Abstract
Objective We estimated the prevalence of potential functional disability among those with systemic lupus erythematosus (SLE), by receipt of disability benefits.
Methods Participants (N = 443, mean age 46.2 years, 91.7% women, 82.6% Black) were recruited from a population-based SLE cohort. Indicators of potential disability included functioning impairments (Short Physical Performance Battery score ≤ 10; age-corrected National Institutes of Health Toolbox Cognition Battery composite score for fluid cognition < 77.5 [1.5 SD below the mean]); activity limitations (physical functioning T-scores < 35 [1.5 SD below the mean]); at least some difficulty performing ≥ 1 of the instrumental activities of daily living (IADLs) or basic activities of daily living (BADLs); and participation restrictions (any vs no reported effect of health on ability to work; restricted community mobility). We performed multivariable logistic regression models predicting potential disability indicators by self-reported receipt of disability benefits and then obtained adjusted prevalence estimates using postestimation margins.
Results Those who reported receiving disability benefits (45.6%) vs not (54.4%) were more likely to have impairments in functioning (physical performance [71.3% vs 50%, P < 0.001]; fluid cognition [35.4% vs 19.2%, P = 0.01]), limitations in activities (self-reported physical limitations [26.7% vs 7.5%, P < 0.001]; IADLs [73.1% vs 42.9%, P < 0.001]; BADLs [60.6% vs 30.8%, P < 0.001]), and restrictions in participation (work [77.8% vs 60.6%, P = 0.09]; community mobility [43.1% vs 22%, P < 0.001]). These associations were not changed with adjustment for personal and SLE factors.
Conclusion Receipt of disability benefits may be an incomplete marker of functioning. A substantial proportion of those not receiving benefits have impairments, limitations, and restrictions that should be addressed.
- activities of daily living
- cognitive dysfunction
- disability evaluation
- disability studies
- physical functional performance
- systemic lupus erythematosus
Although disability1 and associated racial, ethnic, and gender disparities2-4 in systemic lupus erythematosus (SLE) have been well described, many studies have defined disability as the receipt of work disability benefits. In the United States, disability is defined by the Social Security Administration (SSA) as “inability to engage in any substantial gainful activity by reason of any medically determinable physical or mental impairment(s) which can be expected to result in death or which has lasted or can be expected to last for a continuous period of not less than 12 months.”5 This definition may fall short for many with SLE, as they may be less likely to have paid employment or to meet the 12-month expectation, given that SLE is characterized by a systemic, heterogeneous nature; intermittent, unpredictable disease course; often-delayed diagnosis; and unpredictable treatment effects.1 Additionally, for most, paid employment is necessary for financial support and private healthcare coverage (which is tied to employment in the US). Thus, those with SLE might substantially restrict other life activities to prioritize work but still not meet this rigid, work-based definition of disability.
Studies of disability in SLE also often conflate impairments in function, limitations in activity, and restrictions in participation, which are the core components of the World Health Organization’s International Classification of Functioning, Disability and Health (ICF).6 Using broader definitions of disability that include measures across these components could help identify individuals across the spectrum of ability and ensure that individuals receive needed support and access interventions to prevent further loss of, or even improve, ability. Such efforts could, in turn, maximize the independence of individuals with SLE as they age.
Given that the current definition of disability may be incomplete in the setting of SLE, we hypothesized that the presence of functioning impairments, activity limitations, and/or participation restrictions would be high regardless of the receipt of disability benefits in this population. We used data collected from a diverse, population-based cohort of individuals with validated SLE (Approaches to Positive, Patient-centered Experiences of Aging with Lupus [APPEAL] study) in order to (1) estimate the prevalence of multiple functioning impairments, activity limitations, and participation restrictions that may represent disability in this population, by whether individuals were receiving disability benefits; (2) assess whether participant factors modified the associations of the receipt of disability benefits with these indicators; and (3) explore individual patterns of these indicators of potential disability.
