Relationship among symptoms of anxiety, psychosomatic, and coping in adults with rheumatoid arthritis

Abstract

There is limited research on the interplay between psychosomatic symptoms, anxiety, and coping strategies among individuals with Rheumatoid Arthritis (RA). This study examined their frequency and relationships in 97 RA patients assessed at a public healthcare institution in Rio de Janeiro, Brazil. Participants were predominantly female (94%), with a mean age of 59.3 years (SD = 13.10), and completed sociodemographic, anxiety (ASQ), psychosomatic (PSS-Bra), and coping (Brief COPE) measures. Most participants primarily used emotion-focused coping strategies (M = 28.08, SD = 5.85), and 81.3% exhibited severe anxiety symptoms. Network analysis showed that anxiety symptoms occupied a central position, connecting with most psychosomatic symptoms. Psychosomatic symptoms significantly predicted anxiety symptoms, F(3, 93) = 27.5, p < .001, adjusted R² = 0.45. The frequency of gastrointestinal symptoms (β = 0.31, p < .001), pain (β = 0.29, p = .003), and pseudoneurological symptoms (β = 0.27, p = .002) showed the greatest association with anxiety. These findings underscore the close link between psychosomatic and anxiety symptoms in RA, highlighting the need to address both psychological and somatic aspects in clinical management. Future studies should explore these relationships longitudinally and across diverse contexts to inform comprehensive interventions.

Introduction

Rheumatoid arthritis (RA) is a chronic, systemic autoimmune disease characterized by inflammation of the peripheral joints, with pain representing one of its primary symptoms^1–3. Beyond its physical symptoms, RA is frequently associated with psychological conditions such as depression and anxiety, which are particularly prevalent among patients reporting higher levels of perceived pain^4–6. Elevated levels of anxiety and depressive symptoms not only reflect the psychological burden of pain but may also contribute to the onset and exacerbation of psychosomatic symptoms⁷.

Studies indicate that stress and anxiety symptoms not only coexist with RA but may also exacerbate physical manifestations, such as pain perception and motor disability, thereby worsening patients’ quality of life (QoL)^5,8,9. This persistent state of hypervigilance – a state of intense, chronic alertness and scanning for danger – and worry can intensify emotional distress, creating a reciprocal cycle in which anxiety heightens pain perception, and pain, in turn, amplifies anxiety levels^4,10. In this, anxiety heightens the perception of existing RA pain, while also triggering distinct psychosomatic symptoms, such as increased muscle tension and inflammatory activity, which further aggravate joint stiffness^9,11.

Evidence indicates that anxiety may exacerbate pain perception and intensify psychosomatic symptoms, contributing to a cycle of physical and emotional distress that negatively affects individuals’ QoL⁷. Psychosomatic symptoms are prevalent among patients with RA and these psychosomatic symptoms may include cardiovascular, gastrointestinal, dermatological, and musculoskeletal complaints^12,13. Specifically, research shows a significant link between higher anxiety and a greater number of physical symptoms, suggesting that psychological factors play a key role in the somatic experience of this population^4,10. This relationship is further supported by neurobiological mechanisms, including heightened activation of the Hypothalamic-Pituitary-Adrenal (HPA) axis and increased inflammatory responses, which together intensify physical discomfort¹⁴.

While these biological processes define the physical experience, coping strategies, the behavioral and cognitive efforts used to manage stress, have an important role in psychological adjustment and treatment¹⁵. Both adaptive and maladaptive responses influence chronic pain, emotional regulation, and overall QoL¹⁶. From a psychosomatic perspective, ineffective coping can intensify physiological stress, leading to increased pain sensitivity and anxiety^17,18. In contrast, adaptive strategies like optimism and acceptance are associated with better emotional adjustment and an improved sense of control over pain^10,19.

