| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Departments of Cardiology (A.I., H.S., P.L., K.R.), Clinical Chemistry (Y.F.), and Internal Medicine B (R.J.), University Hospital of North Norway, N-9038 Tromsø, Norway; and Institute of Community Medicine (H.S.), Institute of Medical Biology (Y.F.) and Institute of Clinical Medicine (K.R., R.J.), University of Tromsø, N-9037 Tromsø, Norway
Address all correspondence and requests for reprints to: Amjid Iqbal, Department of Cardiology, University Hospital of North Norway, 9038 Tromsø, Norway. E-mail: amjid.iqbal{at}unn.no.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Objective: The objective was to assess the relation between serum TSH level and cardiac function.
Design: We conducted a cross-sectional epidemiological study and a nested case-control study.
Setting: The study was performed at a university hospital.
Subjects: A total of 2035 subjects were included in the epidemiological study and 204 subjects in the nested case-control study (serum TSH < 0.50, 0.50–3.49, and 3.50–10.0 mIU/liter in 20, 118, and 66 subjects, respectively, all with normal serum free T4 and free T3 levels).
Main Outcome Measures: Left ventricular mass by body surface area (LVMI) and indices of left ventricular function, as assessed by conventional and pulsed-wave tissue Doppler (PWTD) echocardiography, were recorded.
Results: No significant relation was found between serum TSH level and LVMI. In the nested case-control study, the subjects with serum TSH 3.50–10.0 mIU/liter had no signs of cardiac dysfunction. However, the PWTD data showed higher velocities at all measurement sites in the subjects with serum TSH less than 0.50 mIU/liter as compared with the euthyroid group.
Conclusions: With the possible exception of overt hypo- and hyperthyroidism, there is no significant association between serum TSH level and LVMI. Subjects with subclinical hypothyroidism, in whom the mean serum TSH level is slightly above the reference range, appear to have normal cardiac function, whereas subjects with serum TSH levels less than 0.5 mIU/liter appear to have changes in myocardial velocities detected by PWTD.
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
In the city of Tromsø, Northern Norway, there have been several large population-based studies focusing on cardiovascular diseases (15, 16), and in the fourth Tromsø study in 1994–1995, serum TSH was measured and conventional transthoracal echocardiography performed in more than 3000 subjects. Furthermore, in the fifth Tromsø study in 2001 serum TSH was measured in nearly 8000 subjects, and from this cohort, subjects with subclinical thyroid dysfunction and euthyroid controls were examined with transthoracal echocardiography including the new tool pulsed-wave tissue Doppler imaging (PWTD) echocardiography. Accordingly, a large database was available for evaluating the relationship between thyroid status and cardiac function.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
In the fourth Tromsø study in 1994–1995, all men and women who were older than 24 yr of age and who lived in the municipality of Tromsø, Northern Norway, were invited to participate in a health survey that was conducted in a manner similar to the previous Tromsø studies (16). All subjects aged 55–74 yr and random 5–10% samples of the other age groups were invited to a second visit for more extensive screening.
Trained nurses checked a questionnaire that included medical history and past and present medication. Height and weight were measured while the subjects wore light clothing and no shoes. Blood pressure was measured with an automatic device (Dinamap Vital Signs Monitor 1846; Critikon Inc., Tampa, FL.).
Echocardiography, using a VingMed CFM 750 (VingMed Sound A/S, Horten, Norway), was performed in a subgroup as previously described in detail regarding method and reproducibility (17). In short, the subjects were examined in a supine, left lateral position with a combined 3.25-MHz mechanical and 2.5-MHz Doppler probe. Left ventricular diastolic dimensions were measured on-line from standard two-dimensional guided M-mode registrations according to the leading edge to leading edge convention, using EchoPAC software. Only one heart cycle was measured per subject. Left ventricular mass (LVM) was calculated using the correction of the cube formula proposed by Devereux et al. (18) for leading edge to leading edge measurements [LVM = 0.832 x ([interventricular septal thickness + posterior wall thickness + end diastolic diameter]3 – [end diastolic diameter]3) +0.6] and divided by the body surface area to give the LVM index (LVMI). Blood samples were drawn in the nonfasting state and stored frozen at –70 C.
