Pappa, Helen M.*; Langereis, Eveline J.†; Grand, Richard J.*; Gordon, Catherine M.‡
*Center for Inflammatory Bowel Disease, Children's Hospital Boston, Boston, MA
†University of Amsterdam School of Medicine, Amsterdam, The Netherlands
‡Bone Health Program, Children's Hospital Boston, Boston, MA.
Address correspondence and reprint requests to Dr Helen M. Pappa, Division of Gastroenerology and Nutrition, Children's Hospital Boston, Hunnewell Ground, 300 Longwood Ave, Boston, MA 02115 (e-mail: Helen.firstname.lastname@example.org).
Received 22 November, 2010
Accepted 18 May, 2011
The authors report no conflicts of interest.
Vitamin D plays an important role in bone metabolism and serves as an immunomodulator (1). Children with inflammatory bowel disease (IBD) have decreased bone mineral density in comparison with healthy peers. The role of vitamin D in improving and protecting bone health or improving disease outcomes in children with IBD has not been established; moreover, the lowest vitamin D level that would promote bone health and regulate the immune system has not been defined. Based on rickets prevention, a serum 25-hydroxyvitamin D (25OHD) concentration >20 ng/mL has been defined as the level of sufficiency in children (2). Maintaining sufficient vitamin D levels in children with IBD may be difficult because of malabsorption and other disease-related factors. Direct comparison of the prevalence of vitamin D insufficiency between children with IBD and their healthy peers has not been reported, although a recent study reported lower vitamin D levels in children with newly diagnosed IBD in comparison with age-, sex- and ethnicity-matched controls (3). We reported a prevalence of serum 25OHD concentration ≤15 ng/mL in 34% of children with IBD in our center (4). We report the prevalence and risk factors of hypovitaminosis D in children with IBD in a study of this population.
PATIENTS AND METHODS
Serum 25OHD concentration of patients with IBD, 8 to 22 years, cared for at the Children's Hospital Boston Division of Gastroenterology, was measured as part of their clinical care and these measurements were used for the screening phase of interventional prospective studies in this population involving vitamin D supplementation. The Children's Hospital institutional review board approved all of the phases of these studies including retrospective chart review. We retrospectively recorded all patients’ 25OHD serum concentration measured between January 1, 2006 and January 1, 2009. Through a retrospective chart review, we collected information on age; sex; ethnicity; diagnosis of Crohn disease (CD), ulcerative colitis (UC), and indeterminate colitis (IC); and vitamin D supplementation in the patients whose 25OHD serum concentration was measured. Serum albumin concentration (grams per decilitre), erythrocyte sedimentation rate (ESR, millimetres per second), and patients’ height and weight were also recorded if measured within 1 month of the serum 25OHD concentration measurement. In the case of ESR, measurements were not recorded if through chart review there was sufficient evidence that an elevated value was caused by a concurrent illness unrelated to IBD (ie, upper respiratory infection). For statistical purposes, ethnicity was reported as white or nonwhite, with African American and Hispanic as nonwhite. Data on serum albumin concentration and ESR were selectively collected because of previous results from our group linking serum albumin concentration with serum 25OHD concentrations (4). Weight (kilograms) and height (centimeters) were recorded as reflecting patients’ nutritional status. Patients were considered as taking vitamin D supplementation if the clinician's report on the day of the serum 25OHD concentration measurement notes regular intake of any supplement containing vitamin D. Serum 25OHD concentration was measured in nanograms per milliliter with the DiaSorin (Stillwater, MN) Liaison assay, a direct competitive immunochemiluminescent assay that accurately measures both 25OHD2 and 25OHD3(5). ESR and albumin were measured using the Excyte-40 automated ESR analyzer (Vital Diagnostics, Lincoln, RI) and the Cobas 6000 chemistry analyzer (Roche Diagnostics, Basel, Switzerland), respectively. Height and weight were measured using a Harpenden stadiometer (Veeder-Root counter) (Holtain Ltd, Crymych, UK), and a Scaletronix scale (accuracy 100 g) (White Plains, NY), respectively. Body mass index (BMI), as well as weight, height, and BMI z scores, was calculated using the EpiInfo nutrition program, version 3.5.3, with the Centers for Disease Control and Prevention 2000 reference. Respective z scores for patients older than 20 years were calculated using an age of 20 years.
Serum 25OHD concentration was classified as optimal (>32 ng/mL), suboptimal (≤32 ng/mL), insufficient (≤20 ng/mL), or deficient (≤15 ng/mL) (1,6).
