Background/Purpose: Ankylosing spondylitis (AS) is characterized both by inflammation of the axial skeleton and systemic inflammation, and may also involve joints, entheses and other organs. For other rheumatic chronic inflammatory diseases, such as rheumatoid arthritis, an increase in mortality compared to the background population has consistently been shown, whereas for AS, information on survival is scarce. The aims of the present study were to: 1) determine mortality in AS vs the general population, overall and by gender, and 2) to investigate factors associated with death in the AS cohort. 

Methods: From the National Patient Register (NPR) we identified a nationwide cohort of patients who were diagnosed with AS at a rheumatology or internal medicine department at least once between Jan 2001 and Dec 2009. A general population comparator, matched on year of birth, gender at the first registered diagnosis year of the AS patient, was identified from the census register, with 5 matched controls per index-patient. Socioeconomic variables and comorbidities (prior to the start of follow up) were identified from Statistics Sweden, NPR and the national Drug Prescription Register. The period of risk began on Jan 1^st 2006 or at the first date of registered diagnosis thereafter in those with previously undiagnosed AS and extended until death, emigration, or Dec 31^st 2012 (end of observation). Incidence Rate Ratios (IRR) were calculated when comparing mortality in AS and control cohorts and Cox regression models were used to determine predictors for death in the AS cohort.

Results: Among the 8 600 AS patients and the 40 460 controls, there were 496 and 1533 deaths, respectively (34% vs 31% due to cardiovascular disease – CVD). Mortality was increased in AS with an overall IRR of 1.71 (95% CI: 1.55 – 1.90), as well as for men (IRR = 1.65, 95% CI: 1.47 – 1.86) and women (IRR = 1.89, 95% CI: 1.55 – 2.29) in separate analyses. Male gender and higher age predicted death in the AS cohort. In addition, lower level of education and several comorbidities, both general (CVD, diabetes, pulmonary and malignant diseases) and AS-related (previous small or large joint surgery) were associated with increased risk of death in age/-sex adjusted analyses.

Conclusion: Mortality in this national, population-based AS cohort was increased both in men and women compared to matched controls from the general population. Both general and AS-related comorbidities predicted death suggesting that both traditional and AS-specific risk factors may affect survival.