https://orcid.org/0000-0002-5604-9378
Figures
Abstract
Background
Thyroid hormones act on the cardiovascular system directly by modulating its function and indirectly by transcriptional regulation of gene expression in the heart and the vasculature. Studies have shown associations between overt and subclinical thyroid disorders and cardiovascular outcomes. The aim of this study was to perform a systematic review and meta-analysis to assess the potential relationships between subclinical hyper- and hypothyroidism and risk of atrial fibrillation (AF), and post-operative AF.
Methods
MEDLINE and Scopus databases were searched from inception to 18th February 2023 for randomised controlled trials, case-control studies, and cohort studies which assessed the relationship between subclinical thyroid dysfunction and incident AF events. Risk of bias and the quality of evidence were assessed using the RoBANS tool and GRADE approach, respectively. Meta-analysis was conducted in Review Manager 5.4 using the Mantel-Haenszel statistical method and a random-effects model. Data are presented as risk ratios with 95% confidence intervals. Statistical heterogeneity amongst studies was assessed by the chi-squared (χ2) test and I2 statistic. p≤0.05 were considered significant.
Results
A total of 6467 records were identified, of which 10 cohort studies met the inclusion criteria. Both subclinical hyperthyroidism and subclinical hypothyroidism were associated with an increased risk of incident AF (risk ratio (RR), 1.99; 95% confidence interval (CI), 1.43–2.77; n = 5 studies; p<0.0001 and RR, 1.19; CI, 1.03–1.39; n = 7 studies; p = 0.02, respectively). Subgroup analysis for post-operative AF revealed marked heterogeneity between studies (I2 = 84%) and association with subclinical hypothyroidism was not significant (RR, 1.41; CI, 0.89–2.22; n = 3 studies; p = 0.15).
Citation: Singh H, Shahid MZ, Harrison SL, Lane DA, Lip GYH, Logantha SJRJ (2024) Subclinical thyroid dysfunction and the risk of incident atrial fibrillation: A systematic review and meta-analysis. PLoS ONE 19(1): e0296413. https://doi.org/10.1371/journal.pone.0296413
Editor: Gaetano Santulli, Albert Einstein College of Medicine, UNITED STATES
Received: September 7, 2023; Accepted: December 12, 2023; Published: January 2, 2024
Copyright: © 2024 Singh et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the manuscript and its Supporting information files.
Funding: The author(s) received no specific funding for this work.
Competing interests: I have read the journal’s policy and the authors of this manuscript have the following competing interests: SLH has received an investigator-initiated grant from Bristol-Myers Squibb (BMS). DAL has received investigator-initiated educational grants from BMS and Pfizer, has been a speaker for Bayer, Boehringer Ingelheim, and BMS/Pfizer and has consulted for BMS and Boehringer Ingelheim; all outside the submitted work. GYHL has received consultancy and speaker fees for BMS/Pfizer, Boehringer Ingelheim, and Daiichi-Sankyo. No fees are received personally. This does not alter our adherence to PLOS ONE policies on sharing data and materials.
Introduction
Thyroid dysfunction is a common endocrine disorder known to exacerbate several cardiovascular pathologies including arrhythmia, dyslipidaemia, and heart failure [1]. Overt hyperthyroidism is a well-established risk factor for the development of atrial fibrillation (AF) and overt hypothyroidism is associated with ventricular arrhythmias [2, 3]. However, there is a lack of consensus on the associations between mild, early forms of thyroid dysfunction (subclinical) and AF.
Subclinical thyroid dysfunction is a primary disease of the thyroid glad defined by abnormal serum thyroid stimulating hormone (TSH) and normal free tetraiodothyronine (T4) and free triiodothyronine (T3) levels [4]. Subclinical hyperthyroidism (low TSH and normal T4) and subclinical hypothyroidism (high TSH and normal T4) affect 4–10% and 2% of the general population, respectively [5]. Women and older adults are primarily affected [6]. About 5% of the European population with subclinical thyroid dysfunction remains undiagnosed [7, 8]. The National Institiute for Health and Care Excellence (NICE) guidelines in the UK recommend treatment for subclinical hypothyroidism with levothyroxine if TSH levels are consistently >10 mIU/L or if consistently above normal reference range with symptoms [9]. Currently, no standard treatment is recommended for subclinical hyperthyroidism [9]. Additionally, suboptimal treatment for thyroid dysfunction is common and increases risk of incident AF, stroke, and cardiovascular mortality [10, 11]. In this context, a systematic review and meta-analysis to identify potential associations between subclinical thyroid dysfunction and AF would be valuable for framing guidelines and evolving clinical practice.
AF is the most common arrhythmia and in 2017 there were 37.57 million prevalent cases and 3.05 million incident cases of AF globally [12]. AF is associated with an increased risk of mortality [13], dementia [14], and stroke and heart failure [15, 16]. Post-operative AF affects 40% of cardiac surgery patients, more than doubles stroke risk, and increases the length of hospital stay [17, 18]. Mortality during admission and 6 months post-operatively is significantly increased in patients with post-operative AF [19]. Hence, there is a significant health and financial burden attached to developing AF per se, and post-operatively.
