| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Editorial |
Division of Endocrinology, Diabetes, and Metabolism, Center for Clinical Epidemiology and Biostatistics, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania 19104-6021
Address all correspondence and requests for reprints to: Anne R. Cappola, M.D., Sc.M., Division of Endocrinology, Diabetes, and Metabolism, Center for Clinical Epidemiology and Biostatistics, University of Pennsylvania School of Medicine, 718 Blockley, 423 Guardian Drive, Philadelphia, Pennsylvania 19104-6021. E-mail: acappola{at}mail.med.upenn.edu.
There is biological plausibility to the premise that subclinical thyroid dysfunction may cause adverse cardiac consequences. Thyroid hormone has both inotropic and chronotropic effects, and it is clear from studies of individuals with overt thyroid disease that hyperthyroidism leads to increased heart rate, increased cardiac output, and decreased systemic vascular resistance, whereas hypothyroidism has the opposite effects (1). In the continuum of thyroid dysfunction, therefore, it is reasonable to propose that a dose-response effect exists, with more subtle cardiac impairment being present in less extreme degrees of thyroid dysfunction.
The key challenges are the measurement of these more subtle cardiac effects and, more importantly, determining their clinical relevance. Addressing these issues requires larger numbers of study participants and more precise measurement of the cardiac phenotype than in overt thyroid disease. It also requires careful consideration of the age and underlying cardiac status of the population studied. Mild perturbations in circulating thyroid hormone levels could either provide a tipping point for older individuals with marginal cardiac reserve or have negligible effects in the face of stronger, competing risk factors for cardiac dysfunction.
Multiple studies have examined the relationships between endogenous (2, 3) or exogenous subclinical hyperthyroidism (4, 5, 6), endogenous subclinical hypothyroidism (7, 8, 9, 10, 11, 12, 13), and echocardiographic measures of systolic and diastolic function. These data have consistently shown measurable adverse parameters, such as increased left ventricular mass index, in subclinical hyperthyroidism. There has been greater heterogeneity in studies of subclinical hypothyroidism. Two studies have shown no difference in left ventricular mass or function between individuals with and without subclinical hypothyroidism (7, 8), and, of the remaining studies that have detected systolic and/or diastolic abnormalities using echocardiography, no pair of these studies report the same pattern of abnormal parameters (9, 10, 11, 12, 13).
A related issue is the laboratory assessment of thyroid function and the definition of a normal TSH. Currently, TSH levels that define subclinical thyroid dysfunction are outside of population reference ranges and imply pituitary "dissatisfaction" with the ambient concentration of thyroid hormone. Given that the clinical consequences of these biochemical abnormalities are not well-defined for the cardiovascular system and there are no data from large, randomized trials to guide management of these conditions (14, 15), another approach is to define a treatment threshold at which clinical benefit would be expected from altering thyroid status, rather than aiming to "normalize" levels to reference range in everyone. This is particularly relevant given the frequency of overreplacement with thyroid hormone in the community, which is a common occurrence due to the narrow therapeutic window of levothyroxine (16, 17). To attain greater certainty about gradations of risk and the optimal TSH treatment threshold, it is useful to subdivide the broader categories of subclinical hyperthyroidism and hypothyroidism into more extreme (TSH < 0.1 mU/liter, TSH > 10 mU/liter, respectively) or less extreme TSH derangements.
In the current issue of the Journal of Clinical Endocrinology and Metabolism, Iqbal et al. (18) have examined the relationship between thyroid status and cardiac function in two analyses of data from the Tromsø studies, a series of five health surveys focusing on cardiovascular epidemiology in selected birth cohorts in this Norwegian municipality. The first analysis was performed in a subgroup of 2035 men and women in the fourth Tromsø study who had TSH measurement and echocardiography. This represents the largest published epidemiological study of the relationship between thyroid function and left ventricular mass index, and it demonstrates no statistically or clinically significant association, when examined by thyroid function category or in linear models. In this large study, there was a tradeoff between sample size and phenotypic precision because there was no measurement of free T4 levels in the study group, and echocardiographic measurements were determined from only one cardiac cycle. However, if anything, the thyroid categorization without free T4 levels would be expected to bias results toward positive findings, given the admixture of subclinical and overt thyroid dysfunction.
