Vitamin D deficiency is associated with higher disease activity and the risk for uveitis in juvenile idiopathic arthritis - data from a German inception cohort

Objective The objective was to evaluate the 25(OH) vitamin D (25(OH)D) status of patients with juvenile idiopathic arthritis (JIA) and determine whether the 25(OH)D level is associated with disease activity and the course of JIA. Methods Patients ≤ 16 years of age with recently diagnosed JIA (< 12 months) were enrolled in the inception cohort of patients with newly diagnosed JIA (ICON), an ongoing prospective observational, controlled multicenter study started in 2010. Clinical and laboratory parameters were ascertained quarterly during the first year and half-yearly thereafter. Of the 954 enrolled patients, 360 patients with two blood samples taken during the first 2 years after inclusion and with follow up of 3 years were selected. The serum 25(OH)D levels were determined and compared with those of subjects from the general population after matching for age, sex, migration status and the month of blood-drawing. Results Nearly half of the patients had a deficient 25(OH)D level (< 20 ng/ml) in the first serum sample and a quarter had a deficient level in both samples. Disease activity and the risk of developing JIA-associated uveitis were inversely correlated with the 25(OH)D level (β = − 0.20, 95% CI − 0.37; 0.03, hazard ratio 0.95, 95% CI 0.91; 0.99, respectively). Conclusion In this study, 25(OH)D deficiency was common and associated with higher disease activity and risk of developing JIA-associated uveitis. Further studies are needed to substantiate these results and determine whether correcting 25(OH)D deficiency is beneficial in JIA.


Introduction
Over the last decade, it has become clear that vitamin D is more than the "bone vitamin" that regulates calcium homeostasis and bone mineralization. In addition to its pleiotropic functions in different cells and tissues, the differentiation, polarization and activity of immune cells are affected by 1,25 (OH) 2 vitamin D3 [1]. This is the biologically active form that results from the double hydroxylation of cholecalciferol that is ingested or produced in the skin by ultraviolet sunlight. Several studies have shown an association between low vitamin D, mostly measured as 25(OH)vitamin D3 (25(OH)D), and higher incidence and severity of autoimmune disorders, such as type 1 diabetes mellitus, chronic inflammatory bowel disease, rheumatoid arthritis (RA) and multiple sclerosis [1]. Nisar et al. conducted a meta-analysis of several studies evaluating patients with JIA with reported serum vitamin D levels [2]; however, the parameters that were measured (25(OH)D, 1,25(OH) 2 D3), clinical outcomes and the cutoff values that were used differed substantially. Currently, the measurement of 25(OH)D is considered standard because it is stable and reflects the vitamin D provision of the last weeks to months. In our study, we wanted to (1) evaluate the 25(OH)D level in patients with newly diagnosed JIA and compare it with that in individuals from the general population; (2) analyze whether disease activity is correlated with the 25(OH)D level and (3) determine whether the 25(OH)D level might predict the disease course.

Patients
The inception cohort of newly diagnosed patients with JIA (ICON) has been described in detail elsewhere [3]. Briefly, patients diagnosed with JIA according to the International League of Associations for Rheumatology (ILAR) criteria [4] over the previous 12 months were included in this prospective, multicenter, observational cohort study. On the basis of standardized questionnaires and pediatric rheumatologic assessments, clinical and demographic data and information on medication use were collected every 3 months during the first year and every 6 months thereafter. Vitamin D supplementation was not ascertained. The active joint count, the physician's global assessment of disease activity on a numeric rating scale (NRS, 0-10, 0 = best), the parents' global assessment of wellbeing (NRS) and the antinuclear antibody (ANA) status (negative/positive based on the standards of the local laboratory that carried out this assay) were recorded. A slit lamp examination was performed by an ophthalmologist every 3-6 months according to the screening guidelines by the German study group "Uveitis in childhood" [5], and the findings were documented on a specific uveitis documentation form. Patients 8 years of age and older and their parents provided informed assent/consent for participation. The study was approved by the ethics committee of the Charité -Universitätsmedizin Berlin.