Several studies have demonstrated the effects of subclinical thyroid dysfunction on adverse cardiovascular outcomes, including coronary heart disease [20], atherosclerosis and myocardial infarction [21], and heart failure [22]. Given the associations between thyroid dysfunction and cardiovascular outcomes and the potentially modifiable nature of thyroid dysfunction, we performed a systematic review to determine whether subclinical thyroid dysfunction is associated with a higher risk of incident AF per se and in post-operative patients.
Methods
This systematic review was prospectively registered on PROSPERO (CRD42020221565) and reported in accordance with Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.
Information sources and search strategy
MEDLINE and Scopus databases were searched from inception to 18th February 2023 without language restrictions. The search terms used in the databases as keywords were: “subclinical hyperthyroidism”, “subclinical hypothyroidism”, “subclinical thyroid”, “thyroid dysfunction”, “thyroid”, “atrial fibrillation”, and “arrhythmia”. These were combined using the Boolean operators “AND” and “OR”. The full search strategy for both databases is provided in S1 Table. Reference lists of included studies were searched for additional relevant literature. Results were imported into EndNote X9 (Clarivate Analytics), and duplicate records removed.
Eligibility criteria
Studies which measured AF incidence in patients with subclinical thyroid dysfunction were included. Systematic reviews were screened for additional studies relevant to the study question. The following inclusion criteria applied: 1) cohort studies, case-control studies, RCTs, systematic reviews and 2) thyroid function tests, including serum TSH and T4, taken at baseline to determine subclinical thyroid dysfunction. The exclusion criteria were: 1) studies involving non-human subjects 2) no measurement of serum TSH and T4 at baseline prior to AF diagnosis 3) AF diagnosed prior to index hospitalisation 4) not including AF as an outcome and 5) review articles or cross-sectional studies.
Studies were included if it could be determined which participants did not have prevalent AF at baseline. Cross-sectional studies and observational studies where thyroid hormone levels were measured after AF was diagnosed were excluded as incidence of new-onset AF could not be ascertained.
Study selection
Two reviewers (HS, MZS/SJRJL) independently screened a 25% sample of the records, 1266 records in total, based on their title and abstract. Agreement at this stage was 99.6% and disagreements were resolved through discussion. As per the PROSPERO protocol, with agreement >80%, the primary reviewer (HS) screened the remaining records to determine inclusion for full text screening. Studies from previous systematic reviews were extracted during the selection process. The same two reviewers (HS, MZS/SJRJL) independently screened the full texts to determine final inclusion for synthesis. Agreement at full-text screening was 94.1% and disagreements were resolved through discussion. Studies in which a euthyroid control group was included and hence a risk ratio (RR) could be calculated were selected for meta-analysis.
Data extraction
The primary reviewer (HS) independently extracted data from the selected studies into a predesigned data extraction form on Microsoft Excel. The second reviewer (MZS) verified the extracted data; there were no disagreements. Data extracted included authors, year of publication, journal name, study design, number of participants, follow-up period, participant age, participant gender, TSH values used, method of AF ascertainment, previous history of cardiovascular disease, incidence of AF and the other outcomes measured. We contacted authors of six studies for further information; however, no data was provided.
Risk of bias of included studies
The risk of bias was assessed by two reviewers (HS, SJRJL) independently using the Risk of Bias Assessment tool for Non-Randomised Studies (RoBANS) [23]. Each study was assessed under the six RoBANS domains (selection of participants, confounding variables, exposure measurement, blinding of outcome assessment, incomplete outcome data and selective outcome reporting) and then the risk of bias was classified as low, high, or unclear as per pre-set criteria (S2 Table). The overall risk of bias for the studies was determined as low if ≤1 domain was classified as high or unclear risk, medium if 2 domains were classified as high or unclear risk, and high if ≥3 domains were classified as high or unclear risk. Non-consideration of the effect of baseline thyroid medication was considered a confounding factor. The risk of bias was assessed at the outcome level of AF and not for the entire study.
Assessing the quality of the evidence
Quality was independently assessed by two reviewers (HS, SJRJL) using the Grading of Recommendations, Assessment, Development and Evaluations (GRADE) [24]. This assessed the risk of bias, imprecision, inconsistency, indirectness, and publication bias of the studies. Outcomes were graded as high, moderate, low, or very low certainty. Summary of findings tables were generated for AF, non-post-operative AF, and post-operative AF, where applicable, using GRADEpro GDT [25].
