Absolute risk estimation of new-onset proteinuria in patients with systemic lupus erythematosus – a Danish nationwide cohort study

Abstract

Abstract Background Kidney involvement in systemic lupus erythematosus (SLE) affects approximately 40% of patients and is associated with increased mortality and morbidity. The risk of renal involvement has primarily been reported as hazard ratios (HR) which may be challenging to interpret on a patient level. Additional data reporting such as absolute risk estimates may strengthen risk stratification and compliance. This study provides absolute risk estimations of risk of new-onset proteinuria among SLE patients. Methods Danish SLE-centres provided clinical data on first time observations of proteinuria and other clinical parameters listed in the 1997 American College of Rheumatology Classification Criteria for SLE. Time from first occurring non-renal manifestation to new-onset proteinuria or censoring defined time at risk. Cox-regression models were used to identify risk factors for new-onset proteinuria and to calculate risk of proteinuria stratified by risk factor debut age, duration and sex. Models were reduced using a backwards elimination process for p>0.01. Potentially relevant interaction covariate terms were added to the model in a forward selection procedure using p<0.01. Results Patient population consisted of 586 patients with SLE, mainly Caucasian (94%) women (88%), mean age at inclusion of 34.6 years (standard deviation, SD = 14.4 years), observed for a mean of 14.9 years (SD =11.2 years). The cumulative prevalence of proteinuria was 40%. Male sex, HR = 1.35 (p=0.009), lymphopenia HR = 1.77 (p=0.005) were associated with new-onset proteinuria. Male patients with lymphopenia had the highest predictive risks of proteinuria with a 1-, 5- and 10-year risk of proteinuria ranging from 9-27%, 34-75% and 51-89 %, depending on the age at presentation (debut at 20, 30, 40 or 50 years). The corresponding risk profiles for women with lymphopenia were 3-9%, 8-34% and 12-58%, respectively. Conclusions Large differences in absolute risk estimates for new-onset proteinuria were identified. The differences may aid risk stratification and patient compliance among high-risk individuals.