Emotional repression in patients with chronic inflammatory rheumatism

: Background: A person’s psychological background may support and direct the inflammatory evolution of a disease toward a specific type of chronic inflammatory rheumatism (CIR). Aim: This study aimed to identify a particular emotional profile of patients with CIR, particularly rheumatoid arthritis (RA) and spondyloarthritis (SpA), based on psychological profile assessments between patients with and without CIR. Emotional repression, that is, a tendency to inhibit the expression of negative feelings and/or unpleasant thoughts, was particularly studied. Methods: This monocentric observational pilot study included patients from the rheumatology department of a university hospital. These patients were systematically assessed for different psychological parameters by an experienced psychiatrist, and their clinical and biological characteristics were collected accordingly. Data analysis was performed using the Chi-squared test or Fisher’s exact test. Results: Fifty-nine patients were assessed: 47 patients with CIR (i.e., 27 with RA and 20 with SpA) (CIR group) and 12 non-CIR patients (i.e., nine with osteoarthritis, one with viral disease, one with osteoporosis, and one with osteomalacia) (control group). Severe emotional repression and early life events were both significantly higher in the CIR group than in the control group ( P = 0.02). In contrast, severe psychological and somatic complaints were significantly higher in the control group than in the CIR group ( P < 0.01 and P = 0.01, respectively). Conclusion: Our findings suggested that emotional repression from traumatic life events could aggravate the etiology and/or course of CIR. Therefore, appropriate psychological care should have a relevant place within the current therapeutic options for the clinical management of CIR. Relevance for Patients: The management of CIR should include psychological support as learning coping mechanisms can facilitate the recovery of CIR patients.


Introduction
The development of chronic inflammatory rheumatism (CIR) is multi-factorial (e.g., genetic, environmental, hormonal, infectious, and/or psychological), and the degree of implication of these different factors in CIR onset has not been determined [1,2].Several studies on traumatic life events and stressful situations have suggested that such situations could be implicated in triggering CIR onset and its symptomatic progression [3,4].
Non-pharmacological interventions have recently been used more frequently for disease management, especially regular physical activity, educational therapeutic workshops with support groups, and mindfulness meditation, among others.In addition, self-reported evaluation and patient-reported outcomes have become increasingly important for the assessment of diseases, suggesting that the patients' psychological well-being is essential in routine practice for rheumatologists to evaluate the efficacy of disease management with better precision [5].Likewise, psychological support is increasingly integrated into the management of patients with CIR as they are more prone to mental and overall health disorders.
Emotional repression refers to the tendency to inhibit the expression of negative feelings and/or unpleasant thoughts and is a common manifestation in CIR patients.Hence, an assessment of the patients' emotional profile could provide a more relevant and effective approach to managing emotional repression and treating CIR.In this regard, we hypothesized that a particular psychological profile, namely, emotional repression, could be associated with autoimmune deregulation.
Herein, this preliminary and exploratory study primarily investigated the emotional function of CIR patients by assessing their psychological profiles to explore a possible association between autoimmunity and emotional regulation.In addition, we further evaluated the emotional profiles for potential associations with other conditions (i.e., rheumatoid arthritis [RA] and spondyloarthritis [SpA]) and different rheumatological characteristics (i.e., biological inflammatory syndrome, sacroiliitis, structural involvement (i.e., osteoarticular destruction of the joint due to the inflammatory activity of rheumatism), and positive rheumatoid factor [RF], anti-citrullinated peptide antibody [ACPA], and human leukocyte antigen [HLA]-B27 allele).

Population and study design
This monocentric cross-sectional observational pilot study was conducted in the rheumatology department of a university hospital.During a routine follow-up, the CIR patients were offered an evaluation based on the global CIR management framework, which included dermatological, dental, and gynecological consultations, as well as rheumatologist-requested psychiatric assessments (anxiety disorders, depressive disorders, sleep disorders, conjugal violence, and childhood traumas).None of the recruited 59 patients refused to participate in the evaluation.The exclusion criteria were: age <18 or >65 years old, cognitive disorders, substance dependence, or acute somatic or any psychiatric disorders requiring emergency hospitalization.
The participating patients were accompanied to the rheumatology department by seven different rheumatologists.CIR patients were classified as having RA (based on the 2010 classification by the American College of Rheumatology/European League Against Rheumatism) SpA (based on the 2009 criteria by the Assessment of SpA International Society), or psoriatic arthritis (based on the ClASsification criteria for Psoriatic ARthritis [CASPAR]).We identified 12 patients with ankylosing SpA and eight with psoriatic arthritis, and these patients were categorized under SpA to facilitate statistical analyses.Although their clinical characteristics might differ slightly, their pathologies share common pathophysiological mechanisms and genetic backgrounds, such as the presence of HLA-B27.
Non-CIR patients constituted the control group, and these patients were in the rheumatology department for other rheumatological diseases.