Statistical analysis was performed using SPSS version 15.0 (SPSS Inc, Chicago, IL). To evaluate differences between the 2 major diagnosis groups (CD and UC), a χ2 test, an independent sample t test, or a Mann-Whitney test was performed, depending on the nature and distribution of the data. Continuous variables were tested for normality using the Shapiro-Wilk test. If normality was not fulfilled, then nonparametric tests were used. Data for patients with IC were not reported separately because of their small number. The influence of various factors on serum 25OHD concentration was analyzed through simple regression analysis. For the latter, the dependent variable was the serum 25OHD concentration, and the independent variables were ethnicity, season of measurement as (a dichotomous value winter and spring vs summer and fall), height, weight, BMI z score, diagnosis, vitamin D supplementation (when available) and ESR, and serum albumin concentration if measured within 1 month of serum 25OHD concentration measurement. Because serum 25OHD concentration was not normally distributed, the values were log-transformed. We then applied multiple regression analysis to examine the effect of diagnosis, ESR, and albumin on serum 25OHD concentration, independently of confounding by other factors in this cohort. BMI z score, ethnicity, season of measurement, and vitamin D intake were entered in the regression model as known factors that independently affect serum 25OHD concentration in healthy individuals. Serum albumin concentration, ESR, and diagnosis were then entered one by one and their regression coefficient and significance of their relation to serum 25OHD concentration was tabulated after this adjustment. The difference in log-transformed serum 25OHD was converted to difference in percentage units using the following formula: 100% × [10changeinlogserum25OHDconcentration − 1] = percent change. P ≤ 0.05 was set to indicate statistical significance.
A total of 448 individual patients had their serum 25OHD concentration measured at least once during the specified study interval. In this group, 288 patients (64.3%) carried the diagnosis of CD, 143 (31.9%) that of UC, and 17 (3.8%) that of IC. Further characteristics are shown in Table 1. Of note is that more patients with CD had their serum 25OHD concentration measured during the summer and fall months than during the winter and spring, whereas patients with UC had their serum 25OHD concentration measured evenly throughout the year. Mean weight, height, and BMI z score were lower, and mean ESR was higher in patients with CD. Patients with UC had lower mean serum 25OHD concentrations than patients with CD, and vitamin D insufficiency was more prevalent among patients with UC (Table 1).
In simple regression, serum 25OHD concentration was found to be 10.9% higher in summer and fall than in winter and spring (confidence interval 3.6–17.4, P = 0.004), 22% higher in “white” patients than in patients with darker skin complexion (confidence interval 13.9–29.6, P < 0.001), 18.9% higher among patients who were taking vitamin D supplements (confidence interval 6.7–32.7, P = 0.002), 10.9% higher for every grams per deciliter increase in serum albumin concentration (confidence interval 1.6–21.1, P = 0.02), 3.4% lower for every 1 standard deviation increase in weight z score (confidence interval −7.1 to −0.5, P = 0.03), and 11.3% lower in patients with UC than in those with CD (confidence interval −18.3 to −3.8, P = 0.004).
After adjustment for ethnicity, season of measurement, BMI z score, and vitamin D intake, higher ESR was independently significantly associated with lower serum 25OHD (3.8% lower for every 10 mm/s of rise in ESR, confidence interval −7.1 to −0.5, P = 0.03). The significance level of serum albumin was 0.06 after adjustment (serum 25OHD concentration 13.2% higher for every grams per deciliter increase in serum albumin concentration, confidence interval −0.7 to 28.8), and diagnosis lost its significance as a predictor of serum 25OHD level after adjustment (serum 25OHD concentration 8.2% lower in UC than in patients with CD, confidence interval −17.8 to 2.3, P = 0.12).
DISCUSSION AND CONCLUSIONS
In this sample from a tertiary care center in Boston, MA (42° N), 58.3% of pediatric patients with IBD had suboptimal serum 25OHD concentration (25OHD ≤32 ng/mL), 14.3% had serum 25OHD level at or below 20 ng/mL, and 5.8% had serum 25OHD level at or below 15 ng/mL. Although the prevalence of vitamin D insufficiency (25OHD <20 ng/mL) in our study does not appear to be higher than that encountered among healthy children, even at lower latitudes in the United States. (7), the racial distribution differences among the reporting studies do not allow for direct comparisons and generalizations. As expected, serum 25OHD concentration showed a strong relation to skin complexion. Kumar et al (8) found a prevalence of serum 25OHD concentration <15 ng/mL in 9% of healthy children throughout the United States, which is higher than that encountered in our study population. This discrepancy can be explained by the fact that the percentage of children with darker skin complexion is higher in the group of healthy children examined in the study above than in our study (about 32% vs 16.7% in our study population), reflecting the racial distribution of the diagnosis of IBD in the general population. Another study performed by Gordon et al (9) among healthy adolescents in the Boston area showed a prevalence of serum 25OHD concentration <15 ng/mL of 24%. In that study of healthy youth, white subjects comprised only 16% of the sample.
