Assessment of the relationship between thyroid function and panic disorder: A mendelian randomization study

doi:10.21203/rs.3.rs-3577312/v1

Download PDF

Research Article

Assessment of the relationship between thyroid function and panic disorder: A mendelian randomization study

https://doi.org/10.21203/rs.3.rs-3577312/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background

Multiple observational studies have indicated a correlation between thyroid function and the risk of panic disorder (PD). Nevertheless, the causality surrounding this association remains unclear. Our objective was to evaluate the causality between thyroid function and the risk of PD by employing Mendelian randomization (MR).

Methods

We employed publicly available genome-wide association studies (GWAS) to select single nucleotide polymorphisms (SNPs) that are associated with various aspects of thyroid function (hyperthyroidism, hypothyroidism, FT4, TSH, TPOAb, and thyroid nodules). The statistical data on panic disorder were obtained from the FinnGen consortium. To assess causality, we utilized the inverse variance weighted (IVW) method, MR-Egger method and weighted median (WM) method for the MR estimates. Sensitivity analyses were conducted using Cochran’s Q test, MR-Egger intercept, MR-Pleiotropy Residual Sum and Outlier method, leave-one-out analysis, and funnel plot.

Results

The genetically predicted presence of hyperthyroidism showed an inverse association with PD as evident from the IVW OR of 0.93 (95% CI: 0.87–0.98; P = 0.01).However, our findings did not indicate any causal effects of variation in FT4 (OR: 0.78, 95%CI: 0.78–1.27; P = 1)、TSH (OR: 1.03, 95%CI: 0.83–1.28; P = 0.77)、TPOAb (OR: 0.9, 95%CI: 0.47–1.72; P = 0.75)、hypothyroidism (OR: 0.57, 95%CI: 0.01–50.54; P = 0.81) and thyroid nodules (OR: 1.02, 95%CI: 0.91–1.14; P = 0.76) on PD risk.

Conclusions

panic disorder

thyroid function

thyroid hormone

hyperthyroidism

Mendelian randomization study

Panic disorder (PD), a prevalent and severe psychological ailment within the anxiety continuum, is distinguished by frequent and unforeseeable episodes of panic, accompanied by a variety of physiological manifestations. It often results in substantial impairment, particularly when accompanied by agoraphobia, and is associated with notable functional morbidity and diminished quality of life. Diagnosis based on International Classification of Diseases 10 (ICD-10) necessitates recurrent panic attacks, accompanied by either apprehension regarding future episodes or the emergence of phobic avoidance. Panic attacks are abrupt and sometimes unforeseen episodes of intense anxiety, often accompanied by physical manifestations such as cardiovascular, otoneurological, gastrointestinal, or autonomic symptoms [1]. A comprehensive meta-analysis revealed that the most common age of onset is during late adolescence or early adulthood [2]. The National Comorbidity Survey Replication (NCS-R) provides prevalence estimates of 2.7% within a 12-month period and 4.7% over the course of a lifetime [3, 4].

Research has shown that risk factors for anxiety disorders include alcohol use, smoking, avoidance behaviors, marijuana use, occupational factors, and negative evaluations of stressful life events. Protective factors that produce some support include physical activity, sports participation, social support, and coping skills. The most researched risk or protective factor was cigarette smoking, which was found to increase the risk of PD development in many prospective studies [5], while much less is known about the role of thyroid function.

At present, the etiology of PD remains unclear, among which the hypothalamic-pituitary-thyroid (HPT) axis is an etiological factor. A large amount of studies have shown that the endocrine system responds to various stressors. Many researchers have explored changes in HPT axis hormones in stress responses [6]. The regulation of thyroid function is a multifaceted process that encompasses not only the thyroid gland, but also the pituitary gland, hypothalamus, and feedback mechanisms. The hypothalamus releases thyrotropin releasing hormone (TRH) acts on the pituitary, rereleasing thyroid stimulating hormone (TSH) acts on the thyroid, which releases free thyroid hormone (FT4). Studies of panic disorder have found evidence of a dulled response of thyroid stimulating hormone (TSH) to thyroid stimulating hormone (TRH) stimulation [7, 8]. Thyroid hormones are essential for neurocognitive development and function, so the link between thyroid dysfunction and panic disorder has been studied over the past few decades [6, 9]. Indeed, both TSH and FT4 have been associated with PD in an observational study [6]. Some researches found initial support for an increased prevalence of thyroid disorders in patients with PD [10, 11]. Because of the high incidence of PD patients, a better understanding of its underlying mechanisms made essential to further improve its prevention and treatment.