Psychological assessment tools
An experienced psychiatrist collected information during a semi-structured, semi-directed interview lasting 1.5 -2 h.The psychological parameters assessed were (i) depressive symptoms and their severity (based on the Montgomery-Asberg Depression Rating Scale) [6]; (ii) history of major depression episodes; (iii) anxiety symptoms using a 21-question multiple-choice self-report inventory (i.e., the Beck Anxiety Inventory) [7]; (iv) alexithymia, a behavioral trait of individuals who are unable to identify and describe their interior feelings, have limited imaginative capacity, and tend to focus their thoughts externally rather than resorting to introspection [8,9], and its severity using the 20-item Toronto Alexithymia Scale [10]; (v) social desirability intensity, that is, the tendency to seek approval of others and preserve one's self-image, using the Marlowe-Crowne Social Desirability scale (MCSD), a 33-item self-report questionnaire [11]; (vi) the severity of emotional repression, that is, the tendency to inhibit the expression of negative feelings or disagreeable thoughts, in accordance with the Weinberger classification; and (vii) the combination of State-Trait Anxiety Inventory (STAI) [12] and MCSD as emotionally repressed individuals typically have a low STAI score and high MCSD.
The psychiatrist used a Likert-type scale to assess the severity of the psychological parameters in a specific questionnaire, which assessed (i) the severity of somatic and psychological complaints (i.e., severe complaints referred to personalized responses to most of the questions); (ii) emotional-expressivity intensity (i.e., graded as "mild" when the patient had several emotional moments without excessive reactions, "moderate" when the patient's facial expressions were accompanied by fluctuating emotions as the discussion progressed, or "severe" when the patient had a tendency for hyper-expressivity throughout the interview); (iii) life events (i.e., type 1 for events occurring before 15 years old and type 2 for events occurring within 3 years preceding the rheumatological disease onset) and their intensities (i.e., classified as "mild" when stressful events were non-existent or mild, "moderate" when stressful events did not incur traumatic stress [e.g., emotional deprivation concerning socio-economic difficulties, severe or disabling diseases of close friends/relatives], or "severe" for a traumatic event [e.g., sudden death of a person caring for a child, physical or sexual assault, patient witnessed a death, or thought he/she would die]); (iv) somatic escalation (i.e., the occurrence of a series of several somatic disorders or another chronic pathology before disease onset); (v) actual stress level; (vi) impact of the rheumatological disease on the patient's professional activities, corresponding to the occupational repercussions experienced (i.e., classified as "mild" for minor impact, "moderate" for notable impact, or "severe" for a painful experience [e.g., loss of professional environment and activity, feeling of injustice, or elevated fear of losing one's job]); (vii) manual labor; and (viii) physical activity enjoyment before and after disease onset.
For the different psychological parameters, we combined the "mild" and "moderate" intensities for comparison against the "severe" intensity.The mild-to-moderate psychological disorders were easily identified in patients exhibiting severe anxiety or depressive comorbidities and emotional experiences (often affected by chronic pain), but retaining the "severe" intensity enabled a more significant difference.In addition, this approach accounted for the population recruitment bias.
During the same hospitalization, various patient characteristics, mostly rheumatological, were collected from the electronic medical files using the DxCare (medical information and prescription software) program.The collected data included: (i) Biological inflammatory syndrome, defined as an erythrocyte sedimentation rate ≥30 mm/1 st h and/or C-reactive protein ≥5 mg/L; (ii) positive ACPA, defined as 7 U/mL by enzyme-linked immunosorbent assay (ELISA); (iii) RF >15 IU/mL by ELISA; (iv) magnetic resonance imaging (MRI) detection of grade-2 bilateral or grade-3 unilateral sacroiliitis, with active inflammatory (edema) or chronic (erosion, bone condensation, bone bridges, and fat conversion) lesions.

Statistical analysis
Continuous variables were expressed as median (first quartile [Q1] -third quartile [Q3]).Categorical variables were expressed as number (percentages), where the percentages were calculated after excluding missing data, and were compared by Chi-squared test or Fisher's exact test accordingly.Missing data were not replaced.For all analyses, P < 0.05 was considered statistically significant.Type I error was not adjusted for multiplicity because of the exploratory character of the comparisons.All statistical analyses were performed with SAS release 9.4 (SAS Institute Inc, USA).

