Introduction
Persistent, excessive, and uncontrollable feelings of anxiety, fear, or apprehension about everyday activities are the common characteristics of generalized anxiety disorder (GAD) [1, 2]. It is a significant public health concern that affects personal, professional, educational, and social lives [3]. Severe emotional distress is the prime characteristic of GAD [4, 5, 6].
Diagnosis of GAD is based on prolonged restlessness, irritability, distress, perspiration, and tremors over something consistently for over 6 months [7]. Despite having only a 3%–6% prevalence rate among the world population [8, 9, 10], the number is rising every day due to the complexities of modern lifestyles of people in today's society [11]. Underdiagnosed and untreated excessive worry can debilitate cognitive and mental functions, which ultimately leads to a poorer quality of life [2, 10].
Moreover, suicidal tendencies [12] and cardiovascular disease risks [13, 14] are higher among GAD patients. The pathophysiology of GAD is still not fully understood, which makes it difficult to treat the disease effectively. Different factors like stress, genes, comorbidities, childhood trauma, and environmental elements may play roles in the etiology of GAD.
Multiple studies suggest a link between GAD and altered monoamine neurotransmission in the limbic system [7]. Moreover, decreased GABAergic activity and increased glutamatergic activity in the amygdala are indicated to be responsible for the pathogenesis of GAD [7, 10].
Two primary therapeutic approaches exist for GAD patients: pharmacotherapy and cognitive behavioral therapy. Combining these therapies may yield the most beneficial outcomes. Among the pharmacotherapies, selective serotonin reuptake inhibitors, pregabalin, and serotonin norepinephrine reuptake inhibitors are given as the first‐line treatments [9, 10].
However, 50% patients do not respond to the first‐line pharmacotherapies [15]. This lack of efficacy suggests limitations in the monoamine neurotransmitter hypothesis of GAD pathophysiology, indicating a need to explore additional pathophysiological mechanisms. The absence of specific biomarkers for early GAD diagnosis complicates its clinical management.
Various research indicated the role of neuroinflammation in the etiology of different psychiatric disorders [16, 17, 18, 19]. Moreover, dysregulated immune responses have been implicated as a possible developmental pathway of GAD in many recent studies [15, 20, 21, 22]. Proinflammatory cytokines administration in mice models can induce anxiety‐like behaviors, which in turn became normal after providing antagonists of these cytokines [23, 24, 25].
Circulating pro‐inflammatory cytokines can trespass our CNS and eventually stimulate microglia or astrocytes to release additional inflammatory mediators, such as interleukins and TNF‐α. A similar inflammatory cycle can impair the neurotransmission of dopamine, noradrenaline, and serotonin, which are associated with mood and behavior. Also, inflammation‐causing cytokines can modify the neurotransmission pathways, which results in either decreased monoamine neurotransmitters inside the brain or increased glutamatergic excitatory activity to produce oxidative stress, damaging neural circuits to produce anxiety and depression [1, 17].
In a recent study, it has been demonstrated that SSRIs can decrease circulating cytokines, which are inflammatory, and the anti‐inflammatory actions of SSRIs could be a cause of relieving anxiety [26].
TNF‐α is a pro‐inflammatory cytokine that is synthesized by various cells in our body including the neuronal cells in CNS [27]. TNF‐α stimulates inflammatory cascades to develop pain and inflammation. TNF‐α interacts with two receptors, TNF‐R1 and TNF‐R2; its binding to TNF‐R1 initiates cellular events that activate nuclear factor κ‐B (NF‐κB) [28], a transcription factor crucial for the expression of several pro‐inflammatory cytokines, such as TNF‐α, IL‐1β, and IL‐6 [29].
This TNF‐α/NF‐κB interaction establishes a positive feedback loop, enhancing the inflammatory response. Peripheral inflammation can enhance the expression of TNF‐α, which in turn may initiate neuropathy [30]. It is a crucial cytokine for the initiation and progression of neuroinflammation in the CNS.