Problem-focused strategies, such as planning and active problem-solving, have been associated with improved psychological adaptation and a reduction in anxiety^4,20. Similarly, the use of active coping strategies, which involve directly addressing stressors, is linked to better treatment outcomes and a higher QoL^21–25. Although these adaptive approaches generally lead to successful adaptation, their effectiveness is not universal and depends on individual patient characteristics¹⁵. These findings suggest that clinical interventions designed to build these specific coping skills can directly reduce anxiety and improve physical health. Finally, understanding these factors allows providers to do a more effective and personalized disease management^26,27.

In contrast, emotion-focused coping strategies such as denial and avoidance may exacerbate anxiety, worsen physical symptoms, and reduce patients’ QoL^4,28. These strategies, typically characterized by efforts to minimize or disregard the problem rather than confront it directly, are associated with increased physiological and emotional stress, potentially creating a negative feedback loop between psychological distress and clinical deterioration^4,28. Furthermore, studies suggest that dysfunctional emotional coping strategies not only contribute to heightened pain perception but may also influence inflammatory processes, thereby intensifying psychosomatic symptoms^17,18. Active coping is often associated with lower levels of anxiety, depression, and somatic pain, whereas disengagement coping is linked to higher levels of anxiety, depression, and somatic complaints²⁹.

There is evidence suggesting that maladaptive coping strategies can exacerbate emotional distress, leading to heightened anxiety and more severe psychosomatic symptoms. This underscores the importance of understanding the underlying mechanisms of this relationship^20,30. These findings highlight the need to explore how emotional factors, coping strategies, and anxiety contribute to the subjective experience of health, which could inform the development of more effective interventions aimed at alleviating the negative effects of chronic pain and enhancing well-being. Furthermore, understanding the interplay between anxiety, psychosomatic symptoms, and coping strategies in patients with RA is critical for comprehending the complexity of the disease symptomatology^31,32.

Current research shows that while RA’s main symptom is joint inflammation¹; Smolen et al., 2016), it is also linked to a set of psychosomatic symptoms, such as heart, skin, or stomach issues, that are often triggered by anxiety^4,12,13. The literature suggests that how patients cope is important; active problem-solving helps manage these symptoms^4,20, while avoiding the problem tends to worsen both the psychological burden and physical discomfort^4,28.

In this context, the present study aimed to understand the relationship between psychosomatic symptoms, anxiety, and coping in adults with RA. Using network analysis, we initially sought to explore how these variables interact. Subsequently, a multiple linear regression analysis was conducted to identify which psychosomatic symptoms (cardiovascular, musculoskeletal, respiratory, gastrointestinal, dermatological, and pain) and coping strategies best predicted anxiety symptoms in patients.

The study was based on the hypotheses that a significant positive relationship exists between anxiety and psychosomatic symptoms in adults with RA^5,9; furthermore, a greater number of psychosomatic symptoms would predict higher levels of anxiety among patients. Additionally, it was hypothesized that emotion-focused coping strategies would be positively associated with higher levels of both anxiety and psychosomatic symptoms in RA patients^4,28. Understanding these connections is essential for developing better treatments that address both the physical and emotional challenges of living with rheumatoid arthritis.

Method

Design

This study employed a quantitative predictive methodology, designed to efficiently and reliably present a large amount of data³³. Given the aim of exploring complex, interrelated issues, the quantitative approach facilitates a broader numerical scope, providing enhanced analytical possibilities³⁴.

Participants

A total of 97 adults with RA were evaluated, including 91 women and 6 men, between February and August 2024. Participants were selected via convenience sampling while awaiting routine outpatient consultations. All participants were receiving care through the Brazilian public health system (Sistema Único de Saúde - SUS). A sample size calculation conducted using G*Power³⁵ indicated that at least 74 subjects would be required to detect an effect size of dz = 0.15, with α = 0.05 and statistical power of 0.95. The inclusion criteria were as follows: (1) meeting the classification criteria for rheumatoid arthritis established by the American College of Rheumatology (ACR)³⁶ and/or the European Alliance of Associations for Rheumatology (EULAR)³⁷; (2) being 18 years of age or older; and (3) undergoing regular follow-up at the rheumatology outpatient clinic of the institution involved in the study.

