To analyze whether changes in serum 25-hydroxyvitamin D (25[OH]D) levels affect activity, irreversible organ damage, and fatigue in systemic lupus erythematosus (SLE).
To analyze whether changes in serum 25-hydroxyvitamin D (25[OH]D) levels affect activity, irreversible organ damage, and fatigue in systemic lupus erythematosus (SLE).
We performed an observational study of 80 patients with SLE included in a previous cross-sectional study of 25(OH)D, reassessed 2 years later. Oral vitamin D3 was recommended in those with low baseline 25(OH)D levels. The relationship between changes in 25(OH)D levels from baseline and changes in fatigue (measured by a 0–10 visual analog scale [VAS]), SLE activity (measured by the Systemic Lupus Erythematosus Disease Activity Index [SLEDAI]), and irreversible organ damage (measured by the Systemic Lupus International Collaborating Clinics/American College of Rheumatology Damage Index [SDI]) were analyzed.
Sixty patients took vitamin D3. Mean 25(OH)D levels increased among all treated patients (P = 0.044), in those with baseline vitamin D levels <30 ng/ml (P < 0.001), and in those with baseline vitamin D levels <10 ng/ml (P = 0.005). Fifty-seven patients (71%) still had 25(OH)D levels <30 ng/ml and 5 (6%) had 25(OH)D levels <10 ng/ml. Inverse significant correlations between 25(OH)D levels and the VAS (P = 0.001) and between changes in 25(OH)D levels and changes in the VAS in patients with baseline 25(OH)D levels <30 ng/ml (P = 0.017) were found. No significant correlations were seen between the variation of the SLEDAI or SDI values and the variation in 25(OH)D levels (P = 0.87 and P = 0.63, respectively).
Increasing 25(OH)D levels may have a beneficial effect on fatigue. Our results do not support any effects of increasing 25(OH)D levels on SLE severity, although they are limited by the insufficient 25(OH)D response to the recommended regimen of oral vitamin D3 replacement.
Vitamin D is a hormone implicated in calcium homeostasis, with additional effects on other organs and systems such as the muscles, endothelium, and immune cells (1). Apart from effects on bone metabolism, vitamin D deficiency has been related to an increased risk of cardiovascular disease and a higher risk of all-cause mortality in the general population (2). Also, an increased risk for autoimmune diseases, including type 1 diabetes mellitus, multiple sclerosis, rheumatoid arthritis, and systemic lupus erythematosus (SLE), in relation to low vitamin D levels and/or intake, has been reported (3, 4). In the case of lupus, vitamin D deficiency has been advocated as one of the causes explaining the increased prevalence and severity of SLE among African Americans (5). However, the relationship between vitamin D levels and the clinical course of SLE is still controversial.
Between January and December 2006, we performed a cross-sectional study including 92 patients from the Lupus-Cruces cohort in whom serum 25-hydroxyvitamin D (25[OH]D) levels were measured. Our results showed a high prevalence of vitamin D insufficiency and deficiency that was related to sun avoidance. A relationship between vitamin D deficiency and fatigue, but not with lupus activity or irreversible organ damage, was found (6).
Thereafter, patients were treated with oral vitamin D3 (in the form of tablets containing fixed combinations of calcium and cholecalciferol or with calcidiol in liquid form) at the discretion of the attending physician. After 2 years of therapy and clinical followup of our patients, we aimed to further define the role of vitamin D in modulating the clinical expression of SLE.
We designed a longitudinal observational study with the objective of analyzing whether changes in 25(OH)D levels relate to changes in SLE activity, irreversible organ damage, and fatigue. Specifically, we aimed to prospectively confirm our previous results, with the hypothesis of an improvement in fatigue following the increase in 25(OH)D levels.
