Article Text

Stress, coping strategies and social support in patients with primary Sjögren’s syndrome prior to disease onset: a retrospective case–control study
  1. D Karaiskos1,
  2. C P Mavragani1,2,
  3. S Makaroni1,
  4. E Zinzaras3,
  5. M Voulgarelis1,
  6. A Rabavilas4,
  7. H M Moutsopoulos1
  1. 1
    Department of Pathophysiology, University of Athens School of Medicine, Athens, Greece
  2. 2
    Hospital for Special Surgery, New York, NY, USA
  3. 3
    Department of Biomathematics, University of Thessaly School of Medicine, Larissa, Greece
  4. 4
    Department of Psychiatry, University of Athens School of Medicine, and University Mental Health Research Institute, Athens, Greece
  1. Clio P Mavragani, MD, Hospital for Special Surgery, 535 East 70th St, New York, NY, 10021, USA; mavraganik{at}


Objectives: Previous evidence suggests the role of psychological stress in triggering the onset of autoimmunity. We aimed to investigate whether stress following major and minor life events could precede the onset of primary Sjögren’s syndrome (pSS). The role of coping strategies and social support, as compensating buffering mechanisms, was also explored.

Methods: 47 patients with pSS were compared with two control groups: 35 patients with lymphoma (disease controls, DC) and 120 healthy controls (HC) with disease onset within the previous year. All subjects completed questionnaires assessing the occurrence of major and minor stressful events, coping strategies and social support prior to disease onset. Data analysis was performed by univariate and multivariate logistic regression models.

Results: A higher number of patients with pSS reported the occurrence of negative stressful life events prior to disease onset compared with patients with lymphoma and HC, while the number and impact of daily hassles did not differ between the three groups. Coping strategies were defective and the overall social support was lower in patients with pSS compared with DC and HC groups. In the multivariate model, pSS status was associated with maladaptive coping and lower overall social support relative to DC and HC, as well as with an increased number of negative stressful life events compared with HC but not DC.

Conclusions: Prior to disease onset, patients with pSS experience high psychological stress following major negative life events, without developing satisfactory adaptive coping strategies to confront their stressful life changes. Lack of social support may contribute to the relative risk of disease development.

View Full Text

Statistics from

Sjögren’s syndrome (SS) is a systemic autoimmune disease characterised by oral and ocular dryness, as a result of lymphocytic infiltration and destruction of the salivary and lachrymal glands. It mainly affects middle-aged women and can occur either alone (primary SS, pSS) or in the context of an underlying connective tissue disease (secondary SS). pSS is a quite common autoimmune disorder with prevalence between 0.5 and 1.0%. Although the exact pathogenesis of the disease remains unknown, genetic, viral, hormonal and environmental factors have been implicated.1 Among the latter, stress as a trigger of autoimmunity had attracted considerable attention in the psychoimmunology literature quite early on.2 3 However, no data are so far available for pSS.

Stress can be defined as a real or perceived imbalance between environmental demands required for survival and an individual’s capacity to adapt to these requirements, resulting in both psychological and biological changes that could place a person at risk of illness.4 Anything that causes stress is called a stressor and can include disasters, life crises, life changes or daily hassles. Human beings survive by constantly adapting to the demands of an ever-changing environment, so stress is assumed as a normal process in the organisms’ attempt to maintain homeostasis.5 Three main approaches of assessing the role of stress in disease risk can be identified. The environmental approach focuses on the assessment of environmental events or experiences that are objectively associated with substantial adaptive demands. The psychological approach is based on individual subjective evaluation of the stressor itself as well as of his ability to cope with the demands, posed by specific events or experiences. Finally, the biological approach focuses on activation of specific physiological systems that have been repeatedly shown to be modulated by both psychologically and physically demanding conditions.6

In clinical practice, autoimmune patients frequently experience periods of intense stress prior to the disease onset or near exacerbation of illness.710 However, this hypothesis has not been systematically explored so far for pSS. The present study seeks to determine whether psychological stress11 following either major stressful life events or microstressors of daily life could precede the onset and serve as a trigger of pSS. Both major and minor stressful life events were assumed to be components of total stress, based on an integrative model.12 In addition, the coping strategies13 14 developed by these individuals as well as social support as compensating buffering mechanisms against stress have been evaluated.15