Data collection procedures

The study was conducted in person at the Rheumatology Outpatient Clinic of a public university in Rio de Janeiro, Brazil. Individual interviews were carried out by two trained researchers to clarify any potential doubts regarding the completion or understanding of the research instruments. Each interview lasted approximately 30 min and was conducted while participants were waiting for their routine outpatient medical appointments. Participants received a link to a Google Forms questionnaire, which first presented the Informed Consent Form (ICF) outlining all ethical aspects of the study. Upon providing informed consent, participants gained access to the research instruments.

Measures and instruments

1. Sociodemographic and Health Conditions Questionnaire.

2. This questionnaire was developed by the researchers to collect participants’ sociodemographic and health-related information, including age group, socioeconomic conditions, and social characteristics. It also contained items addressing disease duration, daily routines, and general health status. Data were collected through interviews and self-report questionnaires.

3. Anxiety Symptoms Questionnaire – ASQ³⁸.

The Brazilian version of the ASQ was adapted by Ujihara and Rodrigues³⁸from the original instrument developed by Baker et al.³⁹,. The ASQ assesses the severity and intensity of cognitive, physical, behavioral, and emotional anxiety symptoms experienced over the past week. It comprises 17 Likert-type items rated on a 0–10 scale. Severity scores are classified as follows: 0 = no distress; 1–3 = mild; 4–6 = moderate; 7–9 = severe; and 10 = extreme distress. Frequency is rated as 0 = never; 1–3 = rarely; 4–6 = sometimes; 7–9 = frequently; and 10 = always. The total score represents the overall severity of symptom intensity and frequency (up to 170 points each), yielding an overall anxiety score (up to 340 points). Overall anxiety levels are interpreted as: absent (< 72), mild (72), mild to moderate (72–132), moderate (132), moderate to severe (132–192), and severe (> 192). Symptom intensity is categorized as absent (< 35), mild (35), mild to moderate (35–66), moderate (66), moderate to severe (66–98), and severe (> 98). Symptom frequency is classified as infrequent (< 33), mildly frequent (33), mildly to moderately frequent (33–62), moderately frequent (62), moderately to severely frequent (62–92), and severely frequent (≥ 92). The scale demonstrated good criterion validity and excellent internal consistency in Brazil (Cronbach’s α = 0.975). In the current sample, Cronbach’s α values were 0.898 for intensity, 0.894 for frequency, and 0.935 for the overall score.

• d)Psychosomatic Symptom Scale – PSS-Bra⁴⁰; Basler & Freitas, 2024)

The PSS-Bra is the Brazilian Portuguese adaptation of the original Psychosomatic Symptom Scale⁴⁰, designed to screen for psychosomatic symptoms. The instrument enables a comprehensive assessment of symptom frequency and its impact on daily functioning across somatic, emotional, cognitive, and behavioral domains. It comprises 35 symptom items and seven additional questions addressing relevant health indicators. Responses are rated on a 4-point Likert frequency scale (1 = never, 2 = a few times per month, 3 = a few times per week, 4 = almost every day) and a 3-point scale for interference with daily activities (1 = not at all, 2 = somewhat, 3 = very much). Scoring is based on symptom frequency and severity across seven systems: pseudoneurological, cardiovascular, musculoskeletal, respiratory, gastrointestinal, dermatological, and pain/weakness. Mean scores were calculated for each system. Interpretation follows a decile-based classification: deciles 1–4 indicate low or no risk of somatic disorders; deciles 5–6 suggest low risk with the need for monitoring; deciles 7–8 indicate moderate risk requiring medical evaluation; and deciles 9–10 denote high risk necessitating clinical diagnosis and follow-up. Participants also rated their perceived pain intensity on a 5-point scale (1 = no pain; 5 = very severe pain) and responded to an open-ended question: “What is most bothersome about this pain?” Additional qualitative items assessed general health status and were interpreted within each participant’s clinical context. In the present sample, the PSS-Bra demonstrated excellent reliability (Cronbach’s α = 0.935).