Serum samples
Serum samples were taken as part of routine laboratory controls in ICON and were collected, frozen and stored at the ICON biobank at the University of Münster, Germany, until they were shipped to Berlin, Germany, to perform the 25(OH)D assay for all samples at the same time at a medical laboratory. For our analysis, we selected patients for whom a pair of serum samples was available. The first sample was collected at a visit that occurred between baseline and the 9-month follow up (94% at baseline or at 3-month follow up), and the second sample was collected at a visit that occurred between the 3-month follow up and the 36-month follow up (92% until the 1-year follow up), with a mean interval of 7 months (SD 5) between blood draws. The time points for serum collection in male and female subjects were equally distributed in winter (October-March) and summer (April-September).

25(OH) vitamin D measurement and cutoff values
The LIAISON 25 OH Vitamin D TOTAL Assay (direct competitive chemiluminescence immunoassay, DiaSorin Inc., Stillwater, MN, USA) was used to quantitatively determine the 25(OH)D level in the serum samples. The measurement range of this assay ranges from 4 ng/ml (10 nmol/l) to 150 ng/ml (375 nmol/l). To compare the 25(OH)D levels of patients with healthy subjects, we used data from the German National Health Interview and Examination Survey for Children and Adolescents (KIGGS), which was conducted from May 2003 to May 2006, to obtain representative data on health status and selected laboratory values of children and adolescents aged 0-17 years across Germany (for details, please refer to [6]). Valid measurements of serum 25(OH)D, also performed with the LIAISON 25 OH Vitamin D TOTAL Assay, from 10,015 participants of this study were available from public-use files. Control subjects were matched with a ratio 1: 1 for age, sex, month of blood draw and migration background. The latter was defined as "mother's and/or father's country of origin not Germany". This variable integrates several factors like skin pigmentation but also cultural habits like type of clothing and veiling that possibly influence the vitamin D production in the skin and thereby the 25(OH)D serum level [7].

Outcome and statistics
The disease activity of patients with JIA was measured by the clinical Juvenile Arthritis Disease Activity Score (cJADAS-10), a composite score (0-30, 0 = best) based on the physician's global assessment of disease activity, the parents' global assessments of wellbeing, and the number of active joints up to a maximum of 10, according to McErlane et al. [14]. For the present analysis, (a) the disease activity, (b) the development of the extended form of oligoarthritis and (c) the occurrence of uveitis up to 3 years after inclusion in ICON in relation to the 25(OH)D level were investigated.
Serum 25(OH)D levels were compared by the paired t test, and the categories of serum 25(OH)D levels between patients and controls were compared by conditional logistic regression analyses. The association between disease activity and 25(OH)D level was assessed by linear regression analysis. The association between the likelihood of uveitis and serum 25(OH)D, including the covariates of prior methotrexate (MTX) therapy and established uveitis risk factors, such as age at JIA onset, female sex, oligoarticular JIA onset, and ANA positivity, were analyzed using a multivariable Cox proportional hazard model. Multivariable logistic regression analysis was used to investigate the proportion of oligoarthritis patients who progressed to extended oligoarthritis until the 3-year follow up, with serum 25(OH)D as a predictor. P values < 0.05 were considered statistically significant. The data were analyzed using SAS software version 9.4 (SAS Institute, In der Neckarhelle 162, 69,118 Heidelberg).

Results
The subgroup of patients whose 25(OH)D levels were analyzed did not differ from the whole ICON study population in terms of the male-to-female ratio, age at study inclusion or symptom onset and ILAR category ( Table 1). Migration background was found in 28.1% of the study population.

25(OH)D level, disease activity and therapy
We hypothesized that an effect of the vitamin D status later in the disease course might be abrogated by already established therapy. Thus, we used the 25(OH)D level from the first measured sample to analyze possible correlation with disease activity because at this time point, the patients were in an early phase of disease (median disease duration at that time 6.6 (IQR 3.8-11.4) months) and any treatment was initiated only briefly (212 (59%) treated with disease modifying anti-rheumatic drugs (DMARDs), 195 (54%) with MTX, and 18 (5%) with biologic disease-modifying anti-rheumatic drugs (bDMARDs); median therapy duration 1.0 (IQR 0.5-3.5) month). We identified negative correlation between disease activity and the 25(OH)D level (Fig. 1a), which became more pronounced when we analyzed DMARD-naïve patients (Fig. 1b)   this time than patients with 25(OH)D > 30 ng/ml (n = 2, 2.6% and n = 44, 56.4%, respectively); these results were not statistically significant.  Table 2).