Nested case-control study after the fifth Tromsø study
The fifth Tromsø study was performed in 2001. A total of 10,419 subjects from the Tromsø municipality were invited to participate, and 8,128 attended. Blood samples were drawn in the nonfasting state for serum TSH measurements.
Subjects with serum TSH level between 3.50 and 10.0 mIU/liter, subjects with serum TSH less than 0.50 mIU/liter, and a control group with serum TSH in the range 0.50–3.49 mIU/liter were invited to a follow-up examination at the Clinical Research Unit at the University Hospital of Tromsø. Those who reported a history of coronary infarction, angina pectoris or stroke in the questionnaire, those participating in other follow-up studies, those using thyroid medication, and those above the age of 80 yr were not invited. The hospital records were also controlled to exclude subjects with serious diseases not reported in the questionnaire.
The subjects were examined at the outpatient clinic at the Medical Department and the Clinical Research Unit, University Hospital of Tromsø, approximately 6–12 months after the fifth Tromsø study. A general health examination was performed. Blood samples for serum TSH, free T4, and free T3 were drawn. Those who did not use blood pressure medication, had a serum free T4 and free T3 within the reference range, and serum TSH within the same range as in the fifth Tromsø study were examined further at a second occasion. Height, weight, and blood pressure were measured as in the fourth Tromsø study.
Echocardiographic evaluation in the nested case-control study
The echocardiography method used in the nested case-control study has previously been described in detail (19). In short, transthoracal echocardiography was performed by a cardiologist with several years of experience in cardiac ultrasound with the subject lying in the supine left lateral position using commercially available equipment (C256 Sequoia; Acuson, Palo Alto, CA) with integrated software for PWTD analysis. A multiphased (2.5–4 MHz) array transducer was used for echo Doppler and PWTD examination. Recordings were done during the end-expiratory phase, and the mean of three measurements on adjacent heartbeats in a recording of adequate quality was used in the analyses. Conventional apical four-chamber, apical two-chamber, and parasternal short axis view were used as recommended by the American Society of Echocardiography (20). From the apical four-chamber view, the following variables were measured: peak flow velocity in early diastole (E wave) and during atrial contraction (A wave), peak E to A ratio, deceleration time of the E wave, duration of the A wave, pulmonary peak flow velocity and time-velocity integral during forward systolic and diastolic flow, and peak velocity and duration of flow reversal during atrial systole and left ventricular isovolumetric relaxation time. From the parasternal short axis view, an M-mode tracing was used to measure the left ventricular end-systolic and end-diastolic diameters, interventricular septum thickness, and posterior wall thickness. From these variables we calculated LVMI and functional parameters such as left ventricular volume, cardiac output, ejection fraction, and fractional shortening.
PWTD of the mitral annulus was obtained from the apical four-chamber and apical two-chamber view. The Doppler sample volume was placed at the corners of the mitral annulus. Efforts were made to keep the annular motion in parallel with the Doppler cursor. The measurements were thus obtained from the septal, lateral, anterior, and posterior mitral annulus. A Doppler velocity range of –20 to +20 cm/sec and a sample volume of 6–10 mm with low reject and minimal optimal gain setting were used. PWTD velocities were recorded at a sweep speed of 100 mm/sec. All measurements were done in three cardiac cycles and averaged. Systolic velocity, early diastolic velocity, and late diastolic velocity were measured (Fig. 1
). We also estimated the mean velocity of the four mitral annulus velocities as a tissue Doppler variable expressing the global left ventricle PWTD function.
|
Laboratory analyses
Serum TSH, free T4, and free T3 were analyzed at the Department of Clinical Chemistry, University Hospital of North Norway, using a Modular E automated clinical chemistry analyzer (Roche Diagnostics, Mannheim, Germany). The TSH assay was a third-generation assay with total analytic precision (coefficient of variation) of 7.2, 3.2, and 3.3% at serum TSH levels of 0.04, 0.92, and 9.37 mIU/liter, respectively. The reference ranges for TSH, free T4, and free T3 were 0.2–4.0 mIU/liter, 9–22 pmol/liter, and 2.8–7.1 pmol/liter, respectively.
