Diabetologia

, Volume 57, Issue 5, pp 902–908

Prevalence of vitamin D deficiency in pre-type 1 diabetes and its association with disease progression

Authors

  • Jennifer Raab
    • Forschergruppe Diabetes e.V.
    • Institute of Diabetes Research, Helmholtz Zentrum München, and Forschergruppe Diabetes, Klinikum rechts der IsarTechnische Universität München
  • Eleni Z. Giannopoulou
    • Forschergruppe Diabetes e.V.
    • Institute of Diabetes Research, Helmholtz Zentrum München, and Forschergruppe Diabetes, Klinikum rechts der IsarTechnische Universität München
  • Simone Schneider
    • Forschergruppe Diabetes e.V.
  • Katharina Warncke
    • Forschergruppe Diabetes e.V.
    • Institute of Diabetes Research, Helmholtz Zentrum München, and Forschergruppe Diabetes, Klinikum rechts der IsarTechnische Universität München
    • Department of Pediatrics, Klinikum rechts der IsarTechnische Universität München
  • Miriam Krasmann
    • Institute of Diabetes Research, Helmholtz Zentrum München, and Forschergruppe Diabetes, Klinikum rechts der IsarTechnische Universität München
  • Christiane Winkler
    • Forschergruppe Diabetes e.V.
    • Institute of Diabetes Research, Helmholtz Zentrum München, and Forschergruppe Diabetes, Klinikum rechts der IsarTechnische Universität München
    • Forschergruppe Diabetes e.V.
    • Institute of Diabetes Research, Helmholtz Zentrum München, and Forschergruppe Diabetes, Klinikum rechts der IsarTechnische Universität München
Article

DOI: 10.1007/s00125-014-3181-4

Cite this article as:
Raab, J., Giannopoulou, E.Z., Schneider, S. et al. Diabetologia (2014) 57: 902. doi:10.1007/s00125-014-3181-4

Abstract

Aims/hypothesis

Vitamin D deficiency is common in people with type 1 diabetes, but its role in disease progression is unclear. Our aim was to assess the prevalence of vitamin D deficiency in prediabetes (defined as the presence of multiple islet autoantibodies), and investigate whether or not progression to type 1 diabetes is faster in children with vitamin D deficiency and multiple islet autoantibodies.

Methods

Levels of 25-hydroxyvitamin D [25(OH)D] were measured in 108 children with multiple islet autoantibodies within 2 years of islet autoantibody seroconversion, in 406 children who remained islet autoantibody-negative and in 244 patients with newly diagnosed type 1 diabetes. Children with multiple islet autoantibodies were prospectively followed for a median of 5.8 years (interquartile range 3.4–8.6 years) to monitor progression to type 1 diabetes.

Results

In the cross-sectional analysis, 25(OH)D levels were lower and the prevalence of vitamin D deficiency (<50 nmol/l) was higher in children with prevalent multiple islet autoantibodies than in islet autoantibody-negative children (59.9 ± 3.0 vs 71.9 ± 1.5 nmol/l; p < 0.001; 39.8% vs 28.3%; p = 0.021). The differences in vitamin D levels between the groups were greatest in summer. The cumulative incidence of type 1 diabetes at 10 years after seroconversion was similar between children with vitamin D deficiency and those with sufficient vitamin D levels (51.8% [95% CI 29.3, 74.3] vs 55.4% [95% CI 35.5, 72.3], p = 0.8).

Conclusions/interpretation

Vitamin D levels were lower in children with multiple islet autoantibodies and in children with type 1 diabetes than in autoantibody-negative children. However, vitamin D deficiency was not associated with faster progression to type 1 diabetes in children with multiple islet autoantibodies.