Risk of developing extended oligoarthritis or uveitis
We  Fig. 2). In the multivariable regression analysis, we showed that the 25(OH)D level was significantly associated with the risk of developing uveitis (Table 3). For every 1 ng/ml increase in 25(OH)D > 22.1 ng/ ml (the mean of the cohort), there was a 5% reduction in the risk of developing uveitis. This correlation persisted when we limited the examination to patients with uveitis who had blood drawn for 25(OH)D determination prior to developing uveitis (HR 0.95, 95% CI 0.92; 1.00, p = 0.03).

Discussion
In this study, in which vitamin D level was repeatedly prospectively measured in patients observed as standard, with recent-onset JIA, the following clinically relevant findings were ascertained: -Vitamin D deficiency was common in patients with JIA and even more common in their peers from the general population -25(OH)D deficiency was associated with higher JIA disease activity -The progression from oligoarthritis to an extended disease course occurred more often in patients with 25(OH)D deficiency than in those with 25(OH)D sufficiency -The 25(OH)D level was inversely correlated with the risk of developing JIA-associated anterior uveitis Pelajo et al. also identified a high prevalence of vitamin D deficiency when they measured 25(OH)D in children and adolescents with rheumatologic diseases and in children with non-autoimmune disorders (e.g., noninflammatory disorders, infectious diseases, pain amplification syndrome) as controls [15]. The risk of vitamin D deficiency was significantly higher in patients with autoimmune disorders (OR 2.3) than in patients with non-autoimmune diseases; however, no significant differences in the mean 25(OH)D levels between these two patient groups were observed. Our finding that vitamin D deficiency is even more common in control subjects than in patients with JIA is somehow unexpected and cannot be easily explained because we controlled for the possible influencing factors by matching for age, sex, migration background and even the month of blood sampling.
Possibly patients with JIA are screened more frequently for vitamin D deficiency in the course of clinical routine and might consequently receive vitamin D supplements. Pepmueller et al. assessed parameters of bone mineralization in patients with JIA and also found that mean 25(OH)D in patients (28.2 ng/ml) was significantly higher than in healthy controls (22.0 ng/ml) [16].
We identified significant negative correlation between the 25(OH)D level from the first serum sample and JIA activity, as measured by the cJADAS-10. Pelajo et al. did not find this association in their cross-sectional study of 154 patients with JIA but noted that the included patients had established ongoing disease and that the majority had received some kind of disease-modifying treatment [17]. In a subset analysis of newly diagnosed patients with JIA, they identified negative association between the serum 25(OH)D levels and the JADAS-27, but this result was not statistically significant. In our study, there was an inverse relationship between 25(OH)D and disease activity that was even more pronounced in DMARD-naive subjects. Van Hamburg et al. showed that TNF-alpha blockade alone was insufficient  to suppress the production of interleukin (IL)-17A and IL-22, which are two key cytokines involved in the pathophysiologic interaction of T helper 17 (Th17) cells and synovial fibroblasts in RA. Only in combination with 1,25(OH) 2 D 3 did TNF-alpha blockade synergistically suppress inflammatory mediators in RA synovial fibroblast cocultures [18]. We did not observe significant differences in mean 25(OH)D between the JIA categories, but the proportion of 25(OH)D-deficient patients differed among the subgroups. Similar to the work by Stagi et al. [19], we identified the highest percentage of vitamin D deficiency in patients with systemic and polyarticular JIA. Whether this finding is somehow specific to a particular JIA category or whether it is the result rather than the cause of disease severity cannot be deduced from our data.
Analyzing the disease course and outcomes of JIA, we observed a higher proportion of stable 25(OH)D-deficient patients with oligoarthritis who developed an extended disease course during the first 3 years in ICON than of patients with oligoarthritis and a stable sufficient 25(OH)D level. To our knowledge, this association has not been described before, and it requires further investigation.
