Deng HX, Liu YZ, Yang L, Liu JJ, Jia FN, Zhao XL, Xia XZ, Zhang XY, Du XD. Thyroid-stimulating hormone levels and suicide attempts in Chinese patients with first-episode drug-naïve major depressive disorder. World J Psychiatry 2026; 16(6): 116435 [DOI: 10.5498/wjp.v16.i6.116435]
Corresponding Author of This Article
Xiang-Dong Du, Chief Physician, Professor, The First Clinical College, Xuzhou Medical University, No. 209, Tongshan Road, Xuzhou 221004, Jiangsu Province, China. xiangdong-du@163.com
Research Domain of This Article
Psychiatry
Article-Type of This Article
research-article
Open-Access Policy of This Article
This article is an open-access article which was selected by an in-house editor and fully peer-reviewed by external reviewers. It is distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/
Baishideng Publishing Group Inc, 7041 Koll Center Parkway, Suite 160, Pleasanton, CA 94566, USA
Share the Article
Deng HX, Liu YZ, Yang L, Liu JJ, Jia FN, Zhao XL, Xia XZ, Zhang XY, Du XD. Thyroid-stimulating hormone levels and suicide attempts in Chinese patients with first-episode drug-naïve major depressive disorder. World J Psychiatry 2026; 16(6): 116435 [DOI: 10.5498/wjp.v16.i6.116435]
Han-Xu Deng, Department of Psychopathy and Psychiatry, Xuzhou Medical University, Xuzhou 221004, Jiangsu Province, China
Yao-Zhi Liu, School of Psychiatry, North Sichuan Medical College, Nanchong 637000, Sichuan Province, China
Lin Yang, School of Psychiatry, Medical College of Soochow University, Suzhou 637100, Jiangsu Province, China
Jun-Jun Liu, Department of Psychiatry, Nanjing Meishan Hospital, Nanjing 210039, Jiangsu Province, China
Feng-Nan Jia, Xing-Zhi Xia, Department of Psychiatry, Suzhou Guangji Hospital, The Affiliated Guangji Hospital of Soochow University, Suzhou 215004, Jiangsu Province, China
Xue-Li Zhao, Department of Sleep Disorders, Suzhou Guangji Hospital, The Affiliated Guangji Hospital of Soochow University, Suzhou 215004, Jiangsu Province, China
Xiang-Yang Zhang, Hefei Fourth People’s Hospital, Anhui Mental Health Center, Affiliated Psychological Hospital of Anhui Medical University, Hefei 230022, Anhui Province, China
Xiang-Dong Du, The First Clinical College, Xuzhou Medical University, Xuzhou 221004, Jiangsu Province, China
Xiang-Dong Du, Suzhou Guangji Hospital, The Affiliated Guangji Hospital of Soochow University, Suzhou 215004, Jiangsu Province, China
Co-corresponding authors: Xiang-Yang Zhang and Xiang-Dong Du.
Author contributions: Deng HX and Liu YZ have played important roles in the manuscript preparation as co-first authors; Deng HX, Liu YZ, Yang L, Liu JJ, Jia FN, Zhao XL, and Xia XZ performed the investigation; Deng HX, Liu YZ, Liu JJ, Jia FN, and Xia XZ analyzed and interpreted the data; Deng HX, Liu YZ, and Zhao XL drafted the manuscript; Deng HX, Zhang XY, and Du XD designed the study; Zhang XY and Du XD critically revised the manuscript as co-corresponding authors; all authors have read and approved the final version for publication.
AI contribution statement: We used ChatGPT (GPT-4) solely for language polishing, and Grammarly for basic grammar checking. The authors have manually verified every sentence. All authors take full responsibility for the originality and accuracy of the manuscript.
Supported by Medical Research Key Project of Jiangsu Provincial Health Commission, No. K2023015; Suzhou Major Diseases Clinical Multi-Center Research Project, No. DZXYJ202414; Scientific and Technological Key Program of Suzhou, No. SYW2024008; and Key Discipline of Psychiatry in Suzhou, No. SZXK202521.
Institutional review board statement: The study was approved by the Institutional Review Board of the First Hospital of Shanxi Medical University (No. 2016-Y27) and performed in accordance with the Declaration of Helsinki.
Informed consent statement: All participants provided informed consent.
Conflict-of-interest statement: All authors declare no conflict of interest in publishing the manuscript.
STROBE statement: The authors have read the STROBE Statement – checklist of items, and the manuscript was prepared and revised according to the STROBE Statement – checklist of items.
Data sharing statement: The datasets generated and/or analyzed during the current study are available from the corresponding author on reasonable request.