Keywords

25(OH)DIslet autoimmunityType 1 diabetesVitamin D

Abbreviations

25(OH)D

25-Hydroxyvitamin D

IQR

Interquartile range

Introduction

Type 1 diabetes is a chronic autoimmune disease that usually manifests during early childhood and puberty [1]. The development of type 1 diabetes is preceded by a preclinical period that includes seroconversion to islet autoantibodies/autoimmunity and progression to diabetes. We recently reported that most children with multiple islet autoantibodies ultimately progress to type 1 diabetes within a few months to 20 years after the appearance of autoantibodies [2]. Disease progression is likely to be influenced by genetic and environmental factors [3]. Vitamin D deficiency is one environmental factor that has long been discussed to influence the risk of type 1 diabetes [4, 5]. In fact, we previously reported that people with newly diagnosed type 1 diabetes have low vitamin D levels [6], a finding that was confirmed in most other studies [79]. We also reported that people with type 1 diabetes are more sensitive to ultraviolet light and have less skin pigmentation than non-diabetic controls [10], which suggests that reduced sun exposure could be an explanation for the lower vitamin D levels observed in patients with type 1 diabetes. Other data that reinforce the role of vitamin D in the pathogenesis of type 1 diabetes include the important functions of this hormone in immunoregulation and immunoactivation [11, 12], the genetic association of vitamin D receptor with type 1 diabetes [13, 14] and the results of epidemiological studies showing that lower serum vitamin D levels and inadequate supplementation during infancy increase the risk of type 1 diabetes [1517]. Two nested case–control studies among healthy US military personnel showed that the prevalence of type 1 diabetes was highest in individuals with low vitamin D levels before the onset of diabetes [15, 16].

Currently, however, little is known about vitamin D status in prediabetes (defined as the presence of multiple islet autoantibodies) or whether or not vitamin D deficiency influences the rate of progression to type 1 diabetes. To date, only one prospective study has investigated the associations between islet autoimmunity or progression to diabetes and vitamin D intake or 25-hydroxyvitamin D [25(OH)D] levels in childhood [18]. However, the investigators found no association between vitamin D and progression to diabetes. The outcome measures evaluated in that study included the presence of any islet autoantibodies; a large proportion of participants had single islet autoantibodies, which carries a low risk of progression to diabetes [2, 18]. The use of vitamin D supplementation for the prevention of type 1 diabetes has been proposed [19] and, therefore, it is relevant to extend the current data on pre-type 1 diabetes.

In the present study, we focused on children who developed multiple islet autoantibodies, which carries a high risk of progressing to diabetes. We determined the prevalence of vitamin D deficiency in children with multiple islet autoantibodies, children with newly diagnosed type 1 diabetes and children with a family history of type 1 diabetes but without islet autoantibodies. We also determined the association between vitamin D deficiency and the rate of progression to type 1 diabetes in children with multiple islet autoantibodies.

Methods

Study cohorts

Levels of 25(OH)D were measured on one occasion in children enrolled in the prospective BABYDIAB, BABYDIET and TEENDIAB studies [2022]. These studies were undertaken to investigate the natural history of islet autoimmunity and type 1 diabetes in children with familial risk for type 1 diabetes. The studies followed a total of 3,140 offspring or siblings of patients with type 1 diabetes from birth (BABYDIAB/BABYDIET, n = 1,650/1,168) or from 6 years of age (TEENDIAB, n = 322). Of these children, 126 developed multiple islet autoantibodies at a median age of 2.3 years (interquartile range [IQR] 1.8–5.3 years; Fig. 1, electronic supplementary material [ESM] Table 1). Levels of 25(OH)D were measured in 108 of the 126 children with multiple islet autoantibodies (43 girls, 65 boys), where sufficient serum was available, within 2 years after islet autoantibody seroconversion. Levels were compared with those of 406 islet autoantibody-negative children. This control group included 108 children who were matched for age at blood sampling, sex, season of blood draw and study type to the children with multiple islet autoantibodies, and 298 children (150 girls, 148 boys) from the TEENDIAB study in whom measurements were taken at study enrolment. In addition, 25(OH)D levels were determined in 244 consecutive children (112 girls, 132 boys) with newly diagnosed type 1 diabetes who were registered in the Bavarian diabetes register cohort (the DiMelli study [23]) between December 2009 and December 2011 (ESM Table 1).
https://static-content.springer.com/image/art%3A10.1007%2Fs00125-014-3181-4/MediaObjects/125_2014_3181_Fig1_HTML.gif
Fig. 1

