analysis, subclinical hypothyroidism was associated with ischemic heart disease independent of age, systolic blood pressure, body mass index, cholesterol, smoking, erythrocyte sedimentation rate, or presence of diabetes mellitus [odds ratio (OR), 2.5; 95% confidence interval (95% CI), 1.1-5.4 in total subjects and OR, 4.0; 95% CI, 1.4-11.5 in men] but not in women. However, there was no association with cerebrovascular disease (OR, 0.9; 95% CI, 0.4-2.4). We were unable to detect an influence of thyroid antibody presence on the association between subclinical hypothyroidism and ischemic heart disease. In a 10-yr follow-up study until 1998, increased mortalities from all causes in yr 3-6 after baseline measurement were apparent in men with subclinical hypothyroidism (hazard ratio, 1.9-2.1) but not in women, although specific causes of death were not determined. Our results indicate that subclinical hypothyroidism is associated with ischemic heart disease and might affect all-cause mortality in men.
Risk for Ischemic Heart Disease and All-Cause Mortality
in Subclinical Hypothyroidism
MISA IMAIZUMI, MASAZUMI AKAHOSHI, SHINICHIRO ICHIMARU, EIJI NAKASHIMA,
AYUMI HIDA, MIDORI SODA, TOSHIRO USA, KIYOTO ASHIZAWA, NAOKATA YOKOYAMA,
RENJU MAEDA, SHIGENOBU NAGATAKI, AND KATSUMI EGUCHI
Departments of Clinical Studies (M.I., M.A., S.I., A.H., M.S., R.M.) and Statistics (E.N.), Radiation Effects Research
Foundation, Nagasaki 850-0013, Japan; First Department of Internal Medicine, Graduate School of Biomedical Sciences
(M.I., T.U., K.E.) and Department of Molecular Medicine, Atomic Bomb Disease Institute (K.A.), Nagasaki University,
Nagasaki 852-8523, Japan; Nagasaki Saiseikai Hospital (N.Y.), Nagasaki 850-0003, Japan; and Japan Radioisotope
Association (S.N.), Tokyo 113-8941, Japan
We investigated possible associations between subclinical hy-
pothyroidism and atherosclerotic diseases (ischemic heart
disease and cerebrovascular disease) and mortality. Of 2856
participants (mean age 58.5 yr) in a thyroid disease screening
between 1984 and 1987, 257 subjects with subclinical hypo-
thyroidism (TSH > 5.0 mU/liter) and 2293 control subjects
(TSH range 0.6 –5.0 mU/liter) were analyzed. In the baseline
cross-sectional analysis, subclinical hypothyroidism was as-
sociated with ischemic heart disease independent of age, sys-
tolic blood pressure, body mass index, cholesterol, smoking,
erythrocyte sedimentation rate, or presence of diabetes mel-
litus [odds ratio (OR), 2.5; 95% confidence interval (95% CI),
1.1–5.4 in total subjects and OR, 4.0; 95% CI, 1.4 –11.5 in men]
but not in women. However, there was no association with
cerebrovascular disease (OR, 0.9; 95% CI, 0.4 –2.4). We were
unable to detect an influence of thyroid antibody presence on
the association between subclinical hypothyroidism and is-
chemic heart disease. In a 10-yr follow-up study until 1998,
increased mortalities from all causes in yr 3– 6 after baseline
measurement were apparent in men with subclinical hypo-
thyroidism (hazard ratio, 1.9–2.1) but not in women, although
specific causes of death were not determined. Our results
indicate that subclinical hypothyroidism is associated with
ischemic heart disease and might affect all-cause mortality in
men. (J Clin Endocrinol Metab 89: 3365–3370, 2004)
S
UBCLINICAL HYPOTHYROIDISM IS defined by the
presence of elevated TSH levels but normal free T
4
level,
and is reasonably common. Its prevalence in the general
population is 1–10% but approaches about 15% in women
who are over 60 yr of age (1–3). Clinical manifestations of
subclinical hypothyroidism include abnormal lipid metab-
olism (46), cardiac dysfunction (7, 8), and neurological and
mental dysfunction (9), and several cross-sectional studies
have suggested that it confers an elevated risk of atheroscle-
rosis and coronary heart disease (10–12). However, neither
of these associations has been confirmed by others (3, 13).
This discrepancy may reflect the small size of the studies or
participation in studies limited to one sex. Also, few longi-
tudinal studies have been conducted. The relationship be-
tween subclinical hypothyroidism and cardiovascular dis-
ease is therefore controversial, and possible outcomes of the
condition remain unclear. Importantly, several previous
studies suggesting that thyroid autoimmunity is a risk factor
for coronary heart disease (10, 14, 15) remain surrounded by
controversy (13, 16).
The Nagasaki Laboratory of the Radiation Effects Research
Foundation (RERF, formerly the Atomic Bomb Casualty
Commission) has been conducting biennial health examina-
tions of atomic bomb survivors in Nagasaki since 1958. Using
this cohort, we examined the possible associations between
subclinical hypothyroidism and atherosclerotic disease
(either ischemic heart disease or cerebrovascular disease)
in 2856 subjects of both sexes who agreed to participate
in thyroid disease screening between 1984 and 1987 (17).
We followed up the subjects until 1998 to screen for a possi-
ble association between subclinical hypothyroidism and
mortality.
Subjects and Methods
Subjects and definitions
A total of 7564 subjects (3374 men and 4190 women) have undergone
biennial health examinations in Nagasaki since 1958 in the follow-up of
atomic bomb survivors by RERF. A detailed description of this program
has been published elsewhere (18) and in the Atomic Bomb Casualty
Commission and Radiation Effects Research Foundation (Research Plan
for RERF Research Protocol 2-75, 1975).
