Purpose To analyze the causes and identify predictive factors of mortality of Systemic Lupus Erythematosus (SLE), and to assess the time evolution and chronological changes in Spain.

Methods We performed a cross-sectional and retrospective study analyzing data from the RELESSER cohort (Spanish Registry of SLE of the Spanish Society of Rheumatology). Sociodemographic, clinical and serological variables, comorbidities and treatments, as well as indicators of disease activity, damage and severity were recorded. We excluded patients with lost information about the death variable and analyzed the differential features of deceased patients in comparisons with survivors through different time stages according to the date of diagnosis: until the 1980's; the 1990's and the first decade of the 21st century. Variables associated with mortality in univariate analysis were entered into different multivariate models to determine which ones were independently associated with the outcome of the disease in each decade.

Results A total of 3665 patients were included, mostly caucasian female with similar general features regardless of the different time stages analyzed. The 18.4% until the 1980’s, the 5.97% in the 1990’s and up to 2.84% of the individuals in the first decade of the 21st century, had died. The main age of death was similar in the different groups, around 55-58 years old (Table). The vascular events were the leading cause of death until the 1980’s, while in the last two decades, were SLE activity followed by infections.

The older age at diagnosis was predictor of mortality in our cohort. Neither gender nor delay in diagnosis was independently associated with mortality, with the exception of the female sex, which behaved as a protective factor until the 1980’s.

The mortality predictors in our cohort were the presence of hypocomplementemia, organ damage, ischemic disease and hospitalizations until the 1980’s; thrombocytopenia, thrombosis, antiphospholipid syndrome and valve disease in the 1990’s; nephritis, organ damage, depression, severe infections and ischemic disease in the first decade of the 21st century. Conversely, skin involvement was related to greater survival over the last two decades of the study.

The use of high doses of corticosteroids was predictor of mortality in each time stage, as well as the use of cyclophosphamide and rituximab from the year 2000. Antimalarial treatment was linked to improved survival in all the decades analyzed as well as the use of mycophenolate in the 1990’s.

Conclusions In the RELESSER cohort, the main causes of death were disease activity and infections, with the exception until the 1980’s, which were vascular events.

The older age at diagnosis, the use of corticosteroids and comorbidities, were associated with a significant increase in mortality in SLE, while antimalarial treatment was linked to improved survival. Data indicate that organ damage is a risk factor and skin involvement is a protective factor against mortality. Differentially, female sex until the 1980’s was independently associated to improved survival, and depression at the beginning of the 21st century was linked to mortality.