In a previous study conducted in our center, we reported serum 25OHD concentration <15 ng/mL in 34% of children with IBD, with racial distribution similar to that of the present study (4). This significant difference could be explained by the difference in assays used for measurement of serum 25OHD concentration in the 2 studies: In our previous study, we used the Nichols competitive binding assay, whereas in our present study, we used the DiaSorin Liaison assay. The later measures accurately both vitamin D2 and D3, whereas the former may not detect vitamin D2 levels as accurately, thus underestimating its contribution to the serum 25OHD concentration. Of note is that the percentage of patients taking vitamin D supplementation was actually higher in our previous study compared with the present study (77% vs 57%).
BMI z score was negatively associated with serum 25OHD level in our study. This finding is in agreement with findings in healthy children (10). Sequestration of vitamin D in adipose tissue and its resulting decreased bioavailability in other tissues is a leading hypothesis for this observation. Doses of vitamin D for both treatment of vitamin D deficiency and supplementation for the maintenance of optimal vitamin D stores may have to be higher, to account for increased adipose tissue in both healthy children and children with IBD; however, the threshold above which adipose tissue interferes with vitamin D bioavailability and appropriate dosing of vitamin D to overcome this have not been defined.
ESR is used as a surrogate marker of intestinal inflammation in IBD. It was found to be independently associated with a lower 25OHD serum level in both our present and previous studies (4). Intestinal inflammation could theoretically affect vitamin D status through impaired absorption and/or loss of circulating protein-bound vitamin D through the inflamed intestinal mucosa. Serum albumin concentration is a surrogate marker of protein-losing enteropathy and of nutritional status. The association of serum albumin concentration with serum 25OHD concentration did not reach significance in our study when adjusted for other factors (P = 0.06), including nutritional status representatives such as BMI z score, but it may have if the number of patients had been larger. Clearly, further study is needed to define the mechanism through which intestinal inflammation affects vitamin D homeostasis.
Limitations of the present study include its retrospective design, the lack of comparable healthy controls, and the nonsystematic collection of data such as vitamin D intake in all of the patients. In addition, data related to disease severity and course were limited in the present study.
Vitamin D may play an important role in bone health and control of inflammation in children with IBD; however, direct comparison of the prevalence of hypovitaminosis D in this population and in healthy children with similar racial distribution has not been undertaken. Our present study highlights the fact that hypovitaminosis D is prevalent among children with IBD. Risk factors for hypovitaminosis D in this population are similar to those identified in healthy children: darker skin complexion, winter season, higher BMI z score reflecting increased adipose tissue mass, and lack of vitamin D supplementation. In addition, higher ESR is associated with lower vitamin D levels in this population. The decision to screen vitamin D status in children with IBD should not depend on diagnosis because both children with CD or UC appear to be at risk for this problem. Further studies are needed to directly compare healthy children with children with IBD in terms of their vitamin D status, uncover the effect of intestinal inflammation on vitamin D homeostasis, quantify the effect of adipose tissue on vitamin D bioavailability, and define appropriate dosing of vitamin D supplementation based on this effect.
1. DeLuca HF. Overview of general physiologic features and functions of vitamin D. Am J Clin Nutr 2004; 80:1689S–1696S.
2. Misra M, Pacaud D, Petryk A, et al. Vitamin D deficiency in children and its management: review of current knowledge and recommendations. Pediatrics 2008; 122:398–417.
3. El-Matary W, Sikora S, Spady D. Bone mineral density, vitamin D, and disease activity in children newly diagnosed with inflammatory bowel disease. Dig Dis Sci 2011; 56:825–829.
4. Pappa HM, Gordon CM, Saslowsky TM, et al. Vitamin D status in children and young adults with inflammatory bowel disease. Pediatrics 2006; 118:1950–1961.
5. Wagner D, Hanwell HE, Vieth R. An evaluation of automated methods for measurement of serum 25-hydroxyvitamin D. Clin Biochem 2009; 42:1549–1556.
6. Holick MF. Vitamin D deficiency. N Engl J Med 2007; 357:266–281.
7. Rovner AJ, O’Brien KO. Hypovitaminosis D among healthy children in the United States: a review of the current evidence. Arch Pediatr Adolesc Med 2008; 162:513–519.
8. Kumar J, Muntner P, Kaskel FJ, et al. Prevalence and associations of 25-hydroxyvitamin D deficiency in US children: NHANES 2001–2004. Pediatrics 2009; 124:e362–e370.
9. Gordon CM, DePeter KC, Feldman HA, et al. Prevalence of vitamin D deficiency among healthy adolescents. Arch Pediatr Adolesc Med 2004; 158:531–537.
10. Dong Y, Pollock N, Stallmann-Jorgensen IS, et al. Low 25-hydroxyvitamin D levels in adolescents: race, season, adiposity, physical activity, and fitness. Pediatrics 2010;125:1104–11.