Factors associated with quality of life in cutaneous lupus erythematosus using the Revised Wilson and Cleary Model

Abstract

Objectives

The purpose of this study was to characterize the impact of cutaneous lupus erythematosus (CLE) in adults and identify the clinical and non-clinical factors associated with quality of life (QoL), using the Revised Wilson and Cleary Model.

Methods

101 patients diagnosed with CLE were included in this cross-sectional study. QoL was measured with the Cutaneous Lupus Erythematosus Quality of Life (CLEQoL) scale and disease activity and damage with the Cutaneous Lupus Activity and Severity Index (CLASI). Patient demographics, clinical, and disease characteristics were also collected. Descriptive statistics were calculated, and multiple regression was employed to determine significant (p<0.05) predictors of overall QoL. Data were analyzed using SPSS v24.

Results

The overall regression QoL model was significantly different from zero, (F=24.96; df=14, 76; p=<0.001). Disease activity (β=0.13), pain (β=0.13), fatigue (β=0.24), body image (β=0.62), and side effects (β=−0.13) were significant predictors of overall QoL while controlling for other predictor variables. Patients who experienced higher levels of disease activity, fatigue severity, pain levels, and greater degree of body dissatisfaction had significantly poorer QoL. Fewer side effects experienced from CLE medications were significantly associated with higher QoL.

Conclusions

Study findings support the considerable burden associated with CLE. Several modifiable variables such as pain, fatigue, body image, disease activity were associated with QoL. Therefore, interventions that incorporate these variables may reduce negative impacts on QoL life and improve health outcomes in CLE patients. Furthermore, given the chronic and recurring nature of the condition, strategies focused on improving QoL are needed for this vulnerable population.

INTRODUCTION

Cutaneous lupus erythematosus (CLE) is a rare, chronic dermatologic autoimmune disease marked by photosensitive lesions that can vary in appearance depending on the subtype.^{1, 2} Causes of CLE have been postulated to be either genetic, hormonal, immunological abnormalities (e.g., cytokine, B-cells, and T-cells dysfunction) or environmental factors (especially ultraviolet irradiation).^{3, 4} CLE can be categorized into three subtypes: acute cutaneous lupus (ACLE), subacute cutaneous lupus (SCLE), and chronic cutaneous lupus (CCLE), with CCLE being the most common subtype. Signs and symptoms associated with CLE include skin pain, itching, rash, and photosensitivity.^{5, 6} The rashes seen with CLE vary based on the CLE subtype, but many patients suffer from hair loss (alopecia) as well.^{5, 6} There is little mortality impact for CLE patients; however, potentially severe co-morbidities such as cancer or systemic lupus erythematosus may develop⁷ Compared to other dermatologic conditions such as eczema, non-melanoma skin cancer, and psoriasis, CLE negatively impacts QoL in patients causing industrial disability and vocational handicap.² In addition to coping with signs and symptoms associated with CLE, patients with CLE have been reported to be additionally distressed with how others perceived their physical appearance.⁸ Fatigue and decreased vitality are additional symptoms associated with CLE that limit the patients’ ability to engage in recreational or desired household activities.⁹

Several factors impact quality of life (QoL) in CLE patients, especially on physical and mental health domains.^{2, 5–8, 10–17} Increasing pain, fatigue, and skin disease activity, female gender, younger age, current smoking status, low income, and low education level, presence of systemic lupus erythematosus (SLE), and photosensitivity are factors associated with poor QoL.^{2, 6, 17, 18} The literature reports conflicting results on the relationship between ethnicity and QoL. While being African American has been reported to have lower QoL,¹⁰ other studies have found no such association.^{2, 7}

Furthermore, the studies examining QoL in CLE patients did not make use of CLE-specific validated instruments in assessing QoL.^{2, 5, 6, 8, 10–16} As a result, the cutaneous lupus erythematosus quality of life (CLEQoL) scale was developed to include additional, salient domains relevant to the experiences of CLE patients, such as body image and photosensitivity, to the SKINDEX-29 questionnaire, which is a generic skin disease QoL questionnaire.^{13, 19} The CLEQoL is a 36-item QoL measure for CLE, consisting of five domains: symptoms, emotions, and functioning, body image/cosmetic issues, and photosensitivity. The CLEQoL assesses how often (Never, Rarely, Sometimes, Often, All the time) during the previous four weeks, the patient experienced the effect described in each item. Scores of 0 (never), 25 (rarely), 50 (sometimes), 75 (often), and 100 (all the time) are assigned to each question and averaged to determine each domain score from 0–100, with higher scores indicating poorer QoL. A recent cross-sectional study that examined the psychometric properties of the CLEQoL found it to be a valid and reliable measure for assessing QoL in patients with CLE.¹³ Summary of key literature findings are presented in Table 1.