• e)Brief COPE – Adapted Version^17,41.
• f)The Brief COPE is a multidimensional inventory designed to assess coping strategies, originally developed by Carver et al.¹⁷, and later adapted to Portuguese by Brasileiro⁴¹. The abbreviated version consists of 28 items and measures three overarching coping dimensions: problem-focused, emotion-focused, and avoidant coping. The questionnaire includes 14 subscales: active coping, planning, instrumental support, emotional support, religiosity, positive reinterpretation, self-blame, acceptance, venting, denial, self-distraction, behavioral disengagement, substance use, and humor. Items are rated on a 4-point ordinal scale ranging from “I never do this” to “I always do this.” Subscale scores are calculated by summing item responses, with higher scores indicating greater use of that coping strategy. In the present study, Cronbach’s α = 0.736, indicating acceptable reliability.

Ethical procedures

All methods were performed in accordance with the relevant guidelines and regulations.

The study was approved by the Research Ethics Committee of the hospital where data collection was conducted. All participants read and signed the Informed Consent Form (ICF) before participation, validating their voluntary agreement to participate and to allow publication of the anonymized data.

Confidentiality, privacy, and participant anonymity were guaranteed throughout the study. Participants were informed that the instruments posed minimal risk to their physical and psychological well-being. At the end of the survey, they were invited to provide an email address to receive a summary of the study findings as an appreciation for their participation.

Data analysis

All data were entered into Microsoft Excel, and statistical analyses were performed using JASP software (JASP Team⁴²,. Network analysis was conducted using the EBICglasso method to eliminate spurious or weak correlations, evaluate the strength of associations in the regression model, and explore their relationships with coping strategies (avoidant, emotion-focused, and problem-focused). The Glasso method was estimated using a regularized solution based on the Extended Bayesian Information Criterion (EBIC). This approach allows an exploratory examination of the structure of associations among variables, without prior specification of relationships, enabling the identification of emergent empirical patterns⁴³.

In network analysis, each variable is represented as a node, and associations among variables are depicted as edges. Edge thickness and color indicate the strength and direction of connections, respectively. Partial correlations were used to determine the direction of associations: blue edges represent positive associations, red edges represent negative associations, and the absence of an edge indicates no statistically significant relationship between nodes after controlling for all other variables. In the network, relationships between variables are shown as edges, where thicker edges and shorter distances between nodes indicate stronger associations.

To determine which factors were most influential, centrality measures were examined. The strength value reflects how strongly a specific node is connected to others; in this sample, pain frequency and anxiety symptoms showed the highest strength values. Centrality also helps identify which factor acts as a primary bridge within the overall network. Network modeling enables the visualization of interdependencies among multiple variables simultaneously, allowing the identification of clusters of symptoms or psychological factors⁴⁴.

The associations identified through network analysis were further tested using multiple linear regression models via the stepwise method. This method sequentially introduces predictor variables based on their explanatory power for the dependent variable⁴⁵. In these analyses, the subscales of the PSS-Bra (pseudoneurological, cardiovascular, pain, dermatological, musculoskeletal, gastrointestinal, and respiratory symptoms) served as predictor variables, and the total anxiety symptom score from the ASQ was the outcome variable. These analyses aimed to identify which variables most strongly influenced the intensity and frequency of anxiety symptoms in the sample.

Results

The sample was predominantly female (94%) with a mean age of 59.3 years (SD = 13.10). Most participants were married and had completed secondary education. Regarding disease-related characteristics, 67% tested positive for rheumatoid factor, 14% were in remission, and the majority had been diagnosed with RA for over 10 years. Additionally, 60% of the participants had no prior psychiatric diagnosis (Table 1).

Table 1.