Cases and controls

This study was designed as a retrospective case–control study. Cases consisted of patients with pSS according to the revised European-American criteria;16 as controls two groups of similar age range and sex ratio with the patients with pSS were recruited: the first included consecutive patients diagnosed with lymphoma and served as disease controls (DC) and the second consisted of healthy controls (HC), recruited during the same time period. To lessen the difficulty of recalling the occurrence of previous stressful life events, only patients with disease onset within the previous year were included. The onset of the disease was defined as the presence of dry eyes, dry mouth and/or parotid enlargement for the pSS group and fever, weight loss, night sweats, fatigue and/or malaise for patients with lymphoma. These symptoms and the exact time of their onset were self-reported by the patients. All patients presented for evaluation in the Department of Pathophysiology, University of Athens School of Medicine (Greece) from September 2004 to February 2007. Exclusion criteria for both DC and HC groups included the presence of autoimmune disease in them or in first-degree relatives, neuropsychiatric history, use of psychotropic drugs or alcoholism and drug abuse. The study cohort consisted of 47 pSS, 35 DC and 120 HC of similar sex ratio and age range with the pSS group.

Cases and controls were required to complete a validated1720 battery of questionnaires after they have been previously informed by phone and had given written consent for their participation. A thorough interview was performed after the completion of the questionnaires and subjects were asked to describe all their stressful experiences—either major life events or daily hassles—for the year prior to disease onset for the patients with pSS and lymphoma and the entrance into the study for the HC. All stressful experiences were registered by the interviewer and compared with the results of the questionnaires. The study was consistent with current laws referring to the conduction of human medical studies.


The first part of the questionnaire included information about demographics, smoking habits, physical activity and alcohol use. Patients with primary SS and DC were screened for the 1-year period before the onset of the symptoms of their disease, while HC for the 1-year period before the completion of the questionnaire.

The second part of the questionnaire was designed to provide information about major life events, daily hassles, coping profiles and the level of social support the individuals experienced in the preceding 12 months from disease onset for the patients and time of enrolment in the study for the HC. It consisted of the Greek versions of Life Experience Survey (LES),17 the Daily Hassles Scale (DHS),18 the Cope Inventory (COPE)19 and the Medical Outcomes Study Social Support Survey (MOS SSS).20 All these psychometric scales have been previously validated in the Greek population. The number and the impact of major stressful life events (expressed as scores) were evaluated using the LES, while DHS was used for assessment of the presence of microstressors. The coping strategies against stress and the presence of social support were explored through the COPE and MOS SSS, respectively. Maladaptive coping was assumed to be the combination of denial, focus on and venting on emotions and behavioural disengagement (supplementary file 1).

The survey explored the stressors and the patients’ psychosocial adjustment prior to the onset of pSS and lymphoma symptoms; therefore, possible biases arising from patients’ somatic symptoms attributed to the disease itself were minimised. Specific instructions were given to patients with pSS and lymphoma in order to exclude possible symptoms of the disease itself in their stressful experience, as our survey aimed at exploring the role of stress prior to the onset of the symptoms and not prior the diagnosis.

Statistical analysis

In continuous variables statistically significant differences between the three groups (pSS, DC and HC) were detected using a Kruskal–Wallis test followed by a Mann–Whitney U test. Categorical variables were compared using a Fisher’s exact test. Associations of major and minor stressful life events, coping strategies and social support with pSS status were assessed as odds ratios (OR) and their respective 95% confidence interval (CI) using a univariate logistic regression model. In a multivariate logistic regression model, OR were adjusted by demographic variables, smoking and drinking habits (which produced statistical significant associations in the univariate analysis), negative stressful life event number >1, daily hassles number >10, maladaptive coping and social support score <50. The analysis was performed using the SPSS 10.0 software (Statistical Package for the Social Sciences) for Windows.


Sociodemographics and habits in all three groups as well as the presenting symptoms in patients with pSS and DC are presented in table 1. Associations of stressful life events, daily hassles, coping strategies and social support with pSS are presented in tables 2, 3, 4 and 5 respectively. Loss of beloved persons (death of close family member or spouse) represented 17.1% of negative life events reported by patients with pSS compared with 2.3% of DC and 2.7% of HC (p<0.01 and p<0.0001 respectively). In addition, DC reported a higher incidence of stressful negative life events related to children leaving home compared with HC (16% vs 5%, p<0.01) (supplementary table 1). The prevalence of all the other major negative stressful life events did not differ significantly between the three groups.