Thyroid dysfunction can cause symptoms similar to those seen in people with anxiety disorders, especially panic disorder. Some of the physical symptoms of hyperthyroidism, such as palpitations, shortness of breath, and increased sweating, overlap with what people with panic disorder experience during panic attacks. A link has been suggested between thyroid disease and panic attacks and phobias [12]. However, the extent to which PD patients suffer from thyroid dysfunction remains unclear. Learning more about such dysfunction first means obtaining an estimate of whether a person with PD has (subclinical) thyroid disease. Besides, this means studying subtle changes in individual thyroid function parameters, namely TSH, FT3, and FT4. Doing so may shed light on the complex relationship between PD and thyroid function and help determine whether the aforementioned panic symptoms may also occur as an epiphenomenon of subclinical thyroid dysfunction. Since observational studies are sometimes vulnerable to selection bias, residual confounding, and reverse causality, it is unclear if the reported relationships are causal [13, 14].

Overall, the previous studies point to a connection between thyroid function and panic disorder, but it's important to clarify whether this connection is causative or not. In recent years, with the increasing popularity of genome-wide association studies (GWAS) databases, Mendelian randomization (MR) has attracted a lot of attention. MR can avoid some types of confounding by using genetic variants, which are fixed at conception, to support causal inferences about the effects of changeable risk factors [15]. MR can also yield insights into causality in situations where randomized controlled trials (RCTs) are impractical or insufficient [16], its level of research evidence lies at the junction of RCTs and observational studies [15], as it has the ability to emulate RCTs, and offers a reliable statistical approach utilizing instrumental variables (IVs) to elucidate the causal relationship between exposure and disease [17].

Therefore, in our current investigation, we executed a two-sample MR analysis using the GWAS database to scrutinize the genetic causality interlinking thyroid functionality and panic disorder. The thyroid function we studied included hyperthyroidism, hypothyroidism, FT4, TSH, thyroid peroxidase antibody (TPOAb), and thyroid nodules.

Two-sample Mendelian Randomization

We conducted two-sample MR analyses using the data from the most recent GWAS on thyroid function as exposures, and PD as summary-level statistics. There was no necessity for further ethical consent as this involved a reevaluation of preexisting summary-level data. The MR analysis was conducted utilizing Two-sample MR (version 0.5.5) and R (version 4.2.1).

The MR analysis hinges on three fundamental assumptions: 1) The chosen IVs must have a significant correlation with the exposure (thyroid function) [18]. We employ the F-statistic to evaluate the potency of each genetic tool. The subsequent formula calculates the F-statistic: F = beta²/se² (beta for the exposure association (beta); variance (se)) [19]. If the F-statistic exceeds 10, the IV possesses a robust capacity to forecast PD. 2) The chosen IVs cannot influence the outcome (panic disorder) via alternative routes, only through the specified exposure [20]. 3) Confounding elements are not linked to the selected IVs. The schematic representation of the research design is depicted in Fig. 1.

GWAS data

We procured single nucleotide polymorphisms (SNPs) from the GWAS database, serving as genetic IVs [18]. Associations between diverse thyroid characteristics and instrument exposure among individuals of European descent. The significant SNPs associated with FT4、 TSH、TPOAb、hypothyroidism、hyperthyroidism (P < 5×10^− 8) and thyroid nodules (P < 5×10^− 6) were procured from the correspondingly latest and most extensive GWAS database (Table 1). We utilized PhenoScannerV2 (www.phenoscanner.medschl.cam.ac.uk) to delve deeper into the potential correlation of IVs with PD risk confounders [21]. To circumvent any potential skewness instigated by intense linkage disequilibrium (LD), we opted for SNPs with an LDr² less than 0.001.