At the baseline period, between October 1984 and April 1987, 2856
people (1119 men, 1737 women, mean age 58.5 yr) agreed to participate
in thyroid disease screening among 4693 people who were available for
the health examinations at RERF (1863 men, 2830 women, mean age, 60.2
yr). Among the 2856 participants, those with a history of treatment with
thyroid hormone or antithyroid medication (n 73) were excluded from
our study subjects. Subjects with TSH levels greater than 5.0 mU/liter
and normal free T
4
levels [0.8–2.5 ng/dl (10.3–32.3 pmol/liter)] were
defined as having subclinical hypothyroidism (n 257). Subjects with
TSH levels of 0.6 –5.0 mU/liter and normal free T
4
levels were treated
as controls (n 2293). Our study therefore involved 2550 subjects.
Among 257 subjects with subclinical hypothyroidism, 240 had TSH
Abbreviations: BMI, Body mass index; CI, confidence interval; ESR,
erythrocyte sedimentation rate; HR, hazard ratio; ICD9, International
Classification of Diseases, ninth version; OR, odds ratio.
JCEM is published monthly by The Endocrine Society (http://www.
endo-society.org), the foremost professional society serving the en-
docrine community.
0021-972X/04/$15.00/0 The Journal of Clinical Endocrinology & Metabolism 89(7):3365–3370
Printed in U.S.A. Copyright © 2004 by The Endocrine Society
doi: 10.1210/jc.2003-031089
3365
levels between 5 and 10 mU/liter and 17 had levels greater than 10
mU/liter.
Among the subjects with subclinical hypothyroidism, people who
were treated with thyroid hormone after baseline measurement (n 7)
were excluded from the analysis of the 10-year follow-up study. The
Dosimetry System 1986 (19) was used in estimating the radiation doses
of individual subjects. Both the thyroid disease screening study and a
mortality study were reviewed and approved by Research Protocol
Committee and Human Investigation Committee in RERF.
Clinical examination and laboratory methods at the
baseline study
Participants visited the RERF Nagasaki Laboratory for clinical ex-
amination between 1984 and 1987. A trained nurse recorded information
on current and past health including thyroid disease, medication, and
lifestyle. Current and past smokers were both categorized as smoker.
Body mass index (BMI) (in kilograms per square meter) was body
weight divided by the square of standing height. Sitting blood pressure
(in millimeters of mercury) was measured on the left arm after an
adequate sedentary period. A standard 12-lead electrocardiogram was
obtained by the regular procedure.
A fasting blood sample was drawn for a hematological test, bio-
chemical test, thyroid function test including free T
4
and TSH levels, and
thyroid autoantibody test. Commercial kits for RIA were used to mea-
sure TSH levels (Eiken RIA, Eiken Chemical Co. Ltd., Tokyo, Japan) (20)
and free T
4
levels (Gammacoat RIA, Travenol Lab Inc., Tokyo, Japan).
Serum thyroid antibodies (antithyroglobulin and antimicrosome anti-
bodies) were measured by passive hemagglutination methods (FUJIRE-
BIO, Inc., Tokyo, Japan). Subjects were described as positive for thyroid
autoantibodies if tests proved positive for either antithyroglobulin an-
tibody or antimicrosome antibody at a dilution in excess of 1:100. Thy-
roid function and thyroid antibodies were not measured after the base-
line measurement during a follow-up period, although subjects had
undergone biennial health examinations. Diabetes mellitus was reeval-
uated according to 1998 World Health Organization criteria (21).
Confirmation of ischemic heart disease and cerebrovascular
disease at baseline
Diagnosis of ischemic heart disease (myocardial infarction and an-
gina pectoris) and cerebrovascular disease (intracranial hemorrhage and
cerebral infarction) at baseline period was based on self-report, electro-
cardiographic findings, and information provided by general practitio-
ners. An experienced cardiologist reviewed the electrocardiographic
findings of all participants. The presence of myocardial infarction was
confirmed by the presence of one or more of the following conditions:
typical electrocardiographic evidence of myocardial infarction, com-
pared with previous electrocardiographic findings; chest pain with typ-
ical acute electrocardiographic changes; and elevation of myocardial
enzymes. The presence of angina pectoris was confirmed by evidence
from a positive result in an exercise electrocardiogram or diagnosis by
a physician and medical treatment. Cerebrovascular disease was defined
as rapid onset of a new neurological deficit in the absence of underlying
potentially important nonvascular causes and lasting at least 24 h. Brain
imaging and other diagnostic tests were used to determine the type of
cerebrovascular disease in most patients. An expert panel of internists
reviewed the diagnoses of ischemic heart disease and cerebrovascular
disease while unaware of the subjects thyroid status.
Confirmation of end points for mortality study
RERF followed the vital status of all participants using Japans family
registration system. We collected all death certificates of subjects who
died between the start of the thyroid disease screening study and De-
cember 31, 1998. The primary cause of death was classified according to
the International Classification of Diseases, ninth version (ICD9).
Statistical analysis
The Statistical Analysis System package for personal computers (SAS
Institute, Cary, NC) was used for statistical analysis. Differences in
baseline characteristics between controls and subjects with subclinical
hypothyroidism were assessed by the
2
and Wilcoxon rank sum tests.
General linear-model procedures were used to compare the age-
adjusted baseline characteristics. Multivariate logistic regression anal-
ysis was used to evaluate the association between subclinical hypothy-
roidism and ischemic heart disease or cerebrovascular disease, adjusting
for sex, age, systolic blood pressure, BMI, total cholesterol, smoking
status, erythrocyte sedimentation rate (ESR), and the presence of dia-
betes mellitus. Kaplan-Meier analysis was performed to assess survival
rates. Cox proportional hazards regression was used to compare the
mortality of subjects diagnosed as having subclinical hypothyroidism
with controls, after adjustment for age, sex, and smoking status. Hazard
ratios (HRs) with 95% confidence interval (CI) were used to estimate the
adjusted relative risk of subjects with subclinical hypothyroidism. The
relative risk for death from all causes and cause-specific death associated
with subclinical hypothyroidism was calculated at 3, 4, 5, 6, and 10 yr
after baseline measurement. Analyses conducted at 3, 4, 5, 6, and 10 yr
after the baseline measurement included all events observed before the
respective years. All significance tests were two sided and P 0.05 was
considered significant.