The 92 patients from the Lupus-Cruces cohort that were included in our previous study (6) were contacted and invited to participate in this second, longitudinal part of the study. Eighty patients were available for clinical evaluation during the 2 months (October and November 2008) in which the study took part. All of the patients fulfilled at least 4 of the American College of Rheumatology (ACR) classification criteria for SLE (7) and signed the informed consent form. The local institutional review board approved the study protocol and the informed consent form (study code CEIC E07/38) in compliance with the Helsinki Declaration.
At the time of signing the informed consent form, the following variables were recorded for this study: age, sex, ethnicity, disease duration, smoking status, current treatment for SLE (prednisone and dose, hydroxychloroquine, immunosuppressive drugs), and the dose and duration of treatment with oral vitamin D3 since the previous study (maximum 24 months). The Systemic Lupus Erythematosus Disease Activity Index (SLEDAI) score (8) and the Systemic Lupus International Collaborating Clinics/ACR Damage Index (SDI) score (9) were calculated at that point. In addition, each patient was asked to reflect the degree of fatigue using the same 0–10 visual analog scale (VAS) used in the previous study (where 0 = no fatigue and 10 = intense fatigue) (10). Levels of 25(OH)D were determined using the Liaison 25OH Vitamin D Total Assay (DiaSorin), as previously described (6). This is a competitive 2-step chemiluminescence assay, with a measuring range of 4.0–150 ng/ml. The reported analytical sensitivity is <1.0 ng/ml and the functional sensitivity is <4.0 ng/ml. The reported specificity is 104% for 25(OH)D2 and 100% for 25(OH)D3. The reliability of this assay in our laboratory has been regularly assessed by participating in the Vitamin D External Quality Assessment Scheme, the largest vitamin D quality assurance program worldwide (11).
Herein, we designate T1 as the time when the previous study was done in 2006, and T2 as the time of the current evaluation in October and November 2008. The clinical assessment of patients was done at T1 by GR-I and M-VE and at T2 by GR-I and SG. The adherence to treatment with oral vitamin D3 was addressed by the patients self-reporting. We analyzed whether the 12 patients unavailable for this longitudinal part of the study were different from the remaining cohort in terms of the variables measured at T1.
The proportion of patients who achieved optimal levels of 25(OH)D at T2, i.e., at or above 30 ng/mg, was calculated. Likewise, we calculated the proportion of patients with 25(OH)D levels below 20 ng/ml and 10 ng/ml at T2. The variations in 25(OH)D levels, the VAS, and the SLEDAI and SDI scores between T1 and T2 were analyzed.
The relationship between vitamin D and the final clinical variables (fatigue, activity, and damage) was analyzed using 2 different strategies. First, we assessed the association between 25(OH)D levels and each of the corresponding scores (VAS, SLEDAI, and SDI), all measured at T2. Second, we tested the correlation between the changes in 25(OH)D levels and the changes of each of the scores from T1 to T2. Subgroup analysis according to baseline (i.e., at T1) levels of serum vitamin D and according to whether or not treatment with oral vitamin D3 was received was also performed.
The normality of continuous variables was established by means of the Kolmogorov-Smirnov test. According to normality, continuous variables were summarized by the mean ± SD or the median and range. Univariate comparisons between nominal variables were done by chi-square test with Yates' correction or Fisher's exact test, as appropriate. Comparisons of continuous variables between 2 groups were done using a Student's 2-tailed t-test in the case of normal variables and the Mann-Whitney U test in the case of non-normal variables. Variations in variables between T1 and T2 were analyzed using McNemar's test, the Student's paired t-test, or Wilcoxon's test, as required. The associations between continuous variables were tested by linear regression. Adjustment for potential confounders was performed by means of multiple linear regression.
All of the statistical calculations were done using SPSS software, version 11.0.4, for Mac OS X (SPSS).
