Factors associated with pain coping and catastrophising in patients with systemic lupus erythematosus: a cross-sectional study of the LuLa-cohort.

OBJECTIVE: The aim of this study was to identify factors associated with pain coping and catastrophising in patients with systemic lupus erythematosus.
METHODS: All patients were participants of the lupus erythematosus long-term study, which is based on patient-reported data assessed among members of the German Lupus Erythematosus Self-Help Organization. Assessments were performed by means of a questionnaire. Among self-reported clinical data the Pain-Related Self Statements Scale (PRSS) was included. To depict significant differences univariable analyses were carried out using non-parametrical rank tests. To examine factors influencing our outcome variables, we performed a multivariable stepwise regression model including variables that presented significantly in the univariable analysis.
RESULTS: 447 cases (94.9% female) were analysed showing a mean catastrophising score of 1.1 (SD 0.8) and a mean coping score of 2.8 (SD 0.9) in the PRSS subscales. Higher catastrophising quartiles went along with higher experienced pain, lupus activity, fatigue, damage and decreased health related quality of life, whereas they presented inversely for coping. In our multivariable model, factors associated with catastrophising were: number of lupus-specific drugs (p value 0.004), pain in the last 7 days (p value 0.034), the Short Form 12 Health Survey Mental Component Summary (p value <0.001) and disease activity measured by the Systemic Lupus Activity Questionnaire (p value 0.042). Social participation reflected by performed leisure activities such as dancing or bowling had a positive association with coping (p value 0.006). In contrast, other health related physical activities and their extent had no impact on coping. A direct association between the amount of pain coping and catastrophising, as well as a great impact of the catastrophising, respectively, coping level on physical and mental functioning could be shown.
CONCLUSIONS: Reduction or increase of detected factors might lead to a modification of pain coping and catastrophising and offer an approach to more effective care in patients with SLE.

More pain, damage and lupus-specific drugs as well as worse mental health go along with higher levels of pain catastrophising.

Social participation/activity might enhance pain coping and reduce catastrophising.

Introduction

Patients with systemic lupus erythematosus (SLE) suffer from possible involvement of numerous organ systems, and often from pain,1–3 fatigue,1 4–6 sleep disorders,1 5 7 8 fear,9 depression9–13 and cognitive deficits.14–19 These complaints, either as an independent condition or associated or aggravated by the disease, are frequently linked to physical and mental restraints. Coping describes a set of intentional, goal-directed efforts people engage in to minimise physical, psychological or social harm of an event or a situation.20 It encompasses behavioural and psychological strategies. These strategies can help in dealing with stress caused by the disease and are associated with a better health related quality of life (HRQoL) in patients with SLE.21 In this work we subsume the beneficial effects of coping with pain under the term ‘coping’. In contrast catastrophising represents a maladaptive cognitive style employed by patients and is associated with an irrationally negative forecast of future events.

Coping is more efficient in an existing reliable social network, which may provide socioemotional support.22 The latter has a high impact on disease activity, damage and quality of life.23 As patients with SLE frequently report a poorer social support than healthy controls,24 it represents a modifiable option to enhance coping behaviour. Former studies demonstrated positive influence of physical activity on fatigue in SLE25–27 and pain as well as physical function in other rheumatic diseases.28 Referring to these results we hypothesised that physical activity might have an influence on coping strategies as well.

In contrast it is known that catastrophising may have serious impact on chronically ill patients in general and patients with SLE in particular.29 Maladaptive coping characterised by catastrophising is associated with increased pain experience30 and predicts higher levels of pain in patients with chronic rheumatic diseases (eg, fibromyalgia syndrome and rheumatoid arthritis).31 Furthermore, catastrophising and maladaptive coping strategies are linked to higher levels of functional impairment and depression in rheumatoid arthritis32 and SLE.33 Numerous other studies verified the negative influence of catastrophising, respectively, suboptimal coping strategies on the outcome of various other chronic diseases by occurrence of depressive symptoms,34 35 reduced cognitive performance36 or even an increased risk for suicide.37

As psychological interventions and education are able to increase coping abilities in patients with SLE and can thus improve their quality of life,38 it is of major importance to explore the main stressors that affect coping behaviour.

