Illness Perceptions and Fatigue in Systemic Vasculitis

Authors


Abstract

Objective

To compare illness perceptions among patients with different forms of vasculitis, identify risk factors for negative illness perceptions, and determine the association between illness perceptions and fatigue.

Methods

Participants were recruited from an online vasculitis registry to complete the revised Illness Perception Questionnaire (IPQ-R). The mean scores on each IPQ-R dimension were compared across different types of vasculitis. Cluster analysis and stepwise regression identified predictors of negative illness perception. Fatigue was measured using the general subscale of the Multidimensional Fatigue Inventory (MFI-20). Patient-reported measures of disease activity and IPQ-R dimensions were assessed in relation to MFI-20 scores using linear regression in sequential, additive models with model-fit comparisons.

Results

In total, 692 participants with 9 types of vasculitis completed the IPQ-R. For 6 of the 8 IPQ-R dimensions, there were no significant differences in mean scores between the different vasculitides. Scores in the identity and cyclical dimensions were significantly higher in Behçet's disease compared with other types of vasculitis (13.5 versus 10.7 for identity and 4.0 versus 3.2 for cyclical [P < 0.05]). Younger age (odds ratio [OR] 1.04, 95% confidence interval [95% CI] 1.02–1.06), depression (OR 4.94, 95% CI 2.90–8.41), active disease status (OR 2.05, 95% CI 1.27–3.29), and poor overall health (OR 3.92, 95% CI 0.88–17.56) were associated with negative illness perceptions. The sequential models demonstrated that the IPQ-R dimensions explained an equivalent proportion of variability in fatigue scores compared with measures of disease activity.

Conclusion

Illness perceptions are similar across different types of vasculitis, and younger age is a risk factor for negative illness perceptions. Illness perceptions explain differences in fatigue scores beyond what can be explained by measures of disease activity.

INTRODUCTION

The idiopathic vasculitides are a group of rare, systemic diseases involving inflammation of arteries and other tissues with resulting organ- and life-threatening disease courses. The different forms of idiopathic vasculitis are typically classified based on the predominant size of the arteries affected in each condition, including small vessel vasculitis (e.g., granulomatosis with polyangiitis [Wegener's]), medium vessel vasculitis (e.g., polyarteritis nodosa), and large vessel vasculitis (e.g., giant cell arteritis). Although patients with each type of vasculitis manage disease-specific aspects of illness, there are substantial disease and treatment burdens common to patients with vasculitis. For each type of idiopathic vasculitis, the disease course is often chronic, relapse is common and unpredictable, organ and tissue damage can accrue over time, new symptoms can occur late in the disease course, and treatment is often associated with toxicity and serious side effects.

Illness perceptions, also called illness representations, are the organized beliefs that patients have about their illness. On the basis of Leventhal's Self-Regulatory Model ([1]), it is theorized that patients consistently structure illness perceptions into common and measurable dimensions. These dimensions include the beliefs about the symptoms that belong to the disease, the perception of the course the disease has and will have, the beliefs about the impact of the illness on functioning and emotional well-being, the beliefs about the degree of personal and treatment-based controllability of the disease, and the overall understanding of the condition.

Illness perceptions have been found to be important determinants of health-related behavior and have been associated with important outcomes such as treatment adherence and functional recovery across a spectrum of diseases ([2]). Despite the substantial burdens of disease and treatment that patients with vasculitis incur, illness perceptions have not been comprehensively assessed in systemic vasculitis. Understanding the illness perceptions of patients with diverse forms of vasculitis may provide insight into the similarities and differences in the cognitive representations of disease and, as seen in other chronic diseases, may help explain important health-related behaviors and outcomes.