Sociodemographic and health characteristics of the participants (n = 97).

Marital Status	N (%)
Married	42 (42.27)
Single	30 (31.96)
Widowed	13 (13.4)
Divorced	6 (6.19)
Separated	2 (2.06)
Cohabiting	4 (4.12)
Educational Level
Completed high school	28 (30.93)
Completed elementary school	22 (23.71)
Incomplete elementary school	14 (14.43)
Completed higher education	14 (12.37)
Incomplete high school	7 (7.22)
Incomplete higher education	6 (6.19)
Illiterate	3 (3.09)
Postgraduate education	3 (2.06)
Religious Affiliation
Evangelical	41 (41.24)
Catholic	36 (37.11)
Agnostic	8 (9.28)
Spiritist	3 (3.09)
Umbanda	3 (3.09)
Atheist	2 (2.06)
Jehovah’s Witness	2 (2.06)
Candomble	2 (2.06)
Duration of RA Diagnosis
More than 10 years	70 (72.17)
Between 5 and 10 years	18 (17.53)
Between 2 and 5 years	5 (5.16)
Between 1 and 2 years	4 (4.12)
Disease Activity
High	24 (24.74)
Low	23 (23.71)
Moderate	27 (27.84)
Remission	14 (14.43)
Not reported	9 (9.28)
Rheumatoid Factor
Positive	66 (67.35)
Negative	19 (19.39)
Not available	12 (12.27)
Having a formal Psychiatric Diagnosis
No	59 (59.79)
Yes, for 1 to 5 years	13 (13.4)
Yes, for more than 5 years	21 (21.65)
Yes, for less than 1 year	4 (4.12)

To address the specific objectives of this study, the first step involved examining the frequency distribution of participants’ responses to the study instruments (Table 2). The data indicated that most participants predominantly employed emotion-focused coping strategies (M = 28.08, SD = 5.85).

Table 2.

Mean, standard deviation, and minimum and maximum scores of participants (n = 97) on the anxiety, psychosomatic, and coping scales.

ASQ Intensity	Mean	SD	Min	Max
	80.19	34.15	7.00	159.00
ASQ Frequency	77.96	34.02	2.00	153.00
ASQ Total	158.14	67.75	9.00	309.00
PSS Pseudo Frequency	1.32	0.35	0.89	2.33
PSS Pseudo	1.64	0.50	1.00	3.22
PSS Cardio Frequency	1.77	0.71	1.00	4.00
PSS Cardio	1.63	0.61	1.00	3.00
PSS Musculo Frequency	2.34	0.99	1.00	4.00
PSS Musculo	1.93	0.66	1.00	3.00
PSS Resp Frequency	1.47	0.56	1.00	3.67
PSS Resp	1.42	0.51	1.00	3.00
PSS Gastro Frequency	1.58	0.46	0.89	3.56
PSS Gastro	1.42	0.38	1.00	2.67
PSS Derma Frequency	1.56	0.63	1.00	4.00
PSS Derma	1.40	0.51	1.00	3.00
PSS Pain Frequency	2.25	0.57	1.00	3.33
PSS Pain	1.97	0.40	1.17	2.83
PSS decile	4.55	2.15	1.00	9.00
PSS decile Frequency	6.10	2.57	1.00	10.00
PFC Mean	2.40	0.65	1.00	4.00
AC Mean	1.99	0.54	1.00	3.38
EFC Mean	2.34	0.49	1.33	3.42

Note. ASQ = Anxiety Symptoms Questionnaire; PSS = Psychosomatic Symptoms Scale; Pseudo = Pseudoneurological; Cardio = Cardiovascular; Musculo = Musculoskeletal; Resp = Respiratory; Gastro = Gastrointestinal; Derma = Dermatological; PFC = Problem-Focused Coping; AC = Avoidant Coping; EFC = Emotion-Focused Coping.