The aim of this study was to identify factors that are associated with our outcomes pain coping and catastrophising in patients with long-standing SLE and to detect possible susceptible targets for intervention. Therefore, we analysed several demographic parameters, disease related outcomes, physical activity, physical and mental functioning, social participation and their impact on pain coping and catastrophising.

Methods

The data was collected within the lupus erythematosus long-term study (LuLa-study), a prospective, patient-centred study investigating the long-term management and course of disease as well as quality of life in patients with SLE. Data collection started in 2001 with annual postal questionnaires among members of the German Lupus Erythematosus Self-Help Organisation. Inclusion criteria were a reported diagnosis of SLE, being a member of the German Lupus Erythematosus Self-Help Organisation and having returned the completed questionnaire. Prior evaluation of the LuLa cohort showed that its data is comparable to physician-reported data and thereby is representative of patients with SLE in Germany.39 In 2009 we surveyed for 18 comorbidities (hypertension, myocardial infarction, stroke, chronic kidney damage, diabetes, cancer, chronic respiratory disease, chronic liver damage, chronic gastrointestinal disorders, hypercholesterolaemia, mental illness, arthrosis, scarring changes of skin, osteoporosis, fibromyalgia, thrombosis, miscarriages, early menopause), lupus-specific drugs, sociodemographic characteristics, inability to work, degree of disability (%), HRQoL measured by the Short Form 12 Health Survey (SF-12),40 ‘pain in the last 7 days’ (NRS 0–10), ‘impairment in the last 7 days’ (NRS 0–10), disease flares during the last 3 months, disease activity measured by the Systemic Lupus Activity Questionnaire (SLAQ),41 health related physical activity assessed by the Freiburg Questionnaire for Physical Activity (FFkA),42 situation-specific aspects of the patients’ cognitive coping with pain (‘catastrophising’ and ‘coping’), measured by the Pain-Related Self Statements Scale (PRSS),43 fatigue measured by the Vanderbilt Fatigue Score (VFS),44 and the Systemic Lupus International Collaborating Clinics/American College of Rheumatology Damage Index (SLICC/ACR DI).45–47 Most assessments were self-reported, only for the SLICC/ACR DI patients were asked to let their physician complete the questionnaire. All questionnaire items were assessed at the same time. Due to the questionnaire's length specific instruments regarding socioemotional support had to be omitted. Therefore we used a construct using other variables.

Measures

Pain-Related Self Statements Scale

Detection of coping strategies was carried out by the German version of the PRSS, which is intended to assess situation-specific cognitions that either promote or hinder attempts to cope with pain. The PRSS includes 18 items, which are subsumed into two nine-item subscales, termed ‘catastrophising’ and ‘coping’. Items are rated on a numerical rating scale, describing how often a statement enters the patients’ minds when they experience severe pain (0=almost never to 5=almost always). Questionnaire's reliability, validity and sensitivity to change were proven by Flor et al43 in a sample of 415 patients including 120 patients with chronic pain suffering from various rheumatic disorders, 213 patients suffering from chronic back pain, 44 patients with temporomandibular pain and dysfunction (TMPD) and 38 healthy controls.

Both subscales demonstrated to be valid and sensitive to change, and to be closely related to pain intensity and interference from pain experiences. There are no cut-off-values defined for high or low coping, respectively, catastrophising, but results from healthy controls showed a relatively low catastrophising score of 0.9 (SD 0.8) and a relatively high coping score of 3.4 (SD 1.1).43 In comparison patients with chronic back pain depicted a catastrophising score of 2.0 (SD 1.2) and a coping score of 3.0 (SD 0.9) and patients with TMPD a catastrophising score of 2.3 (SD 1.0) and a coping score of 2.8 (SD 0.7).43

Short-Form 12 Health Survey

The 12-item short-form (SF-12) is based on the 36-item short-form (SF-36)40 and is used to survey a population's health status. The SF-12 provides comparable results to the SF-36.48 49 Two subscales can be extracted: The Physical Component Summary (PCS) and the Mental Component Summary (MCS). Because of the questionnaire's size limitations we adopted the MCS as a substitute for a more extensive inquiry of the individual factors affecting mental health (eg, depression, anxiety disorders). The MCS has repeatedly proven to be a valid instrument for identifying and assessing the severity of depression and anxiety.50 Additionally the physical functioning index of the SF-36 (SF-36-pfi)51 was assessed.