Data synthesis and statistical analysis
Results of studies without a control group were expressed as the proportion with subclinical thyroid dysfunction who developed AF. Meta-analysis was conducted in RevMan, Version 5.4 [26]. The Mantel-Haenszel statistical method, random-effects analysis model was used, and the effect size is presented as RR with 95% confidence intervals (CI). Forest plots were used to depict the effect of subclinical thyroid dysfunction on incident AF. Statistical heterogeneity amongst studies was assessed by the chi-squared (χ2) test and I2 statistic using RevMan [26, 27] and p-values≤0.05 were considered significant.
Differences between protocol and review
In the prospectively registered PROSPERO protocol (CRD42020221565) we defined subclinical thyroid dysfunction as a TSH value outside the normal range of 0.45–4.50 mU/l and T3 and/or T4 values within their normal range. During the review we identified varying definitions for subclinical hyperthyroidism and subclinical hypothyroidism between studies. Excluding studies that did not strictly adhere to the TSH range specified in the PROSPERO protocol would have significantly reduced the total number of included studies and meta-analyses would not be possible. Therefore, this review included studies with varying definitions of subclinical thyroid dysfunction, as defined by the respective study.
Pre-planned subgroup analyses examining gender and subclinical thyroid dysfunction and timing of the onset of incident AF could not be undertaken due to lack of sufficient data. A subgroup analysis for incidence of post-operative AF and non-post-operative AF was performed. Sensitivity analysis was performed to identify sources of heterogeneity.
Results
Study selection
MEDLINE and Scopus searches returned 6467 records and after removing 1405 duplicates, 5062 records were screened based on title and abstract, with 5021 (99%) records excluded for either not fulfilling inclusion criteria or not being relevant to the review question. Of the 41 remaining records, seven were systematic reviews, of which one additional study was identified. Thus, full text articles of 42 studies were assessed and 10 studies were included (Fig 1) [28–37].
Study selection process for this systematic review.
Study characteristics
Study characteristics are summarised in Table 1. Of the 602,257 participants in the identified studies, 596,358 participants were included in this systematic review; 6,732 participants had subclinical hyperthyroidism, 13,628 had subclinical hypothyroidism and 575,998 were euthyroid. There were 19,263 cases of incident AF. Follow-up length ranged from post-operative [30, 35, 37] to 13.2 years [36]. Mean age ranged from 48.9 to 75.3 years [33, 34]; mean participant age was >65 years in five studies [29–31, 33, 35]. All included studies were cohort studies. Three studies involved patients undergoing cardiac surgery [30, 35, 37]; patients in Park et al. [30] and Zhao et al. [37] underwent coronary artery bypass graft (CABG) surgery and those in Martínez-Comendador et al. [35] had undergone aortic valve replacement procedure. For subclinical hyperthyroidism, TSH values used in these studies were <0.10 [28] to <0.45 [33] mIU/L and for subclinical hypothyroidism >3.40 [36] to >5 [34] mIU/L. The British Thyroid Foundation normal range for TSH is 0.4–4.0 mIU/L and those for free T4 and free T3 are 9.0–25.0 pmol/L and 3.5–7.8 pmol/L, respectively [38].
Subclinical hyperthyroidism and incidence of AF
The effect of subclinical hyperthyroidism on the incidence of AF was assessed in seven studies [28, 29, 31–34, 36]. Among 6,732 patients with subclinical hyperthyroidism in these studies, 484 cases of incident AF were reported. Two studies [31, 32], did not have a control euthyroid group, 2.94% and 2.78% of patients in these two studies developed AF at follow-up, respectively [31, 32]. The remaining five studies were eligible for meta-analysis and the risk of incident AF in patients with subclinical hyperthyroidism relative to euthyroid (RR) was 1.99 (95% CI, 1.43–2.77; p<0.0001) (Fig 2) [28, 29, 33, 34, 36]. There was substantial heterogeneity between studies (I2 = 67%; see section on sensitivity analysis).
Risk ratios for individual studies and the overall effect sizes are shown. The size of blue squares corresponds with the weighting of the study and bars represent 95% confidence intervals. Risk ratio expresses the risk of atrial fibrillation (AF) occurring in patients with subclinical hyperthyroidism relative to euthyroid.
Subclinical hypothyroidism and incidence of AF
Seven studies assessed the effect of subclinical hypothyroidism on the incidence of AF [29, 30, 33–37]; 727 cases of incident AF were reported among 13,628 patients with subclinical hypothyroidism. Meta-analysis (7 studies) revealed a pooled RR 1.19 (95% CI, 1.03–1.39; p = 0.02) for incident AF (Fig 3), although substantial heterogeneity was evident (I2 = 62%). Subgroup analysis was performed for non-post-operative and post-operative AF. The RR for developing non-post-operative AF in patients with subclinical hypothyroidism was 1.12 and heterogeneity was low (95% CI, 1.04–1.22; n = 4 studies; p = 0.004; I2 = 0%) (Fig 3A). RR for post-operative AF was higher at 1.41 (95% CI, 0.89–2.22; n = 3 studies; p = 0.15; I2 = 84%) (Fig 3B), although this was not statistically significant and with high heterogeneity. The study by Martinez-Comendador et al. involved patients undergoing aortic valve replacement surgery and the reported RR was 1.91 (95% CI, 1.39–2.64; p<0.0001). Park et al. and Zhao et al. involved patients undergoing CABG procedure and their combined RR was 1.18 (95% CI, 0.78–1.80; p = 0.43; I2 = 67%).