Flowchart of the study population included in the 25(OH)D analysis; Abs, autoantibodies; T1D, type 1 diabetes

Children with multiple islet autoantibodies were prospectively followed for the development of type 1 diabetes (median follow-up 5.8 years; IQR 3.4–8.6 years). One child was lost to follow-up. Diabetes onset was defined according to American Diabetes Association criteria, which include unequivocal hyperglycaemia with acute metabolic decompensation or (on at least two occasions) 2 h plasma glucose level >11.1 mmol/l after an oral glucose challenge, or a random blood glucose level >11.1 mmol/l if accompanied by unequivocal symptoms. Since 1997, the diagnostic criteria have included fasting blood glucose levels >7.0 mmol/l on two occasions [24].

The BABYDIAB, BABYDIET, TEENDIAB and DiMelli studies were approved by the ethics committees of Bayerische Landesärztekammer (no. 95357), Ludwig-Maximilian University Munich (no. 329/00), Technical University Munich (no. 2149/08), Medizinische Hochschule Hannover (no. 5644) and Bayerische Landesärztekammer (no. 08043), respectively. The studies were coordinated by the Institute of Diabetes Research in Munich, Germany where all samples were stored and centrally measured. Written informed consent was obtained from all of the participants or their parents.

Laboratory measurements

Plasma 25(OH)D levels were measured using a radioimmunoassay (DiaSorin, Stillwater, MN, USA) in samples that had been stored at −80°C. The interassay coefficient of variation of the assay was 8.1% at a mean level of 61.4 ± 5.0 nmol/l. Levels of 25(OH)D were reported in nmol/l. Vitamin D categories were defined as sufficient (>75 nmol/l), insufficient (50–75 nmol/l) or deficient (<50 nmol/l).

Autoantibodies to insulin, GAD, islet antigen-2 and the islet-specific zinc transporter isoform 8 (ZnT8) were measured using radiolabel-binding assays, as previously described [20, 25], with the upper limit of normal for each assay corresponding to the 99th percentile of control individuals [2628]. Children were considered islet autoantibody positive when two consecutive samples tested positive for these antibodies.

Statistical analysis

The 25(OH)D levels were compared among children with multiple islet autoantibodies, autoantibody-negative children and children with type 1 diabetes using the Mann–Whitney U-test for comparisons between two groups. Linear regression was used to adjust for age, study type and the season in which the blood samples were collected as possible confounders. The χ2 test was used to compare the distribution of 25(OH)D categories (i.e. deficient, insufficient and sufficient) among the three groups of children. In a sensitivity analysis, the main analyses were repeated using the matched control group of autoantibody-negative children with a smaller sample size.

Among children with multiple islet autoantibodies, the Kaplan–Meier method was used to compare progression to type 1 diabetes between different 25(OH)D categories. The follow-up time was calculated from the age when 25(OH)D levels were measured to the age at type 1 diabetes diagnosis or the last contact. A Cox proportional hazard model was used to calculate the HR for the risk of progression to diabetes according to 25(OH)D levels and vitamin D deficiency, with adjustment for potential confounding variables (age, seasonal variation of vitamin D levels and study type). Based on an expected progression rate of 50% within 10 years [2], a study with the number of islet autoantibody-positive individuals that were available for our study has 60% power to detect a 50% increase in the risk of progression in children with the lowest tertile of 25(OH)D levels, approximately corresponding to <50 nmol/l in our cohort.

For all analyses, two-tailed p values of <0.05 were considered significant. Statistical analyses were performed using the Statistical Package for the Social Sciences (SPSS 21.0; SPSS, Chicago, IL, USA).

