Patterns of comorbidity and their impact on axial spondyloarthritis

Zhao, Sizheng
(2021) Patterns of comorbidity and their impact on axial spondyloarthritis. PhD thesis, University of Liverpool.

[img] Text
200512737_Nov2020.pdf - Unspecified

Download (10MB) | Preview


Background and Aim: Axial spondyloarthritis (axSpA) patients have higher risk of co-existing medical conditions – comorbidities – than the general population, yet management guidelines are largely based on randomised controlled trials that often exclude those with comorbidities. The overarching aim of this thesis was to describe the patterns of comorbidity and their impact on treatment using three real-world axSpA populations. Methods: Three datasets were used: one from a specialist axSpA service (Aintree cohort) and one from tertiary hospitals in Boston, Massachusetts (Boston cohort), and a longitudinal national axSpA register – the British Society for Rheumatology Biologics Register for axSpA (BSRBR-AS). Chapter 4 compared the prevalence of 39 comorbidities between two axSpA phenotypes – non-radiographic and radiographic axSpA – using the Boston and Aintree cohorts. Cluster analysis was then used to examine how comorbidities co-exist in the Aintree cohort, and how clusters relate to axSpA severity using multivariable regression models. Chapter 5 applied multivariable regression models to baseline BSRBR-AS data to examine whether axSpA disease assessment using various indices were differentially influenced by 14 comorbidities. Chapter 6 applied conditional models to longitudinal data from the BSRBR-AS to investigate the association between comorbidity and treatment response. Chapter 7 focused on the potential causal association of baseline mental health symptoms on treatment outcomes using marginal models. Results: The Boston cohort included 775 patients (mean age 53 (SD 17) years, 74% male), Aintree 421 patients (46 (SD 14) years, 69% male), and BSRBR-AS 2042 patients (49 (SD 15) years, 67% male). Around half of each cohort had at least one comorbidity. Comorbidity patterns and counts were similar between radiographic and non-radiographic axSpA in Boston (mean 1.5 vs. 1.3, respectively) and Aintree cohorts (1.4 vs. 1.3). Mental health conditions tended to co-exist and were associated with greater axSpA disease severity. In the baseline BSRBR-AS data, each additional comorbidity was associated with higher BASDAI (Bath AS Disease Activity Index) by 0.40 units (95% CI 0.27, 0.52) and ASDAS (AS Disease Activity Score) by 0.09 units (95% CI 0.03, 0.15). Among 994 BSRBR-AS patients starting TNF inhibitors (TNFi), those with multiple comorbidities had similar absolute improvement in disease activity but reduced improvement in function (by 1.0 unit in BASFI at 6 months) and health-related quality of life (2.3 ASQoL units). Compared to those with less than mild depressive symptoms, patients with moderate-severe symptoms had reduced response at 6 months, by approximately 2 BASDAI units and 0.9 ASDAS units. Patients with moderate-severe anxiety symptoms had increased treatment discontinuation (HR 1.59; 95% CI 1.12 to 2.26) than those with less than mild anxiety. Conclusions: Comorbidities are highly prevalent in axSpA, in particular cardiovascular and mental health disorders. Comorbidities are associated with axSpA disease activity at baseline and adverse treatment outcomes, particularly depression and anxiety. These results have important clinical and policy implications for the current approach to disease assessment. Clinicians should record and refer comorbidities for optimal management, particularly mental health disorders.

Item Type: Thesis (PhD)
Divisions: Faculty of Health and Life Sciences > Institute of Life Courses and Medical Sciences
Depositing User: Symplectic Admin
Date Deposited: 08 Mar 2021 15:16
Last Modified: 02 Mar 2022 08:41
DOI: 10.17638/03114178