Thyroid Dysfunction in Patients Admitted in Cardiac Care Unit: Prevalence, Characteristic and Hospitalization Outcomes

Introduction

Cardiovascular disease poses one of the heaviest burdens on any health-care system. It is responsible for more than 17 million deaths worldwide and is one of the main causes of hospital admission. The prevalence of ischemic cardiac disease and its complications is the highest. Thus, the spotlight has been shined on primary and secondary prevention and identifying conditions that increase the risk of cardiac events through both biochemical and clinical data.¹

Thyroid hormone has a major effect on the cardiovascular system.^2–4 Changes in thyroid hormone levels (either up or down) produce effects on heart contractility, oxygen consumption, and systemic vascular resistance.⁵ The diagnosis and management of thyroid disorders are crucial as most of these changes are reversible. Thyroid disorders are common, affecting around 9–15% of the adult population and being more common in females, although the rate becomes more equal between men and women at advanced ages.⁶

Patients with hyperthyroidism usually present with symptoms and signs that are more related to the cardiovascular system.⁷ A state of hyperthyroidism affects the hemodynamics of the heart and increases the heart rate, systolic blood pressure, cardiac output, contractility, and left ventricular ejection fraction (LVEF), while it decreases peripheral vascular resistance. In severe cases of hyperthyroidism, these changes may eventually lead to high-output heart failure. In fact, some use the term “thyrotoxic cardiomyopathy” to describe the damage that may happen because of a severe chronic hyperthyroid state.³

An increased incidence of atrial fibrillation with hyperthyroidism has been well identified. The range of incidence is 2–20%,⁸ with 13.8% occurring in cases of overt hyperthyroidism, as opposed to 2.3% in the general population,⁹ especially in the elderly and those with cardiac disease.^10,11 Hashimoto and Graves diseases are both noted to be more associated with mitral valve prolapse, which might be followed left atrium enlargement and atrial fibrillation.¹² Subclinical hyperthyroidism may increase risk for mortality, coronary artery disease, and atrial fibrillation.¹³ Low thyroid-stimulating hormone (TSH) was linked with more risk for atrial fibrillation,¹⁴ and mortality from cardiovascular diseases.¹⁵

In contrast, hypothyroidism manifests with decreased heart rate, cardiac contractility, and cardiac output, along with a mild increase in diastolic blood pressure due to the effect on the renin-angiotensin-aldosterone system.⁶ Although these changes rarely cause heart failure by themselves, they may lead to worsening clinical conditions in patients with heart failure, as well as worsening hospital outcomes.¹⁶ Furthermore, hypothyroidism has been identified to affect lipid metabolism, and multiple studies have found that it is associated with higher cholesterol and low-density lipoprotein (LDL) levels.¹⁷ The prevalence of subclinical hypothyroidism is 7–10% among older women, and it has an effect on lipid metabolism^17,18 with increased lipid and cholesterol levels, which appear to be parallel with serum TSH,^7,19 accelerated atherosclerosis, and coronary artery disease.²⁰

It has been found that 23.3% of the patients with acute myocardial infarction have Nonthyroidal Illness Syndrome (NTIS), which improves the predictive value of in-hospital cardiovascular deaths in such patients.²¹ It has also been linked in past and recent data with increased risk of all-cause mortality and major adverse cardiac events (MACE).²² Cardiovascular diseases are some of the major causes of hospital admission and the leading cause of death worldwide. Furthermore, thyroid dysfunction is a risk factor for cardiac disease, as well as mortality among patients with cardiac disease.^1,23 Thus, the aim of this study is to investigate the prevalence of thyroid hormone abnormality in cardiac care unit (CCU) patients and its relation to admission diagnosis, clinical, biochemical data, and hospital-related outcomes.

Patients and Methods

A retrospective cohort observational study was conducted in the Cardiology Department at King Abdul Aziz Specialist Hospital in Taif City, Saudi Arabia. A total of 411 patients were considered for enrollment. The patients were selected from those who were admitted to the CCU due to acute cardiovascular events or exacerbation of chronic cardiac condition between December 2015 and December 2018.