Summary data from GWAS for FT4 and TSH were procured from a research study involving 26,089 and 27,916 participants respectively [22]. TPOAb [23] was from GWAS data that the participants was 18297. We procured summary-level GWAS data associated with hyperthyroidism from the FinnGen consortium, comprising 962 instances and 172,976 control subjects. Summary statistics for hypothyroidism from UK biobank comprising 463,010 participants (9674 cases and 453,336 controls). The GWAS data pertaining to thyroid nodule were obtained from a comprehensive meta-analysis, encompassing a total of 1121 cases juxtaposed with 187,684 control subjects [24]. The GWAS data pertinent to the outcome (PD) was culled from the FinnGen consortium, encompassing 2376 cases and 198,110 controls of European lineage. A comprehensive detailing of the GWAS summary data in our investigation is provided Table 1.

Table 1

Details of the GWAS included in the Mendelian randomization.
Trait	DOI	Year	Population	Cases	Controls	Samplesize
hyperthyroidism	www.finngen.fi/en	2021	European	962	172976	173938
hypothyroidism	www.nealelab.is/uk-biobank	2018	European	9674	453336	463010
FT4	10.1038/s41467-018-06356-1	2018	European	NA	NA	26089
TSH	10.1038/s41467-018-06356-1	2018	European	NA	NA	27916
thyroid nodule	10.1186/s13044-021-00094-1	2021	European	1121	187684	188805
TPOAB	10.1371/journal.pgen.1004123	2014	European	NA	NA	18297

FT4, thyroid hormone; TSH, thyroid stimulating hormone; TPOAb, thyroid peroxidase antibody; NA, Not Applicable.

Statistical analysis

Estimations of thyroid function's influence on PD risk via Mendelian randomization were computed utilizing the inverse variance weighting (IVW) approach, weighted median (WM) technique and the MR-Egger methodology. The IVW method, which used a meta-analysis method to integrate the Wald ratio of each SNPs, was the main MR analysis in our work. It was based on the assumption that IVs can only affect outcomes through a specific exposure. When horizontal pleiotropy was not present, the IVW method could produce unbiased causal estimates [25]. Consequently, the IVW method offered the most accurate evaluation [26]. To examine the bias caused by inefficient IV and horizontal pleiotropy effects, the complement of analysis was subjected to the WM method and the MR-Egger method [27]. Due to the impact of outlying genetic variations, the estimations of the MR-Egger technique were likely erroneous [28]. It was conducted to test whether there was evidence of the intercept parameter being different from zero. The WM method's accuracy was lower and its bias was quite moderate, especially for the proportion of IVs with horizontal pleiotropy below 50% [29]. The IV estimates of each SNP should be distributed symmetrically close to the point estimation in the absence of directional pleiotropy, demonstrating that there was no systematic bias in the data.

The necessity of a comprehensive sensitivity analysis to scrutinize potential heterogeneity and horizontal pleiotropy was underscored, encompassing methods such as IVW, MR-Egger regression, weighted median, funnel plots, and leave-one-out analysis, in which a single variant was excluded at each step. To evaluate the variability of impact sizes for certain genetic IVs, the Cochran's Q test was used. To measure vertical pleiotropy, the intercept from MR-Egger regression was used [30]. We employed MR-PRESSO to evaluate the horizontal pleiotropy within the constructed MR model, eliminating any outlier prior to recalculating the causal influence if directional pleiotropy was identified with a p-value less than 0.05. A leave-one-out analysis was carried out to examine how the findings would change if one of the chosen SNPs were not included [31].