Systemic Lupus Activity Questionnaire

In order to screen for current disease activity the SLAQ, which is based on the Systemic Lupus Activity Measure,52 was applied. The SLAQ is a patient-reported questionnaire consisting of 24 items, which aims at detection of disease activity in patients with SLE.41 In large observational, community-based cohorts of people with SLE it demonstrated to have an adequate reliability, construct validity and responsiveness.53 54 Additionally, we used the supplemental SLAQ item 1 which assesses the occurrence and severity of disease flares (no/mild/moderate/severe flare) during the last 3 months.

Systemic Lupus International Collaborating Clinics/American College of Rheumatology Damage Index

In order to detect damage in patients with SLE the SLICC/ACR DI was used.45–47 It records damage in nine different organ systems, which accrued since disease onset and persisted for at least 6 months.

Freiburg Questionnaire for Physical Activity

The cohort's health related physical activity was assessed by the FFkA.42 It consists of 12 questions related to the duration of performed basic activities, recreational activities and sport activities per week. In the context of this study, a calculation of leisure activity was carried out using the items ‘dancing’ and ‘bowling’. As these two social leisure activities, contrary to walking, using exercise machines, and so on, can only be performed in company, we considered them as a proxy measure for patients’ ‘social participation’.

Further questions considered the incidence of sick leave during the past 12 months, the number of weeks on sick leave and the degree of disability. The degree of disability is a standardised official governmental approval of non-temporary bodily or psychological disability that is due to an irregular state. Severity is graded from 20% to 100%.

Statistical analysis

Collected data were analysed with the statistical software programme IBM SPSS Statistics V.19. A descriptive analysis was carried out by the calculation of mean, median, SD, minimum and maximum, where applicable.

By means of the PRSS the ‘catastrophising’ and ‘coping’ subscales were calculated. To assess the association between the two subscales and other outcome parameters quantitatively, we studied the four quartiles of the subscales (lower quartile <25%, 25–50%, 50–75%, upper quartile >75%) in relation to each outcome parameter. Spearman's correlations were used for ordinal-scaled and Pearson's correlations for interval-scaled data. Because of the subscales skew distribution univariable analyses (UVA) were carried out using non-parametrical rank tests to depict significant differences. In cases of two independent samples the Mann-Whitney U test and in cases of multiple independent samples the Kruskal-Wallis test was applied. p Values less than 0.05 were considered statistically significant for all tests. In order to prevent an α error due to multiple testing, adjustment of the p values by means of the Bonferroni correction was performed. Consequently, levels of statistical significance were defined at p<α/k (α=0.05/k=number of tests).

To examine factors influencing our outcome variables ‘catastrophising’ and ‘coping’ among patients, we used a stepwise regression model to reduce the number of variables. Those variables that presented as significant in the UVA, entered this multivariable analysis (MVA) model including clinical and demographic parameters (SF-12 PCS, SF-12 MCS, SLAQ score, VFS, SLICC/ACR DI, number of concomitant diseases, number of lupus-specific drugs, inability to work, degree of disability (%), pain in the last 7 days, impairment in the last 7 days, disease flares during the last 3 months and social participation). Missing values were not imputated. Analysed number of cases may vary due to missing values. In MVA only complete cases were included.

Results

The LuLa 2009 cohort included 620 patients of whom 173 reported no pain (86,1%) or did not answer the PRSS questionnaire (13,9%). This resulted in 447 cases (94.9% female) with a mean age of 52.0 years (SD 12.5) and mean disease duration of 16.1 years (SD 8.1), which were included in the analyses. The analysis of the PRSS subscales showed a mean catastrophising score of 1.1 (SD 0.8) and a mean coping score of 2.8 (SD 0.9). Details and further results of self-reported disease activity, damage, current treatment, employment status, assessments of general health and physical functioning are listed in table 1.

