The frequency of vitamin D deficiency in adults with Crohn’s disease

1Division of Gastroenterology and Division of Endocrinology, Department of Medicine, University of Alberta, Edmonton, Alberta; 2Division of Gastroenterology, University of Toronto, Toronto, Ontario Correspondence: Dr RN Fedorak, Division of Gastroenterology, University of Alberta, 519 Robert Newton Research Building, Edmonton, Alberta T6G 2C1. Telephone 780-492-6941, fax 780-492-8121, e-mail richard.fedorak@ualberta.ca Received for publication October 31, 2002. Accepted April 28, 2003 JS Siffledeen, K Siminoski, H Steinhart, G Greenberg, RN Fedorak. The frequency of vitamin D deficiency in adults with Crohn’s disease. Can J Gastroenterol 2003;17(8): 473-478.

25-OHD serves as the major precursor and most objective measure of the active hormone 1,25-dihydroxyvitamin D 3 (1,25-OHD) undergoing 1-alpha hydroxylation in the kidney.1,25-OHD increases calcium and phosphate absorption and reabsorption in the gastrointestinal tract and kidneys, respectively, thus maintaining normal mineralization of newly formed bone (18,22).A low serum concentration of 1,25-OHD delays osteoid deposition by osteoblasts, increases parathyroid hormone (PTH) release and activity and results in increased bone resorption.Prolonged vitamin D deficiency results in osteomalacia in adults (22).
The main source of vitamin D in healthy individuals is cutaneous ultraviolet irradiation of 7-dehydrocholesterol, with a minor contribution from the intake of foods containing, or fortified with, vitamin D (22,23).Serum vitamin D deficiency is rarely, if ever, reported in healthy North American adults (24,25).Serum 25-OHD levels vary seasonally, reaching the highest levels in the summer months, and are lowest in the late winter months (12,(22)(23)(24)(25)(26).The absence of vitamin D deficiency in North American adults in the winter is most likely due to the ingestion of foods fortified with vitamin D (22,25).When ingested, vitamin D is absorbed mainly at the jejunum of the small intestine (27).Furthermore, enterohepatic recirculation at the ileum normally prevents excessive excretion of vitamin D (12,23).Any, or all, of these sources of vitamin D may be affected in Crohn's disease.
The aim of the present study is to determine, in a Canadian cohort of adult patients with Crohn's disease, the frequency of low serum 25-OHD levels, the associated risk of low bone mineral density and to identify the causes of lowered vitamin D in these patients.

Patients
Between 1998 and 2000, 224 patients with Crohn's disease who had attended the Inflammatory Bowel Disease Research Centre and Clinic, at the University of Alberta Hospital, Edmonton, Alberta, were consecutively enrolled in a prospective longitudinal investigation to assess the clinical efficacy of bisphosponate therapy on bone mineral density in Crohn's disease patients.An additional 18 patients from Mount Sinai Hospital, Toronto, Ontario, were also enrolled, giving a total of 242 patients.Data relevant to vitamin D were gathered at baseline assessment.Informed consent from each patient was received in writing before being enrolled.Crohn's disease and the site of the disease was diagnosed on the basis of endoscopic, radiological and histological examination.Those patients with a serum 25-OHD concentration below 25 nmol/L were classified as being vitamin D deficient and were separately analyzed for the purposes of this study.
The following exclusion criteria were applied at baseline: age less than 18 years; patients with known bone disorders other than osteoporosis (such as hyperparathyroidism, Paget's disease, renal osteodystrophy and documented osteomalacia); patients with abnormal thyroid function; patients with significant renal impairment (serum creatinine twice the normal level); patients with clinical short bowel syndrome; patients on total parenteral or enteral nutrition; and patients with spinal anatomy that did not allow adequate assessment with dual energy x-ray absorptiometry (DEXA, Hologic 4500, Hologic Inc, USA).In addition, patients who had received bisphosphonate or fluoride supplement in the 24 months before the data collection or pharmacological doses of calcium (greater than 1.0 g/day) or vitamin D (greater than 800 IU/day) in the six months before entry were excluded from the study.
Patient demographics and disease information, such as age at baseline, diagnosis date, gender, smoking status, Crohn's disease activity in the last year and corticosteroid use in the last year, were obtained through a questionnaire conducted at baseline.Other baseline information was gathered by reviewing patients' records, and by blood, urine and DEXA analysis.Weight and height were obtained to determine body mass index (BMI) (in kg/m 2 ), and disease site was recorded as colonic, ileal, ileocolonic or jejunal/ duodenal.