Eighty patients, 72 (90%) of them women and 78 (98%) of them white, agreed to participate in the study. The mean ± SD age at the time of inclusion was 43 ± 14 years. The median disease duration was 9.5 years (range 2–29 years). Sixty-seven patients (84%) were taking hydroxychloroquine, 44 (55%) were taking prednisone at a median dosage of 2.5 mg/day (range 0–15), and 16 (20%) were receiving immunosuppressive drugs at T2. The clinical characteristics of the study group at T1 and T2 are shown in Table 1.
|Female sex||72 (90)||72 (90)||N/A|
|Age, mean ± SD years||41.2 ± 13.8||43 ± 14||N/A|
|Disease duration, median (range) years||8 (0–27)||9.5 (2–29)||N/A|
|Photosensitivity||54 (67)||54 (67)||N/A|
|Anti-Ro antibodies||31 (39)||31 (39)||N/A|
|Renal disease||23 (29)†||23 (29)†||N/A|
|Prednisone||42 (52)||44 (55)||0.8|
|Prednisone dosage, median (range) mg/day||2.5 (0–10)||2.5 (0–15)||0.7|
|Immunosuppressive drugs||13 (16)||16 (20)||0.45|
|Hydroxychloroquine||66 (82)||67 (84)||1.0|
|Smoking||20 (25)||23 (29)||0.45|
|SLEDAI score, median (range)||0 (0–18)||2 (0–21)||0.015|
|SDI score, median (range)||0 (0–5)||0 (0–6)||0.044|
|25(OH)D level <30 ng/ml||62 (77)||57 (71)||0.42|
|25(OH)D level <20 ng/ml||38 (48)||26 (33)||0.06|
|25(OH)D level <10 ng/ml||13 (16)||5 (6)||0.06|
|VAS for fatigue, mean ± SD||4.1 ± 3.0||3.3 ± 2.6||0.015|
Twelve patients (13%) included in the first study group were not available for a second clinical and biochemical evaluation within the 2 months in which this study was performed. No significant differences in the baseline variables at T1 were found between active and missing patients, except for a clinically irrelevant difference in the SDI score (data not shown).
Sixty patients (75%) in the entire group took oral vitamin D3 for a median period of 24 months (range 7–24 months). Among the 62 patients with 25(OH)D levels <30 ng/ml at T1, 47 (76%) took vitamin D3 at a median dosage of 800 IU/day (range 400–1,200) during a median period of 24 months (range 5–24 months). Ten (77%) of 13 patients with 25(OH)D levels <10 ng/ml at T1 received oral vitamin D3 at a median dosage of 600 IU/day (range 400–1,200) for a median period of 24 months (range 7–24 months).
Mean 25(OH)D levels of the entire cohort increased from 21.7 ng/ml at T1 to 24.8 ng/ml at T2 (P = 0.044). The same was true for patients with 25(OH)D levels <30 ng/ml at T1 (from 16.6 to 23.6 ng/ml; P < 0.001) and for patients with 25(OH)D levels <10 ng/ml at T1 (from 6.4 to 20.1 ng/ml; P = 0.001). A significant increase in 25(OH)D levels was seen only among patients who took oral vitamin D3 (Table 2).
|Subgroup||25(OH)D level at T1, mean ± SD ng/ml||25(OH)D level at T2, mean ± SD ng/ml||P|
|Total cohort (n = 80)||21.7 ± 12.2||24.8 ± 10.1||0.044|
|Total cohort, treated (n = 60)†||20.7 ± 11.5||25.8 ± 9.6||0.005|
|Total cohort, untreated (n = 20)‡||24.8 ± 13.8||21.8 ± 11.6||0.27|
|25(OH)D level <30 ng/ml at T1, treated (n = 47)†||16.1 ± 7.4||25.1 ± 9.8||< 0.001|
|25(OH)D level <30 ng/ml at T1, untreated (n = 15)‡||18.5 ± 8.0||19.0 ± 7.7||0.84|
|25(OH)D level <10 ng/ml at T1, treated (n = 10)†||5.9 ± 1.9||21.1 ± 13.2||0.005|
|25(OH)D level <10 ng/ml at T1, untreated (n = 3)‡||8.0 ± 2.6||16.9 ± 4.4||0.16|
Despite the significant increase in mean 25(OH)D levels at T2, it is noteworthy that as many as 57 patients (71%) still had vitamin D levels <30 ng/ml, 26 (33%) had 25(OH)D levels <20 ng/ml, and 5 (6%) had 25(OH)D levels <10 ng/ml (Table 1). Among patients with 25(OH)D levels <30 ng/ml at T1 who took oral vitamin D3, 34 (72%) did not accrue optimal levels, with 13 (27%) and 2 (4%) patients in this subgroup having 25(OH)D levels <20 ng/ml and <10 ng/ml at T2, respectively. Only 2 of 10 treated patients with 25(OH)D levels <10 ng/ml at T1 attained optimal levels at T2, with 4 (40%) and 2 (20%) of them still having 25(OH)D levels <20 ng/ml and <10 ng/ml, respectively.
