Do serum biomarker levels correlate with anxiety disorders?

Abstract

BACKGROUND

It is clinically meaningful to differentiate healthy controls from patients with anxiety disorders who have a history of violence. This study evaluated clinical features and alterations in the serum biomarkers of patients with anxiety disorders and a history of violence and compared them with healthy controls.

AIM

To explore serum biomarkers for differentiating patients with anxiety disorders and a history of violence from healthy controls.

METHODS

This retrospective study included 31 patients with anxiety disorders and a history of violence (mean age: 35.0 ± 10.3 years) and 31 healthy controls (mean age: 29.0 ± 10.5 years). ELISA kits were used to test serum biomarkers, followed by analysis of demographic and biomarker data.

RESULTS

Compared with healthy controls, anxiety patients with a history of violence had lower serum levels of brain-derived neurotrophic factor (BDNF), aquaporin 4, and mature BDNF (mBDNF), as well as higher levels of BDNF precursor (proBDNF) and glycogen synthase kinase 3β (all P < 0.05). Receiver operating characteristic analyses indicated strong discriminative performance of the combined diagnosis biomarker profile in this case-control study, with an area under the curve value of 0.9698. Significant positive correlations were found between levels of BDNF and mBDNF, BDNF and proBDNF, and mBDNF and proBDNF.

CONCLUSION

Clinical features and biomarkers differentiate patients with anxiety and a history of violence from healthy controls, suggesting their potential diagnostic value warranting further evaluation.

Key Words: Anxiety disorder; Violence; Brain-derived neurotrophic factor; Aquaporin 4; Glycogen synthase kinase 3β; Mature brain-derived neurotrophic factor; Brain-derived neurotrophic factor precursor

Core Tip: This retrospective study investigated the role of serum biomarkers in distinguishing between healthy controls and patients with anxiety disorders and a history of violence. The case group exhibited reduced levels of serum brain-derived neurotrophic factor (BDNF), mature BDNF, and aquaporin-4, along with elevated levels of BDNF precursor and glycogen synthase kinase-3β. The combined diagnosis biomarker panel demonstrated significant discriminatory capacity (area under the curve = 0.9698). This combined diagnosis holds potential diagnostic value for identifying these patients with anxiety disorders.

INTRODUCTION

In today's society, anxiety disorders are prevalent mental health conditions, often causing patients to feel intense worry and fear[1]. Anxiety disorders are neurological disorders characterized by widespread, persistent anxiety or recurrent panic attacks, often with autonomic symptoms and motor tension[2]. Anxiety can stem from a variety of sources, including family environments, social interactions, physical or psychological trauma, and exposure to violent behaviors such as bullying, discrimination, or violence during one’s youth[3,4].

Brain-derived neurotrophic factor (BDNF) is a protein widely distributed in the brain that is widely studied. It controls neuronal and glial development, neuroprotection, and synapses[5]. BDNF plays a vital role in neuronal survival, differentiation, and development. The BDNF gene encodes a precursor peptide, BDNF precursor (proBDNF). Through intracellular and extracellular proteases, proBDNF is converted to mature BDNF (mBDNF). Studies have reported reduced BDNF levels in the blood and brain in anxiety and depression[6,7]. Aquaporin-4 (AQP4) is the major central nervous system aquaporin regulating water transport and homeostasis[8,9]. Loss of AQP4 impacts calcium channels and glutamate reuptake, resulting in reduced BDNF secretion[10,11]. Changes in AQP4 may play a role in anxiety with a history of violence. Glycogen synthase kinase-3β (GSK3β) regulates pathways for protein synthesis, proliferation, and apoptosis. GSK3β dysregulation increases the risk of mood disorders. Research has linked GSK3β to anxiety with a history of violence[12,13].

Treatment for anxiety involves psychotherapy and drugs. However, no biomarkers have been approved for psychiatric diagnosis, and clinical criteria are subjective[14]. We measured levels of BDNF family members, AQP4, and GSK3β in 31 anxiety patients with a history of violence to explore relationships and potential clues for diagnosis.