Results
The prevalence of subclinical hypothyroidism among all
participants in the thyroid disease screening study was 292
of 2856 or 10.2% overall, with 110 of 1119 (9.8%) in men and
182 of 1737 (10.5%) in women before subjects taking thyroid
medication with increased or decreased free T
4
levels or
decreased TSH levels had been excluded.
The baseline characteristics of controls and subjects with
subclinical hypothyroidism are shown in Table 1. The sex
ratios of controls and subjects with subclinical hypothyroid-
ism did not differ (P 0.53). Men and women with sub-
clinical hypothyroidism were older than controls (P 0.01).
In thyroid function tests, as anticipated, TSH levels were
significantly elevated in men and women with subclinical
hypothyroidism. Free T
4
levels were significantly lower in
women with subclinical hypothyroidism, although they
were within normal range. The positive rates for thyroid
autoantibodies were significantly higher in both men and
women with subclinical hypothyroidism, although the pos-
itive rates were low (less than 20%) because the antibodies
were detected by a hemagglutination method but not by a
highly sensitive method such as RIA. After adjusting the
clinical and laboratory data for age, individuals with sub-
clinical hypothyroidism appeared to have higher ESR results
(if male) or higher BMI results (if female). However, there
were no consistent differences in systolic blood pressure,
heart rate, total cholesterol level, white blood cell count,
hemoglobin level, or presence of diabetes mellitus. Further-
more, no difference was observed in thyroid radiation dose
between subjects with subclinical hypothyroidism and con-
trols (Table 1), and radiation dose had no significant effect on
the presence of subclinical hypothyroidism by logistic anal-
ysis (data not shown).
Possible associations between subclinical hypothyroidism
and ischemic heart disease/cerebrovascular disease were
also examined (Table 2). Subclinical hypothyroidism was
significantly positively associated with the prevalence of
ischemic heart disease after adjustment for age and sex in all
subjects [odds ratio (OR) 2.6; 95% CI, 1.25.6; P 0.02] and
after adjustment for age in men (OR 3.7; 95% CI, 1.410.2; P
0.01). The OR for ischemic heart disease in women with
subclinical hypothyroidism was 1.6, but this was not signif-
icant (P 0.45). Additional adjustments for systolic blood
3366 J Clin Endocrinol Metab, July 2004, 89(7):33653370 Imaizumi et al. Subclinical Hypothyroidism and Heart Disease
pressure, BMI, total cholesterol, smoking status, ESR, and
presence of diabetes mellitus did not affect the association
between subclinical hypothyroidism and ischemic heart dis-
ease (OR 2.5; 95% CI, 1.15.4; P 0.02 for all subjects and OR
4.0; 95% CI, 1.4 11.5; P 0.01 for men). This result indicates
that subclinical hypothyroidism is associated with ischemic
heart disease in this group of subjects independently of com-
mon coronary risk factors. In the subset analysis of mild
subclinical hypothyroidism (baseline TSH levels 10 mU/
liter), mild subclinical hypothyroidism was still associated
with ischemic heart disease independently of coronary risk
factors in all subjects (OR 2.7; 95% CI, 1.2 6.0; P 0.01) and
men (OR 4.5; 95% CI, 1.612.8; P 0.01) but not women (OR
1.7; 95% CI, 0.5 6.1; P 0.70). Radiation dose did not affect
the association between subclinical hypothyroidism and
ischemic heart disease (data not shown). Subclinical hypo-
thyroidism was not associated with cerebrovascular disease
in either men or women (Table 2). We also found no evidence
of an association between subclinical hypothyroidism and
either intracranial hemorrhage or cerebral infarction (data
not shown).
When we sought to determine whether presence of thy-
roid autoantibodies might influence the association between
subclinical hypothyroidism and ischemic heart disease, we
found that the ORs for ischemic heart disease in subjects with
subclinical hypothyroidism was not changed by adjusting
for the presence of thyroid autoantibodies [see Table 2: OR
for all subjects 2.5 (95% CI, 1.15.5); P 0.02 and OR for men
3.8 (95% CI, 1.311.1); P 0.01], This result indicates that the
relationship between subclinical hypothyroidism and isch-
emic heart disease was independent of the presence of thy-
roid autoantibodies.
To investigate possible outcomes of subclinical hypothy-
roidism, we collected death information during an average
follow-up of 12.2 yr after the baseline period. We compared
the survival rates of subjects with subclinical hypothyroid-
ism and those of controls using the Kaplan-Meier method
(Fig. 1). Survival was lower in men with subclinical hypo-
TABLE 1. Baseline characteristics of the subjects
Controls
(n 2293)
Subjects with subclinical
hypothyroidism
(n 257)
P
Men Women Men Women Men Women
No. 903 1390 96 161
Age (yr, mean SD) 57.7 11.1 58.1 9.4 62.7 11.1 60.6 9.8 0.01 0.01
Smoker (%) 79.9 11.3 70.7 7.5 0.06 0.18
TSH (mean
SD, mU/liter) 2.74 1.03 2.88 1.00 7.16 4.82 6.57 2.24 0.01 0.01
Free T
4
(mean SD, ng/dl)
a
1.50 0.36 1.44 0.35 1.45 0.34 1.31 0.33 0.29 0.01
Positive for thyroid autoantibodies (%) 7.1 14.6 10.9 19.3 0.03 0.01
Systolic blood pressure (mean SD, mm Hg) 137 23 137 24 141 22 142 28 0.61
c
0.13
c
Heart rate (mean SD/min) 66 10 66 10 66 11 68 9 0.65
c
0.08
c
BMI (mean SD) 22.0 2.8 22.8 3.2 21.7 3.0 23.5 3.9 0.71
c
0.01
c
Total cholesterol level (mean SD, mg/dl)
b
184.8 32.2 203.4 35.0 184.5 31.8 206.1 35.7 0.99
c
0.59
c
White blood cell count (mean SD, 10
6
/liter)
6.5 1.8 5.7 1.4 6.4 1.5 5.8 1.4 0.66
c
0.21
c
Hemoglobin level (mean SD, mg/dl) 14.4 1.4 12.6 1.2 14.0 1.5 12.5 1.2 0.09
c
0.35
c
ESR (mean SD, mm/h) 12.0 8.8 21.6 9.9 16.1 9.7 23.6 9.4 0.01
c
0.14
c
Diabetes mellitus (%) 8.4 4.6 7.5 7.0 0.76 0.18
Thyroid radiation dose (mean SD, sievert) 0.48 0.73 0.46 0.66 0.47 0.73 0.47 0.66 0.88 0.81
All data were from the examination at the time of thyroid disease screening conducted between October 1984 and April 1987.
a
To convert values to picomoles per liter, multiply by 12.87.
b
To convert values to millimoles per liter, multiply by 0.0259.
c
Adjusted for age.