Fatigue is a highly prevalent and increasingly recognized important manifestation of vasculitis, yet the mechanisms underlying fatigue are poorly understood. One study found that fatigue had a significant effect on the lives of approximately one-half of patients with vasculitis ([3]). In another study, fatigue was the most commonly reported symptom in a cohort of patients with vasculitis and was a principal contributor to impaired quality of life ([4]). As observed in other chronic diseases ([5]), fatigue is not well correlated with disease duration or with physician-derived measures of disease activity in vasculitis ([3]), suggesting that fatigue may be driven by psychosocial measures in response to the disease. Further, illness perceptions may be associated with fatigue in vasculitis, as has been demonstrated in other diseases ([6]). Interestingly, fatigue is considered one of the most important disease burdens by patients with vasculitis and is often rated as more burdensome than most manifestations of vasculitis associated with organ damage ([7]). This finding highlights potential different models of illness between patients with vasculitis and health professionals and suggests the importance of understanding illness from a patient's perspective. The objectives of this study were to 1) compare illness perceptions among patients with a variety of forms of vasculitis; 2) determine clinical and demographic variables associated with negative illness perceptions for vasculitis, defined as less-desirable perceptions about illness; and 3) assess the association between illness perceptions and fatigue in vasculitis.

Box 1. Significance & Innovations

  • Illness perceptions, or patient-held beliefs about illness, are strikingly similar across different types of vasculitis.
  • Younger age, history of depression, active disease status, and poor overall health are risk factors for negative illness perceptions in vasculitis.
  • Fatigue is highly prevalent in vasculitis, strongly associated with illness perceptions, and only partially related to disease activity as currently assessed.

PATIENTS AND METHODS

Study sample

Study participants were recruited online within the Vasculitis Clinical Research Consortium (VCRC) Contact Registry to complete an online questionnaire ([8]). The VCRC Contact Registry is an international resource that can be used to conduct online clinical research in vasculitis. To join the registry, participants self-identify as having a particular form of vasculitis. For this study, participants were excluded if they were ages <18 years.

Data elements

Demographic information on age, sex, race, ethnicity, highest education level, and annual income was recorded. The participants were asked if they ever experienced depressive symptoms or sleep disturbance since the onset of vasculitis and were categorized based on affirmative or negative responses to each of these questions. The disease characteristics were assessed per self-report and included disease status (active versus remission), disease duration (continuous variable), disease severity (severe disease defined categorically as history of at least 1 predefined organ or life-threatening symptom), and remission duration (<1 year, ≥1 year).

Assessment of illness perceptions

All participants completed the revised Illness Perception Questionnaire (IPQ-R) ([9]). The IPQ-R was devised to assess cognitive representations or beliefs about any illness and has been determined to have good internal reliability, discrimination, and predictive validity. The IPQ-R has been used to study numerous patient populations, including those with systemic sclerosis ([10]), renal disease ([11]), type 2 diabetes mellitus ([12]), cystic fibrosis ([13]), and multiple sclerosis ([14]), and can be modified for use in a particular disease of interest. The following 8 dimensions of illness perception are assessed within the IPQ-R: identity (22 items), timeline (acute/chronic; 6 items), timeline (cyclical; 4 items), consequences (6 items), personal control (6 items), treatment control (5 items), emotional representations (6 items), and illness coherence (5 items). High scores on the identity, timeline (acute/chronic), consequences, and timeline (cyclical) dimensions of the IPQ-R represent strongly held beliefs about the number of symptoms attributed to the illness, the chronicity of the condition, the negative consequences of the illness, and the cyclical nature of the condition. High scores on the personal control, treatment control, and illness coherence dimensions represent positive beliefs about the controllability of the illness and increased understanding of the condition. The identity dimension was assessed on a 0–22-point scale; all other dimensions were assessed on a 5-point Likert scale, where 1 = strongly disagree, 2 = disagree, 3 = neither agree nor disagree, 4 = agree, and 5 = strongly agree.