Risk ratios for non-post-operative (A) and post-operative (B) atrial fibrillation (AF) in this meta-analysis and overall effect sizes are shown. The size of blue squares corresponds with the weighting of the study and bars represent 95% confidence intervals. Risk ratio expresses risk of AF occurring in patients with subclinical hypothyroidism relative to euthyroid patients. The black diamonds denote effect sizes for non-postoperative and post-operative AF subgroups.
Sensitivity analysis
Sensitivity analysis was performed using the leave-one-out approach (Table 2). There was substantial heterogeneity in studies examining the relationship between subclinical hyperthyroidism and incident AF (I2 = 67%). The removal of Nanchen et al. [33] reduced the heterogeneity (I2 = 19%) but the results were consistent, albeit with a slightly higher risk of incident AF with subclinical hyperthyroidism, RR was 2.28 (95% CI, 1.96–2.65; n = 4 studies; p<0.00001). Heterogeneity was not observed in the assessment of the relationship between subclinical hypothyroidism and non-post-operative AF (I2 = 0%). In the sensitivity analysis for subclinical hypothyroidism and post-operative AF, removing Zhao et al. [37] significantly reduced the heterogeneity (I2 = 0%) and revealed a significant relationship between subclinical hypothyroidism and post-operative AF, RR was 1.76 (95% CI, 1.36–2.28; n = 2 studies; p<0.0001).
Risk of bias and quality of evidence
Using the RoBANS tool and a classification criterion (S2 Table), six studies were assessed as high risk across three RoBANS domains [28–30, 33–35]. One study each was deemed high risk under ‘selection of participants’ and ‘confounding variables’ domains due to small sample size (<100 participants) [28] and inclusion of patients taking thyroid hormone supplement at baseline, respectively [33]. Two studies were classed as unclear risk under ‘confounding variables’ domain as there was no mention of whether patients were taking thyroid medication at baseline [28, 35]. Five studies were rated high risk under the ‘exposure measurement’ domain as they did not measure free T3 levels [29, 30, 33–35]; and one study was assessed unclear risk as free T4 and TSH were measured, but there was no indication of whether T3 was measured [36]. The overall risk of bias assessment identified seven studies [29–32, 34, 36] with low overall risk and three studies [28, 33, 35] with medium overall risk of bias (Table 3). Funnel plots for studies investigating the relationship between subclinical thyroid dysfunction and incident atrial fibrillation are shown (S1 Fig).
The quality of the evidence regarding the effects of subclinical hyperthyroidism and subclinical hypothyroidism on AF was low as all included studies were observational studies with a prospective or retrospective cohort design. The GRADE summary of findings is presented in Table 4.
Discussion
This systematic review assessing the incidence of AF in patients with subclinical thyroid dysfunction identified 10 studies [28–37], all were cohort studies. Meta-analyses revealed that both subclinical hyperthyroidism and subclinical hypothyroidism increased the risk of incident AF by 99% and 19%, respectively. Subclinical hypothyroidism was not found to significantly increase the risk of incident post-operative AF; however, there was considerable heterogeneity and variation in surgical procedures. The quality of current evidence was low for both subclinical hyper- and hypothyroidism, despite low risk of bias in studies.
Patients with subclinical hyperthyroidism had an increased risk of incident AF compared to patients who were euthyroid. Heterogeneity was significant and sensitivity analyses identified one study [33] as the source of heterogeneity; removal of this study also demonstrated an increased risk of incident AF associated with subclinical hyperthyroidism. This study included patients taking thyroid hormone supplements at baseline, which may lead to a reduced risk of incident AF [33]. All other studies in this meta-analysis had either not reported or had excluded patients taking thyroid medication at baseline.
In a systematic review of prospective cohort studies by Sun et al., the relationship between subclinical thyroid dysfunction and risk of various cardiovascular outcomes were assessed [39]. Subclinical hyperthyroidism was associated with increased risk of incident AF (RR, 1.42; 95% CI, 0.69–2.92; p = 0.025; I2 = 72.9%). However, their meta-analysis was limited by small sample size (3 studies with one study being unpublished data), wider 95% CI and significant heterogeneity [39]. Our analysis is updated and includes a larger number of studies (n = 5 including the two published studies analysed by Sun et al.), with tighter 95% CI (1.43–2.77). Additionally, our finding that subclinical hyperthyroidism increased the risk of incident AF is consistent with a 2012 systematic review by Collet et al. that reported a hazard ratio of 1.68 (95% CI; 1.16–2.43) [40].