Table 1

Sample description (447 patients)

	Per cent	Mean±SD	Range	n
Age in 2009 (years)		52.0±12.5	14–86	447
Disease duration in 2009 (years)		16.1±8.1	5–49	440
Medication				447
NSAIDs	29.3
Cortisone ≤7.5 mg	53.0
Cortisone >7.5 mg	16.1
Antimalarials	41.2
Methotrexate	9.8
Azathioprine	18.8
Ciclosporine	2.2
Cyclophosphamide	1.1
Mycofenolatmofetil (MMF)	13.0
Medication groups				447
NSAIDs or no medication	14.3
Antimalarials and/or steroids, possible NSAIDs	40.3
Immunosuppressants, possible NSAIDs, antimalarials, steroids	45.4
Occupation type				443
Retired	50.1
Employees	29.6
Unemployed	12.9
Worker	3.8
Self-employed	2.9
Pupils/students	0.7
Physical functioning (0–100) (SF-36-pfi)		62.4±27.5	0–100	447
SF-12 PCS		37.9±11.0	11.0–62.9	414
SF-12 MCS		46.1±11.7	18.4–69.5	414
SLAQ score		14.8±7.0	0–37	446
VFS score		26.7±19.2	0–70	424
SLICC/ACR DI		2.8±2.9	0–21	407
PRSS catastrophising		1.1±0.8	0–4.1	426
PRSS coping		2.8±0.9	0–4.8	431

Data are percentages (%), means, SD and ranges.

N, number of cases; NSAIDs, non-steroidal anti-inflammatory drugs; PRSS, Pain-Related Self Statements Scale; SF-12 MCS, SF-12 Mental Component Summary; SF-12 PCS, SF-12 Physical Component Summary; SF-36-pfi, 36-item short-form physical functioning index; SLAQ, Systemic Lupus Activity Questionnaire; SLICC/ACR DI, Systemic Lupus International Collaborating Clinics/American College of Rheumatology Damage Index; VFS, Vanderbilt Fatigue Score.

Table 2 opposes four groups of different catastrophising and coping levels. Higher catastrophising centiles (upper quartile) are associated with higher experienced pain, current lupus activity (SLAQ score), fatigue (VFS score), damage (SLICC/ACR DI scores) and a decrease of HRQoL domains (SF-36-pfi, SF-12 PCS and SF-12 MCS) whereas they present inversely for coping (table 2).

Table 2

Characteristics of PRSS catastrophising and coping groups split by quartiles

	PRSS catastrophising				PRSS coping
	Lower quartile (0–0.4)	25–50% (0.5–1.0)	50–75% (1.1–1.7)	Upper quartile (1.8–4.1)	Lower quartile (0–2.3)	25–50% (2.4–2.9)	50–75% (3.0–3.4)	Upper quartile (3.5–4.8)
n	113	101	114	98	131	106	102	92
Age in 2009	50.8	53.2	51.0	52.4	51.6	52.5	51.4	51.0
Disease duration in 2009	16.0	16.4	16.0	15.8	16.6	15.9	15.6	15.7
Coping	2.8	2.9	2.9	2.7	1.6	2.6	3.2	3.9
Catastrophising	0.2	0.7	1.4	2.2	1.2	1.3	1.1	0.9
SF-36-pfi	73.3	65.5	59.3	51.6	62.8	57.0	63.9	66.5
SF-12 PCS	41.6	38.8	36.0	34.9	38.8	36.5	37.6	38.6
SF-12 MCS	52.2	49.2	43.6	40.0	44.5	43.1	48.3	49.8
SLAQ score	10.5	13.6	17.1	18.2	14.7	16.4	14.4	13.4
VFS score	17.9	24.2	30.6	34.4	27.9	29.1	25.9	22.0
SLICC/ACR DI	1.8	2.6	3.1	3.6	2.5	3.5	3.1	2.2
Pain (last 7 days)	2.6	3.4	4.3	5.1	3.8	4.2	3.8	3.3