MATERIALS AND METHODS

Assessment of history of violence

In this study, history of violence is defined as a personal history of violence victimization, including various interpersonal harms and passive, traumatic experience of violence[15]. Childhood bullying; child abuse; exposure to domestic, intimate partner, or community violence; and other related traumatic experiences are major components. The validated Violence Exposure Scale and Conflict Tactics Scale were employed to measure violence exposure[16].

Participants

We identified and retrospectively reviewed the records of 31 patients with anxiety disorders and a history of violence from a hospital in Hangzhou, China (July 2023 to June 2025). Thirty-one healthy blood donors from a hospital in Hangzhou (China) were selected as control subjects. Informed consent was waived for this study by the Medical Ethics Committee of the Affiliated Hospital of Hangzhou Normal University. The study was conducted in accordance with principles of the Declaration of Helsinki. The case group had a mean age of 35.0 ± 10.3 years (20 women and 11 men). Healthy blood donors with a mean age of 29.0 ± 10.5 years (19 women and 12 men) served as the control group. A review of the medical records confirmed that the participants met the inclusion criteria.

Inclusion and exclusion criteria

Inclusion criteria for case group: (1) Age ≥ 18 years at the time of enrollment in the study; (2) Recurrent or persistent anxiety; (3) Anxiety not entirely situational or arising in an environment with no obvious danger; (4) Feelings of helplessness and unable to take appropriate action to end the anxiety-causing situation; (5) Impaired psychological or physiological function; (6) History of violence; and (7) Anxiety disorders meeting Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) diagnostic criteria[17,18].

Inclusion criteria for control group: (1) Age-and gender-matched with the case group; (2) No lifetime history of violence victimization; and (3) Clear consciousness; able to complete questionnaires independently.

Exclusion criteria for case and control groups: (1) Active inflammatory diseases; (2) Neurological diseases, brain damage, or epilepsy; (3) Long-term use of medications that might interfere with the results; (4) Heavy smoking, excessive alcohol intake, or substance dependence; and (5) Experienced major non-violent trauma in the last 3 months.

Measures

Venous blood was collected and centrifuged (3000 rpm, 10 minutes). Serum levels of BDNF, mBDNF, proBDNF, AQP4, and GSK3β were measured with commercial double-antibody sandwich ELISA kits (MultiSciences, Hangzhou, China) according to the manufacturer's instructions. Absorbance was read at 450 nm, and concentrations were drawn from normal curves.

Statistical analysis

Statistical analyses were performed using GraphPad Prism 11.0 (GraphPad Software Inc., San Diego, CA, United States). Continuous data are presented as mean ± SD and were compared using independent samples t-tests. Categorical data were compared using the χ² test. Correlations among variables were assessed using Pearson correlation analysis. Logistic regression was employed for multivariate analysis. To evaluate the diagnostic potential of the serum biomarkers (BDNF, mBDNF, proBDNF, AQP4, and GSK3β), individually and in combination, for anxiety disorder with a history of violence. Receiver operating characteristic (ROC) analyses were conducted. Area under the curve (AUC), sensitivity, specificity, and cut-off values were calculated. The combined diagnosis was constructed using multivariate logistic regression, and the probability predicted by the model was used as a continuous variable to generate a combined ROC curve. A two-sided P value < 0.05 was considered statistically significant.

RESULTS

Clinical features

The mean age of the 31 patients was 35.0 ± 10.3 years, with 67.7% being married and all having a history of violence. Their anxiety symptoms were recurrent, persistent, and impaired function, thereby meeting DSM-IV diagnostic criteria. There was no significant difference in demographics, such as gender and age, between the case and control groups (P > 0.05).

Biomarker findings

Compared to healthy controls, patients exhibited significantly lower serum levels of BDNF, mBDNF, and AQP4, but higher levels of proBDNF and GSK3β (P < 0.05). The effect sizes for BDNF, mBDNF, proBDNF, AQP4, and GSK3β between the case group and control group were 1.312, 0.953, 0.986, 0.854, and 0.513, respectively. Except for GSK3β, which showed a moderate effect, all other biomarkers demonstrated large effects (Table 1).