Assessment of fatigue

Fatigue was assessed using the Multidimensional Fatigue Inventory (MFI-20). The MFI-20 has been widely used for measuring fatigue in a variety of diseases ([15]). Five domains of fatigue are assessed in the MFI-20; however, the general fatigue domain has been shown to best discriminate between patients with vasculitis and healthy controls ([16]) and was therefore used in this study. The general fatigue domain of the MFI-20 consists of 4 questions scored on a 5-point Likert scale. An MFI-20 score ≥13 on the general fatigue domain has been used as a threshold to define significant fatigue, based on the 95th percentile of the score in a healthy group ([17]).

Analytical methods

Internal reliability for each dimension of the IPQ-R except the identity subscale was calculated using Cronbach's alpha, per the IPQ-R guidelines ([9]). The correlation between mean scores for each IPQ-R dimension and for fatigue scores on the MFI-20 general subscale was calculated using Pearson's correlation coefficient. The mean scores for each IPQ-R dimension were compared across the types of vasculitis for those types of vasculitis in which at least 30 participants were recruited using a one-way analysis of variance with post hoc Scheffe's tests. The mean IPQ-R dimension scores in vasculitis were also compared to IPQ-R scores reported in studies from the literature for a variety of chronic diseases, including diabetes mellitus, hypertension, osteoarthritis, and systemic sclerosis.

A K-means cluster analysis of IPQ-R scores across each dimension was used to define a group of participants with negative illness perceptions. The distance between clusters was calculated using Euclidean distances with the cluster centers based on least-squares estimation (PROC FASTCLUS in SAS statistical software, version 9.3). Stepwise logistic regression (P < 0.1 selection criteria) was used to identify clinical and demographic variables associated with negative illness perceptions. Unless otherwise noted, P values less than 0.05 were used to define statistical significance. All statistical analyses were done using SAS, version 9.3.

RESULTS

Participant characteristics

A summary of participant characteristics is shown in Table 1. In total, 692 participants with 9 types of vasculitis completed the study. The majority of participants were women (70%), white (89%), born in the US (68%), college graduates (66%), and ages >50 years. The distribution of demographic characteristics by type of vasculitis was consistent with known distributions for these diseases. For example, all participants reporting a diagnosis of giant cell arteritis were ages >50 years, the highest proportion of women was seen in Takayasu arteritis, and participants with Behçet's disease and polyarteritis nodosa were more likely to be born outside of the US.

Table 1. Participant characteristics
Type of vasculitisClinical and demographic characteristicsDisease characteristics
No.Age, median (range) yearsSex, % womenRace, % whiteEducation, % college graduateAnnual income, % >$75KPlace of birth, % USDepression, % yesSleep disturbance, % yesDisease duration, median (range) yearsDisease activity, % remissionRemission duration, % >1 yearDisease extent, % severe
Behçet's disease4841.6 (22–71)8890574259596611.6 (2.5–26)152648
Central nervous system vasculitis1248.6 (23–65)558367367345524.4 (0–16)325260
Eosinophilic granulomatosis with polyangiitis (Churg-Strauss)12155.0 (25–80)698474436482647.2 (0–37)176367
Giant cell arteritis3267.4 (51–78)857565446255623.9 (0–15)535019
Granulomatosis with polyangiitis33254.8 (26–86)639263507243548.2 (0–41)536263
IgA vasculitis (Henoch-Schönlein)1243.8 (24–69)679258457525302.7 (0–12)334117
Microscopic polyangiitis4257.7 (25–76)829079506950596.7 (0.5–27)596660
Polyarteritis nodosa3649.7 (18–76)668144284152775.1 (0–18)315944
Takayasu arteritis5750.2 (26–72)948980427343538.6 (0–20)516025
Total69253.8 (18–86)708966386847567.4 (0–41)455646

A substantial number of participants reported a history of depression (47%) and sleep disturbance (56%) since the onset of vasculitis. Most participants reported several years of disease duration at the time of study participation (median 7.4 years [range 0–41 years]). Disease remission was reported in 45% of participants, and among those in remission, 56% reported ≥1 years of sustained remission. A history of severe disease was reported in 46% of participants.