Table 1:

Summary of key articles using patient-reported outcome (PRO) measures in CLE patients

References	Main Objective	Country	Sample size	Type of PROs used	Key findings
Vasquez et al.1	To compare quality of life (QoL) indicators between patients with CLE at the University of Texas Southwestern (UTSW) Medical Center at Dallas and those at the University of Pennsylvania (UPenn)	USA	248	Modified Skindex-29+3, SF-36	Female gender, low annual income levels (<$10,000), low levels of education (high school education), presence of SLE, and increased disease activity were all associated with poor QoL.
Ishiguro et al.2	To assess whether the skin symptoms in CLE are associated with the QoL using the Japanese version of the Skindex-29.	Japan	54	Skindex	Female gender, older age, and longer duration of systemic lupus erythematosus (SLE) were correlated with poor QoL.
Klein et al.3	To determine how CLE affects QoL and which independent variables are associated with poor QoL.	USA	157	Skindex-29, SF-36	Female gender, generalized disease, severe disease, distribution of lesions, and younger age were factors realted to poor QoL.
Teske et al.4	To identify demographic and clinical variables associated with decreased QoL in patients with discoid lupus erythematosus (DLE).	USA	117	Modified Skindex-29+3	Female gender and smoking were correlated with impairment of multiple domains of quality of life in DLE.
Mendez-Flores5	To evaluate the association of pain and pruritus with dermatologic QoL and cutaneous disease activity in patients with 1) specific cutaneous lupus erythematosus (CLE) lesions, 2) nonspecific CLE lesions and 3) both types of CLE lesions.	Mexico	42	Dermatology Life Quality Index (DLQI), pain and pruritus, visual analog scales (VAS)	Pain, not pruritus, correlated with dermatologic QoL and cutaneous activity.

QoL has become an increasingly important endpoint for clinically assessing patient outcomes, especially in those with chronic diseases such as CLE. Of equal importance is assessing theoretical frameworks in which QoL is conceptualized in CLE. Inclusion of a framework will allow for identifying complex relationships between variables associated with QoL and for tailoring interventions specifically to CLE patients to improve their overall QoL. In the current study, the theoretical framework used was the Revised Wilson and Cleary Model (WCM).^{20, 21} The Revised Wilson and Cleary Model, modified by Ferrans et al., is a framework that emphasizes the impact of disease and health on QoL, and it was developed to help explain the relationships among patient outcomes on a continuum from those that are proximal (e.g., clinical variables) to those that are distal to the disease (e.g., QoL), within the context of individual and environmental characteristics [See Figure 1 for study model]. The model links biological and physiological factors, symptom status, functioning, general health perceptions, and overall QoL. Because the ultimate outcome is overall QoL, the model also links non-medical factors such as individual and environmental characteristics. Individual characteristics represent those variables that are intrinsic to the patient, such as a patient’s motivation or value preferences. In contrast, environmental characteristics, such as the amount of social, psychological, or economic support, are extrinsic to the patient. The WCM has been used in other disease conditions to understand the several factors that contribute to QoL^{22, 23} but not in CLE.

Figure 1:

Relationships among Measures of Patient Outcomes Based on Ferrans’ Revised Wilson & Cleary Model

The impact of different clinical and non-clinical factors on QoL in patients with skin diseases has mostly involved univariate and multivariate approaches without a theoretical framework.^{2, 5–8, 10–17} The limitation of this approach is the inability to order the complex phenomenon of the multidimensional impact of health on QoL as well as show how several health measures can be combined to allow for a broad assessment of QoL.²⁴ In addition, the use of a theoretical framework provides a basis to test and develop future interventions.

Therefore, the purpose of this study is to fill the gap in current knowledge of the factors affecting QoL in CLE patients by examining the clinical and personal factors associated with QoL using the Revised Wilson and Cleary’s Model. For the first time, given its impact on the QoL of patients, CLE will be used as the prototypical disease to test the clinical applicability of WCM. The proposed relationships between factors are defined in Figure 1. Understanding these factors may be useful in tailoring care to this patient population.

METHODS

Sample and Study Design

This study was a cross-sectional correlational study design. Participants were recruited from two outpatient university-based dermatology clinics in Dallas, Texas - University of Texas (UT) Southwestern Medical Center and Parkland Health and Hospital System. Patients who met the inclusion criteria were recruited during their clinical visits to the hospital as well as from the CLE patient registry maintained by the clinics. The study included patients who met the following inclusion criteria: clinical diagnosis of CLE, aged 18 years and above, and able to give informed consent. This study was conducted within accordance with the guidelines set forth by the institutional review board (IRB) at the University of Texas Southwestern (STU 082010–241).

Sample size estimations were performed a-priori to ensure adequate power to decrease the probability of accepting an incorrect null hypothesis.²⁵ In estimating appropriate sample sizes, estimates of effect size from past studies using the Wilson and Cleary Model (multiple R ranging from 0.32 to 0.53; i.e., R² of 32.1%– 53%),^{26, 27} a significance level of 0.05, 80% power, and 21 predictor variables were used. The power analysis was conducted using GPower software, and the estimated sample was calculated as a range from 59– 85 subjects.