Results

Prevalence of vitamin D deficiency and 25(OH)D levels in pre-type 1 diabetes

The prevalence of vitamin D deficiency (<50 nmol/l) was significantly greater in children with multiple islet autoantibodies than in islet autoantibody-negative children (39.8% vs 28.3%; p = 0.021). Vitamin D deficiency was even more pronounced in children with newly diagnosed type 1 diabetes (51.2%; p = 0.048 vs children with multiple islet autoantibodies; p < 0.001 vs islet autoantibody-negative children; Table 1).
Table 1

Prevalence of vitamin D deficiency

Vitamin D status

Children with multiple islet autoantibodies

Children with no islet autoantibodies

Children with newly diagnosed T1D

n

108

406

244

Deficient (<50 nmol/l)a

43 (39.8)

115 (28.3)

125 (51.2)

Insufficient (50–75 nmol/l)

33 (30.6)

101 (24.9)

70 (28.7)

Sufficient (>75 nmol/l)

32 (29.6)

190 (46.8)

49 (20.1)

Values are n (%)

aχ2 tests were used to compare the distribution of vitamin D deficiency between islet autoantibody-positive children vs children with no islet autoantibodies (p = 0.021), islet autoantibody-positive children vs children with type 1 diabetes (p = 0.048) and children with no islet autoantibodies vs children with type 1 diabetes (p < 0.001)

T1D, type 1 diabetes

Children with multiple islet autoantibodies had significantly lower 25(OH)D levels than islet autoantibody-negative children (59.9 ± 3.0 vs 71.9 ± 1.5 nmol/l; p < 0.001; Fig. 2). By comparison, the mean 25(OH)D level was 49.7 ± 1.5 nmol/l in children with newly diagnosed type 1 diabetes (p = 0.006 vs children with multiple islet autoantibodies; p < 0.001 vs islet autoantibody-negative children; Fig. 2). The differences among the three groups were statistically significant after adjusting for age, season and study type. In addition, the prevalence of vitamin D deficiency was significantly higher in children with multiple islet autoantibodies (p < 0.001) and in children with newly diagnosed type 1 diabetes (p < 0.001) compared with islet autoantibody-negative children. Furthermore, vitamin D levels were lower in children with multiple islet autoantibodies (p = 0.001) and children with newly diagnosed type 1 diabetes (p < 0.001) compared with islet autoantibody-negative children.
https://static-content.springer.com/image/art%3A10.1007%2Fs00125-014-3181-4/MediaObjects/125_2014_3181_Fig2_HTML.gif
Fig. 2

Comparison of 25(OH)D levels in islet autoantibody negative children (Ab) (p < 0.001 vs Ab+), children with multiple islet autoantibodies (Ab+) (p = 0.006 vs type 1 diabetes) and children with newly diagnosed type 1 diabetes (p < 0.001 vs Ab). The solid line indicates the mean 25(OH)D level. The number of children in each group is shown below the x-axis; Ab, autoantibody negative; Ab+, multiple islet autoantibody positive; T1D, type 1 diabetes

Because vitamin D levels show seasonal variations [6], we compared 25(OH)D levels according to the month in which the blood samples were collected (Fig. 3). Children with multiple islet autoantibodies had significantly lower 25(OH)D levels than islet autoantibody-negative children during months with the longest sunshine duration (May p = 0.002, June p = 0.011, July p = 0.001, September p = 0.005; Fig. 3). However, there were no differences in 25(OH)D levels during the winter months. Children with type 1 diabetes had low 25(OH)D levels throughout the year.
https://static-content.springer.com/image/art%3A10.1007%2Fs00125-014-3181-4/MediaObjects/125_2014_3181_Fig3_HTML.gif
Fig. 3

Levels of 25(OH)D according to the month of blood sampling in multiple islet autoantibody positive children (red line), islet autoantibody negative children (grey line) and in children with newly diagnosed type 1 diabetes (black line). Levels of 25(OH)D are expressed as the mean. The dotted lines show SEM. *Significantly different (p < 0.05) between Ab+ and Ab children; significantly different (p < 0.05) between newly diagnosed T1D and Ab children; Ab, autoantibody negative; Ab+, multiple islet autoantibody positive; T1D, type 1 diabetes

Similar results were obtained in a sensitivity analysis when children with multiple islet autoantibodies were compared with the matched control group of autoantibody-negative children (data not shown).