Statistical analyses
Normal distribution was evaluated by visual inspection of histograms and determination of skewness and kurtosis. There was no interaction between gender and serum TSH levels regarding LVMI or other measures of cardiac morphology or function. However, gender appeared as the strongest predictor of LVMI, and results from the fourth Tromsø study therefore were shown separately for males and females. In the fourth Tromsø study, the subjects were subdivided into six groups: less than the 2.5 serum TSH percentile, quartiles within the 2.5–97.5 serum TSH percentile range, and greater than the 97.5 serum TSH percentile. In the nested case-control study, the cohort was divided in three groups (serum TSH < 0.50, 0.50–3.50, and 3.51–10.0 mIU/liter).
Comparisons between groups were performed with ANOVA with post hoc testing using the Bonferroni correction. Groups were also compared with a general linear model with TSH group (and also gender in the nested case-control study) as factor(s) and age, body mass index (BMI), and systolic blood pressure as covariates. In the fourth Tromsø study, a multiple linear regression model was used to evaluate predictors of LVMI and included age, BMI, systolic blood pressure, and serum TSH as independent variables. Linear trends across TSH groups were analyzed similarly with linear regression.
Unless otherwise stated, all data are expressed as mean ± SD. All tests were done two-sided, and P < 0.05 was considered statistically significant. Statistical analyses were performed with SPSS (version 11.0; SPSS Inc., Chicago, IL).
Ethics
The study complied with the Declaration of Helsinki and was approved by the regional ethics committee, and all subjects gave their written informed consent to participate.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
A total of 27,159 subjects were included in the fourth Tromsø study. A total of 6891 subjects attended the second visit, and 3287 randomly selected subjects were examined by echocardiography. This examination was successful in 2794, and among these, serum TSH was measured in 2694. Six hundred sixty subjects were excluded because of known cardiovascular disease, valvular heart disease, or use of blood pressure medication, leaving 2035 subjects (976 males) for the final analyses.
The relation between serum TSH groups and age, BMI, systolic blood pressure, and LVMI is given in Table 1. In the males the mean LVMI was almost identical in all serum TSH subgroups. Seven male subjects had a serum TSH value less than 0.05 mIU/liter and four male subjects had serum TSH greater than 10.0 mIU/liter. Their mean LVMI was 92.9 ± 24.3 and 99.0 ± 32.8 g/m2, respectively, which was not significantly different from those in the reference range.
|
In the multiple linear regression model, gender was the strongest predictor of LVMI, followed by systolic blood pressure, BMI, and age, whereas serum TSH was not significantly related to LVMI (Table 2).
|
A total of 7954 subjects had serum TSH measurements in the fifth Tromsø study. Among these, 249 subjects with serum TSH 3.50–10.0 mIU/liter, 249 subjects with serum TSH 0.5–3.49 mmol/liter, and 76 subjects with serum TSH less than 0.5 mIU/liter were invited to the nested case-control study, and 167, 162, and 55, respectively, attended. Those on blood pressure medication, those who now had a serum TSH level outside the range from which they were invited, and those with serum free T4 or free T3 outside the reference range were excluded, leaving 66, 118, and 20 subjects in the three groups, respectively. Their demographics are given in Table 3. Those with serum TSH less than 0.50 mIU/liter were significantly younger than those in the control group.
|
) or any of the echocardiographic parameters of systolic and diastolic left ventricle function (Table 5).
|
|
regarding the PWTD data, those with serum TSH less than 0.50 mIU/liter had higher S-wave velocity at all four sites of the mitral annulus, compared with the control group. Even after adjustment for the covariates gender, age, BMI, and systolic blood pressure, the difference in the S-wave velocity was statistically significant at the anterior site and also in the combination of the four sites of the mitral annulus described as the global S. The A-wave velocities were significantly higher at the lateral and anterior sites and in the combined global A-wave velocity. The E-wave velocities were also higher in those with serum TSH less than 0.50 mIU/liter, but the differences did not reach statistical significance. There were no differences in the measured PWTD velocities between those with serum TSH within the 3.50–10.0 mIU/liter range, compared with the control group.