Table 1 Sociodemographic characteristics, presenting symptoms and habits in patients with pSS, DC and HC
Table 2 Associations of stressful life events with pSS
Table 3 Associations of daily hassles scores with pSS
Table 4 Associations of coping strategies with pSS
Table 5 Associations of social support scales with pSS

Table 6 illustrates the associations between pSS status and psychometric characteristics after adjustment for demographic characteristics, smoking and drinking habits, negative stressful life events number >1, daily hassles number >10, maladaptive coping and overall social support index <50 in two set of comparisons; pSS versus DC and pSS versus HC. The associations of negative stressful life events number >1 (adjusted OR = 4.25, 95% CI: 1.57 to 11.49), maladaptive coping (adjusted OR = 10.19, 95% CI: 3.56 to 29.16), overall social support index <50 (adjusted OR = 6.04, 95% CI: 2.10 to 17.37) and pSS status remained significant as in the univariate model, when patients with pSS were compared with HC. In the multivariate analysis of patients with pSS versus DC, significant associations remained for maladaptive coping (adjusted OR = 9.35, 95% CI: 2.36 to 37.09) and overall social support index<50 (adjusted OR = 6.91, 95% CI: 1.79 to 26.70).

Table 6 Adjusted OR and 95% CI showing the associations between negative SLE number >1, daily hassles number >10, maladaptive coping, overall social support score <50 and pSS status in multivariate analysis after adjusting for demographics (residential and occupational status), smoking habits and alcohol overuse

Neither the demographic characteristics, nor smoking and alcohol drinking produced significant associations in the multivariate model.


In this paper, our aim was to evaluate the role of stress as an initiating trigger for the development of a classic autoimmune disorder such as pSS. At the same time we sought to explore the coping strategies and social support mechanisms present in these individuals prior to disease onset. This is the first report applying current psychometric testing on patients with pSS, which is based on a holistic view of the stress–illness process.21 In addition to the use of standardised psychometric scales, we explored the magnitude of stressful experiences using a thorough interview. The use of a second disease control group of a severe chronic disease eliminates possible confounding factors attributed to the distress of the chronic illness per se.

According to our findings, patients with pSS seem to experience not only a greater number of negative stressful life events in the year prior to disease onset but actually higher negative life events scores; therefore, they tend to perceive life changes as extremely undesirable. Experiencing more than one negative stressful life event in the year preceding the onset of pSS increases fourfold the risk of disease onset in comparison with HC. Similar conclusions were drawn in previous reports linking major stressful life experiences with the onset or exacerbations7 8 of other autoimmune conditions such as RA,9 22 polymyalgia rheumatica, temporal arteritis,23 systemic lupus erythematosus,10 24 Graves disease25 or multiple sclerosis.26 Although negative stressful life events seemed to precede the onset of lymphoma too, their number and impact was not as high as in pSS. Loss of beloved persons (death of close family member or spouse) was a major source of distress in patients with pSS compared with DC and HC. While microevents are assumed to be components of total stress and previous reports have implicated them in the onset of autoimmune disorders,9 27 psychological stress following minor life events did not seem to increase the relative risk of pSS onset.

The second major finding in our study was that, premorbid stable coping strategies implemented by patients with pSS in order to confront stressful life experiences tend to be maladaptive.28 These included denial (rejection of the reality),29 mental and behavioural disengagement (increased awareness of emotional distress and withdrawing effort from the attempt to attain the goal with which the stressor is interfering) and focus and venting of emotions (concomitant tendency to discharge those feelings). The above avoidance coping profiles have been considered to have bad outcomes in stress illness studies.3033 In contrast, adaptive coping strategies might offer protection by regulating the negative emotions associated with stress.34 Maladaptive coping in patients with pSS was a constant finding even when this group was compared with patients with lymphoma. A blunted response to biological stress has already been documented through hypothalamic–pituitary–adrenal axis dysfunction35 and this study comes to add a maladaptive response of patients with pSS to psychological stress. This coping profile, especially denial mechanisms, is not state-dependent but reflects a pre-existing trait-dependent personality structure. Our study explores dispositional coping strategies before disease onset as stable stress-coping skill and events occurred in the premorbid period. There is a specific pattern of patients with pSS coping, which can be attributed mainly to personality factors rather than the disease process itself. On the contrary, situational coping after the disease onset has already been studied.36

Furthermore, patients with pSS report a lack of functional and structural social support resources compared with DC and their healthy counterparts. The social support scale that we used allows assessment of the availability of perceived social support and not the received social support. It does not include identification of specific sources of support. Social support is a process that arises through interactions between people and refers to the social network including family, friends, spouse or a person providing protective emotional and material resources.37 Social support may act as a shield, reducing the likelihood of undesirable life events from occurring, lessening the adverse impact of stressful events or helping with the adjustment to the situation.38 39 Social support is classically conceived as a protective factor for health, whereas perceived social isolation is a major risk factor for morbidity and mortality.40

Given that the environment is a key component in stress pathogenesis, we carefully recorded demographic parameters of all patients and controls at the entry of the study. Despite the differences detected between patients with SS and controls in certain characteristics, such as employment or smoking habits in the univariate analysis, these associations did not remain significant in the multivariate model.