It can activate microglia in the CNS for increased production of inflammation‐inducing cytokines like IL‐1β, IFN‐γ, and IL‐6, while other proinflammatory cytokines are dependent on NF‐κB for expression [31, 32]. In addition, TNF‐α is responsible for TNF‐R1‐mediated interference of the blood‐brain barrier to infiltrate the CNS with inflammatory molecules [33]. TNF‐α can transform neurons to a greater excited and barely inhibited condition to regulate the connectivity of synapses as a part of homeostasis [34].
TNF‐α can induce anxiety in mice models by enhancing excitatory glutamatergic neurotransmission and inhibiting inhibitory GABA neurotransmission in BLA (Basolateral Amygdala) [35]. Arresting TNF‐signal transduction in peripheral inflammation can decrease anxiety symptoms, which implies a possible role of this cytokine in developing anxiety [36]. In addition, intracerebroventricular TNF‐α injection in mice intensifies anxiety‐like behaviors [25].
TNF‐α inhibitors, such as infliximab, can decrease anxiety‐like behaviors in chronic mild stress [37]. Another study showed that mice deficient in TNF‐R1 and TNF‐R2 manifested decreased anxiety‐like behaviors [38]. However, while several studies suggest a correlation between TNF‐α and anxiety, research directly examining TNF‐α's role in GAD pathophysiology remains limited.
At present, there are no cost‐effective and clinically validated biomarkers for GAD, though some neuroimaging methods like SPECT (Single‐photon Emission Computed Tomography), MRI (Magnetic Resonance Imaging), and DIT (Diffusion Tensor Imaging) are useful in GAD diagnosis. These imaging modalities can reveal brain abnormalities correlating with GAD symptom severity [10]. However, few of these techniques are practical for ongoing treatment monitoring of brain function.
These sophisticated techniques are not only expensive for therapeutic drug monitoring but also less available in all clinical settings. Presently, clinicians primarily rely on subjective self‐reported symptoms and responses to a seven‐item scale (GAD‐7) to diagnose GAD, a method prone to bias and misdiagnosis [39]. The unavailability of any clinical biomarker makes the disease difficult to diagnose and treat at an earlier stage.
So, finding an effective and inexpensive GAD biomarker is crucial. Though various studies investigated the potential link between proinflammatory cytokines and GAD, data correlating TNF‐α and GAD remains limited. The findings that are available have also been inconsistent.
Some studies have reported raised TNF‐α in patients with GAD, while others have found no difference from healthy controls, and the reasons behind this variation have not yet been settled [10]. Moreover, no studies in Bangladesh have evaluated peripheral TNF‐α levels in GAD. This study aimed to investigate TNF‐α's potential role in GAD development and pathogenesis.
Methods
Study Population
This case‐control study was carried out between March 1, 2024, and June 30, 2024, and has been reported in line with the STROBE statement. The study included 43 GAD patients as cases and 26 healthy controls (HCs). Participants in these study cases were recruited from Department of Psychiatry at Bangladesh Medical University (BMU).
HCs were enlisted from the patients' nearby residential areas based on age and sex. A qualified psychiatrist diagnosed the patients based on the DSM‐5 criteria and assessed symptom severity using the Generalized Anxiety Disorder 7‐item (GAD‐7) scale [40]. GAD‐7 scores can range from 0 to 21 and patients with a score of 5 or more were included in this study.
Participants aged between 18 and 60 years with no psychiatric comorbidities such as MDD, OCD, PTSD, panic disorders, schizophrenia, and social phobia were selected for the study. Additionally, persons with Parkinson's, Alzheimer's, liver or kidney diseases, chronic infection, alcoholism, and substance abuse were excluded from this study to nullify the influence of confounding factors to ascertain the best possible similarities among the study participants. We only included the subjects who were not exposed to antipsychotics, antidepressants, anxiolytics, and immunosuppressants for the last 2 weeks before the study to ensure unaffected cytokine levels of the person.