Study Measures

Below is a description of the measures in the Revised Wilson and Cleary Model:

Biological and Physiological Factors

Items in this domain included CLE subtype, SLE diagnosis, disease activity, damage, and duration. The measures were either assessed by the physician and collected at the time of each patient visit to the clinic or in retrospective chart reviews with the clinical staff. CLE subtype was assessed by whether patients fit the criteria to be categorized as either acute (ACLE), subacute (SCLE), or chronic (CCLE). The clinician-reported outcome measure, the Cutaneous Lupus Erythematosus Disease Area and Severity Index (CLASI)²⁸ assessed both skin disease activity and damage. A clinically validated tool, the CLASI yields two separate scores for total CLASI activity and total CLASI damage, with higher scores indicating more severe disease.²⁸ To assess disease duration, patients’ biopsy date was subtracted from their first clinical visit.

Symptom Status

Corresponding unidimensional scales were used to measure symptom-related issues of pain,²⁹ pruritus,³⁰ and fatigue.³¹ Specifically, the numerical rating scale for pain,²⁹ pruritus,³⁰ and fatigue³¹ measured the intensity of pain, pruritus, and fatigue, respectively, on a single 11-point numeric scale with 0 representing the absence of the symptom and 10 representing the highest intensity.

Functioning

One measure of functional status was body image using the 10-item Body Image Scale (BIS).³² Responses on the BIS range from 0 (not at all) to 3 (very much), yielding a range of possible scores from 0–30 with higher scores indicating a greater degree of body image dissatisfaction. Additionally, depression was measured by a single-item: “Have you been diagnosed (by a professional) with depression?” (“0” for no, and “1” for yes).” Finally, comorbidity was assessed by using the Self-Administered Comorbidity Questionnaire (SCQ).³³ The SCQ addresses 13 major body systems with three binary questions regarding the occurrence of health-related problems (problem score), treatments received or medication (treatment score), and limitations experienced in daily life (limitation score). The sum of all the affirmed items from the three sub-scores makes up the total score of the SCQ, ranging from 0 to 39, with higher scores indicating higher comorbidity scores and limitations.³⁴ Further, two CLE-relevant additional conditions were added to the SCQ,³³ which were hepatitis and HIV. However, depression was taken out of the list of comorbidities, as this was already being measured as a separate variable. As a result, the scores for the final SCQ measure possibly ranged from 0 to 42 (14 items), with higher scores representing higher comorbidity.

General Health Perceptions

Three items represented general health perceptions. Side effects and medication burden were measured with two items from the LupusPRO,³⁵ respectively: “In the past 4 weeks, how often did you experience the following due to your CLE: serious side-effects associated with CLE medications, and concerns about the number of medications you currently take for your CLE?” Items were rated on a 5-point Likert scale from 0 (none of the time) to 4 (all of the time) and “not applicable (recoded as ‘0’ for scoring);” each item was then reverse-coded, with higher scores reflecting lower treatment impact – side effects and medication burden. Finally, skin health perception was measured on an 11-point numeric rating scale with anchors at 0 (worst skin imaginable, i.e., total body burn) and 10 (perfect health), with higher scores indicating better perceptions of skin health.

Characteristics of the Individual

The measures used to characterize participants were age (in years), gender (male vs. female), education (lower than high school, high school, college, or graduate degree), race (African-American/Black, Caucasian, Hispanic, Asian, or other), and smoking status (currently smoke, past smoker, never smoked).

Characteristics of the Environment

The characteristics of the environment selected for inclusion in the study were: marital status (single, married/domestic partner, divorced/widowed/separated), income level, and social support. Income level was calculated from publicly-available data on annual median household income,³⁶ which served as a proxy for individual-level annual income. Finally, social support was measured by a single item from the LupusPRO:³⁵ “Generally, I receive support from friends and/or family?” Response scales ranged from 0 (none of the time) to 4 (all of the time); and “not applicable (recoded as ‘0’ for scoring),” with higher scores reflecting higher social support.

Overall Quality of Life (QoL)

The dependent variable, QoL, was operationalized using the cutaneous lupus erythematosus quality of life (CLEQoL) scale. The CLEQoL is a 36-item QoL scale with five domains (symptoms, functioning, emotions, cosmetic effects, and photosensitivity).¹³ Each item is measured on a 0–100 scale with higher scores indicating poorer QoL.

Data Analysis

All data were entered into SPSS version 24.0 for data preparation, screening, and analysis. Descriptive statistics (frequencies, means, and standard deviations) were calculated for each variable. Inferential statistical tests, including t-test, ANOVA, and correlation analysis, were used to analyze variable relationships. For all continuous interval variables, their z-scores were inspected to identify potential outliers. Because the rate of missing data did not exceed the threshold of more than 5%,³⁷ data were analyzed without imputation of missing values. Reliability statistics via Cronbach’s alpha α were calculated for the multi-item scales – CLEQoL (QoL), BIS (body image), and SCQ (comorbidity), where an acceptable value of internal consistency was an α≥ 0.70.³⁸ To simplify the final model and build a more parsimonious one, independent variables that were not significantly related to the dependent variable were excluded in the final regression analysis. Finally, multiple linear regression analysis was performed to assess the predictive ability of the biological and physiological factors, symptoms status, functioning, general health perceptions on overall QoL. The significance levels for this study were based on an alpha of 0.05.