One of the most relevant extra-articular manifestations of JIA is anterior uveitis, which is frequently associated with complications and can be accompanied by visual impairment. We confirmed the known risk factors for the development of anterior uveitis in patients with JIA, including ANA positivity, young age at JIA onset and no prior treatment with methotrexate [20]. A new finding in our study was that a higher percentage of patients with JIA with stable deficient 25(OH)D developed uveitis until the 3-year follow up than did patients with a stable sufficient 25(OH)D level. The odds of developing uveitis were 5% lower for every 1 ng/ml increase in 25(OH)D. As far as we know, this result has not been previously described in childhood uveitis and may be of great interest in predicting and possibly preventing JIA-associated uveitis. Grotting et al. compared 100 adult patients with and without non-infectious anterior uveitis and demonstrated that hypovitaminosis D (defined as 25(OH)D < 30 ng/ml) was associated with this disease, and the odds of developing uveitis were 4% lower for every 1 ng/ml increase in 25(OH)D [21]. Mitulescu et al. analyzed the influence of vitamin D in patients with ankylosing spondylitis in terms of the development of acute anterior uveitis (AAU) and showed an association between low 25(OH)D and AAU accompanied by increased IL-8 and serum amyloid A as markers of the inflammatory response [22]. Yi et al. showed that in patients with Vogt-Koyanagi-Harada disease, a type of autoimmune pan-uveitis, 1,25(OH) 2 D 3 was significantly lower in patients with active disease than in patients with inactive disease and in controls. Moreover, they demonstrated that the incubation of peripheral blood mononuclear cells (PBMC) with 1,25(OH) 2 D 3 resulted in reduced proliferation of PBMC and reduced production of IFN-and IL-17A in the presence of anti-CD3 and anti-CD28 [23], suggesting that vitamin D supplementation might be an adjunctive treatment option for patients with this form of uveitis.
As a large amount of in vitro data has shown the anti-inflammatory and antiproliferative effects of vitamin D on immune cells [1], a causal link between low 25(OH)D and autoimmune diseases is plausible. However, low 25(OH)D might also be the result rather than the cause of the underlying inflammation in these diseases because it has been shown that 25(OH)D decreases during the acute-phase response parallel to the increase in C reactive protein after surgery [24]. Therefore, Welsh et al. investigated whether the administration of a TNF-α blocker -and thereby reduction of the inflammatory reaction in patients with RA -changes the 25(OH)VD levels. They observed a significant improvement in the Disease Activity Score in 28 joints (DAS28) and a decrease in erythrocyte sedimentation rate (ESR), while median 25(OH)D remained unchanged after treatment with adalimumab [25].
To date, only a few papers have described the effects of vitamin D supplementation in children with autoimmune diseases. Reed et al. examined the effect of 25(OH)D supplementation at 1-2 μg/kg body weight/day over 1 year in 13 children with active polyarticular JIA and described a significant increase in 25(OH)D, but did not observe change in disease activity with this regimen [26]. By contrast, Lima et al. performed a study in adolescents and young adults with systemic lupus erythematosus and showed that the increase in the 25(OH)D was accompanied by reduced disease activity and fatigue and decreased frequency of anti-double stranded (ds)DNA antibodies after 24 weeks of supplementation with 50,000 IU of cholecalciferol per week [27].

Conclusion
Our study revealed that vitamin D deficiency is common in patients with JIA, although the prevalence was lower than in matched subjects from the general population. In patients with JIA, vitamin D deficiency is associated with higher disease activity and a higher risk of developing uveitis and possibly the extended form of oligoarthritis. The strength of this study is the prospective controlled design with standardized data ascertainment. A limitation of this study is that the any use of vitamin D supplementation was not recorded. To further clarify if cholecalciferol substitution in order to achieve a sufficient 25(OH)D level might have the potential to prevent the development of uveitis and the extended form of oligoarthritis, these associations must be confirmed by an interventional study.