Of the patients enrolled, 23 were newly diagnosed and had not received any treatment before recruitment, while 20 had been treated earlier and completed the 2‐week washout described above. Each study subject provided informed written consent to participate in this study. We set an organized pre‐designed questionnaire to record the socio‐demographic data of the study participants.
Every aspect of this study was conducted in compliance with the Declaration of Helsinki.
Collection of Blood Samples, Serum Isolation, and Storage
Following standard collection procedures, 5 ml blood was drawn from each participant. After collecting the blood samples, all were kept unagitated in Falcon tubes for 1 h to facilitate coagulation. The samples were centrifuged to separate the serum at room temperature for 15 min at 3000 rpm, and the serum was then stored at −80°C in an Eppendorf tube.
Samples were held at this temperature for a maximum of 6 months before analysis. Each aliquot was thawed only once, immediately before the assay, so that no sample went through more than a single freeze‐thaw cycle.
Serum TNF‐α Estimation
The level of TNF‐α in serum was estimated by ELISA using a commercially available kit (Boster Bio, USA), following the manufacturer's guidelines. First, 100 µl of standards, samples, and controls were added to each well of a 96‐well pre‐coated microplate, sealed, and incubated at 37°C for 1.5 h. After discarding the contents of each well, 100 µl of biotinylated anti‐TNF‐α antibody was added.
Following that, a microplate was placed in the incubator at 37°C for 60 min. In the next step, wells were washed with a 300 µl buffer thrice. Following washing, 100 µl avidin‐biotin‐peroxidase complex was added to each of the wells and subjected to incubation at 37°C for 30 min.
After incubation, the microplate was washed five times with 300 µl buffer. Then, 90 µl color‐developing reagent (TMB) was included in each well and incubated for 30 min in the absence of light. We added 90 µl stop solution to terminate the reaction.
The absorbance of each well of the microplate in a microplate reader was measured at a wavelength of 450 nm. A standard curve was prepared by extrapolating the OD absorbance values and cytokine levels in pg/ml. Serum TNF‐α levels of the samples were then calculated using the standard curve equation.
The whole procedure is performed twice to confirm the reliability and reproducibility of the values.
Statistical Analysis
This study used SPSS version 25 (IBM Corp., Armonk, NY, USA) to analyze the data. Descriptive analysis was used to demonstrate the clinical variables and sociodemographic characteristics between groups. An independent sample t‐test (two‐tailed) was used to determine the statistical significance of differences between patients and controls in continuous variables, while a chi‐square test was applied for categorical variables.
To differentiate the levels of tumor necrosis factor‐α (TNF‐α) between patients and HCs, a box plot graph was generated for visual illustration, and a scatter plot graph was used to show the sex‐specific correlation between the levels of TNF‐α and GAD scores. Receiver Operating Characteristic (ROC) analysis was also conducted to evaluate the effectiveness of serum TNF‐α levels in differentiating GAD patients from HCs. All tests were two‐sided, and a significant in every instance, p‐value less than < 0.05 was considered statistically significant.
Results
Sociodemographic Features of Study Participants
A comprehensive comparison of sociodemographic features between individuals with GAD and HCs is shown in Table 1. Regarding education level, age, BMI sex, and marital status, GAD patients and HCs were very similar. Around 40% of the study subjects (41.86% for GAD patients and 42.31% for HCs) were between 26 and 35 years in both groups.
Most of the subjects participating in this study had normal BMI (81.39% for GAD patients and 73.08% for HCs). No significant differences were found regarding occupation, financial status, living area, and smoking history when comparing patients with HCs. However, 13.95% of patients had a family history of the disease whereas no family history was found in HCs (p = 0.003).
Biophysical Characteristics, Clinical Features, and Laboratory Findings
Analysis of the biophysical characteristics, clinical features, and laboratory outcomes are displayed in Table 2. The GAD patients had a mean age of 31.89 ± 10.92 years and HCs had 33.07 ± 11.22 years of mean age (p = 0.644). The mean values of BMI were 22.41 ± 3.67 and 22.66 ± 3.95 for patients and controls respectively (p = 0.773), indicating marked similarity between the two cohorts.