IPQ-R dimension scores

Correlations among the illness perception dimensions and the internal reliability scores for each dimension are shown in Table 2. A significant correlation was observed between most of the illness perception dimensions. Cronbach's alpha scores indicated good to excellent internal reliability for each IPQ-R dimension (range 0.83–0.90) except for treatment control, which had acceptable reliability (0.67). The mean IPQ-R dimension scores are shown by type of vasculitis in Table 3. For the identity dimension, participants experienced a mean ± SD of 10.7 ± 5.1 symptoms (of a possible 22 symptoms), which were directly attributed to vasculitis. High scores in the timeline (acute/chronic) and consequences dimensions across all types of vasculitis indicated beliefs that the disease will last a long time and has negatively impacted the ability to function. There were no statistically significant differences in mean scores by type of vasculitis for the timeline (acute/chronic), consequences, personal and treatment control, emotional representations, and illness coherence dimensions. The mean dimension scores in the identity and timeline (cyclical) dimensions were significantly higher in Behçet's disease compared with some of the other types of vasculitis, indicating that participants with Behçet's disease attributed more symptoms to their illness and more strongly endorsed a cyclical component to illness. Although there were too few participants with central nervous system vasculitis for comparative analyses, these participants reported the highest scores in the consequences and emotional representations dimensions and the lowest scores in the personal control, treatment control, and illness coherence dimensions, reflecting a high degree of perceived psychological burden of illness in this type of vasculitis.

Table 2. Correlation between revised Illness Perception Questionnaire dimensions and fatigue scores*
 123456789
  1. α = Cronbach's alpha
  2. aP < 0.001.
  3. bP < 0.01.
1. Identity1.00        
2. Timeline (acute/chronic), α = 0.830.24a1.00       
3. Timeline (cyclical), α = 0.840.48a0.22a1.00      
4. Consequences, α = 0.830.52a0.44a0.48a1.00     
5. Personal control, α = 0.83−0.10b−0.17a−0.14a−0.14a1.00    
6. Treatment control, α = 0.68−0.17a−0.38a−0.26a−0.28a0.46a1.00   
7. Emotional representations, α = 0.890.40a0.21a0.41b0.51a−0.26a−0.32a1.00  
8. Illness coherence, α = 0.90−0.16a0.03−0.24a−0.19a0.20a0.22a−0.35a1.00 
9. Fatigue, α = 0.870.43a0.32a0.42a0.54a−0.24a−0.29a0.43a−0.11b1.00
Table 3. Mean ± SD IPQ-R dimension scores in different forms of vasculitis*
 TotalBDCNS vasculitisEGPAGCAGPAIgAVMPAPANTAK
  1. Analysis of variance was performed to compare mean scores across different types of vasculitis where ≥30 participants were recruited into the study (i.e., central nervous system [CNS] vasculitis and IgA vasculitis [Henoch-Schönlein; IgAV] were excluded from comparative analyses due to small sample sizes). IPQ-R = revised Illness Perception Questionnaire; BD = Behçet's disease; EGPA = eosinophilic granulomatosis with polyangiitis (Churg-Strauss); GCA = giant cell arteritis; GPA = granulomatosis with polyangiitis (Wegener's); MPA = microscopic polyangiitis; PAN = polyarteritis nodosa; TAK = Takayasu arteritis.
  2. aSignificantly different from GCA and TAK at the P < 0.05 level by Scheffe's test in the identity dimension.
  3. bSignificantly different from EGPA, GCA, GPA, MPA, and TAK at the P < 0.05 level by Scheffe's test in the timeline (cyclical) dimension.
Identity10.7 ± 5.113.5 ± 5.3a11.7 ± 4.510.5 ± 4.98.7 ± 4.710.9 ± 5.25.7 ± 3.010.9 ± 5.310.9 ± 4.09.2 ± 5.0
Timeline (acute/chronic)4.1 ± 0.84.2 ± 0.64.3 ± 0.94.3 ± 0.83.7 ± 0.74.1 ± 0.83.3 ± 1.04.0 ± 0.94.2 ± 0.73.9 ± 0.9
Timeline (cyclical)3.2 ± 1.04.0 ± 0.6b3.4 ± 1.13.1 ± 0.93.0 ± 0.93.1 ± 1.03.0 ± 1.03.1 ± 1.13.6 ± 0.93.2 ± 0.8
Consequences3.8 ± 0.84.0 ± 0.74.6 ± 0.43.9 ± 0.83.7 ± 0.83.8 ± 0.83.1 ± 0.83.7 ± 0.94.0 ± 0.83.9 ± 0.8
Personal control3.3 ± 0.83.1 ± 0.72.8 ± 0.83.3 ± 0.93.0 ± 0.83.3 ± 0.83.2 ± 0.83.4 ± 0.73.2 ± 0.83.4 ± 0.7
Treatment control3.3 ± 0.73.0 ± 0.73.0 ± 1.03.2 ± 0.73.5 ± 0.73.4 ± 0.73.3 ± 0.53.4 ± 0.73.2 ± 0.83.4 ± 0.7
Emotional representations3.1 ± 1.03.4 ± 0.83.7 ± 0.93.2 ± 1.03.2 ± 1.03.0 ± 0.92.8 ± 0.93.2 ± 1.03.5 ± 0.93.0 ± 0.9
Illness coherence3.4 ± 1.03.2 ± 1.03.0 ± 1.13.4 ± 1.03.3 ± 1.23.5 ± 0.93.1 ± 1.13.3 ± 1.03.2 ± 1.03.4 ± 0.9