Corresponding author: Xiang-Dong Du, Chief Physician, Professor, The First Clinical College, Xuzhou Medical University, No. 209, Tongshan Road, Xuzhou 221004, Jiangsu Province, China. xiangdong-du@163.com
Received: November 12, 2025 Revised: January 6, 2026 Accepted: February 28, 2026 Published online: June 19, 2026 Processing time: 198 Days and 6.9 Hours
Abstract
BACKGROUND
Thyroid dysfunction is commonly observed in patients diagnosed with major depressive disorder (MDD). Thyroid function can frequently be accompanied by fluctuations in thyroid-stimulating hormone (TSH) levels. However, the association between TSH levels and suicide attempts (SAs) in patients with first-episode drug-naïve (FEDN)-MDD remains poorly elucidated.
AIM
To probe into the relationship between TSH levels and SAs in a Chinese population with FEDN-MDD.
METHODS
A total of 1718 FEDN-MDD patients were recruited at the psychiatric outpatient department of the First Hospital of Shanxi Medical University (Taiyuan, Shanxi Province, China) between September 2016 and December 2018. Their sociodemographic characteristics and serum levels of thyroid hormones were acquired. History of SAs was corroborated by conversational interviews and consultation with the patients’ family members. Depressive and anxiety symptoms were assessed by the 17-item Hamilton Rating Scale and Hamilton Anxiety Scale, respectively. Multivariable logistic regression analysis estimated the association between TSH levels and the risk of SAs. Interaction and stratified analyses were performed based on gender, education, marital status, comorbid anxiety, and psychotic symptoms. Threshold effects were tested using two-piecewise linear regression models.
RESULTS
An independent and positive correlation was found between TSH levels and SAs in patients with FEDN-MDD in the multivariable logistic regression model after adjusting for covariates [odds ratio (OR) = 1.09, 95%CI: 1.01-1.17; P < 0.05]. The smoothing plot illustrated a nonlinear association between SAs and TSH levels, with 5.43 μIU/mL serving as the TSH level’s inflection point. TSH levels and SAs exhibited an elevated correlation on the right side of the inflection point (OR = 1.21, 95%CI: 1.08-1.36, P < 0.001), while no statistically significant relationship was observed on the left side (OR = 0.94, 95%CI: 0.82-1.07, P = 0.339).
CONCLUSION
We found a nonlinear TSH-SA association in FEDN-MDD patients. This finding may offer crucial insights into developing effective therapeutic options for suicide prevention in individuals suffering from depression.
Core Tip: This study investigates the relationship between thyroid-stimulating hormone (TSH) levels and suicide attempts in first-episode drug-naïve major depressive disorder patients. The results show a positive independent correlation between TSH levels and suicide attempts, with a significant increase in the correlation when TSH exceeds 5.43 μIU/mL. This finding provides a new biomarker-based perspective for early suicide risk identification in depressed patients.
Citation: Deng HX, Liu YZ, Yang L, Liu JJ, Jia FN, Zhao XL, Xia XZ, Zhang XY, Du XD. Thyroid-stimulating hormone levels and suicide attempts in Chinese patients with first-episode drug-naïve major depressive disorder. World J Psychiatry 2026; 16(6): 116435
Major depressive disorder (MDD) is a severely debilitating disease characterized by at least one discrete depressive episode lasting at least 2 weeks and involves changes in emotional, neurovegetative, and cognitive symptoms[1]. It impacts both physical and mental health, resulting in diminished quality of life and imposing substantial burdens on the healthcare system and society[2]. Globally, approximately 280 million people suffer from MDD, corresponding to a prevalence rate of about 5%[3]. The global annual prevalence rate of MDD is 4.4%, with a lifetime prevalence rate ranging from 10% to 15%[4]. In China, an epidemiological study based on a multistage clustered-area probability sample of adults reports a lifetime prevalence rate of approximately 3.4%, with a substantial number of patients still experiencing chronic or recurrent states[5]. Projections suggest a significant global increase in the number of MDD cases in the coming years[6].
Suicide represents the most serious outcome associated with MDD, and evidence from multiple epidemiological research efforts suggests that the incidence of suicide in MDD patients is notably greater than in the general population[7,8]. Suicide attempts (SAs), an early warning sign of potential suicidal behavior, frequently occur in patients with MDD and present a critical target for timely intervention[9,10]. It has been reported that the worldwide lifetime prevalence of SAs is roughly 2.7%[11], yet among individuals with MDD, the proportion who have made at least one SA in their lifetime ranges from 16% to 33.7%[12]. Furthermore, an SA is substantially predictive of suicide fatalities, and can result in a variety of catastrophic repercussions including injury, hospitalization, and even loss of liberty[13-16]. It can also impose an extensive financial strain on society[17]. Despite these significant implications, the etiology of SAs in individuals with MDD remains understudied, which urges us to investigate the associated risk factors further.