Fatigue improved in the entire cohort according to the reduction seen in the mean ± SD VAS score between T1 and T2 (4.1 ± 3.0 versus 3.3 ± 2.6; P = 0.015) (Table 3). However, an improvement in VAS score was seen only among patients who took oral vitamin D3, although the differences in the subgroups of treated patients with 25(OH)D levels <30 ng/ml and <10 ng/ml at T1 did not reach the level of statistical significance (P = 0.09 and P = 0.23, respectively), probably due to the lack of power related to the small numbers of patients (Table 3).
|Subgroup||VAS at T1, mean ± SD||VAS at T2, mean ± SD||P|
|Total cohort (n = 80)||4.1 ± 3.0||3.3 ± 2.6||0.015|
|Total cohort, treated (n = 60)†||4.0 ± 3.0||3.0 ± 2.6||0.012|
|Total cohort, untreated (n = 20)‡||4.6 ± 3.0||4.0 ± 2.5||0.53|
|25(OH)D level <30 ng/ml at T1, treated (n = 47)†||3.9 ± 2.8||3.2 ± 2.6||0.09|
|25(OH)D level <30 ng/ml at T1, untreated (n = 15)‡||4.3 ± 3.2||4.4 ± 2.2||0.98|
|25(OH)D level <10 ng/ml at T1, treated (n = 10)†||5.5 ± 2.6||4.6 ± 1.6||0.23|
|25(OH)D level <10 ng/ml at T1, untreated (n = 3)‡||4.6 ± 1.6||4.7 ± 1.4||0.97|
We found an inverse significant association between 25(OH)D levels and the VAS score at T2 (P = 0.001). This relationship was independent of the effect of other variables shown to influence tiredness in our first study, such as age, SLEDAI score, and treatment with hydroxychloroquine (P = 0.001).
Using different cutoff points for 25(OH)D levels at T2, the mean ± SD VAS score was different in those with values above or below 30 ng/ml (2.4 ± 2.5 versus 3.6 ± 2.5; P = 0.057), above or below 20 ng/ml (2.5 ± 2.4 versus 4.8 ± 2.1; P < 0.001), and above or below 10 ng/ml (3.2 ± 2.5 versus 4.5 ± 3.5; P = 0.2).
Changes in the VAS score between T1 and T2 were not associated with changes in 25(OH)D levels in the entire cohort (P = 0.41). However, a significant inverse correlation was found after restricting this analysis to patients with 25(OH)D levels <30 ng/ml at T1 (P = 0.017). The results did not change (P = 0.017) after controlling for potential confounders such as treatment with hydroxychloroquine at T2, dose of prednisone at T2, SLEDAI score at T2, SDI score at T2, SLEDAI score variation between T1 and T2, and SDI score variation between T1 and T2, none of which was statistically related to variations in the VAS score.