TABLE 2. ORs for ischemic heart disease and cerebrovascular disease in subjects with subclinical hypothyroidism
Controls Subjects with subclinical hypothyroidism
OR
(95% CI)
a
OR
(95% CI)
b
OR
(95% CI)
c
Total no.
No. of
events (%)
Total no.
No. of
events (%)
Ischemic heart disease
Total 2293 29 (1.3) 257 9 (3.5) 2.6 (1.25.6)
d
2.5 (1.15.4)
d
2.5 (1.15.5)
d
Men 903 14 (1.6) 96 6 (6.3) 3.7 (1.410.2)
e
4.0 (1.411.5)
e
3.8 (1.311.1)
d
Women 1390 15 (1.1) 161 3 (1.8) 1.6 (0.55.7) 1.6 (0.4 5.7) 1.8 (0.56.6)
Cerebrovascular disease
Total 2293 38 (1.7) 257 6 (2.3) 1.2 (0.52.8) 0.9 (0.4 2.4) 0.9 (0.32.3)
Men 903 21 (2.3) 96 3 (3.1) 0.9 (0.33.3) 0.6 (0.12.8) 0.6 (0.12.8)
Women 1390 17 (1.2) 161 3 (1.8) 1.4 (0.44.9) 1.2 (0.4 4.4) 1.2 (0.34.2)
a
Adjusted for age (and sex for total subjects).
b
Adjusted for age (and sex for total subjects), systolic blood pressure, BMI, total cholesterol level, smoking status, ESR, and the presence
of diabetes mellitus.
c
Adjusted for age (and sex for total subjects), systolic blood pressure, BMI, total cholesterol level, smoking status, ESR, the presence of
diabetes mellitus, and the presence of thyroid antibodies.
d
P 0.05.
e
P 0.01.
Imaizumi et al. Subclinical Hypothyroidism and Heart Disease J Clin Endocrinol Metab, July 2004, 89(7):33653370 3367
thyroidism, but no comparable difference was apparent in
women. Mortality from all causes was significantly higher in
men with subclinical hypothyroidism than in controls ac-
cording to the analyses at 3, 4, 5, and 6 yr after baseline
measurement (Table 3, HR 1.9 2.1). There were also signif-
icantly more deaths from nonneoplastic disease in men with
subclinical hypothyroidism in the assessment 6 yr after base-
line [Table 3, HR 2.2 (95% CI, 1.1 4.2)]. Table 4 shows the
specific causes of death from nonneoplastic disease in men
with subclinical hypothyroidism as recorded 6 yr after base-
line and indicates that no significant cause of death was
observed, although each HR increased and the HR of
ischemic heart disease was suggestive [HR 4.8 (95% CI,
0.829.3)]. In the subset analysis of mild subclinical hypo-
thyroidism (baseline TSH levels 10 mU/liter), HRs for
all-cause mortality in men slightly decreased at all stages, but
a significant increased risk was observed at 6 yr after baseline
measurement (Table 3). The significant increases in mortality
in men with subclinical hypothyroidism were no longer ap-
parent when the assessment was made 10 yr after baseline
measurement (Table 3). There was no evidence of increased
mortality in women with subclinical hypothyroidism at any
stage in the follow-up period (Table 3).
Discussion
The prevalence of subclinical hypothyroidism in our study
population was 10.2%, which is consistent with the preva-
lence reported in the Wickham survey, in which 10.0% of
women 5564 yr of age had TSH levels greater than 6 mIU/
liter (3); the Framingham study, in which 13.6% of women
older than 60 yr had TSH levels greater than 5 mIU/liter (2);
and the Colorado study, in which 9.0% of subjects had sub-
clinical hypothyroidism (1). This suggests that our study
population is likely to be reasonably representative of the
general population. We observed no significant difference in
the frequency of subclinical hypothyroidism between men
and women in our study, although most previous studies
showed a higher frequency of subclinical hypothyroidism in
women (1, 3). This result might be related to the fact the
prevalence of elderly men over the age of 70 yr was higher
than that of women in our study (195 of 999, 19.5% of men,
and 197 of 1551, 12.7% of women).
Overt hypothyroidism, defined as low thyroid hormone
levels accompanied by elevated TSH levels, is associated
with cardiovascular disease, presumably because of hyper-
cholesterolemia and hypertension (22, 23). However, direct
evidence of an effect of overt hypothyroidism on cardiovas-
cular disease is lacking (24). Subclinical hypothyroidism has
been shown to be associated with cardiovascular disease in
several case-control studies (10, 13), but the results are
inconsistent (3, 1013). Interestingly, a Rotterdam cross-
sectional study of elderly women indicated that subclinical
hypothyroidism does appear to be associated with myocar-
dial infarction, even after controlling for coronary risk factors
(11). The results of our study are in broad agreement with the
Rotterdam result, in that we are reporting an association
between subclinical hypothyroidism and the prevalence of
ischemic heart disease that is independent of coronary risk
factors such as blood pressure, BMI, total cholesterol level,
smoking status, ESR, and presence of diabetes mellitus. Fur-
thermore, in a follow-up study, an increase in all-cause and
nonneoplastic disease deaths was also apparent in men when
assessed during the first 6 yr after baseline, although specific
causes of death were not determined. We also found that
mild subclinical hypothyroidism (baseline TSH levels 10
mU/liter) was associated with ischemic heart disease in a
cross-sectional study and is accompanied by an increase in
all-cause mortality in a longitudinal follow-up study at 6 yr
after baseline measurement. This result indicated that un-
treated subclinical hypothyroidism, even mild subclinical
hypothyroidism, might directly or indirectly affect fatal dis-
eases. However, increased mortality that became apparent
during the first 6 yr after baseline had disappeared by the
10-year mark (Table 3). This finding is not unexpected be-
FIG. 1. Kaplan-Meier survival curves show that the survival rate was
lower in men with subclinical hypothyroidism than in the controls,
but no comparable difference was apparent in women. A, Men. B,
Women.