METHODS
Study population and data sources. Study participants (N = 451) were recruited from the population-based Georgians Organized Against Lupus (GOAL) adult SLE cohort.7 Study data were obtained from a series of physical and cognitive performance tests administered by a trained research coordinator and self-administered questionnaires during a single, ~1.5-hour study visit (October 8, 2019 to May 12, 2022), as well as from the closest GOAL survey. The APPEAL visit was a median of 2.7 months before the GOAL measurement used (IQR 24.0 months after to 10.0 months before). Those with potentially invalid survey response patterns (n = 4) or missing assessments of receipt of disability support (n = 4) were excluded (N = 443), with additional exclusions for missing information on individual indicators (Supplementary Figure S1, available with the online version of this article).
Receipt of disability benefits. Participants self-reported the receipt of SSA disability benefits (yes vs no) in GOAL.
Indicators of potential functional disability. Following the ICF framework,6 we examined multiple measures within the existing APPEAL data that could be indicators of potential disability, including measures of impairments in function, limitations in activity, or restrictions in participation.
• Impairments in function. Physical performance was assessed by a trained research coordinator using the reliable and valid Short Physical Performance Battery (SPPB).8,9 The SPPB includes assessments of balance, gait speed, and lower body strength (score range, 0-12; higher scores = better performance). Impairment in physical performance was defined as SPPB score < 10 vs ≥ 10.10,11
Cognitive functioning was assessed with the National Institutes of Health Toolbox Cognition Battery (NIHTB-CB),12-14 which was developed and validated in a healthy population aged 20-85 years and provides an overall composite age-corrected standard score (mean 100, SD 15) for fluid cognition (overall capacity to reason and solve novel problems). Impairment was defined as an overall age-corrected score > 1.5 SD below the population mean (< 77.5).15 Because the NIHTB-CB involves timed measurements, only participants who had in-person visits have composite scores for fluid cognition.16
• Limitations in activity. The Patient-Reported Outcomes Measurement Information System (PROMIS) Short Form Physical Functioning v1.0 12a17 (mean 50, SD 10; higher T-scores = better physical functioning) was used to assess self-reported health-related limitations in physical tasks. Limitation in physical functioning was defined as a T-score > 1.5 SD below the population mean (< 35).
Activities of daily living (ADLs) were consistent with Lawton and Brody (instrumental ADLs [IADLs], such as shopping and housekeeping)18 and Katz et al (basic ADLs [BADLs], such as bathing and dressing).19 Limitations were measured on a Likert scale using the question “how much difficulty do you have doing the following activities on your own?” and defined as any difficulty (“some difficulty,” “a lot of difficulty,” and “unable to do”) vs no difficulty (“no difficulty” or “could do it, but don’t”) performing any IADL or BADL independently.20
• Restrictions in participation. Overall work restriction was assessed with the Work Productivity and Activity Impairment–General Health questionnaire v. 2.0.21 Participants who reported being currently employed were asked about work hours missed due to health and the degree to which their health affected productivity while working over the last 7 days. The responses were converted to a percentage representing overall work impairment due to health. We then defined restriction in work as any (> 0%) vs no (0%) reported overall work impairment due to health, and analyses were limited to those who reported being currently employed.
Community mobility was captured by the University of Alabama Birmingham Life-Space Assessment (UAB-LSA),22 which measures how far respondents have gone (life-space levels: bedroom, other rooms in the home, outside the home, neighborhood, town, and out of town), how often, and with how much help (device or person) over the past 4 weeks. We defined restricted community mobility as the inability to reach the neighborhood level or beyond at least weekly and unassisted by either a device or person; this definition was based on our proposed effect of this level of restriction on access to necessary items and services.
Other variables. Other variables of interest, informed by the ICF model,6 were categorized into personal and SLE-related factors.