Regarding anxiety symptoms as measured by the ASQ, results for the intensity dimension revealed that a large proportion of patients (81.3%) reported severe anxiety symptoms, while an additional 6.2% reported symptoms ranging from moderate to severe. Only 5.2% of participants reported no significant anxiety symptoms, and an equal proportion (5.2%) classified their symptoms as mild to moderate. Moderate-intensity symptoms alone were reported by 2.1% of the sample.

When examining the frequency dimension of anxiety symptoms, 36.2% of participants reported moderate to severe symptoms, whereas 35% reported symptoms occurring at a severe frequency. In contrast, only 11.3% indicated infrequent episodes of anxiety. Finally, with respect to the overall level of anxiety symptoms, most participants (51.5%) reported mild to moderate symptoms. A considerable proportion (41.2%) did not exhibit clinically significant anxiety symptoms, while 6.3% of participants presented moderate to severe symptomatology.

Table 3 presents the frequency analysis of psychosomatic symptoms. In general, participants reported lower-than-average total and severity scores (which measure symptom intensity) and total scores (which combine both intensity and frequency). Nevertheless, when specifically examining the frequency of symptoms (how often they occur), the sample showed average to elevated levels. The categorical cut-offs used to define these levels were benchmarked against normative scores from the Brazilian population (Basler & Freitas, 2024).

Table 3.

Psychosomatic symptoms’ frequency (PSS-Bra) of the participants (n = 97).

Frequency	N(%)
Total Score
Below average	55 (56.52)
Average	19 (19.49)
Above average	19 (19.49)
Extremely high	3 (4.50)
Frequency
Below average	22 (23.11)
Average	31 (32.14)
Above average	21 (21.36)
Extremely high	23 (23.47)
Severity
Below average	57 (59.00)
Average	16 (16.61)
Above average	19 (19.49)
Extremely high	4 (5.10)

In the network analysis, Musculoskeletal symptoms demonstrated the greatest centrality, functioning as the key hub that connects and facilitates the interactions between the other physical symptoms and psychological factors.

The network analysis revealed a structure comprising 11 nodes and 23 non-zero edges out of a possible 55 connections, resulting in a sparsity index of 0.58, indicative of a moderately connected network. For interpretative clarity, the magnitude of associations is represented by the thickness of the edges and the proximity between nodes; thicker edges and shorter distances denote stronger relationships⁴³.

The results indicated that anxiety symptoms, as measured by the ASQ, occupied a central position within the network, showing positive associations with psychosomatic symptoms (PSS), particularly with pain, pseudoneurological, and musculoskeletal domains. This pattern suggests that these physical symptoms may have an important role on anxiety manifestation. Gastrointestinal and respiratory symptoms were also positively related to anxiety, further emphasizing the interaction between psychological and somatic processes. When looking at the weight of these connections, PSS Pain Frequency and ASQ Anxiety symptoms showed the highest values in the network. This means they have the strongest impact on other variables. Specifically, the connection between pain and anxiety represents the heaviest weight, showing that as one increase, the other is very likely to increase as well.

With respect to coping strategies, a positive association was observed between avoidant and emotion-focused coping, indicating that these strategies may be interconnected and collectively contribute to the psychosomatic experience. Overall, the findings underscore an interdependence between anxiety and psychosomatic symptoms, highlighting the multidimensional nature of anxiety expression in this population. Figure 1 illustrates the network structure and the associations identified among the study variables.

Fig. 1.

Network analysis among anxiety, psychosomatic, and coping

Regarding the strength estimates in the network analysis (Fig. 2), the PSS Pain Frequency variable showed the highest strength, followed by anxiety symptoms (ASQ). Meaning that these two factors have the most powerful direct connections with other variables. In contrast, PSS Musculoskeletal symptoms showed the highest Betweenness centrality, acting as a key bridge that connects different parts of the network and helps symptoms interact with each other.