Table 1 Comparison of serum brain-derived neurotrophic factor, mature brain-derived neurotrophic factor, brain-derived neurotrophic factor precursor, aquaporin-4, and glycogen synthase kinase-3β levels between the two groups (mean ± SD).

Characteristics	Case group (n = 31)	Control group (n = 31)	P value	t value	Effect size (Cohen′s d)
BDNF (ng/mL)	13.21 ± 6.18	21.52 ± 6.48	< 0.05	-5.167	1.312
mBDNF (pg/mL)	537.01 ± 188.80	874.42 ± 463.69	< 0.05	-3.752	0.953
proBDNF (pg/mL)	1061.49 ± 451.74	729.10 ± 152.44	< 0.05	3.882	0.986
AQP4 (ng/mL)	39.23 ± 19.65	53.45 ± 12. 99	< 0.05	-3.361	0.854
GSK3β (pg/mL)	939.69 ± 249.02	771.28 ± 391.72	< 0.05	2.020	0.513

AQP4: Aquaporin-4; GSK3β: Glycogen synthase kinase-3β; BDNF: Brain-derived neurotrophic factor; mBDNF: Mature brain-derived neurotrophic factor; proBDNF: Brain-derived neurotrophic factor precursor.

Diagnostic value

ROC curve analysis evaluated the diagnostic potential of individual and combined biomarkers. The AUC for diagnosing anxiety disorder with a history of violence was 0.8200 for BDNF, 0.7211 for mBDNF, 0.7430 for proBDNF, 0.7170 for AQP4, and 0.6452 for GSK3β levels. The AUC of the combined diagnosis was 0.9698, indicating enhanced discriminative capacity within the sample (Table 2 and Figure 1).

Figure 1 Diagnostic performance of serum biomarkers in anxiety disorder with a history of violence. A: Receiver operating characteristic (ROC) analysis yielded the characteristic curves for the diagnostic performance of brain-derived neurotrophic factor (BDNF) precursor, aquaporin-4, and glycogen synthase kinase-3β; B: ROC analysis yielded the characteristic curves for the diagnostic performance of BDNF, mature BDNF, and the combined diagnosis biomarker. AQP4: Aquaporin-4; GSK3β: Glycogen synthase kinase-3β; BDNF: Brain-derived neurotrophic factor; mBDNF: Mature brain-derived neurotrophic factor; proBDNF: Brain-derived neurotrophic factor precursor.

Table 2 Diagnostic value of serum brain-derived neurotrophic factor, mature brain-derived neurotrophic factor, brain-derived neurotrophic factor precursor, aquaporin-4, and glycogen synthase kinase-3β levels in anxiety disorder.

Group	AUC	95%CI	Sensitivity	Specificity	Cut-off value	P value
BDNF	0.8200	0.7189-0.9210	0.8065	0.6452	15.26	< 0.05
mBDNF	0.7211	0.5891-0.8532	0.5484	0.9032	776.20	< 0.05
proBDNF	0.7430	0.6106-0.8754	0.9677	0.5806	938.30	< 0.05
AQP4	0.7170	0.5877-0.8462	1.0000	0.4194	32.84	< 0.05
GSK3β	0.6452	0.5011-0.7892	0.4839	0.9355	640.10	< 0.05
Combined diagnosis	0.9698	0.9360-1.000	1.0000	1.0000	/	< 0.05

Optimal cut-off values were determined by the maximum Youden index (sensitivity + specificity - 1) based on raw serum concentrations. The combined diagnosis was constructed using multiple logistic regression. AQP4: Aquaporin-4; GSK3β: Glycogen synthase kinase-3β; BDNF: Brain-derived neurotrophic factor; mBDNF: Mature brain-derived neurotrophic factor; proBDNF: Brain-derived neurotrophic factor precursor.