Patient characteristics
The study population comprised 59 patients: 47 (79.7 %) CIR patients and 12 (20.3%)non-CIR patients.The patients' characteristics according to their disease are summarized in Table 1.
The control group consisted of 12 women with a median age of 65.0 years [55.0 -72.0] (nine with osteoarthritis, one with viral disease, one with osteoporosis, and one with osteomalacia).In addition, the nine women with osteoarthritis had structural involvement.

Psychological assessments
A comparison between the CIR and control groups displayed significant differences in emotional repression, somatic complaints, psychological complaints, type-1 life-event severity, manual labor, or physical activity enjoyment before disease diagnosis (Table 2).
A comparison of the RA and SpA groups revealed significant differences in prior depressive episodes, professional impact, and manual labor (Table 3).
A comparison between patients with and without biological inflammatory syndromes revealed significant differences in emotional repression (Table 4).
A comparison between patients with and without structural involvement revealed significant differences in somatic complaints, emotional expressivity, professional impact, and physical activity enjoyment before disease onset (Table 5).
In addition, none of the psychological parameters were significantly associated with ACPA.Likewise, the number of patients who were HLA-B27-positive or had sacroiliitis was too low for analysis.Hence, their significance and association could not be established.

Discussion
The objective of this study was to assess the relevance of psychological support for patients with CIR.CIR patients, compared to those with other rheumatological pathologies, presented significantly more frequent severe emotional repression, whereas those with other pathologies had significantly more frequent severe psychological and somatic complaints, potentially attesting to their strong emotional repression.Temoshok [13] previously described the personality profile of RA patients as type C, characterized by submissive behavior, conciliatory approach, repression of hostility, self-effacement of personal needs, and depressive vulnerability.Grossarth-Maticek et al. [14,15] made similar observations in cancer patients.Bayle et al. [16] devised a psychological vulnerability score and obtained converging results close to the type C personality, and the scores were significantly higher for patients with secondary Raynaud's syndrome than a control group with idiopathic or primary Raynaud's syndrome.
Nagano et al. [17] demonstrated that rational and anti-emotional behaviors by RA patients, characterized by an extreme tendency to squelch emotional behaviors and rationalize negative experiences, were associated with poorer prognoses.Ishii et al. [18] found that RA patients, who were easily brought to tears in response to stress, had better responses to treatment and better overall prognoses.
Collectively, previous studies and our present findings highlight the importance of emotional dysregulation, especially emotional repression, on CIR etiology and prognosis.However, the present study should not be generalized as biased toward CIR or propose a secondary coping strategy for CIR.Emotional repression has also been considered a consequence of the CIR diagnosis instead of CIR itself [19,20].
The intensity of key early life events was significantly more severe for CIR patients than for those with other diseases.Cutolo and Straub [21,22] reported that stressful events preceded RA onset for 86% of their patients.
Several studies have associated severe life events with CIR onset or flares, whereas other studies have associated CIR with minor life events and daily stress.In contrast, we found no significant difference in the actual stress level presented by CIR and other diseases, probably because of a lack of statistical power.Rimón and Laasko [23] described that higher stress at RA onset predicted a poorer prognosis for the disease.O'Donovan et al. [24] also found that veterans experiencing trauma and  developing post-traumatic stress disorder (PTSD) could enhance the risk of developing autoimmune diseases, including RA.Based on their study of Vietnam veterans, Boscarino et al. [25] reported that those with RA had more PTSD symptoms compared to those without RA.
The results of those studies are in agreement with a biopsychosocial model linking psychological stress and stressful life events with the etiology of autoimmune diseases.The impact of stress, resulting from an intense life event and aggravated by conditions like emotional repression, would increase vulnerability to autoimmune diseases due to dysregulated immunity [26][27][28][29][30] and enhanced inflammation.
Herein, no significant difference was found between CIR and non-CIR patients regarding the frequency of depression disorders, depression severity, history of prior depressive episode(s), or anxiety symptoms.
Depressive and anxiety disorders are described as the most frequent comorbidities for CIR patients.Reynier-Legarçon et al. [31] found that patients with autoimmune diseases (systemic lupus erythematosus, systemic scleroderma, or primary Sjögren syndrome) presented more severe depressive and anxiety symptoms than the general population.Baerwald et al. [32] found that depressive disorders were significantly more common in RA patients than in the general population.