In the GWAS, we pinpointed 31 SNPs for FT4 within the reference range, 61 SNP for TSH, 36 SNPs for hypothyroidism, 7 SNPs for hyperthyroidism, 9 SNPs for TPOAb concentration and 7 SNPs for thyroid nodules. Each of these exhibited an F-statistic exceeding 80, thereby suggesting the absence of weak instrumental bias. Detailed information regarding specific SNP, inclusive of the effect alleles, other alleles, effect allele frequency, beta (β), standard error, P-value, and F-statistic, can be found in Table S1. The outcomes of MR analyses exploring the correlations between genetically predicted thyroid function levels and PD risk are presented in Table S2 and summarized in Fig. 2.

Hyperthyroidism

The genetic predisposition towards hyperthyroidism was found to have a negative correlation with PD, as indicated by an IVW OR of 0.93 (95% CI: 0.87–0.98; P = 0.01). It was similar to WM (OR = 0.91, 95% CI: 0.84–0.99, P = 0.02) (Fig. 2, Fig. S1A). This MR analysis did not reveal any substantial heterogeneity (Cochran’s Q P = 0.49) or horizontal pleiotropy (P = 0.79) (Tables S3 and S4). The robustness of the results was confirmed by sequentially removing SNPs and reanalyzing with the remaining ones, thereby demonstrating that the positive outcomes are not influenced by a single SNP ( Fig.S1C).

Hypothyroidism

We found no indications of a plausible causal link between hypothyroidism and the likelihood of PD (OR = 0.57, 95%CI = 0.01–50.54, P = 0.81). Analyses of sensitivity employing the MR Egger and WM techniques yielded both negative results (Fig. 2, Fig.S2A). Based on the Egger intercept and Cochran’s Q test, no evidence of directional pleiotropy (p = 0.4) and heterogeneity (p = 0.24) were found (Tables S3 and S4 ). The leave-one-out analysis indicated the robustness of the results (Fig. S2C) and the funnel plot showed it was symmetric (Fig. S2B).

FT4

The association between FT4 and PD was not discovered through the IVW method (OR = 1, 95%CI = 0.78–1.27, P = 1), MR-Egger (OR = 0.63, 95%CI = 0.37–1.07, P = 1) and WM (OR = 0.85, 95%CI = 0.62–1.18, P = 0.34) (Fig. 2, Fig.S3A). The outcomes of Cochran’s Q test, MR Egger regression, MR-PRESSO and the leave-one-out test indicated that the MR estimates were fairly robust (Tables S3 and S4, Fig. S3C).

TSH

Heterogeneity (p = 0.01) was found in TSH and PD (Tables S3). Therefore, we used random effect model for this two-sample MR analysis. There was no statistically significant between TSH and PD (OR = 1.03, 95%CI = 0.83–1.28, P = 0.77) (Fig. 2, Fig.S4A). There was no existence of pleiotropy (p = 0.19) as shown in Tables S4, and the exclusion of potentially pleiotropic variants identified using the MR-PRESSO method did not alter the outcomes. The robustness of the MR estimates was further confirmed by the leave-one-out test (Fig. S4C).

Thyroid nodules, TPOAb

We did not find a genetic association between thyroid nodules, TPOAb with PD, respectively (Fig. 2, Fig. S5-6A). Regarding thyroid nodules, the sensitivity analysis and heterogeneity test failed to reveal any potential horizontal pleiotropy or significant heterogeneity (Tables S3 and S4). The leave-one-out test affirmed the stability of the MR estimate upon the removal of individual SNP (Fig. S5-6C). Lastly, the funnel plots on thyroid function and PD are presented in Fig. S1-6B.

This represents the inaugural study employing the MR methodology to scrutinize the causal correlation between thyroid function and PD. Given the accessible vast GWAS datasets pertaining to thyroid function and PD, the current time is indeed opportune for conducting MR analyses on these parameters. Our findings did not corroborate any causal impact of variation in FT4、TSH、TPOAb、hypothyroidism and thyroid nodules on the susceptibility to PD. On the contrary, we discerned significant inverse relationship between hyperthyroidism and the risk of PD.