Data are means illustrated separately for PRSS catastrophising and PRSS coping. Different levels of catastrophising resp. coping are represented in four quartiles.

n, number of cases; PRSS, Pain-Related Self Statements Scale; SF-12 MCS, SF-12 Mental Component Summary; SF-12 PCS, SF-12 Physical Component Summary; SF-36-pfi, 36-item short-form physical functioning index; SLAQ, Systemic Lupus Activity Questionnaire; SLICC/ACR DI: Systemic Lupus International Collaborating Clinics/American College of Rheumatology Damage Index; VFS, Vanderbilt Fatigue Score.

Nine out of the 18 acquired comorbidities showed significant association to the PRSS subscale catastrophising with higher catastrophising means in presence of comorbidities than in absence (table 3).

Table 3

Univariable analysis: comorbidities (n=447)

Comorbidity	Catastrophising mean±SD (n)		p Value catastrophising
	Comorbidity=yes	Comorbidity=no
Diabetes	1.4±0.8 (8)	1.1±0.8 (375)	0.034
Chronic respiratory disease	1.2±0.8 (36)	1.1±0.8 (349)	0.031
Chronic liver damage	1.5±0.9 (10)	1.1±0.8 (370)	0.007
Chronic gastrointestinal disorders	1.3±0.8 (26)	1.0±0.8 (355)	0.002
Mental illness	1.7±0.8 (38)	1.0±0.7 (342)	<0.001
Arthrosis	1.2±0.8 (86)	1.0±0.7 (295)	0.021
Scarring changes of skin	1.3±0.8 (47)	1.0±0.8 (332)	<0.001
Fibromyalgia	1.4±0.8 (27)	1.1±0.8 (347)	0.014
Stroke	1.9±0.9 (9)	1.1±0.8 (376)	0.023
Hypertension			NS
Myocardial infarction			NS
Chronic kidney damage			NS
Cancer			NS
Hypercholesterolaemia			NS
Osteoporosis			NS
Thrombosis			NS
Miscarriages			NS
Early menopause			NS

Data are means and SD, n: number of cases. Means are depending on existence of comorbidities (comorbidity=yes, means: comorbidity exists; comorbidity=no, means: comorbidity does not exist). The p values for catastrophising were derived from the Mann–Whitney U test. As there were no significant associations regarding coping, it is not listed in the table.

NS, not significant.

A relationship between coping and comorbidities was not observed. Three of the nine comorbidities with significant association remained significant after Bonferroni correction (p value <0.003). These three comorbidities ‘chronic gastrointestinal disorders’ (p value 0.002), ‘mental illness’ (p value <0.001) and ‘scarring changes of skin’ (p value <0.001) were included in the MVA but did not show significant association to catastrophising any more.

In total, 52 patients (11.6%) participated in the bowling and dancing activities, whereby 31 (7.0%) participated in dancing, 13 (2.9%) in bowling and 8 (1.8%) in both. This was considered as social participation.

As shown in table 4, coping presented statistical significance for ‘social participation’ in the UVA (p value 0.001) and the MVA (p value 0.006). Mean coping was significantly higher in participants with social participation than without (mean 3.2 vs 2.8). ‘SF-12 MCS’ was significant for both PRSS subscales in the UVA (p value catastrophising <0.001; p value coping 0.001) but only significant for catastrophising in the MVA (p<0.001). Catastrophising demonstrated a multitude of significant correlations with different disease parameters and conditions among others SF-12, SLAQ, SLICC/ACR DI and disease flares during last 3 months. The variables ‘pain in the last 7 days’ (correlation coefficient 0.420), ‘impairment in the last 7 days’ (correlation coefficient 0.388), fatigue (VFS score) (correlation coefficient 0.367) and disease activity (SLAQ score) (correlation coefficient 0.475) evinced the highest correlation coefficients (table 4).