The association between subclinical hyperthyroidism and AF was recently reviewed [41]. This narrative review included cross-sectional studies and case-controls studies in which thyroid hormone levels were measured after AF diagnosis [41]. Our PROSPERO protocol stipulated within the inclusion criteria that thyroid status was determined at baseline before the AF event to determine whether subclinical thyroid status leads to AF. All included studies were cohort studies, and a causal relationship could not be determined. Overall, our analysis showed that in longitudinal observational studies patients with subclinical hyperthyroidism are more likely to have incident AF than euthyroid patients.
The present systematic review demonstrates that patients with subclinical hypothyroidism also have increased risk of AF, although substantial heterogeneity was evident. A subgroup analysis for non-post-operative and post-operative AF eliminated heterogeneity (I2 = 0) in the non-post-operative AF group and demonstrated a slightly increased risk of incident AF in this group. In contrast, two previous systematic reviews assessing the relationship between subclinical hypothyroidism and non-post-operative incident AF did not report significant associations [39, 42]. One [42] was an individual patient data analysis of prospective cohorts and excluded patients taking thyroid hormone at baseline, whilst the other [39] was limited by inclusion of only two studies. The present analyses included prospective [28–31, 33, 35, 36] and retrospective [32, 34, 37] cohort data and patients taking thyroid hormone at baseline, which may explain the differences in RRs reported.
To date this is the first systematic review and meta-analysis to have assessed the relationship between subclinical hypothyroidism and incident post-operative AF. AF presenting post-operatively is mostly transient [18, 30, 35], but is associated with prolonged hospital stays, an 8-fold increased risk of developing AF, and a doubling of long-term cardiovascular mortality [43, 44]. The present analysis showed no significant difference in the overall risk of post-operative AF associated with subclinical hypothyroidism, however there was significant heterogeneity. Among the different surgical procedures, risk of incident AF was significantly elevated post-aortic valve replacement [35], but two studies of CABG patients [30, 37] reported conflicting results. The prospective cohort study [30] (n = 260) reported a significant risk of post0operative incident AF [30], while the other retrospective cohort study [37], found no difference. The strength of prospective cohort study is the accuracy of data collection (for e.g., exposures, confounders and endpoints), with fewer potential sources of bias and confounders, prospective cohort design is typically ranked higher than retrospective cohort design in the hierarchy of evidence [45]. The conflicting results about subclinical hypothyroidism and post-operative AF might also reflect differences in the patient demographics and geographical location of the population studied. Additionally, participants in the prospective study had multiple thyroid hormone tests before and after CABG surgery [30], whereas only one pre-surgery blood test result was used in the retrospective study [37], which may have resulted in a misclassification bias. Sensitivity analyses identified the retrospective study as a source of heterogeneity; removal of this study revealed a significant association between subclinical hypothyroidism and increased risk of post-operative AF. Short term changes in levels of circulating thyroid hormones occur during cardiac surgery because of the surgical insult and may alter atrial electrophysiology leading to post-operative AF [46]. Some evidence suggests a role for postoperative T3/T4 treatment to prevent post-operative AF in cardiac surgery patients [46–48].
In the current systematic review, all included studies were observational and confounding factors were not consistently adjusted for. Therefore, it is possible that confounders including sex, age, ethnicity, pre-existing co-morbidities, and other unmeasured confounders could have affected AF incidence. Two of the 10 included studies had a mean/median follow-up time >10 years [29, 36]; longer follow-up times can result in higher AF incidence [49]. For subclinical hyperthyroidism, the results of these studies were not dissimilar to those of other studies with shorter follow time periods (2 to 5.5 years) [28, 33, 34]. For subclinical hypothyroidism, one of the two studies with >10-year follow-up [36] demonstrated the highest risk for incident AF (non-post-operative). However, the other study [29] reported AF incidence that was similar to those with shorter follow-up times (3 to 5.5 years) [33, 34]. The difference in mean participant age between the two studies (50.9 vs. 72.7 years) is a likely confounding factor [49, 50]. Currently, there is insufficient enough evidence to determine conclusively if there is increased risk of incident AF in patients with subclinical hypothyroidism over a longer follow-up period.
A major strength of our systematic review is the comprehensive literature search from database inception, without language restrictions, for RCTs, case-control, and cohort studies. Previous systematic reviews have only included prospective cohort studies [39, 42]. Our review was registered with PROSPERO which involved prospective submission and publication of information relating to the design and conduct of a systematic review. We had a large participant size which increased the precision of the results. Robust methodology was employed with two reviewers independently screening for eligible studies and dual data extraction. Further, this review utilised the GRADE approach to assess the quality of the evidence, which demonstrated low quality evidence. Furthermore, we assessed the risk of incident AF in both subclinical hyper- and hypothyroidism using a uniform approach.