3368 J Clin Endocrinol Metab, July 2004, 89(7):33653370 Imaizumi et al. Subclinical Hypothyroidism and Heart Disease
cause the death rates from malignant diseases as well as
atherosclerotic diseases caused by various risk factors in-
crease with aging, irrespective of baseline TSH values. We
thus cannot expect initial high TSH levels and elevated mor-
tality to be so closely associated for such a long time period.
Interestingly, the association between subclinical hypothy-
roidism and ischemic heart disease and the predictive value
of subclinical hypothyroidism for mortality were both only
apparent in men. This may be a reflection of the relative
youth of our female study population and their not having
reached the key end points in the lifetime of the study. The
mean age of women at baseline in our study (58 yr) was much
younger than in the Rotterdam study (69 yr) (11). On the
other hand, cerebrovascular disease is also a major athero-
sclerotic disease, but there was no evidence of an association
between cerebrovascular disease and subclinical hypothy-
roidism in the present study. Further study will be required
for understanding the relationship between subclinical hy-
pothyroidism and other atherosclerotic diseases.
There are several possible mechanisms that might explain
the elevated risk of ischemic heart disease in subclinical
hypothyroidism subjects. One involves the thyroid autoim-
munity found in subclinical hypothyroidism together with
the possibility that coronary vascular stenosis is induced by
local inflammation from the pathological immune reactivity
in autoimmune thyroid disease (10, 14, 15). However, this
suggestion was not confirmed by our study or other studies
(13, 16). We think our results weaken the hypothesis that
organ-specific autoimmunity such as autoimmune thyroid-
itis is associated with coronary artery disease (10).
Our study and most cross-sectional studies have detected
no significant differences in total serum cholesterol levels
between subjects with subclinical hypothyroidism and eu-
thyroid subjects (5, 6, 2527). Lipoprotein (a), a putative
contributing factor in the development of atherosclerosis,
was suggested to be related to subclinical hypothyroidism
(28, 29). Other studies, however, failed to provide confirma-
tion (30, 31), and thus the relationship between abnormal
lipid metabolism and ischemic heart disease in subclinical
hypothyroidism remains controversial. A recent report also
suggests that flow-mediated vasodilatation (a marker of en-
dothelial function), which is considered an early sign of
atherosclerosis, may be impaired in patients with subclinical
hypothyroidism (8). Several studies have also demonstrated
the presence of TSH receptor mRNA in coronary arteries,
atria, and ventricles (32, 33). Elevated TSH may directly affect
coronary arteries.
A limitation of our study is that the numbers of end point
events were small, probably for two reasons. First, the power
of the study may not be sufficient because this study was
performed on people who were examined at RERF during
the period of 1984 through 1987 and agreed to undergo
thyroid examinations. Second, the age of subjects was rela-
tively young to reach the key end points (mean age sd,
58.5 10.2 yr). However, the significant association between
subclinical hypothyroidism and ischemic heart disease in
men was observed even after adjustment for all coronary risk
factors, and we consider the association to be well grounded.
TABLE 3. Risk of death for subjects with subclinical hypothyroidism compared with that of controls
Cause of death (ICD-9 code)
Observed deaths in controls/in subjects with subclinical hypothyroidism, HRs (95% CI)
After 3 yr After 4 yr After 5 yr After 6 yr After 10 yr
All causes
Total 64/14, 1.5 (0.92.7) 80/18, 1.5 (0.92.6) 102/22, 1.5 (0.92.3)
a
125/27, 1.5 (1.02.3)
a
268/42, 1.2 (0.81.6)
Men 37/11, 2.1 (1.14.2)
b
45/13, 2.0 (1.13.8)
b
53/14, 1.9 (1.03.4)
b
66/17, 1.9 (1.13.2)
b
143/23, 1.3 (0.82.0)
Women 27/3, 0.8 (0.22.5) 35/5, 0.9 (0.42.4) 49/8, 1.1 (0.52.3) 59/10, 1.1 (0.62.2) 125/19, 1.1 (0.71.8)
Malignant neoplasms (140208)
Total 28/6, 1.6 (0.73.9) 35/6, 1.2 (0.53.0) 43/8, 1.4 (0.62.9) 56/9, 1.2 (0.62.5) 105/14, 1.0 (0.6 1.6)
Men 16/5, 2.5 (0.9 6.8)
a
20/5, 2.0 (0.75.3) 24/5, 1.7 (0.6 4.6) 30/5, 1.4 (0.53.7) 58/6, 1.0 (0.42.2)
Women 12/1, 0.6 (0.14.4) 15/1, 0.4 (0.13.4) 19/3, 1.0 (0.33.6) 26/4, 1.1 (0.43.0) 47/8, 1.3 (0.62.7)
Nonneoplastic diseases
Total 36/8, 1.5 (0.73.2) 45/12, 1.7 (0.93.3) 59/14, 1.6 (0.92.8) 69/18, 1.7 (1.02.9)
b
163/28, 1.2 (0.81.9)
Men 21/6, 1.9 (0.74.7) 25/8, 2.1 (0.9 4.7)
a
29/9, 2.0 (0.9 4.3)
a
36/12, 2.2 (1.14.2)
b
85/17, 1.5 (0.92.6)
Women 15/2, 0.9 (0.24.0) 20/4, 1.3 (0.4 3.7) 30/5, 1.1 (0.42.8) 33/6, 1.2 (0.52.9) 78/11, 1.0 (0.51.8)
All causes
c
Total 64/12, 1.4 (0.82.7) 80/16, 1.4 (0.92.5) 102/20, 1.5 (0.92.4) 125/25, 1.5 (1.02.3)
a
268/38, 1.2 (0.81.6)
Men 37/9, 1.9 (0.93.9)
a
45/11, 1.9 (1.03.7)
a
53/12, 1.8 (0.93.3)
a
66/15, 1.8 (1.03.1)
b
143/20, 1.2 (0.71.9)
Women 27/3, 0.8 (0.32.7) 35/5, 1.0 (0.42.6) 49/8, 1.2 (0.62.5) 59/10, 1.3 (0.6 2.5) 125/18, 1.1 (0.71.9)
Adjusted for age (and sex for total subjects) and smoking status.