Compared with IPQ-R dimension scores for other chronic diseases reported in studies from the literature (Table 4), higher scores in the consequences and emotional representations dimensions were observed in people with vasculitis than the scores observed in people with diabetes mellitus ([18]), hypertension ([19]), and osteoarthritis ([20]), indicating stronger beliefs that vasculitis negatively impacts the ability to function and emotional well-being. Scores in the consequences and emotional representations dimensions were similar between vasculitis and systemic sclerosis ([10]).

Table 4. Mean ± SD IPQ-R dimension scores in vasculitis and other chronic diseases*
 Vasculitis (n = 692)aDiabetes mellitus (n = 164) (18)Hypertension (n = 514) ([19])Osteoarthritis (n = 241) (20)Systemic sclerosis (n = 49) (10)
  1. IPQ-R = revised Illness Perception Questionnaire; NC = not comparable across diseases because of disease-specific modifications to the IPQ-R; NR = not reported.
  2. aThe vasculitis types included Behçet's disease (n = 48), central nervous system vasculitis (n = 12), eosinophilic granulomatosis with polyangiitis (Churg-Strauss; n = 121), giant cell arteritis (n = 32), granulomatosis with polyangiitis (Wegener's; n = 332), IgA vasculitis (Henoch-Schönlein; n = 12), microscopic polyangiitis (n = 42), polyarteritis nodosa (n = 36), and Takayasu arteritis (n = 57).
  3. bIdentity was scored on a scale from 0–22. All other dimensions were scored on a scale from 1–5.
Identityb10.7 ± 5.1NCNCNCNC
Timeline4.1 ± 0.83.9 ± 0.83.6 ± 0.44.2 ± 0.84.2 ± 0.6
Cyclical3.2 ± 1.02.8 ± 0.73.2 ± 0.83.6 ± 0.83.5 ± 1.0
Consequences3.8 ± 0.82.9 ± 0.62.6 ± 0.62.8 ± 0.83.8 ± 0.8
Emotional representations3.1 ± 1.02.0 ± 0.62.6 ± 0.72.4 ± 0.83.1 ± 1.0
Personal control3.3 ± 0.83.9 ± 0.63.5 ± 0.73.1 ± 0.62.9 ± 0.7
Treatment control3.3 ± 0.74.6 ± 0.93.5 ± 0.62.8 ± 0.63.0 ± 0.6
Illness coherence3.4 ± 1.03.5 ± 0.9NR3.6 ± 0.83.1 ± 1.0