Table 4

Findings of the univariable analysis (UVA) and multivariable analysis (MVA)

Variable name	Catastrophising			Coping
	p Value (UVA)	Correlation coefficient	p Value (MVA)	p Value (UVA)	Correlation coefficient	p Value (MVA)
SF-12 PCS	<0.001	−0.250**		NS
SF-12 MCS	<0.001	−0.443**	<0.001	0.001	0.171**	NS
SLAQ score	<0.001	0.475*	0.042	NS		NS
VFS score	<0.001	0.367*		NS
SLICC/ACR DI	<0.001	0.192*		NS
Number of concomitant diseases	<0.001	0.202*		NS
Number of lupus-specific drugs	<0.001	0.204*	0.004	NS		NS
Inability to work	0.001	0.213*		NS
Degree of disability (%)	0.003	0.169**		NS
Pain in the last 7 days	<0.001	0.420*	0.034	NS		NS
Impairment in the last 7 days	<0.001	0.388*		NS
Disease flares during last 3 months	<0.001	0.280*		NS
Social participation	NSX		NS	0.001X		0.006

Data are p values and correlation coefficients. Correlations were carried out by Spearman's (*) and Pearson's (**) correlations. The p values for catastrophising and coping in the UVA were derived from Spearman's and Pearson's correlation or from the Mann–Whitney U test (x). The p values for catastrophising and coping in the MVA were derived from ANOVA (analysis of variance). Only complete cases (n=323) were included in the MVA.

SF-12 MCS, SF-12 Mental Component Summary; SF-12 PCS, SF-12 Physical Component Summary; SLAQ, Systemic Lupus Activity Questionnaire; SLICC/ACR DI, Systemic Lupus International Collaborating Clinics/American College of Rheumatology Damage Index; VFS, Vanderbilt Fatigue Score.

After Bonferroni correction all significant variables of the UVA were included in the MVA. Four variables reached statistical significance regarding catastrophising (‘number of lupus-specific drugs’ (p value 0.004), ‘pain in the last 7 days’ (p value 0.034), ‘SF-12 MCS’ (p value <0.0001), and ‘SLAQ score’ (p value 0.042)), and one (‘social participation’; p value 0.006) regarding coping.

Parameters for basic, sports and leisure activities, assessed by the FFkA, were also compared with the two PRSS subscales. However, none of the activity parameters or the extent of health related physical activity showed a coherent statistical significance regarding pain coping and catastrophising.

Discussion

Despite the numerous recent advances in research and clinical treatment, SLE keeps being an incurable chronic disease with a huge impact on persons’ lives by affecting patients’ physical and mental functioning. Former studies illustrated that coping capacities help in dealing with the disease and are associated with a better HRQoL.21

The PRSS results from our cross-sectional study in SLE depicted significantly worse catastrophising and coping scores than reported from healthy controls.43 Coping presented worse in our cohort compared with patients with chronic pain but was similar to patients suffering from TMPD.43 In contrast our cohort presented significantly less catastrophising.43 As pain related catastrophising differs in younger and older adults55 these differences may partly be explained by the younger age in Flor's cohort43 (42.4 years vs 52.1 years). Compared with patients with fibromyalgia56 our SLE cohort showed lower amount of catastrophising and coping. This might be attributed to a high rate of depression in the fibromyalgia cohort and to the inclusion of patients participating in pain management programmes.56

We identified valuable parameters that seem to influence the occurrence of either catastrophising or coping in patients with SLE. Yet it should be noted that cross-sectional studies limit the ability to make causal assumptions between the predictors and outcomes.

First, our results demonstrated that patients with SLE with our above mentioned comorbidities catastrophise more than those without, which emphasises the impact of these specific comorbidities in patients with SLE. Studies from other chronic diseases support our findings.57 58 Thus it is of importance that ‘scarring changes of skin’,59 ‘mental illness’60 and ‘gastrointestinal disorders’61 might also be attributable to lupus and are preventable by an early and optimised lupus therapy.

