Background To study the prevalence and risk factors of fragility fractures in patients with SLE.

Methods 383 patients who fulfilled ≥4 ACR criteria for SLE and had a DEXA scan performed (baseline) were longitudinally followed for new fragility fractures. Osteoporosis/fracture was defined as a DEXA T score <-2.5 or Z score <-2.0 at the hip/femoral neck/spine or a history of old fractures. The cumulative incidence of new fractures was studied by Kaplan-Meier’s analysis and risk factors by Cox regression, adjusted for confounders.

Results 383 SLE patients were studied (age 40.5±13 years; 94% women). Patients with osteoporosis/fracture at baseline (n=113) were more likely to have childhood onset disease (<18 years), longer SLE duration and higher prevalence of hematological or neuropsychiatric manifestations than those without. Use of glucocorticoids (GCs) and MMF/AZA, BMI≤18kg/m2, premature menopause (<45 years) were also more frequent in the osteoporosis/fracture group. However, no difference in the SLEDAI scores was observed. Over 153±41 months, new symptomatic fragility fractures developed in 34(8.9%) patients (vertebral [n=19], hip [n=2], limbs (non-hip) [n=6], digital/rib [n=7]; incidence 0.69 per 100 patient-years). The cumulative risk of fragility fractures was significantly higher in the osteoporosis than non-osteoporosis group (p<0.001). Cox regression showed that increasing age (HR1.08[1.03–1.12]), osteoporosis/fracture (HR3.47[1.59–7.59]) and a family history of fracture (HR4.31[1.41–13.2]) were independently associated with new fractures after adjustment for SLE duration, childhood onset disease, other osteoporosis risk factors, clinical manifestations, GCs and immunosuppressive medications. No relationship between the daily dosage of GCs and fractures was observed.

Conclusions In this longitudinal cohort of SLE, new fragility fracture developed in 8.9% of patients. Increasing age and severe osteoporosis at baseline was major risk factors.