Many risk factors are associated with SAs, which can be categorized into biological, psychological, cognitive, and environmental factors[18-20]. Previous studies have suggested that thyroid hormones may serve as potential biomarkers for SAs[21-24]. Notably, extensive research involving large samples from diverse cultural and ethnic backgrounds has established a correlation between thyroid function, encompassing both hyperactivity and hypoactivity, and depression[25-27]. Among the various thyroid function markers, the thyroid-stimulating hormone (TSH) level is considered the most sensitive indicator. However, the findings regarding the relationship between TSH levels and SAs in patients with MDD remain inconsistent. For instance, a study involving 1279 MDD patients revealed a significant positive correlation between TSH levels and SAs[28]. Conversely, another study conducted on an adolescent population indicated that TSH levels were lower in the SA group compared to the healthy control group[29]. Additionally, some research has suggested that TSH levels may not be associated with SAs at all[30]. These discrepancies in previous studies may stem from variations in sample characteristics, such as disease duration, comorbidities, antidepressant use, and other confounding factors, all of which can significantly influence TSH levels. Such factors may obscure the true relationship between TSH levels and SAs.
Measuring the relationship between TSH levels and SAs in patients with first-episode drug-naïve (FEDN)-MDD can help minimize the effects of confounding factors such as antidepressant use, disease duration, and comorbidities. Although previous studies have demonstrated a correlation between TSH levels and SAs, the nature of this correlation remains controversial. Our analysis specifically focused on the TSH’s predictive value for SA in the context of FEDN-MDD, rather than its broader applications to standalone SA or MDD alone. By emphasizing MDD with comorbid SA, we aimed to reduce ambiguity and underscore TSH’s potential as a biomarker for suicide risk in depression, distinct from its roles in non-comorbid conditions. To the best of our knowledge, however, there are no studies specifically evaluating the relationship between TSH levels and SAs in patients with FEDN-MDD. Therefore, the primary goal of this study is to investigate this relationship in a large Chinese sample of FEDN-MDD patients.
MATERIALS AND METHODS
Study procedure and participants
A total of 1718 patients with FEDN-MDD were recruited at the psychiatric outpatient department of the First Hospital of Shanxi Medical University from September 2016 to December 2018. Their symptoms were diagnosed by two professionally trained psychiatrists using the Structured Clinical Interview for Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition.
Inclusion criteria: (1) Current first-episode depressive symptoms; (2) 17-item Hamilton Depression Scale (HAMD-17) score > 23; (3) Age 18-60 years, Han nationality; and (4) No previous history of antidepressant use.
Exclusion criteria: (1) Serious physical illness such as cancer, persistent infection, epilepsy, brain injury, or stroke; (2) Meeting any other Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition axis I disorder, including bipolar disorder, schizophrenia, and schizoaffective or other severe medical conditions; (3) Substance use disorder, except for smoking; and (4) Pregnant or breastfeeding women.
This study was approved by the Institutional Review Board of the First Hospital of Shanxi Medical University (No. 2016-Y27), and all participants signed a written informed consent form before enrollment.
Sociodemographic characteristics and clinical assessments
Sociodemographic characteristics and general information, including age, gender, marital status, education, age of onset, and duration of illness, were collected through a structured, self-designed questionnaire. The SA history of the participants was confirmed by asking, “In your lifetime, have you ever attempted suicide?” during a face-to-face interview, a question adapted from a World Health Organization/European multicenter study[31]. Participants who answered “yes” were classified as having attempted suicide. In cases where patients were unable to provide definitive information, we interviewed their family members to clarify details regarding their SA history. All participants were divided into two groups based on their history of SAs: (1) The SA group (n = 346); and (2) The non-SA group (n = 1372). Additionally, participants were further categorized into three groups based on their TSH levels, the low-TSH group (n = 571), the middle-TSH group (n = 574), and the high-TSH group (n = 573), with classifications made according to tertiles. Two qualified psychiatrists who were blinded to the clinical status of the patients assessed the severity of each participant’s depression and severity of anxiety symptoms using the HAMD-17 and Hamilton Anxiety Scale (HAMA), respectively. HAMD scores range from 0 to 52, with a score > 24 indicating the presence of MDD[32]. The HAMA consists of 14 items, with total scores ranging from 0 to 56. When the score is ≥ 18, it is considered to reflect comorbid anxiety[33]. Higher scores on the HAMA and HAMD indicate more severe symptoms. In repeated assessments using these scales, correlation coefficients for interobserver reliability remained > 0.8.