The inclusion criteria were age older than 18 years, both sexes, and no known thyroid disease. We excluded 37 patients because they either had a history of thyroid disease or received amiodarone or IV contrast before blood extraction. After the exclusion, 374 patients were included. The data collected included diagnosis at admission, past medical history of either underlying cardiac disease or other conditions that are considered as a risk factor for cardiac disease, special habits, length of hospitalization, electrocardiograms, echocardiographic evaluations, clinical and biochemical data, and hospitalization-related outcomes.

Laboratory investigations included triglyceride (TGL), high-density lipoprotein (HDL), LDL, total cholesterol, and thyroid function tests (free thyroxine (FT4), free triiodothyronine (FT3), and TSH). The levels were measured from venous blood samples extracted at admission. The results were interpreted by a board-certified endocrinologist, and patients were categorized according to their thyroid state (Table 1) and as it listed FT4 had no significance to reaching the diagnosis of NTIS as it can be falsely high, normal, or low in ill patients as an effect of received treatment or due to the illness itself.²⁴ Results that did not fit any of the mentioned conditions were classified as uninterpretable.

Table 1 Interpretation of Thyroid Function Test

Echocardiographic evaluation was conducted by a certified cardiologist (Philips CX50 Echocardiography). Information on left ventricular ejection fraction (LVEF), diastolic dysfunction, left ventricular hypertrophy (LVH), and valvular pathology were collected for all patients. Furthermore, we documented the presence of left or right bundle branch block through 12-lead electrocardiograms.

According to the diagnosis upon admission, patients were categorized into an ischemic group and non-ischemic group. We compared these groups’ baseline characteristics, thyroid state, underlying chronic diseases (both cardiac and non-cardiac), biochemical data, echocardiographic findings, and hospitalization-related outcomes. The study proposal was submitted to and approved by the Institutional Review Board (IRB) at Taif Health Affairs and the Medical Research Unit at King Abdul Aziz Specialist Hospital.

All qualitative and quantitative data were collected from patients’ medical records. The data were entered and coded through Microsoft Excel (Microsoft, Redmond, WA, USA) and analyzed using SPSS version 26 (SPSS Inc., Chicago, IL, USA). The clinical characteristics were compared for patients with and without thyroid dysfunction, and we defined statistical significance as p < 0.05. Chi-squared tests were performed for categorical variables, and unpaired t-tests were used for continuous variables.

Results

The 374 patients had a mean age of 62.7+14.7 years, were mostly Saudi males, and had a mean hospital stay of 4.0+4.4 days (Table 2). One-third of the sample was active/former smokers, and around 5% of them had right or left bundle branch block on electrocardiograms. The most common admission diagnosis was acute coronary syndrome, followed by heart failure. The least common diagnosis was valvular, congenital, and acute aortic syndrome. The most common thyroid diagnosis based on the thyroid function test at admission was euthyroidism, followed by NTIS, and the least common ones were T3 toxicosis and subclinical hyperthyroidism.

Table 2 Baseline Characteristics of the Whole Cohort

The most common chronic diseases were hypertension, followed by diabetes mellitus (DM), and the least common were peripheral artery disease and non-ischemic cardiomyopathy. The mean results of the thyroid function test and lipid profile were in the normal/controlled range. Regarding the echocardiographic findings, one-third of the patients had LVH, while more than half of them had diastolic dysfunction and mitral regurgitation. More than half of the patients were referred to a tertiary center for intervention, and the in-hospital mortality rate was 5.9%.

Ischemic changes were observed in 70.6% of the patients based on the admission diagnosis (Table 3). Compared to the non-ischemic group, the ischemic group was more likely to be male (P=0.010), to have a shorter hospital stay (P=0.012), to be active/former smokers (P=0.011), to have higher oxygen saturation (P=0.001), to have DM (P=0.009), to have higher total cholesterol levels (P=<0.001), to have higher LDL levels (P=0.001), to have higher TGL levels (P=0.007), to have diastolic dysfunction (P=<0.001), to have undergone thrombolysis (P=<0.001), and to be referred to a tertiary center (P=<0.001). The ischemic group was less likely to have chronic kidney disease (P=<0.001), heart failure (P=<0.001), atrial fibrillation (P=<0.001), cardiomyopathy (P=<0.001), mitral regurgitation (P=0.030), tricuspid regurgitation (P=0.006), and aortic stenosis (P=0.029) when compared to the non-ischemic group.