In this specific model, a zero value suggests that no single symptom is significantly “closer” to all others than the rest; the influence of one variable does not spread more quickly through the network than another. Additionally, some variables like PFC and PSS Derma show negative values for strength or betweenness. These negative numbers are standardized scores (z-scores), which simply mean these variables are below the network average. For example, a negative strength value for PFC implies it has fewer or weaker connections to the overall symptom group compared to the other variables.

Fig. 2.

Centrality graph

In the regression analysis, three models were tested, with the third model demonstrating the best overall fit while satisfying the multicollinearity assumptions. Multicollinearity was evaluated using the Variance Inflation Factor (VIF) and tolerance values, both of which fell within acceptable thresholds; VIF values ranged from 1.27 to 1.62, and tolerance values exceeded 0.20. The Durbin–Watson statistic indicated independence of residuals (DW = 2.14).

The results revealed a significant association of psychosomatic symptom frequency on anxiety symptoms, F(3,93) = 27.5, p <.001, with an adjusted R² = 0.45. Table 4 presents the standardized coefficients for all predictors. As shown, the most strongly associated variable of anxiety symptoms was the frequency of gastrointestinal symptoms (β = 0.31, p <.001), followed by pain frequency (β = 0.29, p =.003) and pseudoneurological symptoms (β = 0.27, p =.002).

Table 4.

Linear regression models of predictors of anxiety symptoms (n = 97).

Model		B	SD	β	t	p
M₀	(Intercept)	158.14	6.88	22.99		< 0.001
M₁	(Intercept)	4.92	22.9		0.22	0.83
	PSS Pain Frequency	68.23	9.88	0.58	6.90	< 0.001
M₂	(Intercept)	−25.78	23.38		−1.10	0.27
	PSS Pain Frequency	48.19	10.97	0.41	4.39	< 0.001
	PSS Gastro Frequency	48.04	13.76	0.32	3.49	< 0.001
M₃	(Intercept)	−61.61	24.93		−2.47	0.02
	PSS Pain Frequency	34.19	11.33	0.29	3.02	0.003
	PSS Gastro Frequency	45.97	13.14	0.31	3.50	< 0.001
	PSS Pseudo Frequency	53.48	16.63	0.27	3.21	0.002

Note. PSS = Brazilian Psychosomatic Symptom Scale; Pseudo = Pseudoneurological; Gastro = Gastrointestinal; F = Frequency.

Discussion

This study investigated the associations between psychosomatic symptoms, anxiety symptoms, and coping strategies among patients with RA. The results partially supported the initial hypotheses^5,9. While a strong relationship was found between anxiety and physical symptoms, the expected link with coping strategies was not confirmed.

The network analysis revealed that anxiety symptoms exhibited strong connections with various psychosomatic symptoms. Instead of a one-way cause, these findings suggest a bidirectional relationship where anxiety may heighten the focus on physical pain, and in turn, experiencing frequent physical symptoms may increase a patient’s anxiety^4,10. Anxiety was closely linked to the frequency of pain and pseudoneurological symptoms, suggesting that there is a significant association between higher anxiety levels and the reporting of diffuse somatic complaints, which may reflect a state of bodily hypervigilance^46–48.

Respiratory symptoms also occupied a central position within the network, indicating a strong association between anxiety and somatic sensations such as shortness of breath, chest tightness, or dyspnea. These results align with prior research showing that psychosomatic respiratory complaints are prevalent among individuals experiencing elevated anxiety, potentially due to autonomic nervous system activation and hyperventilation effects on bodily perception^46,49.

The regression analysis further demonstrated that the frequency of pain, gastrointestinal, and pseudoneurological symptoms significantly predicted anxiety symptoms, highlighting the complex interplay between physical and emotional experiences in RA. Chronic pain, gastrointestinal disturbances, and pseudoneurological complaints appear to contribute directly to heightened anxiety^50–52. Chronic pain alone has been recognized as a major factor in the development and maintenance of anxiety^53,54. Unpredictable pain flares common in RA may generate concerns about daily functioning, fear of disease progression, and persistent worries about the future⁴⁸. Moreover, pain-related stress can elevate cortisol and pro-inflammatory cytokines, further exacerbating inflammation and somatic symptom severity^55–57; Reyes del Paso et al., 2024).

