Background/Purpose: Lungs can be affected in up to 50% of systemic lupus erythematosus (SLE) patients. Our aim was to assess the prevalence and degree of dyspnea in a SLE cohort using questionnaires and to see if they correlate with lung involvement.

Methods: Consecutive patients with SLE by 1997 ACR criteria in the Lupus Clinical Repository at a single academic center were given the UCSD Shortness of Breath Questionnaire (UCSD SOBQ) and the Dyspnea-12 Index (D-12). These questionnaires have been validated in connective tissue disease-related interstitial lung disease (ILD) cohorts but not in SLE. UCSD SOBQ is a 24-item tool (scores 0-120) assessing dyspnea severity with activities of daily living and its psychological impact. D-12 is a 12-item scale (scores 0-36) incorporating “physical” and “affective” aspects of dyspnea severity. Scores >50 were considered high for UCSD SOBQ and >12 for D-12. Demographics and SLE disease characteristics were recorded. Lung involvement was ascertained with medical record review by two rheumatologists. Correlation between dyspnea scores and lung involvement was determined using the Spearman’s Rank Order Correlation or Wilcoxon rank sum test, as appropriate, with SAS 9.4. A p value of <0.05 was considered significant.

Results: 50 SLE patients completed questionnaires (95% female, mean age 44 ± 14 years, mean disease duration 18 ± 10 years). 42% were White, 16% Latino, 12% Black, 10% Asian, and 20% other. Mean ± SD SLE disease activity (SLEDAI-2K) was 3.8± 4. 98% of subjects were ANA positive, with about 1/3 positive for anti-dsDNA (27%), anti-RNP (33%), and anti-Smith (33%). Fewer were positive for Anti-Ro (17%), anti-La (6%) and anti-centromere (4%). 48/50 (96%) of subjects had abnormal UCSD SOBQ scores consistent with some degree of dyspnea. 37/50 (74%) had both abnormal UCSD SOBQ and D-12 scores. The median score was 24 (8-52) for UCSD SOBQ and 3 (0-8) for D-12. 13 had high UCSD SOBQ (score > 50), 10 had high D-12 (score > 12), and 8 had high scores on both indexes. 8 had lung involvement (7 ILD, 1 pleurisy) with all but 1 having a high UCSD SOBQ score, and 5 of them having both high UCSD SOBQ and D-12 scores. 3 had high scores on both indexes without SLE related lung involvement (2 heart failure, 1 unidentifiable cause). 3 had high UCSD SOBQ without high D-12 scores and no SLE related lung involvement (1 asthma, 2 unindentifiable cause). 2 had high D-12 without high UCSD SOBQ scores and no SLE related lung involvement (1 asthma, 1 unindentifiable cause). High UCSD SOBQ (p = 0.002) and D-12 scores (p = 0.004) were strongly associated with SLE lung involvement. High UCSD SOBQ scores correlated with higher SLE disease activity (SLEDAI-2K) (p = 0.03) and high D-12 scores with positive anti-dsDNA (p = 0.03). There were no significant associations with age, sex, SLE disease duration, or other autoantibodies.

Conclusion: Dyspnea was prevalent in our SLE cohort. UCSD SOBQ and D-12 questionnaires captured the majority of SLE patients with known lung involvement, with UCSD SOBQ being potentially more sensitive than D-12. These questionnaires may be useful tools to screen for dyspnea and possibly identify subclinical shortness of breath in SLE patients.