Reply

Abstract

We read with great interest the article by Katz et al recently published in Arthritis Care & Research (1) investigating sleep disorders in people with rheumatoid arthritis (RA). The study revealed that of patients with RA, 21% had a diagnosis or risk of obstructive sleep apnea (OSA), 30% had a diagnosis or symptoms of restless legs syndrome (RLS), and 43% reported short sleep (SS). An increased level of RA-related pain or Rheumatoid Arthritis Disease Activity Index score was associated with all sleep disorders. However, despite promising results, we would like to address concern about the potential confounding effect of unmeasured covariates on this study.Several craniofacial abnormalities, including narrowing of the lateral peritonsillar, retrognathia, tonsillar hypertrophy, micrognathia, macroglossia, high-arched or narrow palate, enlarged or elongated uvula, nasal septal deviation, and nasal polyps, can narrow the upper airways and appear to pose an increased risk of having OSA (2-4). About 40-60% of patients with RLS have a family history indicating that RLS presents a high pedigree trait (5-7). Furthermore, low iron stores in the brain that can accompany reduced serum ferritin levels have shown to be associated with the risk of RLS, especially in older people or those without a family history of RLS (6-8). Some people with depletion of iron stores in the brain may not cause a decrease in hemoglobin or hematocrit levels, and those with RLS and iron deficiency would not have been anemic (9,10). Parkinson's disease, pregnancy, uremia, multiple sclerosis, and spinal cord disorders have also been reported to carry an increased risk of RLS (7,11). These comorbidities or conditions correlated with OSA or RLS were not evaluated in the present study, which could affect the study results. Thus, this issue needs to be clarified.Finally, we appreciate the impressive work of Katz et al. Because there was limited information based on self-report data in this study, we fear the potential confounding effect could not be fully unraveled. Therefore, we would like to draw the reader's attention to the potential limitations in interpreting the important findings of this study.