The underlying mechanism of the causal relationship between hyperthyroidism and reduced risk of PD is still unclear. Thyroxine, a hormone that individually respond to risk, so that hyperthyroidism may be associated with a higher risk response in individuals. This may mean that the individual does not make an excessively reaction about the dangerous environment, thereby alleviating the panic attack. In addition, previous studies have shown that thyroid hormone receptors are widely distributed in high concentrations in the cerebral cortex, hippocampus and other brain regions [32]. we discerned significant inverse relationship between hyperthyroidism and the risk of PD. Thyroid hormone is involved in neuron migration, differentiation and synaptic formation [33]. Therefore, it is necessary to further study the role of hyperthyroidism in the pathogenesis of PD.

In an observational study [6], significant correlations between the thyroid hormone levels and clinical features were observed in PD patients, as well as a positive correlation between TSH levels and the severity of PD. However, other studies have reported an inverse correlation between TSH response and anxiety severity in the same patient population [8], but no evidence of a causal effect of TSH levels on PD risk was found in our study. Despite higher SNP-heritability estimates for thyroid function (TSH, FT4), there were no significant genetic correlations with PD in our study. Therefore, the negative results of our study suggest that thyroid function changes either do not affect PD risk at all. This may suggest that the genetically determined relationship between thyroid function and PD is not mediated through thyroid hormones directly, but using pathways that underpin disease pathogenesis, such as (general) autoimmunity [34]. Besides, it's worth noting that most of the current researches about thyroid function and PD are observational. Evidence from observational studies should be interpreted with caution as it cannot reveal causality and effect and completely exclude the influence of confounding factors.

In another study, the authors tested the thyroid function of 165 patients diagnosed with PD and found that these subjects reported a higher prevalence of thyroid disease compared to the prevalence of thyroid disease in the general population; however, less than 1% of the subjects currently had thyroid dysfunction. The intensity of panic attacks or phobias did not correspond with thyroid function indicators [12].

Studies have shown that patients with PD have changes in peripheral hypothalamic– pituitary– adrenal (HPA) axis markers, which may indicate abnormal functioning of the central corticotropin-releasing hormone system [35]. In a sample of mostly women hospitalized with panic disorder, abnormal expression of N-methyl- D-aspartate receptors (NMDARs) is closely related to anxiety disorder with thyroid lesions, NMDAR subunits may have various activities and exert diverse effects in thyroid nodules, and the NR2A subunit may be an important regulator in anxiety disorder with thyroid nodules [36].

It's crucial to acknowledge that the bulk of existing research is observational. Findings from such studies ought to be treated with skepticism as they neither establish causality nor entirely eliminate the potential impact of extraneous variables. Therefore, we do MR analysis to exclude the interference of confounding factors, in order to study the causal relationship between the thyroid function and PD. Our research has several strengths. Primarily, the MR design's primary benefit is its ability to significantly mitigate the impact of confounding factors and reverse causality. Furthermore, to guarantee the independence of IVs, we conducted a comprehensive screening of single nucleotide polymorphisms utilizing Plink clumping. Finally, the included genetic instruments were fairly powerful because the F-statistics of the included SNPs were all over 10.

There are still some limitations in our study. First of all, our exposure and outcome were both selected from the FinnGen consortium without the original study, so the sample overlap rate could not be calculated. Therefore, it is probable to have sample size overlap. Additionally, our study may be subject to racial bias as all chosen GWAS database populations were of European ancestry, necessitating further exploration to determine if this causal relationship still remains in other demographics. Furthermore, due to the unavailability of comprehensive demographic and clinical data on participants were not available, we were unable to conduct any subgroup analysis. Lastly, despite utilizing the largest GWAS database, the sample size of this study is comparatively modest when juxtaposed with population-based observational studies.

In summary, Our findings indicated a significant inverse correlation between hyperthyroidism and PD risk ,with no discernible causal impacts of alterations in FT4、TSH、TPOAb、hypothyroidism and thyroid nodules on PD risk. It may suggest that most thyroid function may not be the etiological factor of PD, further studies are needed to verify our results in the real world. In the future, studies should explore the related mechanism of hyperthyroidism leading to reduced incidence of panic disorder, and more RCTS are needed to investigate whether thyroid hormone can treat panic disorder.