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
However, in our epidemiological study that included more than 2000 subjects, we could not find any significant association between serum TSH and LVMI after adjustment for age, BMI, and systolic blood pressure, which is in accordance with the report by Dorr et al. (14). In their study on 1510 individuals, the adjusted LVMI was almost identical in subjects with elevated, normal, and decreased serum TSH levels, whereas those with overt hyperthyroidism had an increased LVMI. Furthermore, in our nested case-control study after the fifth Tromsø study, which included 66 subjects with subclinical hypothyroidism, 118 euthyroid subjects, and 20 subjects with serum TSH less than 0.50 mIU/liter, the mean LVMI was almost the same in each serum TSH subgroup. For subclinical hypothyroidism this is in accordance with several other reports (7, 8, 28). However, some have found a moderately increased LVM (10), and there is even a report on a significantly reduced LVM in subclinical hypothyroidism (11).
On the other hand, most echocardiographic studies in subjects with subclinical hyperthyroidism have found an increased LVM (29, 30, 31). This was not found in our study, but it must be emphasized that only a few of our subjects had a suppressed serum TSH level. Furthermore, in the nested case-control study, those with serum TSH less than 0.50 mIU/liter had significantly higher serum free T4 levels than the euthyroid group, but the difference was only 2 pmol/liter, indicating that the thyroid function was only moderately elevated.
Similar to our findings regarding LVMI, we were not able to detect any significant relation between serum TSH and systolic and diastolic left ventricular function with conventional echocardiography. However, such a relation has been reported by Biondi et al. (28), who found signs of diastolic dysfunction with prolonged isovolumic relaxation time, increased mitral Doppler A wave, and reduced ratio between early and late diastolic mitral flow velocities in 26 subjects with subclinical hypothyroidism. A similar finding was also reported by Monzani et al. (10) in a study on 20 subjects and by Franzoni et al. (7) in a study on 42 subjects with subclinical hypothyroidism. In these three studies, the abnormalities were reversed with T4 substitution therapy. On the other hand, these changes in cardiac function could not be confirmed by Owen et al. (13), Tseng et al. (32), or Arem et al. (33).
In recent years a new tool, PWTD echocardiography, has emerged as a valuable supplement to conventional echocardiography (34, 35). Differences in PWTD-derived data indicating diastolic dysfunction has been reported in subjects with borderline hypothyroidism (36). Franzoni et al. (7) found in their PWTD study on 42 subjects with subclinical hypothyroidism lower E wave, higher A wave, and subsequently a lower E wave to A wave ratio. A similar study by Vitale et al. (9) found no difference in PWTD measured E and A wave but prolonged precontraction time, precontraction time/myocardial contraction time, and prolonged myocardial relaxation time as an indication of impaired cardiac function. Even fewer changes in these parameters were described by Arinc et al. (8), whereas Owen et al. (13) in this respect found no difference between subjects with subclinical hypothyroidism and euthyroid controls. In this context it should also be mentioned that Brenta et al. (37), using a different technique, radionuclide ventriculography, demonstrated left ventricular dysfunction that was reversible with T4 treatment in 10 subjects with subclinical hypothyroidism, whereas Bell et al. (38), using a similar technique, could not confirm these changes.
In our own study with PWTD measurements, we included 66 subjects with subclinical hypothyroidism and could not find any significant differences regarding diastolic or systolic function, compared with the euthyroid controls. Given that overt hypothyroidism is associated with diastolic dysfunction (6), there could be several reasons for the discrepancy between our negative observation and those who have found significant diastolic dysfunction in subclinical hypothyroidism. First of all, our subjects had a mean serum TSH level of 5.4 mIU/liter, which is lower than in most of the other studies on subclinical hypothyroidism cited above. Furthermore, our subjects were recruited from an epidemiological study and not from clinical practice in which the thyroid dysfunction might have been detected in subjects seeking medical help for diseases unrelated to hypo- or hyperthyroidism. Accordingly, our subjects were therefore less likely to have concomitant diseases that could affect the cardiac measurements. It should also be mentioned that the number of subjects included in most of these studies have been low, which makes it possible that publication bias, favoring studies with positive results, may have occurred (39).