Blood samples
Blood samples were collected between 6:00 AM and 9:00 AM after an overnight fast, prior to the administration of antidepressant treatment. The samples were then sent to the hospital’s experimental center for testing before 11:00 AM. The measured blood biomarkers included total cholesterol (TC), triglycerides (TG), low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), fasting blood glucose (FBG), TSH, free triiodothyronine (FT3), free thyroxine (FT4), thyroid peroxidase antibody (TPOAb), and anti-thyroglobulin (TGAb). The lipid profile (comprising TC, TG, HDL-C, and LDL-C) and glucose levels were measured using an auto-analyzer (ARCHITECT c8000 system; Abbott Laboratories, Chicago, IL, United States). Thyroid hormones were quantified using a chemiluminescent microparticle immunometric assay with an automated ARCHITECT Immulite 2000 SR analyzer (Abbott, Longford, Ireland).
Statistical analysis
The Kolmogorov-Smirnov one-sample test was used to assess the normality of the distribution. Normally distributed variables were represented as the mean ± SD, while non-normally distributed variables were represented as the median. Categorical variables were expressed as n (%). To analyze differences between TSH tertiles, the Kruskal-Wallis rank sum test was applied to continuous variables, and the Fisher exact probability test was used for categorical variables. A multivariate logistic regression model was employed to evaluate the relationship between TSH levels and SAs. Both unadjusted and adjusted odds ratios (ORs) with 95%CI were reported. To ensure the robustness of the data analysis, a sensitivity analysis was performed. TSH levels were converted into categorical variables, and the trend P value was calculated. The variance inflation factor was used to assess multicollinearity between independent variables. Any covariates with a variance inflation factor > 5.0 were excluded from the final model. Each covariate was included individually in the model; if a covariate changed the TSH estimate of SAs by > 10% or was significantly associated with SAs (P < 0.10), it was included in the final model as a potential confounder. Three different models were constructed to validate the stability of the results: (1) An unadjusted model; (2) Model I (adjusted for age and gender); and (3) Model II (adjusted for age, gender, HAMA, HAMD, TGAb, TPOAb, TC, HDL-C, LDL-C, body mass index, systolic blood pressure, and diastolic blood pressure). The nonlinear relationship between TSH levels and SAs was assessed using a smoothing plot, and the threshold effect was examined through a two-piecewise linear regression model based on the generalized estimation equation. Interaction and stratified analyses were conducted based on gender, marital status, education, comorbid anxiety, and psychotic symptoms.
RESULTS
Demographic characteristics and clinical parameters
Table 1 presents the characteristics of participants categorized according to serum TSH tertiles. Significant correlations were observed between these tertiles and various variables, including age, duration of illness, age at onset, HAMD scores, HAMA scores, TGAb, TPOAb, FBG, TC, TG, HDL-C, LDL-C, body mass index, systolic blood pressure, diastolic blood pressure, comorbid anxiety, psychotic symptoms, and SAs (all P < 0.05). Figure 1 illustrates the distribution of TSH levels in FEDN-MDD patients, both with and without SAs.
Figure 1 Distribution of thyroid-stimulating hormone levels in first-episode drug-naïve major depressive disorder patients with or without suicide attempt.
TSH: Thyroid-stimulating hormone; SA: Suicide attempt.
Table 1 Sociodemographic and clinical features of the participants (n = 1718), mean ± SD/n (%)/median (tertile low-tertile high).
Thyroid-stimulating hormone index (μIU/mL) tertiles
The fully adjusted data (Table 2) showed a substantial association between a higher serum TSH level and a greater risk of SAs (OR = 1.09 ,95%CI: 1.01-1.17; P < 0.05). To further investigate the relationship between serum TSH levels and SAs, generalized additive models were utilized, as illustrated in Figure 2. A non-linear relationship between SAs and TSH levels was depicted in the smoothing plot after adjusting for disease duration, age of onset, education, HAMD, HAMA, FBG, TGAb, TPOAb, TC, TG, HDL-C, LDL-C, and systolic and diastolic blood pressure (Figure 2). Through a two-segment logistic regression model, the inflection point of TSH was found to be 5.43 μIU/mL in the relationship between TSH levels and SAs in patients with FEDN-MDD. Each unit increase in serum TSH level on the right side of the inflection point substantially increased the likelihood of SAs, by 21% (OR = 1.21, 95%CI: 1.08-1.36, P < 0.001). Nevertheless, there was no statistically significant link between TSH levels and SAs on the left side of the inflection point (OR = 0.94, 95%CI: 0.82-1.07, P = 0.339) (Table 3). A sensitivity analysis was conducted to assess the impact of TSH outliers (n = 12, TSH > 10 μIU/mL) using winsorization at the 95%. This confirmed the original inflection point at 5.43 μIU/mL (OR = 1.21, 95%CI: 1.08-1.36, P < 0.001 for TSH > 5.43). Excluding these outliers shifted the inflection point to 5.1 μIU/mL, but reduced statistical power. The results are detailed in Supplementary Table 1.