RESULTS

Patient Characteristics

The total sample included 101 patients with CLE. The average age of participants was 48 years (SD=13). Most participants (n=88, 87.1%) were female, African-American (n=59, 58.4%), single (n=42, 41.6%), had chronic CLE (n=72, 71.3%), had a concomitant diagnosis of SLE (n=56, 55.4%), and never smoked (n=58, 57.4%). Their average CLASI disease activity and damage were 5.16 and 7.95, respectively, and they had an average age of disease duration of 10 years (SD=11). Average pain, itch, and fatigue scores were 3.23 (SD=3.51), 3.61 (SD=3.47), and 5.11 (SD=3.45), respectively. Descriptive statistics of all variables as well as t-tests and ANOVAs depicting the relationships between the categorical variables and overall QoL are reported in Table 2. Depression, gender, education, and smoking status were significantly associated with overall QoL.

Table 2:

Patient Characteristics and Relationship to QoL (N=101)

Characteristics	Frequency (%)	Mean±SD	t or Fa	P
Biological and Physiological Factors
Predominant CLE Subtype
Chronic CLE	84 (83.2)		0.62	0.647
Subacute Lupus Erythematosus (SCLE)	11 (10.9)
Acute Lupus Erythematosus (ACLE)	6 (5.9)
Concomitant diagnosis of Systemic Lupus Erythematosus (SLE)			1.47	0.229
Yes	56 (55.4)
No	45 (44.6)
Disease Activity		5.16±5.72
Disease Damage		7.95±6.72
Disease Duration (yrs.)		10±11
Symptom Status
Pain		3.23±3.51
Itch		3.61±3.47
Fatigue		5.11±3.45
Functioning
Body Image		10.67±8.50
Depression			26.55	<0.000**
Yes	28 (27.7)
No	68 (67.3)
Comorbidity		3.97±4.08
General Health Perceptions
Side effects from CLE Medications		3.35±1.13
Skin Health Perception		5.24±2.99
Pill Burden		3.14±1.28
Characteristics of the Individual
Age		48±13
Gender			2.53	0.016
Female	88 (87.10)
Male	13 (12.9)
Education			3.65	0.039
Lower than High School	15 (14.9)
High School	32 (31.7)
College	39 (38.6)
Graduate Degree	11 (10.9)
Race
African-American/Black	59 (58.4)		1.50	0.211
Caucasian	26 (25.7)
Hispanic	12 (11.9)
Asian	2 (2.0)
Other	2 (2.0)
Smoking Status			5.04	0.009*
Currently Smoke	31 (30.7)
Past Smoker	11 (10.9)
Never Smoked	58 (57.4)
Characteristics of the Environment
Marital Status			2.11	0.087
Single	42 (41.6)
Married/Domestic Partner	39 (38.6)
Divorced/Widowed/ Separated	15 (14.9)
Annual Incomeb	$52,056.75±$22,621.00
Social Support	3.08±1.30

^aFor categorical variables only

^bIncome level was calculated from publicly-available data on annual median household income, which served as a proxy for individual-level annual income

^*p<.05 (two-tailed)

^**p<.01 level (two-tailed).

Bivariate Relationships

Correlational relationships between predictor variables and overall QoL are shown in Table 3. A moderate correlation was evident between overall QoL and disease activity, comorbidity, side effects, and pill burden, while other predictors had strong correlations with overall QoL. Disease damage, disease duration, age, and income were not correlated with overall QoL.

Table 3:

Correlations of Predictor Variables with CLEQoL (N=101)

	QoL	Disease Activity	Disease Damage	Disease Duration	Pain	Itch	Fatigue	Body Image	Comorbidity	Side effects	Pill Burden	Skin Health Perception	Age	Income	Social Support
QoL	1.00
Disease Activity	0.30**	1.00
Disease Damage	0.17	0.50**	1.00
Disease Duration	-0.08	-0.06	0.17	1.00
Pain	0.52**	0.15	0.31*	0.02	1.00
Itch	0.57**	0.24*	0.28**	−0.09	0.75**	1.00
Fatigue	0.57**	0.16	0.13	−0.11	0.67**	0.54**	1.00
Body Image	0.85**	0.18	0.17	−0.18	0.49**	0.53**	0.49**	1.00
Comorbidity	0.36**	−0.09	0.08	−0.26**	0.47**	0.37**	0.45**	0.21*	1.00
Side effects	−0.34**	−0.17	0.09	0.08	−0.23*	−0.21*	−0.28**	−0.29**	−0.22*	1.00
Pill Burden	−0.30**	0.22*	−0.19	−0.00	−0.23*	−0.24*	−0.22*	−0.36**	−0.29**	−0.53**	1.00
Skin Health Perception	−0.41**	−0.34*	−0.8	0.06	−0.13	−0.21	−0.20	−0.35**	0.03	0.11	0.29**	1.00
Age	−0.16	−0.12	−0.19	−0.35**	−0.03	−0.07	−0.03	−0.13	0.15	−0.01	−0.02	−0.07	1.00
Income	−0.19	−0.00	−0.23*	−0.01	−0.17	−0.15	−0.10	−0.09	−0.12	−0.05	−0.03	−0.02	−0.20*	1.00
Social Support	−0.38**	−0.05	−0.33**	0.01	−0.28**	−0.24**	−0.21**	−0.35**	−0.20	0.07	0.19	0.17	0.07	0.06	1.00