BMD measurements
BMD (g/cm 2 ) of the lumbar spine (L1-L4), the femoral neck, total hip and ultradistal radius was measured at baseline by DEXA using standard protocols.Vertebral BMD values were normalized using the Hologic reference database; hip values were normalized using the National Health And Nutrition Examination Study III database (28).Osteopenia (T-values greater than -1.0 but less than or equal to -2.5) and osteoporosis (T-values less than -2.5) were diagnosed using the lowest value of the lumbar vertebrae (L1-L4 inclusive, unless there was a specific reason to exclude a given vertebra), total hip density or femoral neck density (29).

Statistical analysis
Descriptive statistics were reported as a mean ± SD for continuous variables that were normally distributed, unless otherwise stated.Results for the group of Crohn's disease patients with serum 25-OHD deficiency were compared with those of the patient group with normal levels of 25-OHD.For comparison of means of continuous variables between two groups, an unpaired t-test was performed.If the data was not normally distributed the Mann-Whitney rank test was employed.Comparison of discontinuous variables was performed using a χ 2 test.
To determine if any continuous variables could predict vitamin D status in the Crohn's disease patients, a univariate analysis of the data from the 242 patients was first performed using Pearson's correlation test.All significantly associated variables were then subjected to forward stepwise regression analysis to develop a multivariate model.
For all tests, significance was defined as P<0.05 (95% CI), with normality and variance defined at P<0.01.The statistical program Sigma Stat (V.2.03, SPSS Inc, USA 1998) was used for all statistical procedures.

Serum 25-OHD deficiency
Demographic characteristics of the patients with normal serum 25-OHD and those with serum 25-OHD deficiency are given in Table 1.Of the 242 consecutive patients assessed at baseline, 19 (8%) were found to have a serum 25-OHD of less than 25 nmol/L.Of those patients with 25-OHD deficiency, 11 (57.9%) were osteopenic, one (5.3%)was osteoporotic and seven (36.8%) had normal BMD.These BMD results did not differ significantly from the prevalence of low BMD in the non-25-OHD deficient Crohn's disease cohort (48.6% osteopenia and 14.2% osteoporosis, P>0.05).In the 25-OHD deficient patients, decreased bone mineral density was most pronounced at the femoral neck and lumbar spine (47.4% and 42.1%, respectively).This distribution of decreased bone mineral density did not differ from that of patients with normal levels of 25-OHD (51.9% femoral neck and 46.1% lumbar spine, P>0.05).Age at baseline, disease duration, BMI, corticosteroid use, and disease location did not differ between the deficient and normal 25-OHD group.There were a greater percentage of smokers in the 25-OHD deficient group than in the normal group (62.5% and 35.6%, respectively); however, this difference did not show statistical significance (χ 2 =5.929,P=0.115).

Serum 25-OHD insufficiency
Vitamin D insufficiency has been defined conservatively as a serum 25-OHD level below 40 nmol/L.Long term exposure to 25-OHD at this level may also be of clinical importance, although the effect on BMD has not been determined.Fifty-two patients (22%) demonstrated serum 25-OHD insufficiency (levels below 40 nmol/L).
Biochemical characteristics for the 25-OHD-insufficient group were similar to those in the 25-OHD-deficient group (Table 3).

BMD
The mean T-scores at the lumbar spine, femoral neck, total hip and ultradistal radius for the normal 25-OHD and 25-OHD-deficient groups is shown in Table 2.No significant differences in the T-scores at any of these skeletal sites were found.In the 25-OHD-deficient group, age at diagnosis was the only variable to correlate, negatively, with low T-scores, and only at the lumbar spine and femoral neck (-0.587 and -0.625, respectively, P<0.05) (Figure 1).Furthermore, when   When forward regression analysis was performed, only RBC folate predicted serum 25-OHD (r=0.325,P<0.05) in the male Crohn's disease population.In the female Crohn's disease population, smoking, RBC folate and serum iron predicted serum 25-OHD (r=0.418,P<0.05).PTH was excluded from this analysis because of its known relationship to vitamin D.
Finally, only two of the 19 patients with low serum 25-OHD had elevated serum PTH levels above the upper limit of normal (greater than 6.8 pmol/L) at baseline.Both patients reported being smokers and one of these patients had a BMI of 17.6 kg/m 2 .In both cases, serum alkaline phosphatase levels were within normal range (30 to 130 IU/L).