The patient group had a significantly higher (p < 0.001) mean GAD‐7 score (13.25 ± 5.14) than HCs (4.87 ± 2.31). GAD patients also showed higher TNF‐α levels (69.02 ± 27.09 pg/mL) compared to HCs (51.58 ± 26.14 pg/mL) in laboratory examination of the serum, which is statistically significant (p = 0.010) and displayed in Figure 1a.
Association Between Different Study Parameters
A series of correlation studies were carried out to evaluate the association among higher serum TNF‐α levels, age, BMI, and GAD‐7 scores in patients (Table 3). We found the association of age with neither GAD‐7 scores nor TNF‐α levels (p > 0.05). BMI showed no significant correlation with GAD‐7 scores (r = 0.096, p = 0.497).
Also, no association was found between BMI and TNF‐α levels (p > 0.05). On the other hand, higher TNF‐α levels of GAD patients were notably positively correlated with GAD‐7 scores (r = 0.457, p = 0.004), indicating a role of altered TNF‐α in GAD pathophysiology. We presented the correlation between serum TNF‐α and GAD‐7 scores by a sex‐specific scatter plot graph in Figure 1b.
Analysis of ROC Curve
A diagnostic performance assessment was conducted utilizing ROC curve analysis to evaluate the ability of serum TNF‐α to distinguish between GAD patients and HCs. With an area under the curve value of 0.888, the Serum TNF‐α estimation demonstrated great performance in identifying the patients, supported by a 95% confidence interval of 0.810 to 0.966 and a p‐value of less than 0.001. The analysis also found that serum TNF‐α evaluation had 84.6% sensitivity and 83.3% specificity, with 50.85 pg/mL as the cut‐off value, where a serum cytokine level above this value indicates GAD (Table 4) and curve is displayed in Figure 1c.
Discussion
This study is the first case‐control investigation to explore the potential association between TNF‐ α, a pro‐inflammatory cytokine, and the pathophysiology of GAD within the Bangladeshi population. Our findings indicated that both serum TNF‐α level in serum and GAD‐7 scores were significantly higher in GAD patients compared to HCs. Moreover, the elevation in TNF‐α levels demonstrated a significant positive correlation with GAD‐7 scores, suggesting a possible role of this pro‐inflammatory cytokine in modulating the severity of GAD and its associated symptoms.
Receiver Operating Characteristic (ROC) analysis further demonstrated a high predictive accuracy of serum TNF‐α measurements in distinguishing patients with the disease from HCs. This finding implies that TNF‐α levels may function as a viable biomarker for tracking progression of disorder, monitoring responses to anti‐anxiety medications, and in diagnosis. Nonetheless, additional longitudinal investigations are crucial to establish any causal connection between serum TNF‐α and the severity of the disease or its pathogenesis.
One study found that individuals with major depressive disorders showed significantly elevated serum TNF‐α levels, which then decreased to levels comparable to those of the control group following treatment with antidepressants. Furthermore, several researchers have noted an increased release of TNF‐α in women experiencing anxiety, along with a positive association between the intensity of anxiety symptoms and serum TNF‐α levels [41, 42]. Our findings of a notable elevation in TNF‐α among GAD patients align well with those reported in other research studies [1, 20, 43].
In contrast, our findings differ from previous studies that found no notable differences in TNF‐α concentration between patients and HCs. In one investigation, no irregularities in cytokine levels (TNF‐α, IL‐6, IL‐8) were observed in patients with GAD when compared to controls. This finding suggests that there are no immune system alterations associated with GSAD [44].