Negative illness perceptions

The K-means cluster analysis separated participants into 3 groups (169 participants [25%] in group 1, 192 participants [29%] in group 2, and 309 participants [46%] in group 3). An analysis of the mean IPQ-R dimension scores by clustered group demonstrated differences between the groups. Mean IPQ-R scores >3 in the consequences and emotional representations dimensions represent less desirable illness perceptions because these scores indicate agreement that the illness has had a negative impact on function and emotional well-being. Mean IPQ-R scores <3 in the personal control, treatment control, and illness coherence dimensions also represent less desirable illness perceptions because these scores indicate beliefs that the illness is not controllable or understandable. In group 1, the mean scores for the consequences and emotional representations dimensions were >3 and the mean scores for the personal control, treatment control, and illness coherence dimensions were <3, indicating less desirable illness perceptions across all of these dimensions. Consequently, group 1 was defined as having negative illness perceptions. For group 2, the mean scores for the consequences and emotional representations dimensions were <3 and the mean scores for the personal control, treatment control, and illness coherence dimensions were >3, indicating that group 2 demonstrated positive illness perceptions across all of these dimensions. In group 3, which contained the majority of participants, the mean scores for the consequences and emotional representations dimensions were <3, indicating less desirable illness perceptions, but the mean scores for the personal control, treatment control, and illness coherence dimensions were >3, indicating desirable illness perceptions for these dimensions. Group 3 was defined as having mixed illness perceptions.

Stepwise logistic regression was performed to determine the risk factors associated with inclusion in the negative illness perceptions group. For every year of younger age, the odds of inclusion into the negative illness perceptions group increased by 4% (odds ratio [OR] 1.04, 95% confidence interval [95% CI] 1.02–1.06). A history of depression (OR 4.94, 95% CI 2.90–8.41), active disease status (OR 2.05, 95% CI 1.27–3.29), and poor overall health (OR 3.92, 95% CI 0.88–17.56) were also associated with negative illness perceptions. Sex, race, education, annual income, place of birth, disease duration, disease severity, and remission duration were not significantly associated with negative illness perceptions.

Fatigue and illness perceptions

The median MFI-20 general subscale scores are shown by type of vasculitis in Table 5. The majority of participants across all types of vasculitis scored ≥13 on the MFI-20, which is the threshold used to define significant fatigue. There was variability in fatigue scores (interquartile range 8–18) and the range of results encompassed the entire scale of possible scores. Cronbach's alpha for the MFI-20 general subscale was 0.87, and fatigue scores significantly correlated with each IPQ-R dimension (Table 2).

Table 5. Multidimensional Fatigue Inventory (MFI-20) general subscale scores in vasculitis
Type of vasculitisMedian MFI-20 score (range)% MFI-20 scores ≥13a
  1. aThreshold to define significant fatigue ([17]).
Behçet's disease17.0 (7–20)86
Central nervous system vasculitis16.0 (4–20)83
Eosinophilic granulomatosis with polyangiitis (Churg-Strauss)16.1 (4–20)76
Giant cell arteritis16.0 (4–20)82
Granulomatosis with polyangiitis (Wegener's)15.0 (4–20)72
IgA vasculitis (Henoch-Schönlein)14.5 (4–20)58
Microscopic polyangiitis15.5 (4–20)74
Polyarteritis nodosa16.0 (8–20)84
Takayasu arteritis16.0 (4–20)85
Total16.0 (4–20)76

The results from 3 sequential linear regression models using MFI-20 scores as the outcome measure are shown in Table 6. In the first step, disease activity and remission duration were significantly associated with the outcome and the overall model explained 18% of the variability in MFI-20 scores. In step 2, with the addition of clinical and demographic variables, age, race, a history of depression, and a history of sleep disturbance were significantly associated with the outcome and this model explained an additional 8% of the variability in MFI-20 scores. The IPQ-R dimension scores were added in step 3 and all of the IPQ-R dimensions except for the timeline (acute/chronic) dimension were significantly associated with the outcome. The full model, which included the IPQ-R dimensions, explained an additional 18% of variability in the MFI-20 scores, with an overall R2 value of 0.44. These results indicate that IPQ-R dimensions are significantly associated with fatigue in vasculitis and explain the variability of fatigue scores in vasculitis beyond measures of disease activity and other clinical and demographic characteristics.

