Growing evidence suggests that reductions in circulating thyroid hormone (TH) levels, in particular reductions in biologically active T3, are commonly observed in patients with various acute and chronic cardiac etiologies, which may negatively influence prognosis. Therefore, the recently published report by Kim et al is of great interest. In their study, the investigators identified an independent association between subclinical myocardial injury, as assessed by elevated high-sensitivity cardiac troponin T, and low serum T3 levels in patients with chest pain and without clinically apparent coronary heart disease (CHD). We commend the investigators for their study and certainly recognize the potential clinical importance of characterizing serum thyroid alterations in cardiac patients. In fact, we believe that the aforementioned study nicely complements a previous report from our laboratory that demonstrated that serum T3 levels add prognostic information to conventional clinical and functional cardiac parameters in patients with both postischemic and nonischemic dilated cardiomyopathies. However, in its present form, we find that it is challenging to adequately draw independent conclusions from this report without further clarification.

First, the population examined in this study remains largely undefined from a cardiac perspective as the investigators did not elaborate on the diagnostic screening criteria used to identify the presence or absence of CHD in patients presenting with chest pain. No information was provided regarding electrocardiographic changes (ST elevation or downsloping, T-wave inversion, and Q-wave abnormalities), characteristics of the chest pain (typical vs atypical, acute vs chronic), regional and global left ventricular function, or cardiac event history. It would be enlightening if the investigators clarify whether criteria used to identify CHD was solely on the basis of cardiac enzyme elevations or if more accurate and validated techniques such as invasive coronary angiography or coronary computed tomography angiography were used to confirm the true absence of CHD in patients. Considering that angina is clinically a transitory and completely reversible phenomenon that can occur without myocardial damage or tissue necrosis, it would be invalid to assume that all patients presenting with chest pain symptoms but without elevated cardiac enzymes are truly free from CHD.

A second key clarification is needed as to why the investigators chose to exclude only patients with overt hyperthyroidism from their study population. We find it puzzling that this study did not account for (1) overt or subclinical hypothyroidism, (2) concomitant treatment with synthetic THs or antithyroid medications, or (3) drugs that alter thyroid metabolism (amiodarone, corticosteroids, glucocorticoids, dopamine, dobutamine) and worry that patients with pharmacologic thyroid alterations or severe TH abnormalities may provide disproportionate influence on the study findings and conclusions (selection bias). By only screening for hyperthyroidism, the investigators have likely created an artificial study population characterized by mean serum thyroid-stimulating hormone (TSH) levels greater than those typically observed in the general population and already meeting the diagnostic criteria for subclinical hypothyroidism (TSH 5.97 ± 15.0 mIU/L, median 1.92, interquartile range 1.02 to 3.31). The likelihood of far outlying patients skewing the findings of this investigation becomes even more apparent after the investigators categorize patients according to serum T3 levels (low T3 group: TSH 10.1 ± 20.9 mIU/L). These significant TSH elevations, coupled with concomitantly reduced serum T3 levels in this subgroup are strongly suggestive of a patient population composed of both mixed (overt and subclinical hypothyroidism, low T3 syndrome, exogenous thyroid alterations) and coexistent (primary hypothyroidism and low T3 syndrome) thyroid abnormalities. Moreover, the prognostic importance of the observed TSH elevations should not be overlooked considering the well-demonstrated relationship between elevated TSH values, especially those >10 mIU/L and risk for CHD events and mortality that occur irrespective of serum T3 or T4 alterations. Subsequently, we find it highly unlikely that the investigators can truly elucidate the relative diagnostic value of T3, as an isolated predictor of subclinical myocardial injury, considering the “mixed” and coexistent nature of TH abnormalities in this patient population.

It is essential to recognize that distinct peripheral pathophysiologic mechanisms and processes contribute to the altered serum thyroid profile observed in the low T3 syndrome (reduced T3, normal TSH, and FT4), as are commonly observed in cardiac patients, compared with other forms of primary thyroid illness (overt hypothyroidism: elevated TSH, reduced FT4; subclinical hypothyroidism: elevated TSH, normal FT4). This key distinction is especially important considering that the low T3 syndrome carries a separate and distinct relative risk for cardiac and all-cause mortality compared with other forms of altered serum thyroid homeostasis. Although we certainly recognize the merits and clinic importance of this investigation, it is our belief that more meaningful clinical insight could be drawn from the present study if the investigators comment on and/or reevaluated their data and findings after better controlling for the characteristics of their study population.