a
P 0.1.
b
P 0.05.
c
Analyzed in the subjects with baseline TSH levels 10 mU/liter.
TABLE 4. Risk of death caused by nonneoplastic diseases for
men with subclinical hypothyroidism compared with controls at
first 6 yr during follow-up period
Cause of death (ICD-9 code)
Observed deaths in
controls/in subjects
with subclinical
hypothyroidism, HR
(95% CI)
Circulatory diseases (390 459) 21/6, 1.8 (0.74.6)
Cardiovascular diseases (390 414) 6/2, 2.4 (0.511.8)
Ischemic heart diseases (410 414) 3/2, 4.8 (0.829.3)
a
Cerebrovascular diseases (430 438) 9/3, 1.9 (0.57.3)
Respiratory diseases (460519) 8/3, 2.5 (0.79.6)
Others
b
7/3, 3.2 (0.812.5)
a
Adjusted for age (and sex for total subjects) and smoking status.
a
P 0.1.
b
Othersconsists of infectious diseases (001139), metabolic dis
-
eases (240289), mental disorders (290319), diseases of the nervous
system (320 389), diseases of the digestive system (520579), dis-
eases of the genitourinary system (580 599), and injury (800959).
Imaizumi et al. Subclinical Hypothyroidism and Heart Disease J Clin Endocrinol Metab, July 2004, 89(7):33653370 3369
Further studies using a larger number of subjects or more
elderly subjects will be required to assess particular rela-
tionships between subclinical hypothyroidism and ischemic
heart disease in women, cerebrovascular disease, and cause-
specific mortality.
We have clearly demonstrated that subclinical hypothy-
roidism is associated with ischemic heart disease, especially
evident in men in a cross-sectional study, and is accompanied
by an increase in all-cause mortality in a longitudinal fol-
low-up study. Our study indicates the importance of thyroid-
status screening, especially in people who are close to reach-
ing the age at which the incidence of ischemic heart disease
is high. Optimal patient outcome for people with subclinical
hypothyroidism, including mild subclinical hypothyroid-
ism, in the future is highly likely to be dependent on careful
follow-up. Further prospective studies are necessary to con-
firm the association between subclinical hypothyroidism and
ischemic heart disease, other atherosclerotic diseases, and
cause-specific mortality. Also, the validity of thyroid hor-
mone replacement therapy in subclinical hypothyroidism is
left for future studies.
Acknowledgments
We thank Drs. Donald G. MacPhee and Richard Haber for reviewing
the manuscript. We also thank Mrs. Kaoru Yoshida for general
assistance.
Received June 25, 2003. Accepted March 29, 2004.
Address all correspondence and requests for reprints to: Misa
Imaizumi, Radiation Effects Research Foundation, 1-8-6 Nakagawa, Na-
gasaki 850-0013, Japan. E-mail: misaima@rerf.or.jp.
This work was supported by RERF Research Protocol 01-83. The
RERF, Hiroshima and Nagasaki, Japan, is a private, nonprofit founda-
tion funded by the Japanese Ministry of Health, Labor, and Welfare and
the U.S. Department of Energy, with the latter funding provided through
the National Academy of Sciences.
References
1. Canaris GJ, Manowitz NR, Mayor G, Ridgway EC 2000 The Colorado thyroid
disease prevalence study. Arch Intern Med 160:526 534
2. Sawin CT, Castelli WP, Hershman JM, McNamara P, Bacharach P 1985 The
aging thyroid. Thyroid deficiency in the Framingham Study. Arch Intern Med
145:13861388
3. Tunbridge WM, Evered DC, Hall R, Appleton D, Brewis M, Clark F, Evans
JG, Young E, Bird T, Smith PA 1977 The spectrum of thyroid disease in a
community: the Whickham survey. Clin Endocrinol (Oxf) 7:481 493
4. Pucci E, Chiovato L, Pinchera A 2000 Thyroid and lipid metabolism. Int J Obes
Relat Metab Disord 24(Suppl 2):S109 S112
5. Caron P, Calazel C, Parra HJ, Hoff M, Louvet JP 1990 Decreased HDL cho-
lesterol in subclinical hypothyroidism: the effect of l-thyroxine therapy. Clin
Endocrinol (Oxf) 33:519523
6. Althaus BU, Staub JJ, Ryff-De Leche A, Oberhansli A, Stahelin HB 1988
LDL/HDL-changes in subclinical hypothyroidism: possible risk factors for
coronary heart disease. Clin Endocrinol (Oxf) 28:157163
7. Biondi B, Fazio S, Palmieri EA, Carella C, Panza N, Cittadini A, Bone F,
Lombardi G, Sacca L 1999 Left ventricular diastolic dysfunction in patients
with subclinical hypothyroidism. J Clin Endocrinol Metab 84:2064 2067
8. Kahaly GJ 2000 Cardiovascular and atherogenic aspects of subclinical hypo-
thyroidism. Thyroid 10:665679