Recently, Kang [33] investigated the effect of arthritis on mental health using the 12-item version of the general health survey (GHQ-12) and reported a total of three factors of GHQ-12, that is, GHQ-12A (social dysfunction and anhedonia; six items), GHQ-12B (depression and anxiety; four items), and GHQ-12C (loss of confidence; two items), suggesting that both the global mental health and dimensions of mental health are negatively affected by arthritis.We were not able to replicate those findings, probably because of a lack of statistical power.In contrast, the homogeneity of our patients, in terms of depressive and anxiety symptoms, attenuated any potential bias linked to their emotional dysregulation associated with these psychological comorbidities.
Moreover, no significant differences were found between the RA and SpA groups for the frequency of depression and depression severity.However, RA patients reported significantly more frequent prior depressive episodes than SpA patients, a finding of which has never been reported previously.Nonetheless,  some findings indicated that anxiety symptoms, depression, and perception of the disease impacted the physical quality of life differently for SpA and RA patients.Some authors noted that RA patients had more physical quality-of-life difficulties, while those with psoriatic arthritis and ankylosing SpA had more mental quality-of-life issues [33][34][35][36].Hyphantis et al. [37] reported that SpA patients' quality of life was associated with anxiety and not depressive symptoms, which were associated with RA patients.Our results, consistent with earlier findings, suggest a link between depression and RA patients.Together, these results suggest that CIR (RA or SpA) patients would probably benefit from emotional management from psychotherapists and antidepressants for symptomatic depressive episodes.Moreover, these symptoms can contribute to a poorer prognosis for their respective rheumatological disease [37].Bijsterbosch et al. [38] demonstrated that arthrosis patients' perceptions of their disease were predictive of the functional disability and that cognitive-behavioral therapy could modify the representations of the disease and obtain a better functional result.
A biological inflammatory syndrome was significantly associated with a low disorder of emotional regulation (mild/ moderate emotional repression).However, earlier studies have reported a stronger association between emotional disorders and a biological inflammatory syndrome.
Smoak et al. [39] found increased nuclear factor-κB activity in patients with PTSD, resulting from childhood violence, compared to healthy controls.Howren et al. [40] reported that patients with depression had significantly higher interleukin-1 and -6, tumor necrosis factor-α, and C-reactive protein levels.Goldsmith et al. [41] also observed from a meta-analysis of 68 studies, that patients with schizophrenia, bipolar disorder, or major depressive episodes had increased levels of inflammatory cytokines.In addition, depressive symptoms were more frequent in patients with autoimmune diseases and taking anti-inflammatory drugs.Gobin et al. [42] found that antidepressants lowered the production of inflammatory cytokines, for example, interleukins-1β and -6 and tumor necrosis factor-α.
Several monoclonal antibodies targeting relevant inflammatory pathways for the treatment of CIR were associated with neuropsychiatric adverse events, notably depression, or suicidal  ideation or behavior [43].The development of brodalumab, a molecule targeting interleukin-17, was stopped after suicidal behaviors were observed during clinical trials [44,45].
Our results, consistent with the previous studies, suggested a bidirectional relationship between depressive syndrome and biological inflammatory syndrome, but this finding should be validated in a longitudinal study.It remains to be determined whether systemic inflammation could induce emotional symptoms through neuronal cells or whether these emotional disorders are considered biological inflammatory syndrome that subsequently trigger the onset of CIR.
One of the main limitations of this preliminary study was the lack of statistical power due to a small sample size.However, this pilot study investigated potential associations between CIR and biopsychosocial factors, and its findings are indicative of areas that should be examined in greater depth in future studies.In addition, because this was an observational study, no conclusions could be drawn about any association with causality.As this was a crosssectional investigation, it was not possible to determine whether the primary emotional regulation disorders, possibly contributing to the rheumatological disease onset or secondary disorders, should be interpreted as adaptive or coping modalities to handle a functional handicap or limiting pain caused by the disease.Our study suffered from two selection biases: (i) it was a monocentric study recruiting patients at a university hospital, representing a particular sociodemographic status; and (ii) most patients were assessed in a day for severe pathologies and comorbidities.Furthermore, the clinical psychological parameters chosen for assessment, although not subjected to consensual agreement, corresponded to clinical entities.Nonetheless, participants in this study benefited from combined psychiatric and rheumatological assessments during a day of hospitalization, which is considered novel, enabling the evaluation of a large number of psychological factors and comparing them to concomitant rheumatological findings.

Conclusion
Our study established an association between emotional repression, intense life events, and the etiology and development

Table 5 .
Psychological assessments of patients with versus without articular structural involvement of CIR.Moreover, RA patients had significantly more depressive episodes than SpA patients.Our observations suggested that targeted psychotherapy could complement the clinical management of CIR.