Note: Pearson’s correlations are significant at

^*p<.005

^**p<.001

Overall Quality of Life

The total overall QoL score, as measured with the CLEQoL, was 39.06±25.11 out of a possible score range of 0–100 (higher scores indicating poorer QoL). This indicates that respondents had relatively good QoL. The individual average scores for the CLEQoL questions (a-kk) are presented in Table 4. The highest means±SD per individual items were noted for items referring to the effect CLE has on sun protection efforts (69.70±34.68), worrying about sun flares (61.62±36.30), and skin sensitivity (59.60±35.66) (Table 4).

Table 4:

Means, Standard Deviations, and Frequency Distributions of Individual Questions from the CLEQoL (n=101)

	N	Mean	SD	Never(0)	Rarely(25)	Sometimes (50)	Often (75)	All of the time (100)
a.  Skin Hurts (Sym)	101	37.87	32.52	30 (29.7)	21 (20.9)	28 (27.7)	12 (11.9)	10 (9.9)
b.  Disrupted sleep (Em)	101	33.42	35.76	45 (44.6)	12 (11.9)	20 (19.8)	13 (12.9)	11 (10.9)
c.  Worry about progression (Fxn)	99	52.78	34.42	17 (16.8)	15 (14.9)	29 (28.7)	16 (15.8)	22 (21.8)
d.  Affects work and hobbies (Em)	100	35.00	33.71	34 (33.7)	24 (23.8)	22 (21.8)	8 (7.9)	12 (11.9)
e.  Affects social life (Em)	100	39.75	39.26	40 (39.6)	11 (10.9)	19 (18.8)	10 (9.9)	20 (19.8)
f.  Depressed (Fxn)	99	38.38	36.82	37 (36.6)	14 (13.9)	21 (20.8)	12 (11.9)	15 (14.9)
g.  Skin burns or stings (Sym)	100	42.25	35.12	30 (29.7)	14 (13.9)	27 (26.7)	15 (14.9)	14 (13.9)
h.  Stay at home (Em)	100	36.50	38.01	43 (42.6)	12 (11.9)	16 (15.8)	14 (13.9)	15 (14.9)
i.  Worry about scars (Fxn)	100	58.00	37.75	19 (18.8)	14 (13.9)	14 (13.9)	22 (21.8)	31 (30.7)
j.  Skin itches (Sym)	100	54.50	33.41	16 (15.8)	13 (12.9)	28 (27.7)	23 (22.8)	20 (19.8)
k.  Affects closeness (Em)	99	26.26	34.14	55 (54.5)	10 (9.9)	16 (15.8)	10 (9.9)	8 (7.9)
l.  Ashamed (Fxn)	100	43.00	38.61	34 (33.7)	14 (13.9)	18 (17.8)	14 (13.9)	20 (19.8)
m.  Worry skin worse (Fxn)	100	57.50	33.62	12 (11.9)	17 (16.8)	26 (25.7)	19 (18.8)	26 (25.70
n.  Do things alone (Em)	100	32.50	36.15	46 (45.5)	13 (12.9)	18 (17.8)	11 (10.9)	12 (11.9)
o.  Angry (Fxn)	100	30.00	34.63	46 (45.5)	18 (17.8)	17 (16.8)	8 (7.9)	11 (10.9)
p.  Water bothers (Sym)	100	18.00	29.10	66 (65.3)	11 (10.9)	12 (11.9)	7 (6.9)	4 (4.0)
q.  Difficult showing affection (Em)	99	22.73	31.15	57 (56.4)	11 (10.9)	21 (20.8)	3 (3.0)	7 (6.9)
r.  Side effects from medicine*	100	53.25	35.11	19 (18.8)	12 (11.9)	29 (28.7)	17 (16.8)	23 (22.8)
s.  Skin irritated (Sym)	99	51.26	34.51	19 (18.8)	14 (13.9)	29 (28.7)	17 (16.8)	20 (19.8)
t.  Interaction with others (Em)	100	29.50	34.88	50 (49.5)	11 (10.9)	20 (19.8)	9 (8.9)	10 (9.9)
u.  Embarrassed (Fxn)	100	43.50	39.67	36 (35.6)	10 (9.9)	21 (20.8)	10 (9.9)	23 (22.8)
v.  Problem for loved ones (Em)	100	13.25	23.96	70 (69.3)	14 (13.9)	12 (11.9)	1 (1.0)	3 (3.0)
w.  Frustration (Fxn)	100	46.25	38.66	31 (30.7)	10 (9.9)	26 (25.7)	9 (8.9)	24 (23.8)
x.  Skin sensitive (Sym)	99	59.60	35.66	18 (17.8)	5 (5.0)	26 (25.7)	21 (20.8)	29 (28.7)
y.  Affects desires for others (Em)	100	27.75	35.16	55 (54.5)	7 (6.9)	20 (19.8)	8 (7.9)	10 (9.9)
z.  Humiliated by skin (Fxn)	99	26.52	35.86	57 (56.4)	9 (8.9)	14 (13.9)	8 (7.9)	11 (10.9)
aa.  Skin bleeds (Sym)	99	19.44	27.33	56 (55.4)	22 (21.8)	11 (10.9)	7 (6.9)	3 (3.0)
bb.  Annoyed (Fxn)	99	46.21	36.83	29 (28.7)	8 (7.9)	31 (30.7)	11 (10.9)	20 (19.8)
cc.  Sex life interference (Em)	100	22.75	35.73	64 (63.4)	10 (9.9)	10 (9.9)	3 (3.00)	13 (12.9)
dd.  Tired (Em)	100	29.00	35.13	50 (49.5)	14 (13.9)	17 (16.8)	8 (7.9)	11 (10.9)
ee.  Worry about flares (Ph)	99	61.62	36.30	17 (16.8)	8 (7.9)	18 (17.8)	24 (23.8)	32 (31.7)
ff.  Worry about hair loss (Bod)	100	58.25	40.21	23 (22.8)	12 (11.9)	10 (9.9)	19 (18.8)	36 (35.6)
gg.  Prevents outdoor activities (Ph)	100	54.50	39.15	25 (24.8)	8 (7.9)	22 (21.8)	14 (13.9)	31 (30.7)
hh.  Worry about what others are thinking (Cos)	100	34.00	37.19	44 (43.6)	15 (14.9)	17 (16.8)	9 (8.9)	15 (14.9)
ii.  Influence clothes (Cos)	100	48.25	40.40	32 (31.7)	9 (8.9)	21 (20.8)	10 (9.9)	28 (27.7)
jj.  Affects grooming practices (Cos)	99	52.27	41.20	28 (27.7)	13 (12.9)	12 (11.9)	14 (13.9)	32 (31.7)
kk.  Affects sun protection efforts (Ph)	99	69.70	34.68	10 (9.9)	12 (11.9)	11 (10.9)	22 (21.8)	44 (43.6)
Score Total	98a	39.06b	25.11
Cronbach’s Alphac	0.97

