Hyperthyroidism and its impact on cardiovascular disease : with special emphasis on atrial fibrillation

Abstract

Background: Hyperthyroidism is a common disease throughout the world, affecting 0.5%-2% of women and one-tenth as many men at some time. Graves’ disease and toxic nodular goitre are the two most common etiologies. Cardiovascular symptoms are often prominent, and atrial fibrillation (AF) is a well-known complication. Several large register studies have indicated that patients treated for hyperthyroidism may be at increased risk of death compared to the general population; it appears that this difference in mortality is chiefly attributable to cardiovascular and endocrine diseases. However, these previous studies are quite heterogenous with sometimes contradictory results. Subclinical hyperthyroidism (SH) is a condition afflicting approximately 1% of the population, defined as serum levels of thyroid-stimulating hormone (TSH) below the normal range while levels of triiodothyroxine (T3) and thyroxine (T4) are within the reference range. Symptoms are similar to, but milder than those seen in regular hyperthyroidism. AF, in particular, appears to occur more often when SH is present.

Aims: The aim of this thesis is to further investigate the correlation between hyperthyroidism and cardiovascular disease. Using a very large cohort based on register data, long-term cardiovascular mortality and morbidity (Paper I) were assessed in patients treated for hyperthyroidism. Another study addressed differences in long-term effects between treatments for hyperthyroidism (Paper II). In the last article, we attempted to find out whether subclinical hyperthyroidism might be a common cause of AF (Paper III).

Methods: Papers I-II were both based on essentially the same data on individuals with thyroid disease: Information on patients who had undergone thyroidectomy (complete or partial removal of the thyroid gland) was gathered from the Swedish National Patient Register, and information on patients treated with radioactive iodine was based on a material consisting of local hospital records. Data on cardiovascular outcomes and death was collected from the Patient Register, the Causes of Death Register, and Statistics Sweden. The Swedish Prescription Register was also used for assessment of levothyroxine treatment. Study subjects had undergone either surgery or radioiodine treatment between the years 1976-2000, and the cohort was followed until 2012. In Paper I, patients treated for nontoxic nodular goitre were used as a control group, but comparisons were also made to the general population of Stockholm. Outcomes were assessed in terms of all-cause mortality, cardiovascular mortality, and cardiovascular morbidity. In Paper II, which included only patients with hyperthyroidism, thyroidectomy was compared to radioiodine treatment in terms of all-cause and cause-specific mortality. Three different statistical methods were applied: Cox regression, propensity score matching, and inverse probability matching. Paper III was a cross-sectional observational study in which thyroid status was assessed among patients who underwent electrophysiological ablation for cardiac arrhythmias. Cases of AF were compared to controls with AV-nodal re-entry tachycardia with regard to thyroid hormone levels. The hypothesis was that AF would correlate to a higher prevalence of subclinical hyperthyroidism, defined as a suppression of TSH levels and normal free T4 levels, compared to controls.

Results: In Paper I, an increased risk of all-cause mortality (hazard ratio (HR) 1.27 with a 95% confidence interval (CI) of 1.20-1.35) was found among the 12,239 patients treated for either Graves’ disease or toxic nodular goitre, compared to the 3,685 patients treated for nontoxic goitre. Increased cardiovascular mortality (HR 1.29, CI 1.17-1.42) and cardiovascular morbidity (HR 1.12, CI 1.06-1.18) was also seen, with AF being by far the most common finding at early follow-up. Furthermore, all of these outcomes were significantly more common in comparisons with the general population. The risk of death and cardiovascular disease was most evident in patients treated for toxic nodular goitre while only a weak association was seen among Graves’ disease patients. No decrease in risk was found among subjects included later (after 1990) compared to those included earlier. Paper II included 10,992 subjects with hyperthyroidism; 10,250 had been treated with radioiodine and 742 with thyroidectomy. Surgically treated subjects had lower all-cause mortality as assessed by cox regression (HR 0.82, CI 0.71 -0.96), propensity score matching (HR 0.80, CI 0.68-0.94), and inverse probability weighting (0.85, CI 0.72-1.00, p = 0.044) compared to subjects who received radioiodine. Significantly lower cardiovascular mortality was also found among thyroidectomised individuals in all three analyses while no clear differences were found regarding cancer mortality or other causes of death. In Paper III, 312 patients were included. Of these, 212 had AF, and 100 had AV-nodal reentry tachycardia. Analyses showed that subclinical hyperthyroidism was not more common among subjects with AF than among control subjects. However, it was found that levels of free T4 were significantly higher in the AF group (CI of difference 0.03-1-35, p = 0.039).

Conclusions: Hyperthyroidism is associated with an increased risk of cardiovascular disease and death. The mechanisms behind this observation remain unknown, but some factors – notably hyperthyroidism due to toxic nodular goitre and treatment with radioiodine – seem to represent a stronger association. As for the specific diagnosis of AF, we found no clear evidence to support that subclinical hyperthyroidism might be one of the predominant underlying causes, although we did make the interesting observation that levels of free T4 were significantly elevated among AF patients.