Seasonal variation of 25-OHD levels
Table 4 shows the number of patients with deficient and normal 25-OHD levels, relative to the seasonal time of measurement (ie, winter, October to March, and summer, April to September).Sixteen (84.2%) of the 19 patients with deficient 25-OHD levels were identified during the winter.When the frequency of 25-OHD deficiency in patients during summer and winter periods was compared, patients in the winter period exhibited a significantly higher frequency of vitamin D deficiency than did patients in the summer period (11.9% versus 2.8%, respectively, P=0.02).

DISCUSSION
This study, comprising 242 patients, demonstrates that 8% of adult patients with Crohn's disease exhibit deficient serum levels of 25-OHD.This finding is highly significant, considering that deficient levels of 25-OHD are almost never reported in the healthy North American population (24,25).Nevertheless, the frequency of 25-OHD deficiency in our cohort is lower than the previously reported range of 17% to 68% (3,8,12,(15)(16)(17)(18)(19)(20).This difference may be explained by the fact that some of the previous studies used heterogeneous inflammatory bowel disease populations, others focused on small subsets of Crohn's patients with small bowel involvement, and most were characterized by small study populations.Furthermore, global variation in exposure to sunlight and dietary intake and sources of vitamin D could account for the difference in the reported incidence of serum 25-OHD deficiency between this current study and others that have been reported from Norway (19), Finland (15), Austria (12,17), Denmark (20), Netherlands (3), Germany (8), Wales (18) and Chicago, USA (16).Edmonton, the site of the present research, receives approximately 728 h of sunlight between October and March, and approximately 1577 h of sunlight between April and September, a variation sufficient to alter serum 25-OHD levels between seasons (33).
Many experts in the area of vitamin D research also examine the issue of 25-OHD insufficiency, defined conservatively as a serum 25-OHD below 40 nmol/L.In our study, 52 patients (22%) with Crohn's disease demonstrated serum 25-OHD insufficiency (levels below 40 nmol/L) at the time of analysis.Similar to the 25-OHD-deficient group, the 25-OHD-insufficient group demonstrated no significant differences in T-scores at the lumbar spine, femoral neck, total hip or ultradistal radius, and continued to demonstrate elevated levels of alkaline phosphotase and PTH, but not N-teleopeptide (Table 3).Nevertheless, patients with 25-OHD below 25 nmol/L demonstrated even higher levels of PTH and alkaline phosphatase.
It is conceivable that our results represent an under-reporting of the incidence of 25-OHD-deficiency.Our study observed that the majority of patients exhibiting deficient serum levels had their 25-OHD levels assessed during the winter months (84.1%), yet the total distribution of measurements was equal between the winter and summer months.We would have expected a higher incidence of serum 25-OHD deficiency if all measurements were carried out within a specified time period during the winter months.In fact, when only those patients who had been assessed for vitamin D status during the winter months were considered, almost 12% demonstrated 25-OHD deficiency.Our results further demonstrated that there was a four-fold higher probability of 25-OHD deficiency if patients were assessed in the winter months, rather than in the summer.This would concur with the results of Vogelsang et al (12), who observed that 25-OHD levels in Crohn's disease patients were higher in the summer than in the winter and that of the patients with low 25-OHD levels in the winter, less than half continued to exhibit low levels in the summer.
