Background/Purpose: Frailty, defined as a generalized vulnerability to stressors, has emerged as a relevant concept in SLE¹, although its origins are in geriatrics.  As defined by the Fried Frailty Phenotype (FFP)², frailty is an aggregate of 5 constructs: weight loss, exhaustion, low physical activity, slowness, and weakness.  Other means of identifying frailty have been suggested that may be easier to implement in clinical settings, including a shorter performance-based method, the Short Physical Performance Battery (SPPB)³.  We examined the prevalence of frailty using the SPPB in a well-characterized SLE cohort.

Methods: Data are from the California Lupus Epidemiology Study (CLUES), a longitudinal cohort of individuals with rheumatologist-confirmed SLE.  Data are collected annually through in-person research visits and structured interviews.  At one visit, the SPPB was administered to a subset of the cohort.  SPPB consists of balance tests, time to complete 5 chair stands, and gait speed over 4 meters.  Scores range from 0 – 12; a score of < 10 identifies frailty.  Differences in participant characteristics for frail/not frail were examined using t-tests or chi-square analyses.  Multivariate logistic regression analyses identified independent predictors of frailty, including age, sex, race/ethnicity, l, obesity, SLE duration, SLE disease activity, and steroid use.  Separate models were constructed to include either physician-assessed (SLE Disease Activity Index, SLEDAI) or patient-reported disease activity (Systemic Lupus Activity Questionnaire).  The impact of frailty on patient-reported outcomes (PROs) was examined with multiple regression analyses.

Results: Among the 150 individual who completed the SPPB, mean age was 45 ± 14 years, disease duration 15 ± 10 years, 89% were female, 41% Asian, 25% Hispanic, 26% white non-Hispanic, and 8% African American.  The proportion of individuals classified as frail was 48% (Table 1). Factors associated with frailty in bivariate analyses were low income, obesity, physical inactivity, and patient-reported disease activity (Table 1).  In multivariate analysis, only race/ethnicity remained a significant independent predictor of frailty (Hispanic only when adjusting for SLEDAI, Hispanic and African America when adjusting for SLAQ).  In post-hoc analyses, obesity and inactivity were found to be significantly more common among Hispanics and African Americans (p< .01, not shown). Frailty was associated with worse scores on most PROs, even after adjusting for race/ethnicity, income, obesity, and physician-assessed disease activity (Table 3).

Conclusion: Using the SPPB, nearly half of this sample was identified as frail.  A previous study using the FFP found 20% prevalence of frailty and 50% pre-frailty.  Racial/ethnic differences were noted, but may be due to greater prevalence of obesity and inactivity among Hispanics and African Americans.  Patient-reported outcomes were consistently worse among individuals classified as frail, even after adjusting for covariates.  Interventions aimed at reducing inactivity may reduce frailty among individuals with SLE.

¹ Katz P. Lupus Sci Med 2017;4:e000186
² Fried L. J Gerontol 2001; 56:M146
³ Guralnik J. J Gerontol 1994; 49:M85

« Back to ACR Convergence 2020

ACR Meeting Abstracts - https://acrabstracts.org/abstract/predictors-of-frailty-identified-by-the-short-physical-performance-battery-and-associations-with-patient-reported-outcomes/