Table 6. Sequential multiple linear regression models to assess variability in Multidimensional Fatigue Inventory scores
StepValue
  1. aP < 0.001.
  2. bP < 0.01.
Step 1: disease characteristics 
Disease activity (active vs. remission)β = 1.01a
Disease extent (severe vs. not)β = 0.33
Disease duration (continuous)β = −0.03
Shorter remission duration (<1 year vs. ≥1 years)β = 1.03a
F24.16a
Adjusted R20.18
Step 2: add clinical/demographic variables 
Age (per year)β = 0.01a
Sex (men vs. women)β = 0.43
Race (other vs. white)β = 2.28b
Depression (yes vs. no)β = 1.63a
Sleep disturbed (yes vs. no)β = 0.81b
F15.03a
Adjusted R20.26
R2 change0.08
Step 3: add revised Illness Perception Questionnaire dimensions 
Identity (more symptoms)β = 0.07b
Timeline, acute/chronic (more chronic)β = 0.23
Timeline, cyclical (more cyclical)β = 0.43b
Consequences (negative impact)β = 1.23a
Personal control (more control)β = −0.42b
Treatment control (more control)β = −0.73b
Emotional representations (negative impact)β = 0.38b
Coherence (more understanding)β = 0.40b
F16.03a
Adjusted R20.44
R2 change0.18

DISCUSSION

This study examined beliefs about illness in 692 participants with 9 forms of systemic vasculitis. Participant-held beliefs about vasculitis were strikingly similar across the different forms of vasculitis, except stronger beliefs about the cyclic nature of illness and a greater number of symptoms attributed to illness were reported among patients with Behçet's disease. Currently, efforts are underway to develop patient-reported outcomes in vasculitis ([21]). Given the similarity of illness perceptions and the high prevalence of fatigue across many different forms of vasculitis, these data suggest that different types of vasculitis can reasonably be combined when studying illness perceptions and fatigue as a patient-reported outcome. Compared with other chronic illnesses like diabetes mellitus and osteoarthritis, participants with vasculitis reported stronger beliefs about the negative impact of the disease on their ability to function and emotional well-being. Despite substantial progress that has transformed vasculitis from a frequently fatal disease into a chronic illness, the illness perceptions of participants with vasculitis highlight the need for further improvements in treatment paradigms, including potential interventions that target patient beliefs.

Negative illness perceptions were observed in 25% of the study participants. Younger age, a history of depression, active disease status, and poor overall health were risk factors for negative illness perceptions. Of these, younger age is perhaps the most intriguing finding. Vasculitis can affect people of all ages, and clinicians caring for patients with vasculitis should be mindful of the heightened psychological burden of the disease perceived by younger patients. The association between negative illness perceptions and younger age may reflect differences in the meaning of health problems at different phases of life ([22]). There are few published studies examining the relationship between illness perceptions and age in other rheumatic diseases. In systemic sclerosis, illness perceptions were not significantly associated with age, with the exception that younger patients reported more personal control over illness ([10]). In gout, younger age was associated with a belief in stronger negative emotional impact of disease, but age was not associated with other dimensions of illness perception ([23]).