• Personal. Age, sex, race (Black, White, other), ethnicity (Hispanic, not Hispanic), and education (high school graduate or lower, some college/associate degree, and college graduate or higher) were self-reported. Physical activity was assessed with the International Physical Activity Questionnaire Short Form.23 Depression was assessed via the validated 8-item PROMIS Depression Short Form 8a24 and perceived stress was assessed using the 10-item Perceived Stress Scale (PSS-10).25
• SLE-related factors. Current SLE activity was assessed during the APPEAL visit through the Systemic Lupus Activity Questionnaire (SLAQ; range 0-44; higher scores = greater SLE-related disease activity).26 The Brief Index of Lupus Damage (BILD) score (range, 0-46; higher scores = greater cumulative SLE-related organ damage)27 was obtained from GOAL. Higher BILD scores (1, 2-3, and 4+ vs 0) are associated with worse self-rated health and functional status, higher unemployment and work disability, and higher healthcare utilization.27 Current steroid use was reported by the participant at the study visit.
Statistical analysis. Characteristics of participants were described, and summary scores (means and medians [for skewed distributions]) and percentages with impairments, limitations, or restrictions by our definitions were calculated, overall and by receipt of benefits. Multivariable logistic models were used to assess associations between receipt of disability benefits (independent variable) and indicators of potential disability (dependent variables). Adjusted percentages with the indicators among those with and without disability benefits were then obtained using postestimation marginal predictions. For modeling, potential confounders (grouped as personal and SLE-related factors and included separately and together) were identified a priori following the ICF model6 and included if they were not considered mediators. We used complete case analysis to handle missing data. Sensitivity analyses included the following: (1) making an additional adjustment in the multivariable models for depressive symptoms, to provide results under the assumption that depression acted as a confounder rather than mediator (as it was considered in primary models); (2) restricting models to those with assessments of all indicators (excluding work restrictions), to allow direct comparisons across the same populations; (3) using multiple imputations with chained equations to handle missing data, to explore whether complete case analysis magnified any bias due to nonrandom missing data; and (4) stratifying models on potential effect modifiers (age, race, education, and disease activity and burden) and assessing multiplicative interaction via interaction terms of receipt of disability benefits with each of these variables in our adjusted models. In secondary analyses, we assessed the total numbers of indicators with impairments, limitations, and restrictions (= sums of all indicators). We also estimated the percentages of individuals with each observed combination of potential disability indicators and displayed them using UpSet plots. The statistical significance threshold was set at 0.05. All analyses were conducted using Stata v.18.5 (StataCorp).
RESULTS
Characteristics of study participants. Participants had a mean age of 46.2 years; most were female (91.7%) and Black (82.6%; Table 1). Nearly one-quarter (23.3%) had a high school degree or less. The median SLE duration was 14.8 years, with median SLAQ and BILD scores of 11.0 and 2.0, respectively. Compared to those not receiving disability benefits, those receiving these benefits (45.6%) were older (48.3 vs 44.4 years), more likely to be Black (91.6% vs 75.1%), and less likely to have a college education or higher (22.8% vs 51.5%). Those receiving disability benefits also had higher median SLAQ (13.0 vs 9.0) and BILD (3.0 vs 2.0) scores (Table 1).
Selected characteristics of study participants with SLE.
Indicators of potential functional disability among participants with SLE by receipt of disability benefits.
• Impairments in function. Mean SPPB scores were lower in those receiving disability benefits vs not (8.4 vs 9.6, P < 0.001; Table 2), and 71.3% vs 50% had potential physical performance impairment (Figure 1). Adjustment for personal and SLE-related factors did not substantially change these percentages (Table 3). Results were similar with additional adjustment for depressive symptoms, with the population limited to those with complete data only, and with multiply imputed data, except that only the latter showed statistically significant differences (Supplementary Table S1, available with the online version of this article). These associations also persisted in analyses stratified by age, race, education, and SLAQ scores. However, in those with lower BILD scores, the fully adjusted percentage with physical performance impairment for those with vs without benefits was 72.1% vs 43%; whereas in those with higher BILD scores, the difference was in the opposite direction (40.7% vs 74.1%, Pinteraction = 0.002; Supplementary Table S2).