One limitation of this review was the varying definitions used for subclinical thyroid status in the included studies would likely have resulted in some participants being assigned to euthyroid groups in certain studies but to subclinical thyroid dysfunction groups in other studies. An individual patient data analysis would have avoided this problem but was not conducted due to non-availability of data. AF can be paroxysmal, and it is plausible that in patients with subclinical thyroid dysfunction, undiagnosed AF may have pre-existed, and the recorded incident events may not have been new-onset AF.
Conclusions
This systematic review and meta-analysis suggests that both subclinical hyperthyroidism and subclinical hypothyroidism are associated with increased risk of incident AF; however, the quality of the available evidence is low. Future studies should have consistency in the definition of subclinical hyper- and hypothyroidism, RCT design, large sample size and long follow-up to allow assessment of the effect of treatment on the incidence of AF and other cardiovascular outcomes in patients with subclinical thyroid dysfunction.
Supporting information
S1 Table. Search strategy for Scopus and MEDLINE databases.
https://doi.org/10.1371/journal.pone.0296413.s001
(PDF)
S2 Table. Risk of bias classification criteria for RoBANS domains.
https://doi.org/10.1371/journal.pone.0296413.s002
(PDF)
S1 Fig. Funnel plot for studies on the relationship between subclinical thyroid dysfunction and incident atrial fibrillation.
(A) Subclinical hyperthyroidism and incident atrial fibrillation (AF), (B) subclinical hypothyroidism and incident AF and post-operative AF. The x-axis represents the risk ratio (RR) and y-axis points to the standard error of the RR on a logarithmic scale. Data for non-post-operative AF (black circle) and post-operative AF (red diamond) are shown.
https://doi.org/10.1371/journal.pone.0296413.s003
(PDF)
Acknowledgments
This work was supported by the MRes Clinical Sciences programme in the Institute of Life Course and Medical Sciences at The University of Liverpool.
References
- 1. Cappola AR, Desai AS, Medici M, Cooper LS, Egan D, Sopko G, et al. Thyroid and cardiovascular disease: research agenda for enhancing knowledge, prevention, and treatment. Circulation. 2019;139(25):2892–2909. pmid:31081673
- 2. Reddy V, Taha W, Kundumadam S, Khan M. Atrial fibrillation and hyperthyroidism: a literature review. Indian Heart J. 2017;69(4):545–550. pmid:28822529
- 3. Marrakchi S, Kanoun F, Idriss S, Kammoun I, Kachboura S. Arrhythmia and thyroid dysfunction. Herz. 2015;40(2):101–109. pmid:24990773
- 4. Jones DD, May KE, Geraci SA. Subclinical thyroid disease. Am J Med. 2010;123(6):502–504. pmid:20569751
- 5. Gharib H, Tuttle RM, Baskin HJ, Fish LH, Singer PA, McDermott MT. Subclinical thyroid dysfunction: a joint statement on management from the american association of clinical endocrinologists, the american thyroid association, and the endocrine society. J Clin Endocrinol Metab. 2005;90(1):581–5. pmid:15643019
- 6. Kim YA, Park YJ. Prevalence and risk factors of subclinical thyroid disease. Endocrinol Metab (Seoul). 2014;29(1):20–29. pmid:24741450
- 7. Mendes D, Alves C, Silverio N, Batel Marques F. Prevalence of undiagnosed hypothyroidism in europe: a systematic review and meta-analysis. Eur Thyroid J. 2019;8(3):130–143. pmid:31259155
- 8. Garmendia Madariaga A, Santos Palacios S, Guillén-Grima F, Galofré JC. The incidence and prevalence of thyroid dysfunction in europe: a meta-analysis. J Clin Endocrinol Metab. 2014;99(3):923–231. pmid:24423323
- 9.
National institute for health and care excellence (NICE, UK). Thyroid disease: assessment and management. NICE guidance. 2019 [cited 2023 August 22]. https://www.nice.org.uk/guidance/ng145/chapter/Recommendations.