^*Item 18 on side effects of treatment is not scored nor included in the final CLEQoL scale.

^aTotals do not equal 101 due to missing responses.

^bThe composite score for the overall scale calculation based on 101 responses, possible scale range 0 to 100.

^cCronbach’s alpha based on 36 items.

Abbreviations: Sym = Symptoms; Em = Emotions; Fxn = Functioning; Cos = Cosmetic effects; Ph = Photosensitivity

Mean scores on the photosensitivity and cosmetic effects domains were the highest, 62.33 and 48.30, respectively. The internal consistency of the CLEQoL scale was acceptable (Cronbach’s α = 0.97). Cronbach’s alpha of the five domains ranged from 0.67 – 0.95, with the domain of the cosmetic effects being the lowest (Table 5).

Table 5.

Mean CLEQoL Scores, Standard Deviations, and Cronbach’s Alpha

Subscale/Domains	N	Mean	Median	SD	Min	Max	Cronbach’s α
Emotions	95	43.00	37.50	29.78	0	100.00	0.93
Symptoms	97	39.69	35.71	24.45	0	100.00	0.87
Functioning	97	29.32	22.92	27.37	0	100.00	0.95
Cosmetic effects	99	48.30	50.00	28.29	0	100.00	0.67
Photosensitivity	98	62.33	66.67	31.32	0	100.00	0.81
Total Score	98a	39.06b	32.29	25.11	0	95.83	0.97c

^aTotals do not equal 101 due to missing responses

^bThe composite score for the overall scale calculation based on 101 responses, possible scale range 0 to 100

^cCronbach’s alpha based on 36 items

Model Predictors

Disease activity (β=0.13), pain (β=0.23), fatigue (β=0.24), body image (β=0.62), and side effects (β=−0.13) accounted for 85% of the variance explained in overall QoL after adjusting for several variables (F(14,76) = 24.957, p < 0.001). Patients who experienced higher levels of disease activity, higher fatigue and pain severity, and greater degree of body dissatisfaction had significantly poorer QoL. Fewer side effects experienced from CLE medications were significantly associated with better QoL. The result of the multiple regression analysis is displayed in Table 6.

Table 6.