Scores for indicators of potential disability, overall and by receipt of disability benefits, among study participants with SLE.
Prevalence of indicators of potential disability among individuals with SLE, by whether they were receiving disability benefits. BADL: basic activities of daily living; IADL: instrumental activities of daily living.
Unadjusted and adjusted estimated percentages of participants with indicators of potential disability (impairments in functioning, limitations in activity, and restrictions in participation), by receipt of disability benefits.
Those receiving disability benefits also had lower mean composite scores for fluid cognition than those without benefits (83.5 vs 90.9, P < 0.001; Table 2). Potential fluid cognition impairment was nearly twice as high in those receiving vs not receiving benefits (35.4% vs 19.2%, P = 0.01; Figure 1). Adjusting for personal and SLE-related factors (Table 3) and additionally adjusting for depressive symptoms, limiting the population to those with complete data, and using multiply imputed data (Supplementary Table S1, available with the online version of this article) did not affect this difference. The difference in percentage with potential fluid cognition impairment by receipt of disability benefits was not different by age, race, SLAQ, or BILD, but there was a marginally significant (Pinteraction = 0.048) difference by education: those with the lowest educational attainment (high school or less) had a smaller difference by receipt of disability benefits (52.2% vs 43.8%) than those with higher attainment (eg, 30.8% vs 8.2% for college or higher; Supplementary Table S2).
• Limitations in activity. T-scores for self-reported physical functioning were lower in those with disability benefits vs those without (39.3 vs 46.8, P < 0.001; Table 2). The percentage with limitations in physical functioning was > 3-fold higher in those with vs without disability benefits (26.7% vs 7.5%, P < 0.001; Figure 1). This difference persisted with adjustment for personal and SLE-related factors (Table 3), and with additional adjustment for depressive symptoms, with the sample limited to those with complete data, and with multiply imputed data (Supplementary Table S1, available with the online version of this article). There was a larger but nonstatistically significant difference by disability benefits among those with the highest (28.3% vs 3.2%) vs lowest (26.2% vs 14.3%) educational attainment; there was no difference by age, race, SLAQ, or BILD (Supplementary Table S2).
Receipt of disability benefits was associated with a median of 3 IADLs with limitations, vs 0 (P < 0.001) in those without these benefits (Table 2), and the distribution of a number of IADL limitations was more right-skewed among those with disability benefits (Supplementary Figure S2A, available with the online version of this article). Those with vs without benefits were more likely to have any limitations in any of the IADLs (73.1% vs 42.9%, P < 0.001; Figure 1). This difference was not substantially changed by adjustment, limiting to those with complete data, or using multiply imputed data (Table 3; Supplementary Table S1). Those with the lowest educational attainment had no difference in percentage with potential IADL limitations by receipt of disability benefits (63.3% vs 66.7%), whereas those with the highest attainment showed about a 2-fold difference (75% vs 35.8%, P = 0.001; Supplementary Table S2).
BADL limitations were less frequent than IADL limitations, with receipt of disability benefits being associated with a median of 1 IADL with limitations, vs 0 (P < 0.001) in those without these benefits (Table 2). The distribution of a number of BADL limitations was more right-skewed among those with disability benefits (Supplementary Figure S2B, available with the online version of this article). Those with vs without benefits were about twice as likely to have any limitations in any of the BADLs (60.6% vs 30.8%, P < 0.001; Figure 1), and this association was not substantially changed by adjustment, limiting to those with complete data, or using multiply imputed data (Table 3; Supplementary Table S1). Those with the lowest educational attainment had a much smaller difference in percentage with potential BADL limitations by receipt of disability benefits (51.7% vs 45.2%), whereas those with the highest attainment showed a nearly 3-fold difference (60% vs 21.8%, P < 0.001; Supplementary Table S2).