9. Monzani F, Del Guerra P, Caraccio N, Pruneti CA, Pucci E, Luisi M, Baschieri
L 1993 Subclinical hypothyroidism: neurobehavioral features and beneficial
effect of l-thyroxine treatment. Clin Investig 71:367371
10. Tieche M, Lupi GA, Gutzweller F, Grob PJ, Studer H, Burgi H 1981 Bor-
derline low thyroid function and thyroid autoimmunity. Risk factors for cor-
onary heart disease? Br Heart J 46:202206
11. 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 132:270278
12. Dean JW, Fowler PB 1985 Exaggerated responsiveness to thyrotrophin re-
leasing hormone: a risk factor in women with coronary artery disease. Br Med J
(Clin Res Ed) 290:15551561
13. Heinonen OP, Gordin A, Aho K, Punsar S, Pyorala K, Puro K 1972 Symp-
tomless autoimmune thyroiditis in coronary heart-disease. Lancet 1:785786
14. Bastenie PA, Banhaelst L, Bonnyns M, Neve P, Staquet M 1971 Preclinical
hypothyroidism: a risk factor for coronary heart-disease. Lancet 1:203204
15. Bastenie PA, Vanhaelst L, Neve P 1967 Coronary artery disease in hypothy-
roidism. Lancet 2:12211222
16. Vanderpump MP, Tunbridge WM, French JM, Appleton D, Bates D, Clark
F, Grimley Evans J, Rodgers H, Tunbridge F, Young ET 1996 The develop-
ment of ischemic heart disease in relation to autoimmune thyroid disease in
a 20-year follow-up study of an English community. Thyroid 6:155160
17. Nagataki S, Shibata Y, Inoue S, Yokoyama N, Izumi M, Shimaoka K 1994
Thyroid diseases among atomic bomb survivors in Nagasaki. JAMA 272:364
370
18. Commission ABC 1962 Research plan for joint ABCC-NIH Adult Health Study
in Hiroshima and Nagasaki. Hiroshima and Nagasaki, Japan. ABCC Technical
Report 1162, Hiroshima, Japan: Atomic Bomb Casualty Commission
19. Roesch WC, ed 1987 U.S.-Japan Joint Reassessment of Atomic Bomb Radiation
Dosimetry in Hiroshima and Nagasaki. Hiroshima, Japan: Radiation Effects
Research Foundation
20. K Shizume HD, Takano K 1978 Studies of TSH radioimmunoassay kit, Eiken.
Horumon To Rinsho 26:10631066
21. Alberti KG, Zimmet PZ 1998 Definition, diagnosis and classification of dia-
betes mellitus and its complications. Part 1: diagnosis and classification of
diabetes mellitus provisional report of a WHO consultation. Diabet Med 15:
539553
22. Vanhaelst L, Neve P, Chailly P, Bastenie PA 1967 Coronary-artery disease in
hypothyroidism. Observations in clinical myxoedema. Lancet 2:800 802
23. Steinberg AD 1968 Myxedema and coronary artery diseasea comparative
autopsy study. Ann Intern Med 68:338 344
24. Klein I, Ojamaa K 2001 Thyroid hormone and the cardiovascular system.
N Engl J Med 344:501509
25. Bigos ST, Ridgway EC, Kourides IA, Maloof F 1978 Spectrum of pituitary
alterations with mild and severe thyroid impairment. J Clin Endocrinol Metab
46:317325
26. Parle JV, Franklyn JA, Cross KW, Jones SR, Sheppard MC 1992 Circulating
lipids and minor abnormalities of thyroid function. Clin Endocrinol (Oxf)
37:411414
27. Bogner U, Arntz HR, Peters H, Schleusener H 1993 Subclinical hypothyroid-
ism and hyperlipoproteinaemia: indiscriminate l-thyroxine treatment not jus-
tified. Acta Endocrinol (Copenh) 128:202206
28. Kung AW, Pang RW, Janus ED 1995 Elevated serum lipoprotein(a) in sub-
clinical hypothyroidism. Clin Endocrinol (Oxf) 43:445 449
29. Yildirimkaya M, Ozata M, Yilmaz K, Kilinc C, Gundogan MA, Kutluay T
1996 Lipoprotein(a) concentration in subclinical hypothyroidism before and
after levo-thyroxine therapy. Endocr J 43:731736
30. Caraccio N, Ferrannini E, Monzani F 2002 Lipoprotein profile in subclinical
hypothyroidism: response to levothyroxine replacement, a randomized pla-
cebo-controlled study. J Clin Endocrinol Metab 87:15331538
31. Tzotzas T, Krassas GE, Konstantinidis T, Bougoulia M 2000 Changes in
lipoprotein(a) levels in overt and subclinical hypothyroidism before and dur-
ing treatment. Thyroid 10:803 808
32. Sellitti DF, Hill R, Doi SQ, Akamizu T, Czaja J, Tao S, Koshiyama H 1997
Differential expression of thyrotropin receptor mRNA in the porcine heart.
Thyroid 7:641646
33. Drvota V, Janson A, Norman C, Sylven C, Haggblad J, Bronnegard M,
Marcus C 1995 Evidence for the presence of functional thyrotropin receptor in
cardiac muscle. Biochem Biophys Res Commun 211:426 431
JCEM is published monthly by The Endocrine Society (http://www.endo-society.org), the foremost professional society serving the
endocrine community.