Regarding the relationship between 25(OH)D levels and disease severity, we obtained results akin to our previous study. Serum levels of 25(OH)D at T2 did not correlate with either SLEDAI or SDI values (P = 0.5 and P = 0.3, respectively). Likewise, median values for both the SLEDAI and SDI did not significantly differ whether the patients had 25(OH)D levels below or above the different cutoff points (Table 4).
|Median (range)||P||Median (range)||P|
|25(OH)D level <30 ng/ml||2 (0–21)||0.41||0 (0–5)||0.89|
|25(OH)D level >30 ng/ml||2 (0–8)||1 (0–5)|
|25(OH)D level <10 ng/ml||2 (0–17)||0.88||2 (0–5)||0.35|
|25(OH)D level >10 ng/ml||2 (0–21)||0 (0–5)|
Both the SLEDAI and SDI scores significantly increased between T1 and T2, although this increase was small in clinical terms (SLEDAI: median 0 [range 0–18] versus median 2 [range 0–21]; P = 0.015, and SDI: median 0 [range 0–5] versus median 0 [range 0–6]; P = 0.044). However, no significant association was found between the variation of SLEDAI or SDI values and the variation in 25(OH)D levels between T1 and T2 (P = 0.87 and P = 0.63, respectively). This was also true for the subgroups of patients with 25(OH)D levels at T1 <30 ng/ml (P = 0.84 and P = 0.64 for the SLEDAI and SDI, respectively), <20 ng/ml (P = 0.58 and P = 0.64, respectively), and <10 ng/ml (P = 0.64 and P = 0.85, respectively).
In this observational longitudinal study, we have found a high prevalence of suboptimal 25(OH)D serum levels despite therapy with standard doses of oral vitamin D3. Changes in serum 25(OH)D levels were inversely associated with fatigue, as measured using a 0–10 VAS. On the other hand, no relationship was found between absolute values or variations in 25(OH)D serum levels and scores measuring SLE activity (SLEDAI) or damage accrual (SDI).
Low serum 25(OH)D levels in patients with SLE have been demonstrated in a number of studies performed in different populations from countries at variable latitudes (6, 12–17). Rather unsurprisingly, photosensitivity has been found to predict vitamin D deficiency (6, 13).
Besides its well-known effects on calcium homeostasis, vitamin D exerts several additional effects. Vitamin D receptors (VDRs) have been found in muscle cells, with a well-recognized relationship between vitamin D deficiency and muscle weakness (3, 18), as well as with cardiovascular disease (19). With respect to autoimmunity, the presence of VDRs is protean among immunologic cells, including T and B lymphocytes, macrophages, and dendritic cells (1). An immunoregulatory role of vitamin D has been proposed, with an increased risk of developing autoimmune diseases among vitamin D–deficient individuals (20). Moreover, “prophylactic” treatment with this compound has been advocated by some authors in order to prevent the development of SLE (21).
Yet, the specific effects of vitamin D on SLE are far from clear. Four cross-sectional studies found an inverse relationship between 25(OH)D levels and lupus activity, measured by means of the SLEDAI and/or Systemic Lupus Activity Measure (13, 22–24). However, other authors have not found such an association (14–17). Our previous cross-sectional study did not show any relationship between 25(OH)D levels and SLEDAI scores, either as a continuous variable or taking into account different cutoff points. Moreover, damage accrual, as measured by the SDI, was not associated with vitamin D deficiency (6).
It is important to note that 1,25-dihydroxyvitamin D (1,25[OH]2D) is the active form of the hormone, which has been shown to inhibit proliferation and differentiation and enhance apoptosis of activated B cells in in vitro studies (14). In one study, it was 1,25(OH)2D, but not 25(OH)D, that correlated with SLE activity (14). The fact that this active form of the hormone is not routinely measured due to fluctuations of its serum levels further complicates the interpretation of the clinical effect of vitamin D on SLE activity. Moreover, conversion of 25(OH)D into 1,25(OH)2D is reduced by drugs such as hydroxychloroquine as well as by renal disease (3, 25).