Figure 2 Association between thyroid-stimulating hormone levels and suicide attempt.
A non-linear association between thyroid-stimulating hormone levels and suicide attempt was found in the multivariable logistic regression model. Solid line represents the smooth curve fit between variables. Dotted lines represent the 95%CI from the fit. All adjusted for disease duration, age of onset, education, Hamilton Depression Scale, Hamilton Anxiety Scale, fasting blood glucose, anti-thyroglobulin, thyroid peroxidase antibody, total cholesterol, triglycerides, high-density lipoprotein cholesterol, low-density lipoprotein cholesterol, systolic and diastolic blood pressure. TSH: Thyroid-stimulating hormone.
Table 2 Relationship between thyroid-stimulating hormone levels and suicide attempt in different models.
Thyroid dysfunction prevalence and relationship between FT3/FT4 and SA
The prevalence of subclinical hypothyroidism was 1034/1718 (60.19%), and the prevalence of hypothyroidism was 0/1718 (0.00%) as shown in Table 4. FT3 and FT4 were also evaluated in logistic regression models for SA. In the fully adjusted model, neither FT3 (OR = 0.970, 95%CI: 0.801-1.175, P = 0.758) nor FT4 (OR = 0.980, 95%CI: 0.938-1.024, P = 0.360) showed a significant association with SA (Table 5).
Figure 3 reveals significant relationships between gender, education, marital status, comorbid anxiety, and psychotic symptoms in relation to SAs. However, some factors did not appear to significantly influence the risk of SAs in conjunction with each other. Notably, individuals exhibiting psychotic symptoms demonstrated a markedly increased risk (OR = 1.73, 95%CI: 1.47-2.03, P < 0.001), highlighting the necessity for clinicians to prioritize mental health assessments. Further research is warranted to elucidate the complex role of TSH levels among these critical variables and to enhance intervention strategies.
Figure 3 Subgroup analysis of the association between thyroid-stimulating hormone and suicide attempt.
The odds ratio (95%CI) was derived from the logistic regression model (age, sex, education, duration of illness, Hamilton Depression Scale, Hamilton Anxiety Scale, anti-thyroglobulin, thyroid peroxidase antibody). OR: Odds ratio.
DISCUSSION
In this cross-sectional study utilizing a population-based sample, we investigated the relationship between serum TSH levels and SAs, accounting for various covariates. The results indicated a significant association, showing that elevated serum TSH levels were associated with SAs. Furthermore, we found a non-linear correlation between TSH levels and SAs, with the inflection point occurring at 5.43 μIU/mL. On the right side of the inflection point, TSH levels were positively correlated with SAs, while on the left side of the inflection point, the correlation was not remarkable. These findings were consistent across categories, including gender, education level, marital status, and concomitant anxiety.
Several previous studies have yielded findings consistent with the present results, reinforcing the potential link between elevated TSH levels and SAs in patients with MDD. For example, Zhang et al[23] reported that TSH levels were significantly higher in adolescent suicide attempters compared to non-attempters within an MDD population. Similarly, a recent systematic review and meta-analysis examining the association between peripheral hormone levels and suicidal behavior concluded that suicide attempters tend to exhibit higher TSH concentrations than non-attempters[21]. Additionally, a large-scale clinical study involving 1279 first-time MDD patients from Chinese outpatient clinics demonstrated that individuals with SAs showed notably increased serum TSH levels relative to those without such behaviors[28]. Collectively, these findings suggest that elevated TSH levels may serve as a promising biological marker for identifying individuals at heightened risk of SAs among MDD patients. Based on existing literature and physiological understanding, several potential mechanisms may underlie this association. First, compensatory failure of the thyroid axis in chronic depression is a significant factor. In patients with long-standing or severe depressive episodes, the thyroid gland may lose its compensatory regulatory capacity, resulting in subclinical or overt hypothyroidism[34]. Hypothyroidism has long been associated with various psychiatric manifestations, including suicidal ideation and behavior[35]. A striking case report detailed a woman with severe hypothyroidism and a TSH level of 152 mIU/L (reference range: 0.20-5.10), who exhibited multiple violent SAs[36]. Moreover, in a cohort study involving 1410 patients with primary MDD, the prevalence of hypothyroidism was 13.2%, and those with comorbid hypothyroidism exhibited more severe depressive symptoms, greater psychotic features, and higher rates of psychiatric hospitalization[37]. These findings suggest that elevated TSH may increase suicide vulnerability in MDD through central neurobiological pathways and MDD-specific dysregulation. First, elevated TSH, especially when FT3/FT4 remain largely within the reference range, may indicate a shifted central set point of the hypothalamic-pituitary-thyroid (HPT) axis. This reflects altered hypothalamic thyrotropin-releasing hormone (TRH) drive or impaired negative-feedback sensitivity rather than frank peripheral hypothyroidism. Such central HPT-axis perturbations can influence key suicide-relevant neural systems: (1) Thyroid hormone signaling modulates serotonergic (serotonin) neurotransmission, including receptor sensitivity and downstream signaling involved in mood regulation, impulsivity, and stress