Correlation analysis findings

Pearson correlation analysis revealed significant positive correlations between BDNF and mBDNF (r = 0.6575, P < 0.05), BDNF and proBDNF (r = 0.5750, P < 0.05), and mBDNF and proBDNF levels (r = 0.7865, P < 0.05; Table 3 and Figure 2).

Figure 2 Correlations of serum biomarkers in anxiety disorder with a history of violence. A: Correlation between brain-derived neurotrophic factor (BDNF) and mature BDNF (mBDNF) levels; B: Correlation between BDNF and BDNF precursor (proBDNF) levels; C: Correlation between mBDNF and proBDNF levels. BDNF: Brain-derived neurotrophic factor; mBDNF: Mature brain-derived neurotrophic factor; proBDNF: Brain-derived neurotrophic factor precursor.

Table 3 Correlation analysis of serum brain-derived neurotrophic factor, mature brain-derived neurotrophic factor, and brain-derived neurotrophic factor precursor levels in patients with anxiety disorder.

	r value	P value
BDNF-mBDNF	0.6575	< 0.05
BDNF-proBDNF	0.5750	< 0.05
mBDNF-proBDNF	0.7865	< 0.05

BDNF: Brain-derived neurotrophic factor; mBDNF: Mature brain-derived neurotrophic factor; proBDNF: Brain-derived neurotrophic factor precursor.

DISCUSSION

Anxiety disorders include separation anxiety, selective mutism, social anxiety, panic, and general anxiety, with a history of mental and physical symptoms and violence[19]. These conditions impact women more, often occurring alongside depression and substance abuse, which can lead to relapse if not addressed[20-22].

Current research reports that the BDNF family, AQP4, and GSK3β have roles in mental illness[12,13]. Similar to existing studies, we observed notable alterations in multiple biomarkers that may be associated with the biological underpinnings of anxiety in the context of violence exposure. We observed lower serum BDNF, mBDNF, and AQP4 levels, and higher levels of proBDNF and GSK3β, in the case group than in healthy controls. The effect sizes for BDNF, mBDNF, proBDNF, AQP4, and GSK3β between the case group and control group were 1.312, 0.953, 0.986, 0.854, and 0.513, respectively. Except for GSK3β, which showed a moderate effect, all other biomarkers demonstrated large effects. The significant associations between BDNF and mBDNF, BDNF and proBDNF, and mBDNF and proBDNF levels highlight their potential involvement in the interconnected regulatory mechanisms underlying anxiety[23]. Our results suggest that longitudinal monitoring of these biomarkers may contribute to a more comprehensive evaluation framework and inform individualized management strategies. Future research should validate these biomarkers in larger cohorts and explore the potential of artificial intelligence to integrate biomarker data with clinical profiles for improved prediction and more precise, individualized diagnosis.

This case-control study illustrates the heterogeneity of biomarker alterations in anxiety patients sharing a history of violence. The representative cases demonstrated individual variations, reinforcing the potential utility of a combined biomarker panel (AUC = 0.9698, confidence interval: 0.9360-1.000) for supporting diagnosis and differentiating anxiety disorder patients with vs without a history of violence. Alterations in these biomarkers are unlikely to be disorder-specific and may instead reflect transdiagnostic biological processes, such as impaired neuroplasticity and stress-related neurobiological adaptation[24]. Although our logistic regression model achieved a high combined AUC of 0.9698, the relatively small sample size (n = 62) and multiple predictors may have inflated the performance due to overfitting. Future studies with larger, independent cohorts are needed to validate these findings.

CONCLUSION

Anxiety disorders are primarily managed with psychotherapy and pharmacotherapy, and combined treatment yields superior efficacy. Specialized interventions for anxious patients with a history of violence remain limited. Serum biomarkers may aid auxiliary diagnosis, guide targeted therapy, and improve clinical prognosis.

ACKNOWLEDGEMENTS

The authors thank associate professor Lai-Ling Du, a statistics expert, for her contributions to the statistical review and revision of all original data in this paper.