- 10. Evron JM, Hummel SL, Reyes-Gastelum D, Haymart MR, Banerjee M, Papaleontiou M. Association of thyroid hormone treatment intensity with cardiovascular mortality among US veterans. JAMA Netw Open. 2022;5(5):e2211863. pmid:35552725
- 11. Papaleontiou M, Levine DA, Reyes-Gastelum D, Hawley ST, Banerjee M, Haymart MR. Thyroid hormone therapy and incident stroke. J Clin Endocrinol Metab. 2021;106(10):e3890–e3900. pmid:34137866
- 12. Dai H, Zhang Q, Much AA, Maor E, Segev A, Beinart R, et al. Global, regional, and national prevalence, incidence, mortality, and risk factors for atrial fibrillation, 1990–2017: results from the global burden of disease study 2017. Eur Heart J Qual Care Clin Outcomes. 2021;7(6):574–582. pmid:32735316
- 13. Conen D, Chae CU, Glynn RJ, Tedrow UB, Everett BM, Buring JE, et al. Risk of death and cardiovascular events in initially healthy women with new-onset atrial fibrillation. JAMA. 2011;305(20):2080–2087. pmid:21610240
- 14. Ott A, Breteler MM, de Bruyne MC, van Harskamp F, Grobbee DE, Hofman A. Atrial fibrillation and dementia in a population-based study. The rotterdam study. Stroke. 1997;28(2):316–321. pmid:9040682
- 15. Stewart S, Hart CL, Hole DJ, McMurray JJ. A population-based study of the long-term risks associated with atrial fibrillation: 20-year follow-up of the renfrew/paisley study. Am J Med. 2002;113(5):359–364. pmid:12401529
- 16. Wolf PA, Dawber TR, Thomas HE, Kannel WB. Epidemiologic assessment of chronic atrial fibrillation and risk of stroke: the framingham study. Neurology. 1978;28(10):973–977. pmid:570666
- 17. Creswell LL, Schuessler RB, Rosenbloom M, Cox JL. Hazards of postoperative atrial arrhythmias. Ann Thorac Surg. 1993;56(3):539–549. pmid:8379728
- 18. Dobrev D, Aguilar M, Heijman J, Guichard JB, Nattel S. Postoperative atrial fibrillation: mechanisms, manifestations and management. Nat Rev Cardiol. 2019;16(7):417–436. pmid:30792496
- 19. Hossein Almassi G, Schowalter T, Nicolosi AC, Aggarwal A, Moritz TE, Henderson WG, et al. Atrial fibrillation after cardiac surgery: a major morbid event? Ann Surg. 1997;226(4):501–513. pmid:9351718
- 20. Rodondi N, den Elzen WPJ, Bauer DC, Cappola AR, Razvi S, Walsh JP, et al. Subclinical hypothyroidism and the risk of coronary heart disease and mortality. JAMA. 2010;304(12):1365–1374. pmid:20858880
- 21. Hak AE, Pols HA, Visser TJ, Drexhage HA, Hofman A, Witteman JC. Subclinical hypothyroidism is an independent risk factor for atherosclerosis and myocardial infarction in elderly women: the rotterdam study. Ann Intern Med. 2000;132(4):270–278. pmid:10681281
- 22. Gencer B, Collet T-H, Virgini V, Bauer Douglas C, Gussekloo J, Cappola Anne R, et al. Subclinical thyroid dysfunction and the risk of heart failure events. Circulation. 2012;126(9):1040–1049. pmid:22821943
- 23. Kim SY, Park JE, Lee YJ, Seo HJ, Sheen SS, Hahn S, et al. Testing a tool for assessing the risk of bias for nonrandomized studies showed moderate reliability and promising validity. J Clin Epidemiol. 2013;66(4):408–414. pmid:23337781
- 24. Balshem H, Helfand M, Schünemann HJ, Oxman AD, Kunz R, Brozek J, et al. GRADE guidelines: 3. Rating the quality of evidence. J Clin Epidemiol. 2011;64(4):401–406. pmid:21208779
- 25.
GRADEpro GDT: GRADEpro Guideline Development Tool [Software]. McMaster University and Evidence Prime, 2022. gradepro.org.
- 26.
Review Manager (RevMan) [Computer Program]. Version 5.4, The Cochrane Collaboration, 2020.
- 27.
Deeks J, Higgins J, Altman D. Analysing data and undertaking meta-analyses. In: Higgins JPT, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, Welch VA, editors. Cochrane handbook for systematic reviews of interventions version 6.4 (updated August 2023): Cochrane, 2023; www.training.cochrane.org/handbook.
