Background/Purpose: Systemic Lupus Erythematosus (SLE) can lead to a number of neuropsychiatric manifestations including cognitive impairment (CI). Previous meta-analyses have reported the prevalence of CI at 38% (95% confidence interval: 33-43%), with a range of 15-79%. Comprehensive neuropsychological battery (NB) of tests is considered the gold standard when diagnosing CI. We aim to compare a subjective questionnaire, the perceived deficits questionnaire (PDQ-20) to the NB and also to other patient reported outcomes, in an SLE cohort.

Methods: This is a cross-sectional study of consecutive consenting patients, aged 18-65 years, who attended a single center (Jul 2016 – Mar 2019). Each patient completed a comprehensive NB evaluating six cognitive domains including: manual motor speed and dexterity, simple attention and processing, visual spatial construction, verbal fluency, learning and memory (verbal and visuospatial) and executive function. They also completed the 20 item PDQ-20 questionnaire (subjective cognitive function) along with other patient reported outcome questionnaires such as the Beck anxiety score, Beck depression score, fatigue severity score (FSS), Short Form Health Survey (SF-36) domains [Role Physical, Mental Health, Bodily Pain, General Health, Physical Functioning, Emotional, Social Functioning, and Vitality domains].

The variable of main interest was the total score of 19 tests in NB, along with patients’ demographics, lupus disease activity, organ damage, treatment and other PROs. Mean ± std., median (interquartile range) and count (%) were calculated for these variables. Univariate and multivariable linear regressions were performed to evaluate the associated factors with total PDQ-20 scores. Least Absolute Shrinkage and Selection Operator (LASSO) method was used in the variable selection in multivariable model building process. Linear model assumptions were tested by residual density plots and quantile-quantile plots.

Results: Data on 238 patients was analysed; 89.9% were females with an average age and SLE duration at baseline visit of 41.1 ± 12.1 and 14.3 ± 10.0 years, respectively. In the univariate analysis, PDQ-20 was associated with the NB, SDI, fatigue severity score (FSS), BECK anxiety and depression score and all SF-36 domains (Table 1). In the multivariate analysis, PDQ-20 was significantly associated with female gender, SDI, fatigue, BECK depression and anxiety scores, SF-36 Role Emotional domain and glucocorticoid dosage. PDQ-20 was not associated with the NB. There was also no association with age at first visit, SLE disease duration, ethnicity, education level, SLEDAI score, anti-malarial or immunosuppressive treatment with the PDQ-20 score.

Conclusion: Subjective cognitive report by PDQ-20 was associated with the NB in univariate analysis but not multivariate analysis. While PDQ-20 was associated with all SF-36 domains in the univariate analysis, this association was significant only with Role Emotional in the multivariate analysis. There was a clear association of PDQ-20 with other subjective patient reported outcomes (such as fatigue, anxiety and depression, and SF-36 Role Emotional domain).

« Back to ACR Convergence 2021

ACR Meeting Abstracts - https://acrabstracts.org/abstract/association-of-subjective-cognitive-report-using-pdq-20-to-a-neuropsychological-battery-in-a-cohort-of-sle-patients/