The second analysis was performed using data from a nested case-control study conducted after the fifth Tromsø study, in which there were important refinements of the study subjects’ phenotypes: participants’ thyroid function tests had to persist within their thyroid categories (subclinical hyperthyroid, euthyroid, or subclinical hypothyroid) on two occasions separated by 6 to 12 months. This requirement is a strength of the study design because it minimizes misclassification of thyroid status and improves clinical relevance, as practitioners are less interested in the cardiac effects of transient thyroid functional abnormalities. Moreover, the echocardiographic technique was more refined, with the inclusion of detailed analyses of diastolic function through pulsed wave tissue Doppler echocardiography, which permits assessment of the velocity of a selected myocardial region to provide quantitative measures of regional and global ventricular function. Again, no association was found between thyroid functional status and left ventricular mass index, or with other standard echocardiographic measures of left ventricular systolic and diastolic function. Using pulsed wave tissue Doppler echocardiography, study participants with persistent subclinical hypothyroidism (defined as a serum TSH level between 3.5 and 10.0 mU/liter with normal free T4 and free T3 levels) did not differ from the euthyroid group in any of the measures. However, subjects with persistent subclinical hyperthyroidism (defined as a TSH < 0.5 mU/liter and normal free T4 and free T3 levels) had statistically significantly higher global S wave and A wave velocities than the euthyroid group, although the magnitude of this difference was only 10%.
What are the clinical implications of these findings? Theoretically, the pulsed wave tissue Doppler echocardiography findings suggest impaired diastolic function in individuals with subclinical hyperthyroidism, which could represent the pathophysiological link to adverse clinical cardiac events, such as atrial fibrillation (19, 20). However, the clinical significance of a 10% difference in pulsed tissue wave Doppler findings is unknown. Furthermore, there was substantial overlap in the Doppler measurements between the subclinically hyperthyroid and euthyroid groups. At the level of the individual, therefore, there is unlikely to be a meaningful cutoff that would enable a physician to stratify patients who might be more likely to benefit from treatment to prevent adverse clinical sequelae.
The lack of echocardiographic findings in individuals with subclinical hypothyroidism in this study is somewhat surprising, given the many studies showing echocardiographic abnormalities, including some that employed pulsed wave tissue Doppler (12, 13). The study of Iqbal et al. (18) has excellent internal validity, with precise definitions of both the predictors and the outcomes, and good statistical power. The question then is, how generalizable are these results? Nearly 8000 subjects were initially screened to obtain the 66 individuals meeting all inclusion criteria for the subclinical hypothyroidism arm of the nested case-control study. The mean TSH level of study participants with subclinical hypothyroidism was 5.4 mU/liter; their mean age was 61 yr; and they were selected to have no underlying cardiac disease, including no use of antihypertensive medication. For individuals with these characteristics, the data suggest that there are no adverse effects of persistent subclinical hypothyroidism on cardiac function. In other studies that included participants with higher mean serum TSH levels and younger age, there were measurable differences in echocardiographic parameters between subclinically hypothyroid and euthyroid subjects (9, 10, 11, 12, 13). This discrepancy highlights the importance of refining the "at risk" population and the need to link abnormalities in these surrogate echocardiographic markers with adverse clinical cardiac events, such as congestive heart failure (21).
Whether to treat individuals with subclinical thyroid dysfunction remains problematic. Large, definitive randomized clinical trials of individuals with subclinical hyperthyroidism and subclinical hypothyroidism are ultimately the only way to fully address this controversy. In the absence of such data, clinical studies such as the Tromsø study should be conducted with the goal of characterizing the appropriate target populations for these trials and quantifying the expected benefits, cardiac or otherwise, that these specific subpopulations would be expected to receive from treatment.
Footnotes
Disclosure Statement: The author has nothing to disclose.
Received July 16, 2007.
Accepted July 20, 2007.
References
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]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