reactivity[38]; (2) HPT-axis activity interacts with dopaminergic pathways and limbic reward circuits, potentially linking higher TSH states to anhedonia, motivational dysregulation, and stress-related vulnerability[39]; and (3) The HPT axis is tightly coupled to hypothalamic stress systems (including the HPA axis and circadian regulation), which are commonly disrupted in MDD[40]. In parallel, elevated TSH may also index systemic health deterioration, as higher TSH has been associated with cerebrovascular disease, hypertensive crises, acute myocardial infarction, dyslipidemia, and thyroid malignancies[41-43]. These conditions can directly or indirectly exacerbate psychological distress and functional burden, thereby amplifying suicidality[44,45]. Consistent with impaired HPT-axis feedback, reduced TSH responsiveness to TRH has been observed in depression and suicidality[46]. Moreover, a blunted TSH response to TRH administration has been proposed as a biomarker of suicide risk in depressed individuals[47]. Importantly, MDD may amplify these effects because monoaminergic and stress-regulatory circuits are already sensitized; thus, subtle shifts in HPT-axis feedback may translate into disproportionately larger behavioral consequences than in non-depressed populations. This MDD-specific vulnerability is further supported by evidence from FEDN-MDD cohorts, in which TSH differences are more consistently observed when comparing patients with and without SAs. Such findings suggest that TSH may index a suicide-risk–relevant biological phenotype within MDD rather than merely reflecting nonspecific psychological distress[48]. Finally, our results that FT3/FT4 were not independently associated with SAs, together with the very low prevalence of overt hypothyroidism in this cohort, argue against a simple peripheral thyroid hormone deficiency explanation; instead, they are more consistent with altered central HPT-axis regulation/feedback as a plausible mechanistic substrate for the observed nonlinear threshold relationship between TSH and SAs.
Nevertheless, some studies have reported findings that diverge from the present results. For instance, a comprehensive meta-analysis found no statistically significant difference in TSH levels between individuals exhibiting suicidal behavior and those without such behavior[35]. Similarly, De Sousa et al[49] identified a negative correlation between TSH levels and suicidal behavior among patients experiencing first-episode schizophrenia, indicating that lower TSH levels were associated with increased suicide risk in that population. In another study, Duval et al[47] compared 122 hospitalized patients with MDD and comorbid suicidal behavior disorder to 50 healthy controls, revealing that violent suicide attempters exhibited significantly lower TSH levels than their healthy counterparts. These inconsistencies across studies may be attributable to several interrelated factors. One of the most prominent confounding variables is the influence of pharmacotherapy. Many psychotropic medications, especially antidepressants, have been shown to alter thyroid function. Common classes of antidepressants, such as tricyclic antidepressants, monoamine oxidase inhibitors, selective serotonin reuptake inhibitors, and serotonin-norepinephrine reuptake inhibitors, impact TSH secretion and peripheral thyroid hormone metabolism[50]. Seven literatures suggest that antidepressant use can either suppress or enhance thyroid activity, depending on the specific drug and dosage employed[51-57]. Furthermore, a systematic review and meta-analysis encompassing 17 observational studies found that antidepressant exposure significantly elevated the risk of suicide and SAs among children and adolescents compared to non-exposed controls[58]. Thus, discrepancies in medication use across studies may explain divergent outcomes in the association between TSH levels and suicidality. Additionally, not all antidepressants exert equal effects on suicide risk. For example, treatment with mirtazapine has been associated with a significantly higher suicide risk ratio compared to citalopram, a commonly prescribed selective serotonin reuptake inhibitor[59]. The varying impact of specific medications may lead to inconsistent findings in studies that do not stratify participants by drug class or treatment duration. Moreover, differences in sample characteristics, including psychiatric diagnosis (e.g., MDD and schizophrenia), disease phase (first-episode vs recurrent), symptom severity, and history of treatment, can contribute to variation in outcomes. For instance, FEDN-MDD patients may exhibit distinct neuroendocrine profiles compared to multi-episode, medicated individuals, thus confounding attempts to generalize across populations. Finally, cultural and ethnic differences can also influence both thyroid function and mental health outcomes. Variations in dietary iodine intake, genetic predispositions, healthcare access, and cultural attitudes toward mental illness and suicide may lead to population-level differences in the manifestation of both depression and thyroid dysfunction. In summary, the relationship between TSH levels and SAs is likely multifactorial, shaped by a complex interplay of biological, pharmacological, demographic, and sociocultural variables. These findings underscore the importance of adopting a nuanced, stratified approach when interpreting the role of thyroid function in mental health research. Future studies should aim to control for medication status, utilize homogeneous diagnostic criteria, and explore ethnically diverse cohorts to more accurately delineate the mechanisms linking TSH dysregulation to suicidal behavior.