Regarding our group with serum TSH less than 0.50 mIU/liter, the results from the PWTD measurements with higher velocities at nearly all measurement sites are in accordance with previous reports on hyperthyroidism (12, 29). It therefore appears that subclinical hyperthyroidism has an effect on cardiac function detectable by PWTD, but the clinical relevance of these higher velocities remains to be determined.
PWTD combined with conventional echocardiography is accepted as a good tool for detecting diastolic dysfunction with better sensitivity and specificity, compared with mitral Doppler data alone. It is important to consider diastolic dysfunction because of the increased morbidity and mortality reported in subjects with diastolic dysfunction (40). Further studies with PWTD are therefore needed in subjects with subclinical hyperthyroidism not only to evaluate progression of any diastolic dysfunction but also to see whether treatment may cause regression of the observed changes.
Our study has several weaknesses. If we had included more subjects in the nested case-control study, we might have found statistically significant differences between the subclinical thyroid disease groups and the euthyroid group regarding echocardiographic parameters of cardiac morphology and function. However, most of the nonsignificant differences were less than 3%, and, even if reflecting a true difference, probably of marginal clinical relevance. Similarly, we cannot rule out that we would have found statistically significant differences in the PWTD measurements between those with subclinical hypothyroidism and the euthyroid controls had we included more subjects. However, to include 66 subjects with subclinical hypothyroidism in the nested case-control study, we had to screen almost 8000 subjects, and a larger group would be hard to find. Furthermore, our subjects with subclinical hypothyroidism had a mean serum TSH level of 5.4 mIU/liter, and our negative results do not necessary apply to those with higher serum TSH levels. In addition, we excluded a considerable number of subjects because of concomitant diseases, and our results do therefore mostly apply to a fairly healthy population. It should also be emphasized that our negative findings regarding LVMI in those with reduced serum TSH levels must be viewed with caution because of the low number of subjects in that group. And finally, we measured serum TSH only in the epidemiological part of the study, and additional information might have been gained if free T4 and free T3 had also been included. The blood samples were drawn in the nonfasting state, and because thyroid hormones do have a circadian rhythm (41), some of the subjects may have been misclassified.
On the other hand, our study has considerable strength. The epidemiological study included more than 2000 subjects and is, to our knowledge, the largest so far reported on serum TSH and LVMI. In the nested case-control study, we used strict selection criteria during inclusion, and all subjects had a stable thyroid function with the serum TSH level elevated, normal, or low on at least two occasions before inclusion in the study.
In conclusion, we have found no significant relation between serum TSH and LVMI in an epidemiological study using conventional echocardiography. Furthermore, in subjects with subclinical hypothyroidism in which the serum TSH level is only slightly above the reference range, there appears to be no systolic or diastolic dysfunction when evaluated with PWTD in addition to conventional echocardiography. However, subclinical hyperthyroidism appears to be associated with increased PWTD velocities indicating left ventricle hyperfunction.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Disclosure Statement: The authors have nothing to disclose.
First Published Online June 12, 2007
Abbreviations: A wave, Atrial contraction; BMI, body mass index; E wave, early diastole; LVM, left ventricular mass; LVMI, LVM index; PWTD, pulsed-wave tissue Doppler.
Received March 29, 2007.
Accepted June 1, 2007.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  N. Rodondi, D. C. Bauer, A. R. Cappola, J. Cornuz, J. Robbins, L. P. Fried, P. W. Ladenson, E. Vittinghoff, J. S. Gottdiener, and A. B. Newman Subclinical Thyroid Dysfunction, Cardiac Function, and the Risk of Heart Failure: The Cardiovascular Health Study J. Am. Coll. Cardiol., September 30, 2008; 52(14): 1152 - 1159. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. R. Cappola Subclinical Thyroid Dysfunction and the Heart J. Clin. Endocrinol. Metab., September 1, 2007; 92(9): 3404 - 3405. [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