• Restrictions in participation. The median reported overall work restriction due to health among employed recipients receiving vs not receiving disability benefits was 30% vs 10% (P = 0.08; Table 2). Work restriction was higher in those with vs without disability benefits (77.8% vs 60.6%, P = 0.09; Figure 1). Adjustment for personal and SLE-related factors did not substantially change these percentages (Table 3), nor did additional adjustment for depressive symptoms or analyses with the population limited to those with complete data only or with multiply imputed data (Supplementary Table S1, available with the online version of this article). These observations persisted in analyses stratified by age, race, education, and lower vs higher SLAQ and BILD scores (Supplementary Table S2).
Median UAB-LSA scores in those receiving vs not receiving disability benefits were 42.0 vs 62.0 (P < 0.001; Table 2). Community mobility restriction was about twice as high in those with vs without disability benefits (43.1% vs 22%, P < 0.001; Figure 1). Adjustment for personal and SLE-related factors did not substantially change these percentages (Table 3). Additional adjustment for depressive symptoms and analyses with the population limited to those with complete data only and with multiply imputed data (Supplementary Table S1, available with the online version of this article), as well as stratification by age, race, education, SLAQ, and BILD (Supplementary Table S2), did not change these estimates.
• Observed combinations of indicators. The total number of impairments, limitations, and restrictions (excluding work restrictions) among those with disability benefits was higher than that in those without disability benefits: 8.6% vs 23.3% of those with vs without disability benefits had zero total impairments, limitations, and restrictions, whereas 7.5% vs 4.9% had the maximum (Supplementary Figure S3, available with the online version of this article). For those with disability benefits, the most common combinations were (1) having zero impairments, limitations, or restrictions; (2) having impairments in physical performance, limitations in IADLs and BADLs, and restricted community mobility (8.6% each); and (3) having the maximum number of impairments, limitations, and restrictions (7.5%; Figure 2A). The most common combination among those without benefits (23.3%) was having zero impairments, limitations, or restrictions, leaving 76.7% with at least 1 impairment (16.5% with a single impairment [physical performance], and 60.2% having various combinations of at least 1 impairment, limitation, or restriction; Figure 2B).
UpSet plots showing overlap of impairments, limitations and restrictions, excluding work restriction, for those (A) receiving and (B) not receiving disability benefits. All observed patterns are shown. BADL: basic activities of daily living; IADL: instrumental activities of daily living.
DISCUSSION
Not unexpectedly, we found that individuals with SLE who were receiving work disability benefits were ~1.5-3 times more likely to have functioning impairments, activity limitations, and participation restrictions, independent of personal and SLE-related factors. However, these indicators were far from rare among those not receiving these benefits: 7.5% and 19.2% had limitations in physical and cognitive functioning, and 22% had restrictions in community mobility; limitations in BADLs (30.8%) and IADLs (42.9%) as well as impairment in physical performance impairment (50%) were even more common. Among those currently working, 60.6% reported some level of restriction in their work activity due to their health.
Overall, these results suggest that the receipt of work disability benefits is an incomplete measure of disability in SLE. More than three-quarters (76.7%) of those who were not receiving benefits had at least 1 impairment, limitation, or restriction, including, but not limited to, work restriction. This observation suggests that there are individuals who need disability benefits but are not receiving them. However, we cannot determine whether individuals applied for benefits; this remains an important avenue for future research.
We also found that 8.6% of individuals who were receiving disability benefits did not have any impairments, limitations, or restrictions by our definition. This unexpected observation may reflect unmeasured issues, such as impairment in emotional functioning,28 which could seriously affect their ability to work. Although the time gap between the measurement of receipt of disability benefits and the measurements of the indicators was generally short, it could overlap with the cessation or initiation of benefits for some; additionally, there are conditions under which someone may resume work while receiving benefits.29 Importantly, the fluctuating nature of SLE may play a role. Previous qualitative work among persons with SLE in the United Kingdom showed that many who needed work and welfare benefits were unable to receive them, suggesting that current benefits systems do not serve those with chronic intermittent illnesses like SLE.30,31
Our results also suggest that most persons with SLE, regardless of disability benefits, are likely to have nonwork-related issues that may require other types of tangible support to ensure maintenance or improvement of functioning and continued independence. For example, physical therapy can improve physical performance; physical and occupational therapy can improve the independent performance of day-to-day tasks; and improved transportation options can improve community mobility.