The largest study assessing TNF‐α (n = 1010) demonstrated no significant differences between individuals with GAD and control participants, and no correlation was established between TNF‐α levels and symptoms of anxiety [45]. This conclusion was further corroborated by an investigation involving 93 individuals with GAD and concurrent ischemic heart disease, which utilized a combined inflammatory index including TNF‐α, CRP, and IL‐6, reporting no differences when compared to control subjects [20]. In conclusion, while most of the small cross‐sectional studies that assessed the differences in TNF‐α levels between individuals with GAD and HCs found significantly elevated levels, notably, the largest study found no significant differences.
Several factors may account for these differences. Sample sizes have varied a great deal across studies, and the smaller ones are more prone to unstable estimates. The severity of illness and the proportion of participants already receiving medication also differ between cohorts, and both influence circulating cytokine levels.
Ethnic background and environmental exposure may shift baseline inflammatory activity as well. Finally, the ELISA kits and the sample handling procedures used are not uniform across laboratories, so absolute values from different studies are not always directly comparable.
The immune system and inflammation play a substantial role in the pathological advancement of numerous psychiatric conditions, including schizophrenia, Alzheimer's disease, major depressive disorder (MDD), anxiety disorder, and Huntington's disease. It has been noted that the concentrations of pro‐inflammatory cytokines, such as TNF‐α, along with other biomarkers, are significantly increased in the disorders mentioned above [43]. Even though depression and anxiety occur co‐currently and share correlations with metabolic and cardiovascular diseases, the investigation into neuroinflammation's role in the context of anxiety disorders has not been as comprehensive as it has been for depression.
Research has indicated elevated inflammatory markers in individuals of both genders diagnosed with late‐onset anxiety disorder, but a causal relationship has yet to be established. Moreover, one research also indicates a potential connection between increased serum TNF‐α and GAD, but the causal association between anxiety and inflammation still needs more exploration [46]. Clinical findings suggest that higher cytokine levels can affect neurotransmitters like glutamate and monoamines in brain regions linked to anxiety, for instance, the anterior cingulate cortex, insula, and amygdala [47].
Consistently, heightened TNF‐α levels, an important cytokine for apoptosis and cellular regulation, have been associated with GAD in human subjects [46]. Regarding the development of diagnostic markers, in this study, the measurement of TNF‐α serum levels has shown strong effectiveness in distinguishing individuals with GAD from controls. Additionally, the TNF‐α levels exhibited a significant and positive correlation with the severity of the disease, suggesting that TNF‐α levels in serum could serve as a potential objective biomarker for GAD.
One significant implication of this study's results is that TNF‐α signaling may be a viable target for investigating potential novel therapies. Anti‐TNF‐α medications may alleviate anxiety and depressive symptoms by influencing the alterations in neurotransmitter release or functions induced by elevated TNF‐α levels. The use of etanercept, which inhibits TNF‐α, has been shown to decrease depressive and anxiety like behaviors in mice model that exhibit inflammation and mood changes related to type‐2 diabetes [48].
Inhibiting TNF‐α has been demonstrated to produce an anxiety‐reducing effect in mice subjected to a peripheral immune challenge using lipopolysaccharide [36] as well as in those with experimental autoimmune encephalomyelitis. Additionally, the administration of infliximab, a TNF‐α‐blocking antibody, into the basolateral amygdala has been noted to reduce anxiety‐like responses in mice experiencing pain due to inflammation [41, 46]. Currently, we can suggest that TNF‐α might act as a biomarker for assessing vulnerability to anxiety disorders, aiding in the timely identification of the condition and promoting the adoption of treatment approaches.
The present study was observational, however, and no conclusion about treatment can be drawn from it. Interventional clinical trials will be needed to establish whether targeting TNF‐α has any therapeutic effect in patients with GAD.
GAD remains an under‐researched area in the field of psychiatric neuro‐inflammation when compared to other mental health conditions, such as schizophrenia and MDD. Long‐term investigations are essential to determine whether immune dysregulation serves as a precursor to anxiety or if the relationship is reciprocal. While we are still distant from utilizing inflammatory cytokines as a biomarker or therapeutic target for GAD, recent research indicates inflammatory alterations like those observed in other mental health conditions.