Table 3 Baseline Characteristics Based on the Admission Diagnosis Category

Euthyroidism was observed in 57.8% of the patients based on the thyroid function test at admission (Table 4). Compared to those with thyroid dysfunction, those with euthyroidism were more likely to be active/former smokers (P=0.002), to have lower heart rates (P=0.018), to not have chronic kidney disease (P=0.016), to not have heart failure (P=0.006), to have lower TSH levels (P=<0.001), to have lower tricuspid regurgitation (P=0.042), and to have lower aortic regurgitation (P=0.007). The highest TSH level was observed in those with heart failure and acute aortic syndrome, while the lowest was observed in those with pericardial disease (Figure 1).

Table 4 Baseline Characteristics Based on the Admission Thyroid Function Test

Figure 1 The mean TSH according to the admission diagnosis.

Partial correlation analyses were done with adjustments for gender, age, total cholesterol, HDL, TGL, LDL, systolic and diastolic blood pressure, heart rate, oxygen saturation, DM, hypertension, cerebrovascular accident, peripheral vascular disease, chronic kidney disease, and end-stage renal disease and smoking history. The results showed non-significant positive correlations between TSH and each of the following: 30-day readmission (r=0.15, P=0.815), in-hospital mortality (r=0.026, P=0.679), referral to a tertiary center (r=0.058, P=0.349), and length of hospital stay (r=0.034, P=0.590).

Discussion

This study showed that most patients were male with a mean age of 62.7+14.7 years. A recent study showed similar sex and mean-age distributions to those observed in our study.²⁵ Around 5% of the patients in our study had right and left bundle branch blocks, which is like a previous study.²⁶ The prevalence of the left bundle branch block was higher in the previous study, and this difference is likely due to the lower percentage of the heart failure patients in our study.²⁷

The most common diagnosis on admission was acute coronary syndrome, which had a prevalence of 69.3%. This is consistent with a recently published study.²⁸ NTIS was only present in 15% of the patients which is similar to previous studies,^29,30 while euthyroidism was prevalent in 57.8%. A recent study on critically ill patients showed that 84% of the patients having altered thyroid function.³¹ Most patients had DM and hypertension, and half of the sample had diastolic dysfunction. A recent systemic review showed that 48% of the hospitalized patients had diastolic dysfunction, which is consistent with our findings.³² The in-hospital mortality rate in our study was 5.9%, which is slightly higher than reported in a previous study from the United States, which showed an average annual rate of 3.4%.³³ This difference is likely related to the study design and duration, along with the lack of facilities like cardiac catheterization units at our hospital.

In our study, 70% of the patients had ischemic changes, which were more common among males. This concurs with a previous study with a similar age distribution.³⁴ Our study showed that 53.2% of the patients had diastolic dysfunction, which is also like other studies.^35,36 Valvular disease was more prevalent in the non-ischemic group, as reported previously.³⁷ This is likely explained by those with valvular disease presenting with heart failure rather than ischemic changes.

Our study showed that TSH is positively associated with 30-day hospital readmission, in-hospital mortality, referral to a tertiary center, and length of hospital stay. A recent study showed that TSH more than 4 is correlated with worse hospital outcomes.³⁸ Another recent study showed that women with high TSH were at significantly higher risk of increased mortality from coronary artery disease.³⁹ Most of our patients were male, and this more likely why our result was non-significant.

Conclusion

Thyroid dysfunction is common among patients admitted to the CCU, among which NTIS was been the most common diagnosis. Non-significant positive correlations between TSH and hospitalization length, tertiary center referral, 30-day readmission, and in-hospital mortality when adjusting for potential confounders.