Risk factors and co-morbidities associated with neuro-psychiatric involvement in SLE: a sub-analysis in 413 patiens without antiphosphiolipid antibodies

Background: Neuropsychiatric (NP) involvement represents a major cause of morbidity and mortality in Systemic Lupus Erythematosus (SLE) patients and appears strongly related to the presence of antiphospholipid antibodies. Objectives: To evaluate, by a retrospective cross-sectional case-control multicentric study from 8 italian rheumatology institutions, the distribution of a number of risk factors and co-morbidities potentially related to central nervous system involvement in a large cohort of SLE patients without antiphospholipid antibodies. Methods: From a large Italian multicentric cohort of patients with SLE a sub-analysis including only those patients negative for aPL (aCL, LA and anti-b2GP1) was performed comparing the distribution of a number of ''generic'' (i.e. not strictly SLE related: hypertension, diabetes, atherosclerosis, dyslipidemia, smoking, hyperhomocysteinemia, glucocorticosteroids, ecc.) and ''specific'' (i.e. disease-related: anti-Sm and anti Ro/SSA antibodies, cutaneous vasculitis, Sjogren's syndrome, ecc) risk factors. Results: From the entire cohort of 959 SLE patients, 100 SLE patients with neuropsychiatric (NP) involvement and 313 SLE patients without neuropsychiatric symptoms (SLE) without anti-phospholipid antibodies were judged eligible for this sub-analysis. Among ''generic'' risk factors, hypertension did emerged as more frequently represented in SLE than in NPSLE patients (31,7% vs 19%; p=0,020) and a higher prevalence of familiarity for psychiatric disorders resulted significantly more represented in NPSLE than in SLE (9,3% vs 2,5%; p=0,010). Considering NP events, a more frequent occurrence of hypertension was observed among patients without NP events (33,2%) whilst a cumulative dose of GC >10 grams was significantly more frequently recorded in patients with SLE-related than in patients with SLE-unrelated NP events and in those without NP events at all. The stepwise logistic regression analysis applied to this subset of patients showed that age at disease onset (OR =0,970; 95% CI 0,949 to 0,992) and a cumulative dose of glucocorticosteroids <10 grams (OR =0,426; 95% CI 0,230 to 0,789) were independently and inversely related to NP involvement. When the occurrence of SLE-related NP events was assumed as the dependent variable, again, the same two variables were retained as inversely and independently related. Conclusions: aPLA, LA and APS have been confirmed as the most relevant risk factor associated with NP involvement (especially focal) in the original cohort of 959 SLE patients. In the sub-analysis performed in 413 patients negative for aPL (aCL, LA and anti-b2GP1) or APS, other risk factor such as familiarity for psychiatric disorders, cumulative dose of GC >10 grams and a younger age at disease onset did emerge as related to the NP involvement. Conversely hypertension did prove to be as more frequently represented in SLE than in NPSLE patients.