Multiple Regression Analysis of Overall QoL – CLEQOL Scores (N=101)

Variables	Unstandardized Coefficients		Standardized Coefficients	95.0% Confidence Interval		P-Value
	B	Std. Error	Beta	Lower Bound	Upper Bound
Intercept	14.93	9.00		−3.06	32.92	0.102
Biological & Physiological Factors
Disease Activity	0.79	0.35	0.13	0.10	1.49	0.027*
Symptom Status
Pain	1.75	0.72	0.23	−3.19	−0.30	0.019*
Itch	0.80	0.59	0.11	−0.38	1.98	0.179
Fatigue	1.78	0.55	0.24	0.69	2.88	0.002**
Functioning
Body Image	1.69	0.19	0.62	1.32	2.06	<0.001**
Depressiona	5.70	3.56	0.10	−1.42	12.81	0.115
Comorbidity	0.50	0.43	0.08	−0.36	1.36	0.246
General Health Perceptions
Side Effects	−3.25	1.72	−0.13	−6.70	0.19	0.045*
Pill Burden	1.29	1.39	0.06	−1.50	4.08	0.359
Skin Health	−0.43	0.52	−0.05	−1.47	0.62	0.420
Characteristics of the Individual
Genderb	6.06	4.27	0.08	−2.48	14.59	0.161
Educationc	−0.40	2.89	−0.01	−6.19	5.38	0.890
Smoking statusd	5.42	3.28	0.10	−1.14	11.98	0.104
Characteristics of the Environment
Social Support	−1.18	1.11	−0.06	−3.39	1.03	0.289

F statistic = 24.957; df = 14,76; Model p-value= <0.001; R² = 0.849; Adjusted R² = 0.815

^aNo = reference category

^bMale = reference category

^cEducation was collapsed into two categories: “Lower than high school/high school and college/graduate degree” with lower than high school/high school being the reference category

^dSmoking status was collapsed into two categories: “Currently smoke/past smoker vs. never smoked with never smoked as the reference category

^*Indicates significance at p<0.05

^**Indicates significance at p<0.01.

DISCUSSION

The purpose of this study was to explore QoL of patients with CLE and its correlates using the Revised Wilson and Cleary Model. Overall, CLE patients were most affected by the lupus-specific domains – photosensitivity and cosmetic effects while the functioning domain had the least impact. While the cosmetic effects domain included in the CLEQoL measure is a unique contribution to the literature, similar findings have been reported in other studies regarding photosensitivity.^{3, 7} The effects of sun protection efforts and flares due to sun exposure were the greatest concerns patients had within the photosensitivity domain. These concerns can be disruptive and problematic, especially with engagement in daily activities and social interactions; findings which are consistent with other studies.^{1, 2, 12} Health care providers can help patients improve on this domain with appropriate medications and adequate counseling on effective photo-protective methods.

The Revised Wilson and Cleary model was useful for exploring the relative importance of clinical and non-clinical factors in accounting for QoL in patients with CLE. The combination of variables in the model explained more than 80% of the variance in overall QoL, with body image having the largest effect. Similar to chronic diseases with cosmetic effects, changes in physical appearance as experienced in CLE may cause significant body image issues in patients.³⁹ Specifically in CLE patients, body image issues experienced may be due to permanent and irreversible skin damage due to scarring, skin pigmentation changes (or dyspigmentation), and scarring alopecia, which are not alleviated by current therapies.¹⁸ Given that current QoL measures used in CLE patients do not include body image measures, our study findings highlight the importance of the inclusion of this assessment. Signs of chronic skin damage can lead to self-esteem and body image issues. Patients are also distressed about their looks mostly because of how they are perceived by others.⁸ This could explain why compared to the general population, dermatologic patients, including those with CLE, often have higher rates of mental health issues like anxiety and depression.⁴⁰ Therefore, studies assessing QoL in CLE patients should incorporate other attributes such as body image to capture the disease effect and treatment impact on patients. The goal of including body image measures should be to promote overall body acceptance, which may in turn positively improve QoL.

Increasing disease activity, as measured by the CLASI, was significantly associated with poorer QoL. The CLASI activity score, which is based on erythema, scaling, mucous membrane involvement and non-scarring alopecia, has been reported to affect QoL in most CLE patients compared to the CLASI damage score which is based on dyspigmentation and scarring.^{2, 10, 17} Therefore, our study findings emphasize the association between higher CLE disease activity and poor QoL.

CLE patients who reported high levels of pain had poor overall QoL. This inverse relationship has been well-documented in other studies.^{6, 41} However, studies conducted in other disease conditions such as fibromyalgia, arthritis, and chronic back pain have reported conflicting results with pain and QoL.⁴² An explanation for these contrary findings could be because of how the measure for ‘pain’ was operationalized. Rather than utilizing rating scales to measure the intensity of pain on a spectrum, pain was mostly measured as a dichotomy – the absence or presence of pain. The use of rating scales can help gain a better understanding of the impact of pain. Furthermore, because pain is a subjective outcome associated with poor QoL that can impair functionality,^{6, 41} interventions aimed at improving QoL in CLE patients can incorporate adequate pain management techniques. This is important, as the presence of pain or inadequate pain management has been reported to be associated with QoL.⁴³

Side effects from CLE medications are another significant predictor of QoL that has been identified in other studies.^{4, 7} Some medications used in CLE patients can have dose-limiting toxicities, including low blood counts, GI upset, and vision loss, and patients have reported some concerns regarding these side effects.^{4, 7} Therefore, evaluation of the side effects of CLE medications can be regarded as an important aspect of QoL assessment in clinical trials and outcome studies in this patient population.