- 28. Tenerz A, Forberg R, Jansson R. Is a more active attitude warranted in patients with subclinical thyrotoxicosis? J Intern Med. 1990;228(3):229–233. pmid:2401873
- 29. Cappola AR, Fried LP, Arnold AM, Danese MD, Kuller LH, Burke GL, et al. Thyroid status, cardiovascular risk, and mortality in older adults. JAMA. 2006;295(9):1033–1041. pmid:16507804
- 30. Park YJ, Yoon JW, Kim KI, Lee YJ, Kim KW, Choi SH, et al. Subclinical hypothyroidism might increase the risk of transient atrial fibrillation after coronary artery bypass grafting. Ann Thorac Surg. 2009;87(6):1846–1852. pmid:19463607
- 31. Rosario PW. Natural history of subclinical hyperthyroidism in elderly patients with TSH between 0.1 and 0.4 mIU/l: a prospective study. Clin Endocrinol (Oxf). 2010;72(5):685–688. pmid:20447066
- 32. Poola R, Mathiason MA, Caplan RH. A retrospective study of the natural history of endogenous subclinical hyperthyroidism. WMJ. 2011;110(6):277–280. pmid:22324204
- 33. Nanchen D, Gussekloo J, Westendorp RG, Stott DJ, Jukema JW, Trompet S, et al. Subclinical thyroid dysfunction and the risk of heart failure in older persons at high cardiovascular risk. J Clin Endocrinol Metab. 2012;97(3):852–861. pmid:22238391
- 34. Selmer C, Olesen JB, Hansen ML, Lindhardsen J, Olsen A-MS, Madsen JC, et al. The spectrum of thyroid disease and risk of new onset atrial fibrillation: a large population cohort study. BMJ. 2012;345:e7895. pmid:23186910
- 35. Martínez-Comendador J, Marcos-Vidal JM, Gualis J, Martin CE, Martin E, Otero J, et al. Subclinical hypothyroidism might increase the risk of postoperative atrial fibrillation after aortic valve replacement. Thorac Cardiovasc surg. 2016;64(5):427–433. pmid:26121379
- 36. Langén VL, Niiranen TJ, Puukka P, Lehtonen AO, Hernesniemi JA, Sundvall J, et al. Thyroid-stimulating hormone and risk of sudden cardiac death, total mortality and cardiovascular morbidity. Clin Endocrinol (Oxf). 2018;88(1):105–113. pmid:28862752
- 37. Zhao D, Xu F, Yuan X, Feng W. Impact of subclinical hypothyroidism on outcomes of coronary bypass surgery. J Card Surg. 2021;36(4):1431–1438. pmid:33567099
- 38.
British thyroid foundation. Thyroid function tests 2021 [cited 2023 August 22]. https://www.btf-thyroid.org/thyroid-function-tests.
- 39. Sun J, Yao L, Fang Y, Yang R, Chen Y, Yang K, et al. Relationship between subclinical thyroid dysfunction and the risk of cardiovascular outcomes: a systematic review and meta-analysis of prospective cohort studies. Int J Endocrinol. 2017;2017:8130796. pmid:29081800
- 40. Collet T-H, Gussekloo J, Bauer DC, den Elzen WPJ, Cappola AR, Balmer P, et al. Subclinical hyperthyroidism and the risk of coronary heart disease and mortality. Arch Intern Med. 2012;172(10):799–809. pmid:22529182
- 41. Smedegaard SB, Riis AL, Christiansen MK, Linde JKS. Subclinical hyperthyroidism and the risk of developing cardiovascular disease-a systematic review. Dan Med J. 2020;67(11):A12190701. pmid:33215608
- 42. Baumgartner C, da Costa BR, Collet TH, Feller M, Floriani C, Bauer DC, et al. Thyroid function within the normal range, subclinical hypothyroidism, and the risk of atrial fibrillation. Circulation. 2017;136(22):2100–2116. pmid:29061566
- 43. Ahlsson A, Fengsrud E, Bodin L, Englund A. Postoperative atrial fibrillation in patients undergoing aortocoronary bypass surgery carries an eightfold risk of future atrial fibrillation and a doubled cardiovascular mortality. Eur J Cardiothorac Surg. 2010;37(6):1353–1359. pmid:20138531
- 44. Cerillo AG, Bevilacqua S, Storti S, Mariani M, Kallushi E, Ripoli A, et al. Free triiodothyronine: a novel predictor of postoperative atrial fibrillation. Eur J Cardiothorac Surg. 2003;24(4):487–492. pmid:14500064
- 45. Euser AM, Zoccali C, Jager KJ, Dekker FW. Cohort studies: Prospective versus retrospective. Nephron Clin Pract. 2009;113(3):c214–c217. pmid:19690438
- 46. Szécsi B, Nagy Á, Balog Á, Gál J, Merkely B, Székely A. The importance of thyroid hormone levels after heart surgery-preliminary study. J Cardiothorac Vasc Anesth. 2021;35(1):S14–S15.
- 47. Klemperer JD, Klein IL, Ojamaa K, Helm RE, Gomez M, Isom OW, et al. Triiodothyronine therapy lowers the incidence of atrial fibrillation after cardiac operations. Ann Thorac Surg. 1996;61(5):1323–1327. pmid:8633935
- 48. Mullis-Jansson SL, Argenziano M, Corwin S, Homma S, Weinberg AD, Williams M, et al. A randomized double-blind study of the effect of triiodothyronine on cardiac function and morbidity after coronary bypass surgery. J Thorac Cardiovasc Surg. 1999;117(6):1128–1134. pmid:10343261
- 49. Kalarus Z, Mairesse GH, Sokal A, Boriani G, Średniawa B, Casado-Arroyo R, et al. Searching for atrial fibrillation: looking harder, looking longer, and in increasingly sophisticated ways. An EHRA position paper. EP Europace. 2022;25(1):185–198. pmid:36256580
- 50. Lehto M, Haukka J, Aro A, Halminen O, Putaala J, Linna M, et al. Comprehensive nationwide incidence and prevalence trends of atrial fibrillation in Finland. Open Heart. 2022;9(2):e002140. pmid:36414332