Correlation between serum 25(OH)D values and lupus disease activity: an original article and a systematic review with meta-analysis focusing on serum VitD confounders

Abstract

Notwithstanding that several original studies and some systematic reviews have been undertaken on the subject “correlation between serum values of vitamin D (VitD) and lupus disease activity,” there is still no consensus on the importance of sectional measurement of serum VitD in the prediction of disease activity and important confounders in estimation of serum VitD. Medline, Web of Knowledge, and Scopus databases were searched from 1995 to 2013. The following medical subject heading (MeSH) terms and/or text words were used: “Vitamin D” OR “25OHD” OR “25(OH)D” combined with “systemic lupus erythematosus” OR “lupus” OR “SLE.” References cited in the identified articles were also manually searched. Human studies in any language were included. Original research on this topic was also carried out on 82 lupus patients, considering important VitD confounders according to our systematic review and we included them in the meta-analysis. A total of 35 studies were registered for this study. Only 11 of these pointed to this correlation by Pearson test. The pooled Pearson correlation ( r) of associations between disease activity and VitD was −0.365 (95% CI: −0.536, −0.165) with significant heterogeneity ( p = 0.001 I2 = 93%). Sensitivity analysis resulted in no significant differences. The most important adjustable confounders considered by researchers were drugs, especially hydroxychloroquine, prednisolone and supplementary VitD, body mass index (BMI) and proteinuria or renal function. Only proteinuria was reported to influence VitD concentration strongly. BMI was another probable influencing factor. Our original research presented no correlation between VitD and SLEDAI ( p = 0.68, rs = 0.003). This meta-analysis demonstrated that most of the studies on the relationship between VitD and lupus disease activity that found no correlation did not present the details of the statistics. However, analyzing 11 studies, most of which found a reverse correlation and reported it in detail, and our study found a weak reverse correlation between those two items. Systematic review of confounders showed that BMI, medications and kidney involvement were the most remarkable ones reported by researchers.