Interestingly, in comparison with previous studies, our investigation revealed a non-linear association between TSH levels and SAs. Specifically, we identified an inflection point at 5.43 mIU/L, beyond which the relationship between TSH and SAs became statistically significant. Below this threshold, there was no meaningful correlation, suggesting that within and slightly above the clinically accepted reference range, TSH fluctuations may not exert a significant impact on suicidal behavior. Notably, the standard reference range for TSH is 0.27-4.20 mIU/L. The identified inflection point exceeds the upper limit of this normal range by 1.23 mIU/L, indicating that risk elevation occurs only when TSH levels become pathologically elevated. In the range 4.20-5.43 mIU/L, although some fluctuations in SA rates were observed, these variations remained statistically nonsignificant and clinically subtle. This may be attributed to the homeostatic mechanisms of the human body, which are capable of compensating for mild hormonal disturbances through feedback regulation and adaptive responses. Such mechanisms likely help maintain a dynamic balance between psychological and physiological functions, thereby buffering the impact of hormonal irregularities within moderate deviations from the norm. However, once TSH levels cross the 5.43 mIU/L threshold, this compensatory capacity may become overwhelmed. Our data suggest that for every unit increase in serum TSH beyond this inflection point, the risk of SAs increases by 21%, indicating a clinically meaningful escalation in mental health vulnerability. This finding underscores the importance of considering subclinical or overt hypothyroidism as a potential contributor to suicidal behavior, especially in individuals with borderline or elevated TSH levels. From a clinical perspective, identifying this inflection point is crucial. It provides a quantifiable biomarker for early identification and intervention in at-risk populations. Routine thyroid function screening, particularly in psychiatric assessments, could facilitate timely preventive strategies. Moreover, this finding may prompt a reevaluation of current TSH reference intervals, especially in mental health populations, and support further exploration into individualized thresholds for hormonal risk assessment.
While this study provides valuable insights into the relationship between TSH levels and SAs in patients with MDD, several limitations must be acknowledged, which may impact the generalizability and interpretation of the findings. Firstly, the study was conducted at a single center, specifically at the psychiatric outpatient department of a general hospital in Taiyuan, Shanxi Province, China. All participants were Han Chinese, which introduces a notable lack of ethnic and regional diversity. This homogeneity restricts the external validity of the findings, as the results may not be applicable to populations from different geographical regions or ethnic backgrounds. In addition, SAs were identified based on patient interviews and medical records, and were assessed as lifetime history without capturing the timing (e.g., recent vs remote attempts) relative to the baseline thyroid assessment. This reliance on retrospective data may have introduced recall bias, as patients’ recollections of past SAs may not be fully accurate, and the current TSH level may not coincide with the time window of SA occurrence, thereby precluding causal or temporal inferences. Such bias could lead to either overreporting or underreporting of suicidal behaviors, which could influence the study’s conclusions. Moreover, another limitation lies in the variability of TSH testing methodologies across different hospitals. This study did not account for potential inconsistencies in testing procedures, which can result in varying reference standards for TSH levels. As a result, caution is warranted when interpreting the inflection point of 5.43 mIU/L identified in the study, as this threshold may not be universally applicable. Furthermore, the cross-sectional design of the study prevents the establishment of a causal relationship between serum TSH levels and SAs. Although an association was observed, the inability to establish temporal causality means that we cannot definitively conclude that abnormal TSH levels contribute to suicidal behavior or vice versa. In addition to these design-related constraints, the interpretability of the TSH-SA association may also be influenced by neurobiological heterogeneity within MDD. Our study establishes a nonlinear TSH-SA association in FEDN-MDD. Emerging evidence suggests that this link may vary across neurobiologically distinct MDD subtypes, including those characterized by different functional connectivity patterns (e.g., hypoconnectivity vs. hyperconnectivity) and divergent neurotransmitter profiles. For instance, Li et al[60] highlights heterogeneity in brain functional connectivity, transcriptome, and neurotransmitter profiles in MDD, and Luo et al[61] reviews advances in psychoradiology emphasizing the integration of imaging markers with psychiatric biomarkers. These observations underscore that subtype-related biological differences could contribute to between-study variability and motivate prospective, subtype-stratified investigations to refine the clinical and mechanistic interpretation of thyroid-related signals. Additionally, the study focused on FEDN patients diagnosed with MDD. While this specific population was chosen to control for the effects of medication on TSH levels, there is potential for selection bias. During enrollment, it was not possible to fully exclude patients with bipolar disorder, as initial depressive episodes may be difficult to distinguish from the onset of bipolar disorder. This overlap in diagnostic criteria could have introduced heterogeneity into the sample, as patients with bipolar disorder may exhibit different neurobiological characteristics and TSH responses compared to those with unipolar depression. Lastly, several confounding variables that may influence the relationship between TSH levels and SAs were not assessed in this study. Factors such as non-disorder alcohol use, or broader markers of subclinical inflammation, smoking, social status, personality traits, dietary iodine intake, history of thyroid disease or thyroid-related medication use, menstrual/menopausal status, and other biological markers (e.g., cortisol levels) may significantly impact both TSH regulation and suicidal behavior, which may further contribute to residual confounding. The absence of these variables in the analysis limits the study’s ability to fully account for the complex interplay between biological, environmental, and psychological factors. Future research should take these variables into account to enhance our understanding of the pathophysiological mechanisms linking TSH levels and SAs in patients with MDD.