Competing interests

The authors declare no competing interests.

Ethics approval and consent to participate

Not applicable. Ethical approval and informed consent for studies included in the analyses were provided in the original publications. Each study was approved by the appropriate institutional review board/ethics committee.

Consent for publication

Not applicable.

Data Availability

The datasets generated and analysed during the current study are available in the IEU open gwas project [https://gwas.mrcieu.ac.uk/] and its additional file.

Funding

This work was supported by the Medical Science and Technology Project of Zhejiang Province (2022KY613).

Author contributions

Sijie Yu: Conceptualization, Methodology, Software, Data curation, Writing- Original draft preparation. Chongkai Shen: Visualization, Investigation. Junpeng Zhu: Supervision, Software, Project administration, Validation, Writing- Reviewing and Editing.

Acknowledgements

We gratefully acknowledge all the studies and databases that made GWAS summary data available. We are extremely grateful to Medical Science and Technology Project of Zhejiang Province (2022KY613).

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Roy-Byrne, P.P., M.G. Craske, and M.B. Stein, Panic disorder. Lancet, 2006. 368(9540): p. 1023-32.
Solmi, M., et al., Age at onset of mental disorders worldwide: large-scale meta-analysis of 192 epidemiological studies. Mol Psychiatry, 2022. 27(1): p. 281-295.
Kessler, R.C., et al., Lifetime prevalence and age-of-onset distributions of DSM-IV disorders in the National Comorbidity Survey Replication. Arch Gen Psychiatry, 2005. 62(6): p. 593-602.
Kessler, R.C., et al., Prevalence, severity, and comorbidity of 12-month DSM-IV disorders in the National Comorbidity Survey Replication. Arch Gen Psychiatry, 2005. 62(6): p. 617-27.
Zimmermann, M., et al., Modifiable risk and protective factors for anxiety disorders among adults: A systematic review. Psychiatry Res, 2020. 285: p. 112705.
Kikuchi, M., et al., Relationship between anxiety and thyroid function in patients with panic disorder. Prog Neuropsychopharmacol Biol Psychiatry, 2005. 29(1): p. 77-81.
Roy-Byrne, P.P., et al., Reduced TSH and prolactin responses to TRH in patients with panic disorder. Am J Psychiatry, 1986. 143(4): p. 503-7.
Tükel, R., et al., Thyrotropin stimulating hormone response to thyrotropin releasing hormone in patients with panic disorder. Psychoneuroendocrinology, 1999. 24(2): p. 155-60.
Yeragani, V.K., et al., Thyroid hormone levels in panic disorder. Can J Psychiatry, 1987. 32(6): p. 467-9.
Simon, N.M., et al., Hypothyroidism and hyperthyroidism in anxiety disorders revisited: new data and literature review. J Affect Disord, 2002. 69(1-3): p. 209-17.
Fischer, S. and U. Ehlert, Hypothalamic-pituitary-thyroid (HPT) axis functioning in anxiety disorders. A systematic review. Depression & Anxiety, 2017. 35(1).
Lesser, I.M., et al., Past and current thyroid function in subjects with panic disorder. J Clin Psychiatry, 1987. 48(12): p. 473-6.
Grimes, D.A. and K.F. Schulz, Bias and causal associations in observational research. Lancet, 2002. 359(9302): p. 248-52.
Bunevicius, R. and A.J. Prange, Jr., Thyroid disease and mental disorders: cause and effect or only comorbidity? Curr Opin Psychiatry, 2010. 23(4): p. 363-8.
Davies, N.M., M.V. Holmes, and G. Davey Smith, Reading Mendelian randomisation studies: a guide, glossary, and checklist for clinicians. Bmj, 2018. 362: p. k601.
Teumer, A., Common Methods for Performing Mendelian Randomization. Front Cardiovasc Med, 2018. 5: p. 51.