Second, we found statistically significant associations between the number of lupus-specific drugs, pain, ‘SF-12 MCS’, damage and catastrophising. In detail the high correlation between the ‘experienced pain in the last 7 days’ and catastrophising (Table 4) confirms findings from former studies depicting that the experience of pain in rheumatological disorders is closely related to catastrophising.29 31 62

The moderate correlation between ‘SF-12 MCS’ (a proxy for depression and anxiety) and catastrophising presented in our work substantiates the fact that suboptimal mental health is associated with catastrophising, which has already been demonstrated in former studies.34–37

Besides, we showed that the ‘number of lupus-specific drugs’ significantly increases with a higher catastrophising score. Yet, it has not been clarified whether a rising amount of medication leads to increased catastrophising or vice versa. Consequently, physicians should be aware of possible aggravations and ought to apply objective criteria for the choice of medication.

Third, referring to coping, the variable ‘social participation’ reached statistical significance in the MVA. This is allegeable by the numerous positive effects associated with an existing social network and the provided socioemotional support.22 An obviously anticipated reciprocal correlation between the amount of pain coping and catastrophising could likewise be shown in our work. This indicates that a sufficient manner of coping (eg, by increasing social participation/activity) might be helpful to reduce catastrophising and consequently improve physical and mental functioning in SLE (Table 2).

In addition, it needs to be considered that catastrophising, pain, disability and mood-emotional functioning might bias the questionnaire's response behaviour. Patients might indicate a poorer health condition than it actually is because of their tendency to catastrophise, their experienced severity of pain, their physical functioning and/or psychological condition. Our results depicted in table 2 support this hypothesis, as the group of the strongest catastrophiser (upper quartile) showed poorer values for all mentioned variables than the others (lower quartile, 25–50%, 50–75%). A similar observation was published by Mancuso et al63 who detected that patients with asthma with more depressive symptoms reported worse HRQoL than patients with asthma with similar disease activity but fewer depressive symptoms. These findings emphasise that psychological conditions might have an impact on patients’ response behaviour.

Most of our identified key factors that interact with pain coping strategies (number of lupus-specific drugs, pain, mental health, disease activity) have been identified as disease aggravating factors in the SLE treat-to-target recommendations published in 2014.64 They can be modified by an effective therapy which is regularly adjusted to lupus manifestations and individual needs. Furthermore pain and mental health as well as pain coping and catastrophising can effectively be improved by using psychological or psychoeducational interventions,38 65–68 which among others aim at increasing the self-efficacy of patients.

Limitations

There are certain limitations to this study. In favour of other items there was no implementation of further questions regarding social participation and social environment in the LuLa study. Our variable ‘social participation’ was constructed by use of two variables (‘dancing’ and ‘bowling’), not considering other social activities. It should be mentioned though that other solitary activities (eg, ‘taking a walk’, ‘cycling’), which by the way can be performed also in groups and not only alone, as well as the extent of health related physical activity, assessed by parts of the FFkA,42 did not show statistical significance regarding coping nor regarding catastrophising. For further evaluation future studies should consider other activities in detail as well (eg, membership with active participation in societies or clubs, team sports or regular leisure activities in groups) in order to provide better recommendations for optimal patient care.

As the surveyed participants were all members of the German Lupus Erythematosus Self-Help Organisation, we were not able to evaluate the impact of support groups. A distinction between passive and active members (eg, participants in support group meetings, activities or online discussion groups) could point out further strategies to improve coping and minimise catastrophising.

To our knowledge our sample included predominantly Caucasians and did not include other ethnical groups. In addition we are not able to comment on catastrophising and coping in recently diseased patients. Hence, further studies in an inception cohort are necessary in order to evaluate coping strategies in early disease.

Conclusion

Our work points out that in order to optimise pain coping and catastrophising in the care of patients with SLE and to affect patients’ physical and mental functioning, positive factors should be protected and promoted and negative factors need to be prevented. In this context social participation represents an essential positive factor whereas negative ones include disease activity, pain, impairments and an extensive lupus-specific medication. Further studies are necessary to determine which patients benefit most from a psychoeducation and to identify the impact of revised modifiable parameters on coping strategies.