Our study confirms previous reports that serum levels of 25-OHD do not correlate with, or predict, bone mineral status in patients with Crohn's disease (1,(3)(4)(5)(6)(7)15,19).Furthermore, while 63% of patients with deficient serum 25-OHD exhibited low BMD, this frequency did not differ from the study population with normal levels of 25-OHD (Table 2).This has led to the conclusion that factors other than 25-OHD insufficiency may be primarily responsible for the increased incidence of osteopenia and osteoporosis in patients with Crohn's disease.Nevertheless, it does appear that a deficiency in 25-OHD did have adverse effects on bone metabolism.Patients with low serum 25-OHD experienced a higher mean alkaline phosphatase and serum PTH, when compared to the patients with normal 25-OHD levels.Similar findings have been previously reported in other studies (8,19).It has been recognized that secondary hyperparathyroidism, as a consequence of 25-OHD deficiency, along with an increase in bone formation rate mark the beginnings of 25-OHD deficiency-induced metabolic bone disease, which is often asymptomatic and subclinical (16,(34)(35)(36).Furthermore, a long term borderline 25-OHD serum level will contribute to a nutritional deficiency disorder resulting in osteoporosis (35,36).25-OHD deficiency may occur in patients with Crohn's disease for a number of reasons, including low sunlight exposure, malabsorption and reduced vitamin D intake.In this study, very few patients demonstrated low 25-OHD in the summer months.Indeed, only three of the 108 patients who were assessed between April and September exhibited low serum 25-OHD, two of them being assessed in the first week of April.This observation is in agreement with the accepted notion that skin generation of 25-OHD through sunlight exposure is likely a significant source of 25-OHD in patients with Crohn's disease (12,16,18,23).Low serum RBC folate levels was predictive of 25-OHD deficiency in both males and females.Furthermore, patients with low serum 25-OHD exhibited lower albumin and carotene levels than those with normal levels of circulating 25-OHD.These findings imply that nutritional status, possibly related to the dietary intake of foods fortified with 25-OHD, plays an important role in determining 25-OHD deficiency in our patients with Crohn's disease.
Intestinal vitamin D malabsorption did not appear to be a major factor in determining 25-OHD status in our patients, as disease location did not influence serum 25-OHD levels.This is likely explained by the fact that the site of vitamin D absorption, the proximal jejunum, is not frequently involved in Crohn's disease (27).Only two out of 242 patients in our study reported jejunal disease involvement.In addition, enterohepatic interruption, an important aspect of vitamin D absorption, remained intact since vitamin B 12 , a measure of ileal dysfunction, did not appear to be related to, or influence, serum levels of 25-OHD.
Interestingly, our data observed that smoking in females predicted deficient serum levels of 25-OHD.Whether smok- ing has a 25-OHD-lowering effect in females has never been previously reported in patients with Crohn's disease, but a few mechanisms are proposed.One study has suggested women who smoke are thinner than their nonsmoking counterparts ( 37), yet we did not find any relationship between BMI and 25-OHD in the present study.Smoking has also been suggested to decrease estrogen production in females (38), conceivably leading to decreased intestinal uptake of vitamin D, decreased renal production of 1,25-OHD, and reduced levels of 25-OHD (39).Further studies are required to elucidate the precise mechanism behind the 25-OHDreducing effect of cigarette smoke in Crohn's disease patients, and to determine what implications this has for their BMD.This study did not examine the effects of small intestinal surgery, fat malabsorption or dietary intake on 25-OHD levels.In summary, 25-OHD-deficient Crohn's disease patients exhibit biochemical evidence of metabolic bone disease, although no detectable difference in BMD was observed.Sunlight exposure, nutrition and smoking status were predictors of 25-OHD deficiency in this patient cohort.

Figure 1 )
Figure 1) Relationship of serum 25-hydroxyvitamin D 3 (25-OHD) to T-scores at the lumbar spine, femoral neck, total hip and ultradistal radius in 242 patients with Crohn's disease.Bone mineral density was assessed by dual energy x-ray absorptiometry.*r-values represent Pearson correlation coefficients between 25-OHD and the different skeletal sites assessed.No statistically significant relationship was demonstrated between 25-OHD and T-scores at any skeletal site assessed (P>0.05)

TABLE 1 Demographic characteristics of 19 patients with Crohn's Disease and serum 25-hydroxyvitamin D 3 (25-OHD) deficiency, compared with 219 Crohn's disease patients with normal serum 25-OHD
25-OHD deficiency is defined as a serum 25-OHD concentration below 25 nmol/L.BMI Body mass index; F Female; M Male