3370 J Clin Endocrinol Metab, July 2004, 89(7):33653370 Imaizumi et al. Subclinical Hypothyroidism and Heart Disease
    • "10.1016/j.numecd.2015.09.001 function, as clearly documented by the fact that thyroid function did not predict carotid atherosclerosis after adjusting for confounders. Although not universally accepted, both subclinical hypothyroidism and hyperthyroidism have been reported to affect the risk of ischemic heart disease [26] [27], suggesting the opportunity of treating these patients. Albeit that no long-term prospective analyses are available in literature, some case-control studies focused on the effect of thyroxine supplement therapy in subjects with subclinical hypothyroidism , with divergent results based on small groups of patients [28e30]. "
    [Show abstract] [Hide abstract] ABSTRACT: Background and aims: Increased carotid artery intima-media thickness (IMT) and the presence of plaques have been shown to be predictors of cardiovascular disease. The cardiovascular risk in patients with overt thyroid diseases is related to increased risk of atherosclerosis, but there has been no clear evidence about subclinical disorders. We have assessed whether subclinical thyroid dysfunction is associated with arterial thickening and plaque. Methods and results: The SardiNIA study is a population-based survey on the Italian island of Sardinia. We reviewed data from 5815 subjects (aged 14-102 years), none of whom had overt hyperthyroidism or hypothyroidism or was taking thyroid medication. Serum thyrotropin (TSH), free thyroxine, together with carotid ultrasound IMT and the presence of common carotid plaques were analysed in all subjects. Possible association of IMT and carotid plaques with thyroid parameters was evaluated by univariate and multivariate analyses. IMT was significantly associated with age, sex, smoking, low density lipoprotein cholesterol (LDL), high density lipoprotein cholesterol, pulse pressure (PP), history of arterial hypertension, diabetes, and previous cardiovascular events (p = 0.001 or lower, R(2) = 0.47). Carotid plaques were predicted by age, sex, LDL, PP, history of diabetes, previous cardiovascular events, and the use of statins (p = 0.029 or lower). Thyroid hormone was not predictive of carotid atherosclerosis when adjusted for confounders. Conclusion: Thyroid hormone is not associated with increased IMT or with the presence of carotid artery plaque. Our data do not support the idea that treating subclinical disorders might help to prevent arterial remodelling or carotid atherosclerosis.
    Article · Oct 2015
    • "Nevertheless, some evidence starts to appear. Subclinical hypothyroidism[31] was associated with higher cardiovascular risk in several32333435, but not all[23,36] previous studies. The same controversy appears with the association between metabolic syndrome and TSH, which is found in most[1,3,5,37] but not all[4,38] previous studies. "
    [Show abstract] [Hide abstract] ABSTRACT: Background and aims: Subclinical thyroid conditions, defined by normal thyroxin (T4) but abnormal thyroid-stimulating hormone (TSH) levels, may be associated with cardiovascular and metabolic risk. More recently, TSH levels within the normal range have been suggested to be associated with metabolic syndrome and cardiovascular risk. This work studies the linearity of the relationship between metabolic syndrome and TSH across the euthyroid range. Methods and results: We studied 3533 male participants of the Aragon Workers' Health Study (AWHS) with normal TSH and free T4 levels, across quintiles of these variables, after adjusting for age, alcohol intake, and smoking. Compared with the lowest TSH quintile, the odds ratios for metabolic syndrome at the higher quintiles, which indicate lower thyroid function, were 1.34 (1.04, 1.73), 1.56 (1.21, 2.01), 1.57 (1.22, 2.03), and 1.71 (1.32, 2.21). The lowest free T4 quintile also showed an odds ratio of 1.49 (1.16, 1.90) with respect to the highest quintile. In addition, spline models showed departures from linearity: the risk of metabolic syndrome mostly increases at TSH values below the median (sample half-closest to subclinical hyperthyroidism). Interestingly, glucose also increases with TSH primarily below the median TSH, diastolic blood pressure shows similar changes across the entire TSH range, whereas body mass index, triglycerides, and high-density lipoprotein (HDL)-cholesterol change only at the highest normal TSH values, which are associated with lower free T4 concentration. Conclusions: TSH and free T4 within the normal range are associated with the metabolic syndrome. The sample half-below the TSH median (with probably higher functional thyroid status) exhibited better metabolic and cardiovascular profiles.
    Full-text · Article · Sep 2015
    • "Some, but not all, studies demonstrated a higher prevalence of hypertension in SCH subjects [25,3132. Previous literature demonstrated no difference in levels of fasting glucose or hemoglobin A1C2728, and the prevalence of diabetes mellitus [5,2526 between the SCH and euthyroid subjects. Obesity, hypertension, diabetes, and dyslipidemia had been associated with cancer incidence or mortality3334353637383940. Smoking, alcohol consumption and betel nut chewing were also known as risk factors for lung, pancreatic, kidney, colorectal, or oralesophageal cancer, respectively414243. "
    [Show abstract] [Hide abstract] ABSTRACT: The association between subclinical hypothyroidism (SCH) and cancer mortality is seldom discussed. A total of 115,746 participants without thyroid disease history, aged 20 and above, were recruited from four nationwide health screening centers in Taiwan from 1998 to 1999. SCH was defined as a serum thyroid-stimulating hormone (TSH) level of 5.0-19.96 mIU/L with normal total thyroxine concentrations. Euthyroidism was defined as a serum TSH level of 0.47-4.9 mIU/L. Cox proportional hazards regression analyses were used to estimate the relative risks (RRs) of death from cancer for adults with SCH during a 10-year follow-up period. Among 115,746 adults, 1,841 had SCH (1.6%) and 113,905 (98.4%) had euthyroidism. There were 1,532 cancer deaths during the 1,034,082 person-years follow-up period. Adjusted for age, gender, body mass index, diabetes, hypertension, dyslipidemia, smoking, alcohol drinking, betel nut chewing, physical activity, income, and education level, the RRs (95% confidence interval) of cancer deaths among subjects with SCH versus euthyroid subjects were 1.51 (1.06 to 2.15). Cancer site analysis revealed a significant increased risk of bone, skin and breast cancer among SCH subjects (RR 2.79, (1.01, 7.70)). The risks of total cancer deaths were more prominent in the aged (RR 1.71, (1.02 to 2.87)), in females (RR 1.69 (1.08 to 2.65)), and in heavy smokers (RR 2.24, (1.19 to 4.21)). Subjects with SCH had a significantly increased risk for cancer mortality among adult Taiwanese. This is the first report to demonstrate the association between SCH and cancer mortality.
    Full-text · Article · Apr 2015
Show more

  • undefined · undefined
  • undefined · undefined
  • undefined · undefined

Recommended publications

Discover more