Fatigue is a major symptom reported by patients with all forms of systemic vasculitis, yet the mechanisms underlying fatigue are poorly understood. Physicians and patients with vasculitis may differ in their perspectives about the relationship between fatigue and active disease in vasculitis ([7]). Physician-derived measures of disease activity do not correlate with fatigue ([3]). In this study, patient-reported measures of disease activity and remission duration were significantly associated with fatigue scores, suggesting that patients considered fatigue a manifestation of active vasculitis. Illness perceptions significantly explained variability in fatigue scores beyond what could be explained by measures of disease characteristics and depression. These data suggest that 1) fatigue is a major domain of illness that is only partially related to disease activity as currently assessed, 2) illness perceptions may have a causal and modifiable role in fatigue, and 3) the mechanisms underlying fatigue in vasculitis are complex and multifactorial. Illness perceptions have been associated with fatigue in other chronic illnesses. In chronic fatigue syndrome, illness perceptions were stronger predictors of a fatigue score than mood scores ([24]). In patients who had an acute myocardial infarction, declining beliefs in the controllability of illness predicted the development of chronic fatigue and a lowered quality of life ([25]). In multiple sclerosis, illness perceptions were strongly predictive of fatigue ([9]), and a randomized controlled trial of cognitive–behavioral therapy designed to change cognitive factors contributing to fatigue, such as illness perceptions about symptom identity and the controllability of illness, demonstrated clinically significant and sustained reductions in fatigue ([26]).

There are some important strengths of this study. This study is the first of its kind to systematically assess illness perceptions in idiopathic vasculitis, a group of illnesses characterized by a high degree of disease and treatment burden. The results provide unique insight into vasculitis from a patient's perspective. Further, a substantial number of participants representing 9 different types of vasculitis completed the study, and rarer subtypes of vasculitis, such as polyarteritis nodosa, Behçet's disease, and central nervous system vasculitis, were represented. Additionally, the complete version of the IPQ-R rather than a shortened version known as the Brief Illness Perception Questionnaire ([27]) was used to comprehensively assess illness perceptions, and excellent internal reliability scores for the various illness perception dimensions supported the validity of the IPQ-R for use in vasculitis.

Some potential limitations to this study should be noted. This was an online study in which participants self-reported their diagnosis; however, the distribution of demographic characteristics across each type of vasculitis was consistent with known distributions for these diseases, lending face validity to the accuracy of self-reported diagnosis. Selection bias is an issue inherent to most online studies because the study population may not be representative of the general population. In this study, patients were more likely to be women, well educated, and affluent, which limited the generalizability of the findings. Because of the cross-sectional study design, changes in illness perceptions over time in relation to fatigue and other outcomes such as depression could not be assessed. Variables such as depression and sleep disturbance were assessed per self-report without the use of validated measures and analyzed as categorical variables, leading to the possibility of residual confounding. Understanding the differences in perspective between patients and health care providers could highlight aspects of vasculitis that require further understanding and could help prioritize future research efforts in accordance with patient-specified needs. Given the online nature of the study, it was not possible to directly compare patient-reported measures of disease activity with similar physician-derived measures in vasculitis. Such a comparison might highlight the importance of illness perceptions as a means to further delineate ideological differences in disease understanding between patients and physicians.

For a variety of chronic diseases, illness perceptions are known to be important determinants of health-related behavior ([28]) and are associated with important health-related outcomes, including treatment adherence and functional recovery ([29]). In this study, illness perceptions highlighted a group of participants with vasculitis who had negative illness perceptions and provided novel insight into the complex and highly prevalent symptom of fatigue in systemic vasculitis. Given the potential for a myriad of poor health outcomes and function impairments in vasculitis, illness perceptions may provide an understanding of the patient-driven factors that contribute to these outcomes. Future studies could focus on the illness perceptions of patients with vasculitis in relation to medication adherence, the doctor–patient relationship, relapse rates and survival, utilization of health care, and quality of life. Efforts to modify maladaptive perceptions of illness in relation to these outcomes should also be explored.

AUTHOR CONTRIBUTIONS

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Grayson had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Grayson, Leduc, Fraenkel, Merkel.

Acquisition of data. Grayson, McAlear, Leduc, Shereff, Richesson, Merkel.

Analysis and interpretation of data. Grayson, Amudala, Fraenkel, Merkel.

Ancillary