Population Characteristics, Glucocorticoid Dosage, and Risk Factors for Osteonecrosis of the Femoral Head in Systemic Lupus Erythematosus: A Systematic Review and Meta-Analysis

Jiawen Zhang^1*Jindong Zhang²Yawei Dong³Nongyi Li⁴Junjie Li⁴Yishu Wen⁴Yan Jia⁵Biao Tan⁶Yan Yan⁴

• ¹Beijing University of Chinese Medicine Third Affiliated Hospital, Beijing, China
• ²Ningxia Zhang's Orthopedic Hospital, Yinchuan City, Ningxia Hui Autonomous Region, China
• ³China Academy of Chinese Medical Sciences Wangjing Hospital, Beijing, China
• ⁴Chongqing University of Chinese Medicine, Chongqing, China
• ⁵Shanxi Bethune Hospital, Taiyuan, China
• ⁶Chongqing City Hospital of Traditional Chinese Medicine, Chongqing, China

Background: Osteonecrosis of the femoral head (ONFH) is a severe complication of systemic lupus erythematosus (SLE). However, the demographic characteristics, glucocorticoids (GCs) risks, and other contributing factors remain debated. Objective: To elucidate the population characteristics, GCs-related risks, and other risk factors for ONFH in patients with SLE through a systematic review and meta-analysis, thereby enhancing the clinical identification of high-risk populations and optimizing GCs therapy strategies in SLE. Methods: We searched seven databases, from their inception until July 2025 for relevant cohort and case-control studies. The quality of included studies was assessed using the Newcastle-Ottawa Scale. Meta-analysis was performed using RevMan 5.3. Results: Thirty-five studies involving 11,356 participants were included. Regarding population characteristics, patients with SLE who developed ONFH had a significantly younger age at diagnosis (SMD = -0.19, P < 0.00001) and higher SLEDAI scores (SMD = 0.21, P = 0.002). Among metabolic and immune indicators, elevated triglycerides (SMD = 0.21, P = 0.02), decreased high-density lipoprotein cholesterol (SMD = -0.22, P = 0.03), and positive antiphospholipid antibodies (OR = 2.00, P = 0.04) were associated with ONFH occurrence. Regarding GC therapy, pulse steroid therapy (OR = 2.02, P < 0.00001), an initial dose >60 mg/day (OR = 4.19, P < 0.0001), a maximum daily dose >50 mg (SMD = 0.42, P = 0.0002), and higher average daily GC intake (SMD = 0.32, P = 0.004) significantly increased ONFH risk. In contrast, cumulative GC dose showed no significant association (P = 0.14). Furthermore, vasculitis (OR = 3.17, P < 0.00001), hypertension (OR = 1.48, P = 0.02), Raynaud's phenomenon (OR = 1.60, P = 0.0003), thrombocytopenia (OR = 1.69, P = 0.007), and arthritis (OR = 1.88, P = 0.006) were identified as independent risk factors. Conclusion: Patients with SLE at high risk for ONFH exhibit distinct characteristics. Short-term high-dose GC exposure, rather than cumulative dose, constitutes the core medication-related risk. Enhanced imaging screening and comprehensive, multi-factorial prevention strategies are warranted, particularly for patients receiving high initial doses or pulse therapy. Clinical management should focus on optimizing GC regimens in these high-risk individuals to minimize the occurrence of ONFH.