Previous studies have shown the relationship between fatigue and QoL.^{44, 45}A recent study by Tarazi et al. reported that patients with autoimmune conditions experience higher fatigue levels than controls.⁴⁶ Furthermore, the study showed that patients with SLE had worse fatigue scores than CLE patients. Given that more than half of the respondents in our current study had a concomitant diagnosis of SLE, it could explain why fatigue was a significant predictor of QoL. Beyond just affecting mental domains, fatigue can play a distinct role in causing a decline in physical activity. For example, fatigue could start with a reduction in physical activity but can escalate to a wider range of negative effects that can leave patients feeling out of control and isolated, thereby impacting mental health.⁴⁷ Determining patients’ experience with fatigue could lend an understanding of the relevant domains (both physical and mental) impacting QoL. As such, interventions that reverse the effect or onset of fatigue could have a positive effect on QoL. Therefore, clinical trials can consider incorporating fatigue measures as an assessment of QoL.

Itching, depression, comorbidity, pill burden, skin health, gender, education level, smoking status, and social support were not significant predictors of QoL in this study. Itching, for example, has been reported to affect QoL in other skin diseases, such as psoriasis and dermatomyositis, but not in CLE.⁴¹ While depression has not been extensively studied in cross-sectional studies conducted in CLE patients, a qualitative study reported that self-reported depression was a concern for patients.¹⁹ A plausible reason for the insignificance in the current study, despite the prevalence of depression in CLE^{40, 48}, is that depression is commonly underdiagnosed in other chronic diseases,⁴⁹ and is likely applicable to CLE. Current and past smoking histories have been linked to an increase in systemic vascular inflammatory processes such as IL-6,⁵⁰ which can worsen some symptoms and, subsequently, QoL. The relationship between smoking status and QoL has produced mixed findings.^{2, 7}

Although important conclusions were drawn from this study, several limitations should be noted. One limitation is the generalizability of the findings because this study took place in a tertiary-care referral center (UT Southwestern Medical Center) and a safety-net hospital (Parkland), where patient recruitment took place. Another is that while our study was adequately powered, a more robust sample size could have led to more statistical power to detect additional relationships among variables. Also, due to the study design being cross-sectional, changes, and trends in QoL over time could not be made. Therefore, future longitudinal studies are needed to capture changes in QoL, beginning from the stage of diagnosis through the CLE continuum of care. Also, longitudinal analyses can be combined with causal modeling and mediation analyses via structural equation modeling (SEM) to support the causal relationships implied in the Revised Wilson and Cleary Model.

Despite these limitations, this project has some strengths. One strength of the current study included the addition of some variables (such as body image and cosmetic effects) that have not been formally investigated before in CLE patients. Also, by utilizing the Revised Wilson and Cleary Model, this study advances what was previously known about the determinants of QoL in patients with CLE as well as the importance of utilizing a holistic approach in modeling QoL. Finally, the use of a CLE-specific QoL measure, the CLEQoL, included CLE-specific measures, which increased the variability and sensitivity of scores. Results obtained from this study could be beneficial in designing future studies in this area.

CONCLUSION

Our study findings supports the importance and utility of theoretical frameworks like the Revised Wilson and Cleary Model in better understanding the QoL in CLE patients. Our ability to demonstrate the model’s utility in assessing QoL in prototypical skin diseases such as CLE could lay a foundation for the use of the model in other skin and rare diseases in future explorations of QoL. Several modifiable factors such as disease activity, pain, fatigue, body image, and side effects from CLE medications were predictive of overall QoL in this patient population and could aid health care providers and researchers in interpreting and assessing QoL outcomes in CLE patients. In addition, these factors could be considered when designing specific interventions to improve QoL in CLE patients. Finally, incorporating both patient-centered and clinical measures via the use of the Revised Wilson and Cleary model allowed for a fuller theoretically-based understanding of QoL issues in CLE patients. Because this is the first study utilizing the Revised Wilson and Cleary model in CLE patients, this study provides the foundation for testing the utility and validity of this model for future QoL studies. Given the chronic, recurring, and heterogeneous nature of CLE, continued research into factors affecting QoL, and the development of interventions focused on symptoms management and disease-modifying parameters are needed in this special patient population.

Abbreviations

ACLE

Acute Cutaneous Lupus Erythematosus

BIS

Body Image Scale

CCLE

Chronic Cutaneous Lupus Erythematosus

CLASI

Cutaneous Lupus Erythematosus Disease Area and Severity Index

CLE

Cutaneous Lupus Erythematosus

IRB

Institutional Review Board

QoL

Quality of Life

CLEQoL

Cutaneous Lupus Erythematosus Quality of Life

SCLE

Subacute Cutaneous Lupus Erythematosus

SCQ

Self-Administered Comorbidity Questionnaire

SEM

Structural Equation Modeling

SLE

Systemic Lupus Erythematosus

UT

University of Texas

WCM

Revised Wilson and Cleary Model