VA PUGET SOUND/UNIVERSITY OF WASHINGTON Seattle, Washington, United States
Disclosure information not submitted.
Background/Purpose: RA is associated with premature aging and frailty. Growth differentiation factor-15 (GDF-15) is a biomarker that increases in response to cellular stress and inflammation and is associated with cachexia, frailty, and death in the general population. A recent trial of an anti-GDF-15 antibody demonstrated improved weight gain and lower cachexia symptoms in participants with cancer1. GDF-15 is elevated in RA compared to controls and is associated with disease activity. We aimed to evaluate the association of GDF-15 with frailty and death in a national RA cohort. Methods: Participants from the Veterans Affairs Rheumatoid Arthritis (VARA) Registry, a multicenter prospective RA cohort, enrolled between 1/2003 through 12/2022 with baseline GDF-15 levels were included. Frailty was measured at baseline and annually throughout observation using the VA Frailty Index (VA-FI), a validated deficit accumulation score based on >6000 diagnostic codes. Participants were categorized as frail using the established VA-FI cutoff of >0.2. Vital status was assessed using the National Death Index. Serum GDF-15 was measured on biobanked baseline samples using the MesoScale platform and divided into quartiles for analysis. Multiple imputation with chained equations addressed missing data. Multivariable logistic regression evaluated the cross-sectional association between GDF-15 quartile and frailty. Multivariable Cox modeling also evaluated the association between GDF-15 quartile and incident frailty or death, in those not frail at baseline. Both models controlled for age, sex, race, BMI, disease duration, DAS28ESR, CCP positivity, prednisone use, csDMARD use, b/tsDMARD use and Rheumatic Disease Comorbidity Index (RDCI). Results: In total, 2,866 participants were included in the cross-sectional analysis (mean age of 64.6±11.1, 88% male, 76% White and 29% frail) (Table 1). Compared to the lowest quartile, those in the highest quartile of GDF-15 were older (70.9±9.2 vs. 56.3±11.5 years), more frequently male (95% vs. 74%), White (80% vs 68%), had longer disease duration (12.7±12.0 vs. 9.6±10.0), and had more comorbidities (RDCI 4.1±1.8 vs. 2.6±1.9). Compared to the lowest quartile, the 3rd quartile of GDF-15 was associated with 1.5 times the odds of frailty (aOR 1.50 [95%CI 1.1-2.1], p=0.016) and the highest quartile was associated with 2.2 times the odds of frailty (aOR 2.18 [95%CI 1.57-3.04], p< 0.001) (Figure 1). The incident frailty subcohort was comprised of 2,448 participants (mean age 62.7±11.2), of whom 61% (N=1,493) became frail (N=1130) or died (N=363) over 13,631 person-years of observation. Compared to the lowest quartile, the highest quartile of GDF15 was associated with a 2.6-fold increase in the risk of incident frailty or death (aHR: 2.59, 95%CI 1.68-3.98, p< 0.001) (Figure 2). Conclusion: Elevated GDF-15 was associated with baseline frailty and predicted incident frailty or death in a large RA cohort, independent of important demographic and disease characteristics. Further work is needed to evaluate the effect of GDF-15 on frailty onset to determine its relevance as a therapeutic target in at-risk individuals with RA.
