Background The prevalence of depression and associated factors in systemic lupus erythematosus (SLE) are not well known and there are no longitudinal studies addressing this relevant subject in SLE.

Objectives We aimed to evaluate the prevalence of self-perceived depression in patients with SLE and associated factors in a large, multicenter cohort (RELESSER-PROS).

Methods Prospective longitudinal study of patients with SLE (1997 ACR criteria) answering positively to the depression question of the Lupus Impact Tracker (LIT) questionary (namely: “I was depressed” question number 7) (LITQ7), over 4 years of follow-up (5 annual visits, V1 to V5). Self-perceived depression was addressed as “depression any time” or “depression most of time”, according to the kind of answer to the LITQ7 (answers 1,2,3 or 4 and answers 3 or 4 respectively). Only patients with no missing values in the covariates, making possible run longitudinal models, were included in the multivariable analysis. The following covariates, with potential impact in depression, were considered: SLEDAI, age, duration of the disease, SLICC/AR DI (SDI), fibromyalgia, Charlson index, smoking, BMI, menopause, sedentary lifestyle, marital status, unemployment and glucocorticoid use.

Friedman test was used to test if the change in repeated measures was significative. Generalized estimating equation (GEE) models with binomial response, were built exploring the associations of individual longitudinal determinants with longitudinal assessment of depression. The best model was selected using quasi-likelihood under the independence model information criterion (QIC)

Results A total of 1463 were included. Mean age: 55 (DS±13.59) years, 90% were female. Mean duration of the disease: 14 (±8.59) years. Fibromyalgia was present in 5.7% (76/1343). Corticosteroids use ranged from 49.4% to 57%, depending on the visit. Median SLEDAI ranged from 0 to 2 and SDI ranged from 1 to 2.

Prevalence of “depression any time” was 89.9% (1104/1228) and 34.6% (200/578) were in depression “most of time”. Up to 26.5% (153/578) answered to LITQ7 “depression most of time” in the five visits; 89.7% of the patients which perceived themselves as depressed at least in 2 out of 5 visits. Only 6.9% of the patients with previous diagnosis of depression answered “0” to the Q7 of LIT (“none of the time”).

Only following covariates showed changes, statistically significative, during the follow up: SLEDAI, SDI, Charlson and glucocorticoids use (Friedman test).

Patients with “depression any time” develop more damage at V5 than patients without depression (answer to LITQ7=0) (p = 0.00931, T-test). In the GEE binomial analysis considering all the predefined covariates, that included only patients with no missing values for any of them (namely, 155 patients),

fibromyalgia (OR 2.79; 95%CI: 1.28-6.05), unemployment (OR 1.95; 95%CI 1.02 -3.73), and glucocorticoids use (OR 1.88; 95%CI 1.18-2.99) were significant associated with “depression any time”.

The best model (according QIC) displayed a statistically significant association only with fibromyalgia (OR 2.90; 95%CI: 1.58-5.33) and glucocorticoids use (OR 1.85; 95%CI 1.17-2.93). Neither SDI nor unemployment reached significance here (Table 1). Without entering glucocorticoids, SLEDAI turns significant in the model, suggesting collinearity.

Conclusion The prevalence of self-perceived depression is high in SLE. Longitudinal data analysis suggests a causal relationship between glucocorticoids use, fibromyalgia and self-perceived depression.

QICC: 1006.75

SLICC/ACR DI: SLICC/ACR damage index; BMI: body mass index

Acknowledgements Research Unit of the Spanish Society of Rheumatology. Spanish Foundation of Rheumatology financial supporting, through the research intensification grants program. GSK financial supporting,

Disclosure of Interests None Declared.

Keywords: Systemic lupus erythematosus

DOI: 10.1136/annrheumdis-2023-eular.1811