Bennett, D.A. and M.V. Holmes, Mendelian randomisation in cardiovascular research: an introduction for clinicians. Heart, 2017. 103(18): p. 1400-1407.
Lawlor, D.A., et al., Mendelian randomization: using genes as instruments for making causal inferences in epidemiology. Stat Med, 2008. 27(8): p. 1133-63.
Bowden, J., et al., Assessing the suitability of summary data for two-sample Mendelian randomization analyses using MR-Egger regression: the role of the I2 statistic. Int J Epidemiol, 2016. 45(6): p. 1961-1974.
Bowden, J., G. Davey Smith, and S. Burgess, Mendelian randomization with invalid instruments: effect estimation and bias detection through Egger regression. Int J Epidemiol, 2015. 44(2): p. 512-25.
Kamat, M.A., et al., PhenoScanner V2: an expanded tool for searching human genotype-phenotype associations. Bioinformatics, 2019. 35(22): p. 4851-4853.
Teumer, A., et al., Genome-wide analyses identify a role for SLC17A4 and AADAT in thyroid hormone regulation. Nat Commun, 2018. 9(1): p. 4455.
Medici, M., et al., Identification of novel genetic Loci associated with thyroid peroxidase antibodies and clinical thyroid disease. PLoS Genet, 2014. 10(2): p. e1004123.
Lau, L.W., et al., Malignancy risk of hyperfunctioning thyroid nodules compared with non-toxic nodules: systematic review and a meta-analysis. Thyroid Res, 2021. 14(1): p. 3.
Hemani, G., et al., The MR-Base platform supports systematic causal inference across the human phenome. Elife, 2018. 7.
Sekula, P., et al., Mendelian Randomization as an Approach to Assess Causality Using Observational Data. J Am Soc Nephrol, 2016. 27(11): p. 3253-3265.
Sanderson, E., Multivariable Mendelian Randomization and Mediation. Cold Spring Harb Perspect Med, 2021. 11(2).
Bowden, J., et al., Consistent Estimation in Mendelian Randomization with Some Invalid Instruments Using a Weighted Median Estimator. Genet Epidemiol, 2016. 40(4): p. 304-14.
Verbanck, M., et al., Detection of widespread horizontal pleiotropy in causal relationships inferred from Mendelian randomization between complex traits and diseases. Nat Genet, 2018. 50(5): p. 693-698.
Burgess, S., et al., Sensitivity Analyses for Robust Causal Inference from Mendelian Randomization Analyses with Multiple Genetic Variants. Epidemiology, 2017. 28(1): p. 30-42.
Wu, F., et al., Mendelian randomization study of inflammatory bowel disease and bone mineral density. BMC Med, 2020. 18(1): p. 312.
Williams, G.R., Neurodevelopmental and neurophysiological actions of thyroid hormone. J Neuroendocrinol, 2008. 20(6): p. 784-94.
Ausó, E., et al., A moderate and transient deficiency of maternal thyroid function at the beginning of fetal neocorticogenesis alters neuronal migration. Endocrinology, 2004. 145(9): p. 4037-47.
Foldager, L., et al., Bipolar and panic disorders may be associated with hereditary defects in the innate immune system. J Affect Disord, 2014. 164: p. 148-54.
Fischer, S., The hypothalamus in anxiety disorders. Handbook of clinical neurology, 2021. 180: p. 149-160.
Wang, S., et al., Role of N-methyl-d-aspartate receptors in anxiety disorder with thyroid lesions. Journal of psychosomatic research, 2022. 161: p. 110998.

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Assessment of the relationship between thyroid function and panic disorder: A mendelian randomization study

Status:

Version 1

Abstract

Background

Methods

Results

Conclusions

Figures

Introduction

Materials and Methods

Two-sample Mendelian Randomization

GWAS data

Statistical analysis

Results

Hyperthyroidism

Hypothyroidism

FT4

TSH

Thyroid nodules, TPOAb

Discussion

Conclusion

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1