Effects of vitamin D deficiency and 25(OH)D levels on the rate of progression to type 1 diabetes in children with multiple islet autoantibodies

The cumulative risk of progression to type 1 diabetes in children with multiple islet autoantibodies was 37.9% (95% CI 27.8, 48.0) and 57.5% (95% CI 45.3, 69.7) 5 and 10 years from the time vitamin D was measured, respectively. The rates of progression at 5 and 10 years were 38.3% (95% CI 21.5, 55.1) and 51.8% (95% CI 29.3, 74.3), respectively, in children with vitamin D deficiency. The respective values in children with insufficient vitamin D levels were 38.9% (95% CI 20.6, 57.2) and 64.0% (95% CI 43.7, 84.3), while those in children with sufficient vitamin D levels were 36.8% (95% CI 19.3, 54.3) and 55.4% (95% CI 35.5, 72.3; Fig. 4). The rates of progression at 5 and 10 years were not significantly different between the three groups of children.
https://static-content.springer.com/image/art%3A10.1007%2Fs00125-014-3181-4/MediaObjects/125_2014_3181_Fig4_HTML.gif
Fig. 4

Cumulative risk of progression to type 1 diabetes in multiple islet autoantibody positive children with sufficient (>75 nmol/l; solid line), insufficient (50–75 nmol/l; dashed line) or deficient (<50 nmol/ml; dotted line) 25(OH)D levels. The follow-up time (x-axis) was calculated from the age when 25(OH)D was measured. The number of children included in each follow-up is shown below the x-axis; T1D, type 1 diabetes

There was no significant association found between 25(OH)D levels (adjusted HR 1.00 [95% CI 0.98, 1.03]; p = 0.7) or vitamin D deficiency (adjusted HR 0.68 [95% CI 0.35, 1.31] p = 0.2) and progression to type 1 diabetes in children with multiple islet autoantibodies after adjusting for age, season and study type.

Discussion

Our study demonstrated that vitamin D levels were lower in children with multiple islet autoantibodies and in children with newly diagnosed type 1 diabetes than in islet autoantibody-negative children. However, vitamin D deficiency was not associated with faster progression to type 1 diabetes in children with multiple islet autoantibodies. These findings imply that vitamin D deficiency precedes the onset of type 1 diabetes and is unlikely to be a consequence of metabolic decompensation in patients. Nevertheless, vitamin D deficiency does not appear to have an important role in determining the rate and evolution of the ongoing autoimmune process underlying type 1 diabetes.

Interestingly, vitamin D levels in children with multiple islet autoantibodies were markedly lower during months with the longest duration of sunshine and showed less seasonal variability compared with the levels in autoantibody-negative children. This is consistent with our previous study, which showed that the decrease in vitamin D levels in patients with newly diagnosed type 1 diabetes mainly occurred during summer and was due to a loss of the seasonal rhythm of vitamin D observed in healthy controls [6]. Vitamin D is synthesised in the skin in response to ultraviolet B light. It is possible that reduced ultraviolet light exposure due to higher sensitivity and low skin pigmentation (as previously reported, [10]) endorses a subclinical alteration in vitamin D metabolism that becomes more noticeable during summer in susceptible children.

Our study has several strengths, including the enrolment of a large number of children at high risk of progressing to diabetes based on the presence of multiple islet autoantibodies, measurement of vitamin D levels close to the time of islet autoantibody seroconversion, the prospective study design with a long follow-up and measurement of 25(OH)D at a central laboratory using a standardised, validated method. A limitation of our study is that several children who developed diabetes rapidly after seroconversion were excluded because of sample unavailability, and we do not know whether vitamin D deficiency was more prevalent in these cases. In addition, children with islet autoantibodies were slightly younger than those without islet autoantibodies. Findings were similar, however, when age-matched islet autoantibody-negative children were used as a control group. Moreover, vitamin D levels are usually higher in younger children [18, 29], suggesting that lower age of islet autoantibody-positive children may have led to an underestimation of the difference between islet autoantibody-positive and -negative children. Further limitations were the lack of dietary data on vitamin D supplementation and the lack of samples to measure vitamin D levels before islet autoantibody seroconversion. The latter would have allowed us to determine whether or not vitamin D deficiency precedes islet autoimmunity or develops after autoantibody seroconversion and immunoactivation.