In this study, we offer a dynamic view of the problem. After recommendation to take oral vitamin D3, which was followed by 75% of patients, the mean serum 25(OH)D levels of our population with SLE increased significantly, although a high number of patients still had values below 30 ng/ml. However, an inverse decrease in SLEDAI values was not found; rather, mean SLEDAI scores increased parallel to 25(OH)D levels between T1 and T2, although with no significant correlation. Patients accrued little damage within the study period despite the improvement in vitamin D levels, again without a statistical association between these 2 variables. Therefore, these data confirm our previous results that showed no relationship between 25(OH)D levels and lupus severity, and neither activity nor irreversible damage (6).
On the other hand, our previous finding of improved fatigue with higher 25(OH)D levels has now been confirmed prospectively. We found an independent association between 25(OH)D and the VAS score, both as continuous variables. The median VAS was also higher (reflecting more tiredness) in patients with levels below 30 ng/ml, 20 ng/ml, and 10 ng/ml as compared with those patients with levels above these values, although with a lack of statistical significance in the latter comparison, probably due to the low number of patients in this subgroup (n = 5). Indeed, in patients with vitamin D levels below 30 ng/ml at T1, increasing 25(OH)D levels were inversely and significantly correlated with a decrease in the VAS score. Therefore, with fatigue being a complex and multifactorial manifestation of lupus, our results confirm a potential role of vitamin D deficiency in SLE.
Measuring fatigue in SLE is a difficult task. In 2007, the Ad Hoc Committee on Systemic Lupus Erythematosus Response Criteria for Fatigue reviewed the available instruments to measure fatigue in patients with lupus (26). Fifteen different scales were used in 34 studies. More than 50% of the studies used the Fatigue Severity Scale, a 9-item scale with a 1–7 possible score in each item, which was the instrument recommended by the authors. A VAS similar to that used in this study has also been tested in SLE and employed to measure fatigue in other clinical settings (10). This scale was used in our 2006 cohort and again in this study in order to analyze its variations. However, the actual clinical impact of the reduction of fatigue seen in this study is uncertain and our results should be further validated using more accurate instruments.
The nonrandomized design of this study also limits our findings. Moreover, although most patients were treated with oral vitamin D3, with 25(OH)D levels significantly increasing in the cohort, many of them did not reach optimal levels, which constitutes an additional problem. The possible influence of obesity in the final 25(OH)D levels could not be established because we did not calculate the body mass index of our patients. We did not find any association between 25(OH)D levels and the SLEDAI, either in absolute values or in terms of change over time. It is possible that achieving higher final 25(OH)D levels would have resulted in more striking differences in activity. Likewise, larger variations in SLEDAI scores would have resulted in an increased potency of the study. However, even the trends were not suggestive of an inverse association, since both 25(OH)D levels and the SLEDAI increased significantly over time. On the other hand, despite the same limitation of insufficient vitamin D repletion in many patients, a significant and consistent reduction in fatigue was noted parallel to increasing 25(OH)D levels.
In summary, this study supports the use of dosages of vitamin D3 higher than 800 IU/day for lupus patients with vitamin D insufficiency or deficiency. Dose adjustments should be based on monitoring of serum 25(OH)D levels. In the general population, levels above 30 ng/ml are the recommended goal in order to avoid parathyroid hormone activation (3). However, it is not clear whether these are also the optimal levels in patients with lupus. Increasing 25(OH)D levels may have a beneficial effect in reducing fatigue in patients with SLE. The relationship between vitamin D and lupus activity is complex and still poorly understood, needing further studies to clarify the clinical consequences of vitamin D deficiency and repletion. However, our results do not support any improvement of SLE activity following treatment, and subsequent increase of serum 25(OH)D levels, with oral vitamin D3.
All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be submitted for publication. Dr. Ruiz-Irastorza had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study conception and design. Ruiz-Irastorza, Egurbide, Aguirre.
Acquisition of data. Ruiz-Irastorza, Gordo, Olivares.
Analysis and interpretation of data. Ruiz-Irastorza, Gordo.
We thank Dr. Mugica for his help in defining the reliability of the 25(OH)D assay in our hospital.