Background/Purpose: Health-related quality of life (HRQoL) is an overarching outcome in axial spondyloarthritis (axSpA), and is considered influenced by disease activity and physical function. However, this relationship has never been demonstrated in a longitudinal framework. The impact of disease activity on HRQoL may be either direct, or occur indirectly by worsening physical function. The aim of this study was to investigate whether the effect of disease activity on HRQoL is primarily direct or mediated by physical function over time in early axSpA.

Methods: AxSpA patients from the DESIR cohort were analysed over 10 years of follow-up. Disease activity (assessed by the Ankylosing Spondylitis Disease Activity Score, ASDAS), physical function (Bath Ankylosing Spondylitis Functional Index, BASFI), and HRQoL outcomes (Physical Component Summary [PCS] and Mental Component Summary [MCS] of the SF-36, with higher values representing better HRQoL, and Ankylosing Spondylitis Quality of Life [ASQoL], with higher values corresponding to worse HRQoL) were assessed. Structural equation modeling (SEM) was performed to assess mediation, including direct and indirect (i.e. mediated) effects of disease activity on HRQoL. The hypothesis regarding mediation was that ASDAS and BASFI (assessed at the same time point) impact HRQoL at a subsequent time point (ie one year later). All models were adjusted for significant covariates (eg sex, marital/social status, BMI, NSAID use, bDMARDs use, enthesitis, IBD, comorbidity count) .Results are presented as beta coefficient (95% confidence intervals).

Results: A total of 663 patients were analysed, 46% males, mean age 33.5 (8.6) years. At 10 years, both ASDAS and BASFI were on the low range (ASDAS 2.0 [1.0] BASFI 2.3 [2.2]), and QoL scores were moderate at the group level (PCS 42.9 [10.4], MCS 45.0 [10.5], ASQoL 6.5 [5.5]). BASFI significantly mediated the association between ASDAS and HRQoL (Table 1). The indirect effects through BASFI were consistently larger than the direct effects. For PCS, the indirect effect was -3.09 (-3.38,-2.80) and the direct effect was -1.22 (-1.65,-0.80). For MCS, the indirect effect was -2.09 (-2.50,-1.71) versus a direct effect of -1.55 (-2.11,-1.00). For ASQoL, the indirect effect was 1.58 (1.41,1.75) and the direct effect was 0.59 (0.36, 0.82). Despite the majority of the effect of ASDAS on HRQoL being mediated, about a quarter of the effect was direct (Figure 1). These findings confirm HRQoL as a higher-order outcome, supported by its longitudinal association with both ASDAS and BASFI. We then propose a model of disease outcomes in axSpA, with proven longitudinal relationships (Figure 2).

Conclusion: In early axSpA, disease activity impacts HRQoL mainly through its negative influence on physical function; however a direct, independent effect also exists. These findings reinforce the hierarchical model of disease outcomes and emphasize the central role of targeting disease activity to optimize HRQoL, through its direct and indirect effects.

Table 1. Mediation analysis considering a direct and indirect (mediated by function, BASFI) effect of disease activity (ASDAS) on HRQoL, according to the main mediation hypothesis (ASDAS and BASFI influence HRQoL one year later)

Figure 1. Mediation structure, with estimation of direct and indirect effects

Legend. t = time point; t+1 = time point plus 1 year

Figure 2. Longitudinal framework with Health Related Quality of Life as the overarching outcome

Legend. The shadowed blocks represent the main variables of interest for this analysis. Red arrows represent the longitudinal relationships clarified by this current analysis. Blue arrows represent (longitudinal) associations previously described in the literature.

ACR Meeting Abstracts - https://acrabstracts.org/abstract/physical-function-mediates-most-but-not-all-of-the-effect-of-disease-activity-on-health-related-quality-of-life-in-axial-spondyloarthritis-a-10-year-longitudinal-analysis/