The relationship between stress, anxiety, and gastrointestinal functioning, mediated by the gut-brain axis, may explain the significant impact of gastrointestinal symptoms on anxiety. Heightened stress or anxiety may trigger increased intestinal activity, visceral hypersensitivity, or microbiota dysregulation, resulting in abdominal pain, bloating, nausea, and altered bowel habits^52,58,59.

Pseudoneurological symptoms, including dizziness, balance disturbances, globus sensations, visual or auditory disturbances, fainting, or memory lapses, may imitate neurological disorders, leading patients to become hypervigilant and seek medical attention. This hypervigilance may further amplify anxiety, creating a feedback loop between somatic perception and emotional response^8,21,25,47.

Contrary to our hypothesis, coping strategies did not show a significant relationship with anxiety or psychosomatic symptoms in this study. Although literature often suggests that active coping helps and avoidant coping hurts QoL²¹; de²⁸ our data placed these strategies at the edge of the network. This suggests that for this specific group of RA patients, the internal experience of physical symptoms and anxiety is so intense that the chosen coping style—whether problem-focused or emotion-focused—does not directly change the level of distress. It is possible that the impact of coping is indirect or that these strategies are less effective when physical pain is very severe^16,24.

Together, these findings offer strong evidence for the link between psychosomatic symptoms and anxiety in RA. While network analysis shows the general connections, the regression analysis specifically reveals that physical symptoms—such as pain, stomach issues, and pseudoneurological complaints—predict higher levels of anxiety. This means that for these patients, the frequency of physical distress is a primary driver of their psychological burden. Coping strategies did not show a significant relationship with these symptoms. This suggests that in this group, the physical experience of the disease is overwhelming, and the patient’s specific coping style does not significantly change their level of anxiety or somatic distress. Because mental and physical health are connected in this population, it is important to consider include psychological support and mental health experts within rheumatology clinics. Integrating these services would help provide more effective, comprehensive care for managing RA.

Conclusion

In conclusion, this study demonstrates that the presence and frequency of psychosomatic symptoms are strong predictors of anxiety symptoms in patients with RA. The findings highlight that physical distress, particularly pain and gastrointestinal issues, explains a significant portion of the psychological burden patients face. This emphasizes the need for medical care that addresses both physical and mental health together.

Nonetheless, several limitations warrant consideration. The sample was small, predominantly female, and drawn from a single Brazilian city, limiting generalizability. Socioeconomic factors, social support networks, and access to healthcare may influence symptoms perception and coping strategies. Furthermore, the study did not control for variations in treatment type, time since RA diagnosis, or psychiatric comorbidities, all of which may affect symptoms and coping behaviors.

Future research should aim to include more diverse populations, accounting for different socioeconomic and healthcare contexts, to enhance generalizability. Controlling for treatment type and disease duration in subsequent studies would enable more precise evaluations of their effects on psychological and physical outcomes. Longitudinal designs are also recommended to examine temporal changes and identify potential intervention points for improving quality of life. Finally, future studies should incorporate mediation and moderation analyses to clarify the mechanisms through which coping strategies, psychosomatic symptoms, and emotional factors interact to influence well-being.

Author contributions

All authors contributed to the study’s conception and design. Samantha Castro Teixeira led the development of the study framework, data collection, data analysis and manuscript writing. Geraldo da Rocha Castelar Pinheiro contributed to critical revisions of the manuscript. Jaqueline de Carvalho Rodrigues assisted with data analysis, interpretation of results, and provided important feedback throughout the writing process. All authors reviewed and approved the final version of the manuscript.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Declarations

Competing interests

The authors declare no competing interests.

Other identifying information

There are no additional identifying details that compromise the anonymity of the review process.