Our results are consistent with those of the Diabetes Autoimmunity Study in the Young (DAISY), which also revealed that vitamin D levels measured in childhood were not associated with progression to type 1 diabetes [18]. The mechanisms leading to lower vitamin D levels in children with islet autoimmunity are unclear. We did not measure vitamin D concentrations prior to islet autoimmunity and, therefore, cannot assess whether or not the relative vitamin D deficiency we observed may have preceded the development of islet autoantibodies. In DAISY vitamin D levels were measured before the onset of islet autoimmunity but there was no evidence that vitamin D concentrations influenced the likelihood of developing islet autoimmunity. Nevertheless, a recent analysis in DAISY revealed that single nucleotide polymorphisms of vitamin D metabolism-related genes are associated with plasma vitamin D levels (DHCR7, GC and CYP2R1) and islet autoimmunity (DHCR7) [30]. These findings suggest that vitamin D status may be genetically determined, and that a combination of genetic susceptibility and reduced exposure to ultraviolet light might lead to vitamin D deficiency in such children. Furthermore, it is possible that vitamin D shortage may be a consequence of islet autoimmunity, reflecting changes in metabolism or increased vitamin D expenditure. Indirect support for this theory is provided by the observation that children with type 1 diabetes in our study had more severe vitamin D deficiency and decreased vitamin D levels throughout most of the year.

In conclusion, the results of our study suggest that vitamin D deficiency is more prevalent in children with islet autoantibodies but is not associated with the rate of progression to type 1 diabetes. Assessment of vitamin D status and eventual supplementation may be indicated for children with islet autoantibodies.

Acknowledgements

We thank C. Ried, A. Wosch, L. Lachmann, F. Haupt, M. Scholz, A. Gavrisan, A. Knopff, S. Krause, C. Matzke, V. Dietrich and J. Stock for data collection and expert technical assistance, R. Chmiel and A. Huppert for clinical care of the children, R. Puff for laboratory management and A. Beyerlein for statistical assistance (all from Institute of Diabetes Research, Helmholtz Zentrum München, and Forschergruppe Diabetes, Klinikum rechts der Isar, Technische Universität München, Neuherberg, Germany). We also thank all of the children, paediatricians and family doctors for participating in the BABYDIAB, BABYDIET and TEENDIAB studies and all of the clinics who enrolled children into DiMelli.

Funding

The work was supported by grants from the Kompetenznetz Diabetes mellitus (Competence Network for Diabetes mellitus), funded by the Federal Ministry of Education and Research (FKZ 01GI0805-07, FKZ 01GI0805), from Deutsche Forschungsgemeinschaft (DFG ZI-310/14-1 to -4) and the Juvenile Diabetes Research Fund (JDRF-No 17-2012-16).

Duality of interest

The authors declare that there is no duality of interest associated with this manuscript.

Contribution statement

JR and CW collected and reviewed the data, performed statistical analyses, interpreted the results and drafted the manuscript. SS assisted with follow-up assessments and data collection, and critically reviewed the manuscript for intellectual content. KW, EZG and MK contributed to statistical analyses, interpreted the results, and critically reviewed the manuscript for intellectual content. A-GZ is the principal investigator of the BABYDIAB, BABYDIET, TEENDIAB and DiMelli studies, designed the studies, wrote the manuscript and critically reviewed it for intellectual content. A-GZ had primary responsibility for the final content of the manuscript. All authors read and approved the final version.

Supplementary material

125_2014_3181_MOESM1_ESM.pdf (21 kb)
ESM Table 1(PDF 20 kb)

Copyright information

© Springer-Verlag Berlin Heidelberg 2014