CONCLUSION
We found a nonlinear relationship between TSH levels and SAs in Chinese FEDN-MDD patients, with an inflection point at 5.43 μIU/mL. Specifically, TSH levels on the right side of the inflection point were positively correlated with SAs, whereas no significant correlation was observed on the left side. These findings suggest that TSH levels could serve as a promising biomarker for evaluating SAs in MDD patients. However, further prospective cohort studies are needed to validate our results.
Huang Y, Wang Y, Wang H, Liu Z, Yu X, Yan J, Yu Y, Kou C, Xu X, Lu J, Wang Z, He S, Xu Y, He Y, Li T, Guo W, Tian H, Xu G, Xu X, Ma Y, Wang L, Wang L, Yan Y, Wang B, Xiao S, Zhou L, Li L, Tan L, Zhang T, Ma C, Li Q, Ding H, Geng H, Jia F, Shi J, Wang S, Zhang N, Du X, Du X, Wu Y. Prevalence of mental disorders in China: a cross-sectional epidemiological study.Lancet Psychiatry. 2019;6:211-224.
[RCA] [PubMed] [DOI] [Full Text][Cited by in Crossref: 1960][Cited by in RCA: 1586][Article Influence: 226.6][Reference Citation Analysis (1)]
Kern DM, Canuso CM, Daly E, Johnson JC, Fu DJ, Doherty T, Blauer-Peterson C, Cepeda MS. Suicide-specific mortality among patients with treatment-resistant major depressive disorder, major depressive disorder with prior suicidal ideation or suicide attempts, or major depressive disorder alone.Brain Behav. 2023;13:e3171.
[RCA] [PubMed] [DOI] [Full Text][Cited by in RCA: 26][Reference Citation Analysis (0)]
Nock MK, Borges G, Bromet EJ, Alonso J, Angermeyer M, Beautrais A, Bruffaerts R, Chiu WT, de Girolamo G, Gluzman S, de Graaf R, Gureje O, Haro JM, Huang Y, Karam E, Kessler RC, Lepine JP, Levinson D, Medina-Mora ME, Ono Y, Posada-Villa J, Williams D. Cross-national prevalence and risk factors for suicidal ideation, plans and attempts.Br J Psychiatry. 2008;192:98-105.
[RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)][Cited by in Crossref: 2001][Cited by in RCA: 1848][Article Influence: 102.7][Reference Citation Analysis (0)]
Gokalp G, Berksoy E, Bardak S, Demir G, Demir S, Anil M. Is there a relationship between thyroid hormone levels and suicide attempt in adolescents?Arch Clin Psychiatry. 2020;47:130-134.
[PubMed] [DOI] [Full Text]
Fiore E, Rago T, Provenzale MA, Scutari M, Ugolini C, Basolo F, Di Coscio G, Berti P, Grasso L, Elisei R, Pinchera A, Vitti P. Lower levels of TSH are associated with a lower risk of papillary thyroid cancer in patients with thyroid nodular disease: thyroid autonomy may play a protective role.Endocr Relat Cancer. 2009;16:1251-1260.
[RCA] [PubMed] [DOI] [Full Text][Cited by in Crossref: 190][Cited by in RCA: 151][Article Influence: 8.9][Reference Citation Analysis (0)]
Keen F, Williams DM, Essame J, Udiawar M, Nagarajah K, Witczak J, Mitchem K, Kalhan A. Isolated central hypothyroidism: Underlying pathophysiology and relation to antidepressant and antipsychotic medications-A multi-centre South Wales study.Clin Endocrinol (Oxf). 2024;100:245-250.
[RCA] [PubMed] [DOI] [Full Text][Cited by in RCA: 4][Reference Citation Analysis (0)]