This highlights the importance of further investigation into the impact of inflammation on the development and advancement of GAD. Additionally, well‐structured, longitudinal studies that explore the mechanisms and relationships between inflammation and GAD, considering established factors that affect cytokine production, are crucial.
This investigation suggests that TNF‐α is a significant candidate for future studies exploring the causal relationship between inflammation and anxiety, as well as for potentially novel pharmacological trials. Prior studies focusing on anxiety and inflammation have either been limited to anxiety disorders within small clinical populations, conducted in cohorts with cardiovascular disease or restricted to non‐clinical samples. The strengths of this investigation include the establishment of comprehensive inclusion and exclusion criteria for participant recruitment, which helped us collect data from a homogeneous population.
Additionally, this study assessed the pro‐inflammatory cytokine TNF‐α while considering several variables, such as patient age, occupation, gender, body mass index, smoking habits, etc.
The primary limitation of this research is the limited sample size, with only 43 GAD patients and 26 HCs, which may not adequately represent the broader Bangladeshi population. An equal number of cases and controls would have provided a more balanced comparison. Additionally, the extent to which inflammation levels are independent of variables such as smoking, body mass index, and socioeconomic status remains uncertain.
As our study design is cross‐sectional, it lacks the temporal precedence and covariation required for causal inference, limiting our ability to determine causality. Several factors that are known to influence inflammatory markers were also not recorded, among them recent infection, changes in body weight, physical activity, dietary habits, sleep quality and hormonal status. Since this information was not collected, residual confounding from these sources cannot be ruled out, and we would recommend that future studies record them.
Conclusion
Serum TNF‐α levels were significantly elevated in GAD patients compared to the control groups. Furthermore, there was a notable positive correlation between TNF‐α levels and the severity of anxiety symptoms, suggesting a potential role of this cytokine in the pathophysiology of GAD. These findings highlight the involvement of inflammation in anxiety disorders and suggest that TNF‐α may be a candidate biomarker for anxiety severity.
Given the cross‐sectional design and the single‐center sample, this will require validation in multicentre longitudinal studies with larger and more diverse populations before any clinical use can be considered. Research in the future should focus on investigating the causal relationship between TNF‐α and anxiety, possibly through longitudinal studies that assess changes in TNF‐α levels over time. Additionally, investigating the effects of anti‐inflammatory treatments on anxiety symptoms could provide significant insights into the management of GAD and its underlying mechanisms.
Author Contributions
Tasnu Ara Nur: conceptualization, methodology, data curation, formal analysis, writing – original draft. Zahra Labiba Ahmed: conceptualization, writing – original draft, methodology, formal analysis, data curation. Faisal Abdullah Emon: data curation, investigation, project administration.
Taslima Jamal Urmi: investigation, project administration, data curation. Mohammad Masud Parves: investigation, data curation, project administration. Md.
Rabiul Islam: conceptualization, methodology, supervision, writing – review and editing. Zabun Nahar: conceptualization, writing – original draft, formal analysis, data curation, methodology.
Funding
The authors did not receive any specific funding for this research. No supporting source or financial relationship had any role in the study design, in the collection, analysis and interpretation of data, in the writing of the report or in the decision to submit the report for publication.”
Ethics Statement
The Research Ethics Committee of the University of Asia Pacific (UAP), Dhaka, Bangladesh, approved the study protocol (UAP/REC/2024/115) on February 22, 2024. The study was conducted between March 1, 2024, and June 30, 2024. The informed written consent was obtained from all participants prior to data collection.
The investigations were conducted in accordance with the principles outlined in the Declaration of Helsinki.
Consent
We received written consent from each participant after giving them a thorough explanation about the purpose of the study. The primary caregiver gave the sign on behalf of patients if independent thinking capacity was suspected or from those with no formal education.
Conflicts of Interest
The authors declare no conflicts of interest.
Transparency Statement
Md. Rabiul Islam affirms that this manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned have been explained.