One could also argue that the impaired coping strategies and social support detected in the SS group might be the result of the increased burden of stressful life events reported by these patients. However, regarding coping strategies, the dispositional form of COPE inventory, which we used, allowed the assessment of coping abilities of an individual as a stable trait of reacting to stress irrespectively from stress triggers. Even if there seems to be a rational association between high stress, less active coping and less social support when all these variables were entered in the multivariate model, each of them was correlated significantly with pSS onset as an independent risk factor.

We are fully aware of the limitations of a retrospective study addressing the causal contribution of stressful events in the onset of an autoimmune disease such as SS. The sample might be relatively small to provide substantial evidence of the impact of our findings; however, this limitation resulted from the need to apply our psychometric tests in newly diagnosed patients (within 1 year from onset of symptoms) in order to eliminate possible recall biases. Although LES is an objective scale that allows the documentation of true facts such as death, unemployment or divorce, one could argue that in the context of a chronic disease, patients tend to attribute the onset of their illness to specific adverse events; in this context, both patient groups may have overestimated the actual number of major negative life events. In favour of this hypothesis is the occurrence of increased negative stressful life events before lymphoma onset too, which has not been reported so far. However, the number and the impact of negative stressful life events prior to lymphoma onset are not as high and as severe as reported by patients with pSS. Furthermore, self-reported lack of social support in patients with pSS may be subjective, as due to the mental and behavioural disengagement coping profile, this group of patients avoid help from the social environment. Unfortunately, the measures used in our study cannot specify the exact role of social contributions, but only how patients make use of the available social resources. MOS SSS allows assessment of the availability of perceived social support, which is defined as the support that the individual perceives as available; however, it does not include identification of specific sources of support. Therefore, whether the patients have poor coping mechanisms, or others in society are unfamiliar with the condition and, therefore, do not render support, remains a question. Unfortunately, under the design of this study we could not identify the attitudes of family members, of the spouses as well as of the physicians towards patients with pSS.

Socio-economic status is a risk factor for arthritis41 and paternal low socio-economic status has been previously correlated with pSS.42 Despite that in the univariate analysis there were differences in regard to family, residential and occupational status between patients and controls, when all these parameters were entered in the multivariate model they did not produce significant associations, suggesting that impaired coping and stressful life events represented independent risk factors for pSS onset. Unfortunately, the financial status of our study patients was not directly assessed.

The pathophysiological mechanisms governing the effect of stress in the induction of an autoimmune disorder, such as pSS, are not well delineated. Short-term psychological stress exceeding a threshold induces the activation of the hypothalamic–pituitary–adrenal axis, as well as the sympathetic–adrenal medullary axes leading to the peripheral production of glucocorticoids and catecholamines in order to improve organisms’ homeostasis and increase chances of survival. However, long-term stressful stimuli result in the hypofunction of both axes.43 44 Hypoactivity of the hypothalamic–pituitary–adrenal axis in patients with SS has been described so far45 reflected by the low serum levels of cortisol.46 Altered levels of glucocorticoids and catecholamines induce changes in cellular trafficking, proliferation, cytokine secretion, antibody production and cytolytic activity; these imply a putative modulating role of neuroendocrine interactions on the function of the immune system and the generation of autoimmune disorders.47 In this setting, low-dose treatment with corticosteroids is beneficial in replacing the reduced levels of endogenous cortisol in autoimmune rheumatic diseases.48 While many people confront stressful experiences, only a small proportion of them develop autoimmunity. Most likely, stress interacts with predisposing alleles, hormonal factors such as oestrogens—known to enhance immune responses49—or other neurotransmitters leading to the activation of genes that promote autoimmunity.

In conclusion, prior to disease onset, patients with pSS experience a period of high psychological stress mainly in the form of major negative life events without developing satisfactory adaptive coping strategies to confront their stressful life changes. Lack of perceived social support seems to increase further their relative risk for disease development. These findings highlight the importance of a multidimensional clinical approach for these patients, providing at the same time some clues for the implication of stress in the aetiopathogenesis of autoimmune disorders.


View Abstract

Supplementary materials


Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.