Interestingly, although the prevalence of these indicators of potential disability was higher among those who were older, of Black race, and had higher disease activity, we found no differences in prevalence by receipt of disability benefits across indicators by age, race, or disease activity. However, we did find differences in several of the indicators by educational attainment. Those with higher education may be better able to navigate the complicated process of applying for disability benefits or may be more likely to work at jobs that are less physically taxing32 or offer more flexibility and accommodations. Further, there may be discrimination by providers who identify and certify disability or administrators who approve applications.
Our study has several additional limitations. The cross-sectional design limits causal inference and does not account for the unpredictable course of SLE and changes in impairments, limitations, and restrictions over time. Our cutoffs for our indicators of potential disability may not align with individual needs or preferences for support. Although work impairment is an often-used example of restrictions in activity in SLE,33-36 there are several other domains of participation restrictions that may be just as, or more, important to individuals with SLE (eg, reducing time spent with family or friends,37 or reducing or stopping leisure activities); these domains are related, but not equivalent, to community mobility. The value that an individual places on a function, activity, or participation should also inform the plan to address it.38,39 Apart from objective physical and cognitive performance measures, our primary focus was on patient-perceived indicators of potential disability. Physician-reported indicators, which could have provided valuable comparative information, were not available. There may also be other important indicators of potential disability in SLE that were not measured in our study, notably, fatigue and pain. Variables captured from GOAL (such as BILD) may be misclassified, particularly for those with longer gaps between the GOAL survey and APPEAL assessment. Further, the BILD may incompletely capture disease damage, relative to physician assessments. Finally, residual confounding and selection bias due to unmeasured or unknown factors, such as social, environmental, and socioeconomic barriers, remain a possibility.
In conclusion, although more work is needed to assess the generalizability of our results across populations, our study does suggest that disability in SLE is incompletely captured by the receipt of disability work benefits. Future research aimed at preventing or supporting disability in this population should consider that many who struggle with one or more domains may not be receiving work benefits, and that work benefits are likely only a part of the needed support for this population. Further, disability benefits programs that consider a more holistic picture of disability (including impairments in functioning, limitations in activities, and restrictions in participation), as well as chronic conditions with intermittent presentations, may better serve individuals with SLE.
ACKNOWLEDGMENT
We thank the participants of the APPEAL study; Jessie Black, Aita Akharume, Meaza Girmay, and Sydnei Simpson for completing study visits; and Charmayne Dunlop-Thomas and S. Sam Lim for access to GOAL data and participants.
Footnotes
CONTRIBUTIONS
LCP: conceptualization, formal analysis, funding acquisition, visualization, writing – original draft; ETR: conceptualization, methodology, writing – reviewing and editing; CH: data curation, project administration, writing – reviewing and editing; BDP: funding acquisition, writing – reviewing and editing; CBB: conceptualization, funding acquisition, writing – reviewing and editing.
FUNDING
The research reported in this publication was supported by the National Institute on Aging of the National Institutes of Health (NIH) under Award Number R01AG061179 (LCP). The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.
COMPETING INTERESTS
The authors declare no conflicts of interest relevant to this article.
ETHICS AND PATIENT CONSENT
The Emory University Institutional Review Board approved the APPEAL (IRB00110977) and GOAL (IRB00003656) study protocols. Participants provided informed consent before completing study visits.
- Accepted for publication February 24, 2025.
- Copyright © 2025 by the Journal of Rheumatology
REFERENCES
SUPPLEMENTARY DATA
Supplementary material accompanies the online version of this article.
