Dr. Bernstein is supported in part by a Crohn's and Colitis Foundation of Canada Research Scientist Award and the Bingham Chair in Gastroenterology. This study supported by a research grant from the Manitoba Medical Services Foundation
Homocysteinemia and B vitamin status among adult patients with inflammatory bowel disease: A one-year prospective follow-up study
Article first published online: 20 MAY 2011
Copyright © 2011 Crohn's & Colitis Foundation of America, Inc.
Inflammatory Bowel Diseases
Volume 18, Issue 4, pages 718–724, April 2012
How to Cite
Vagianos, K. and Bernstein, C. N. (2012), Homocysteinemia and B vitamin status among adult patients with inflammatory bowel disease: A one-year prospective follow-up study. Inflamm Bowel Dis, 18: 718–724. doi: 10.1002/ibd.21785
- Issue published online: 19 MAR 2012
- Article first published online: 20 MAY 2011
- Manuscript Accepted: 4 MAY 2011
- Manuscript Received: 5 APR 2011
- inflammatory bowel disease (IBD);
- vitamin B6 deficiency
The aim of this study was to longitudinally study serum homocysteine levels in patients with Crohn's disease (CD) and ulcerative colitis (UC) in relation to disease activity and B vitamin status.
In all, 98 consecutive adult patients (age 25–55 years) with CD (n = 70) and UC (n = 28) were enrolled and assessed at three timepoints over 1 year.
There were no significant differences in levels of homocysteine, B vitamins, or dietary intake by disease type, disease activity, or across visits. 13% of all inflammatory bowel disease (IBD) patients had elevated homocysteine at least once during the study. Nine patients with CD had fluctuating homocysteine levels during the study but these were inconsistent, ranging from within normal range to elevated levels in any individual. Six of these nine patients were persistently in remission. 30% of all IBD patients had vitamin B6 deficiency, 11% had vitamin B12 deficiency, and one patient (CD) had folate deficiency. All vitamins showed a significant correlation between intake and serum levels (B6; r = 0.46, P < 0.001, B12; r = 0.42, P < 0.001, and folate; r = 0.26, P = 0.008). There was an inverse relationship between serum homocysteine in the blood and serum vitamin B12 (r = −0.241, P = 0.017).
Serum homocysteine was mostly normal in patients with IBD and changed minimally over time. There was no association between disease activity and elevation of serum homocysteine. 30% of patients have vitamin B6 deficiency but vitamin B6 is not associated with elevated homocysteine. The routine measurement of homocysteine is not warranted. (Inflamm Bowel Dis 2011;)
Homocysteine is a sulfur amino acid derived from the metabolic conversion of methionine. This metabolic pathway is dependent on vitamins as enzymatic cofactors, specifically, vitamin B12, vitamin B6, and folate.1 While the etiology of hyperhomocysteinemia is considered multifactorial (including genetic, nutritional, and lifestyle factors), a deficiency of one or more B vitamins likely plays a key role. Some studies support hyperhomocysteinemia as an established independent risk factor for both arterial and venous thromboembolism2 but the relationship is controversial.3 Nevertheless, both of these hypercoagulable states may be increased in inflammatory bowel disease (IBD).4, 5 It is plausible that homocysteine may have a role in IBD-associated thrombosis since homocysteine may promote oxidative tissue damage.6 Of particular interest is that IBD is associated with deficiencies in vitamins B12 and B6 that may promote increased homocysteine levels.7 Folic acid, although involved in the metabolism of homocysteinemia, is not a common nutritional deficiency in IBD.7
Deficiencies in B vitamins in IBD may arise from impaired absorption due to inflammation (active disease), medication, a deficiency secondary to poor dietary intake of nutrients or, in the case of Crohn's disease (CD) and vitamin B12, a decreased absorptive area if the ileum is diseased or resected.7 Vitamin B6 has not been extensively studied in IBD. In a recent study by our group, the prevalence of vitamin B6 deficiency was reported at 29%.7 Low levels of vitamin B6 have been associated with hyperhomocysteinemia among IBD subjects.8
Hyperhomocysteinemia has been previously described in IBD in point prevalence studies,1, 9–13 although the causes in IBD specifically are unknown. Further, there have been no longitudinal assessments in IBD that correlate serum homocysteine with disease activity state or even with concurrent measures of B vitamins. The aims of the present study were to investigate serum levels of B vitamins and homocysteine among adult patients with IBD in relation to disease activity status and dietary intake assessments at three different points in time over 1 year.
PATIENTS AND METHODS
A total of 100 consecutive adult subjects with IBD were recruited from the outpatient gastroenterology clinic of a single investigator (C.N.B.) at the University of Manitoba, Health Sciences Centre, between August 2007 and September 2008. The Research Ethics Board of the University of Manitoba approved the study protocol and all subjects gave their informed written consent before the start of the study. Subjects were asked to complete a brief medical history questionnaire to extract information regarding 1) IBD type; 2) year of diagnosis; 3) use of medications; 4) use of vitamin and mineral supplementation; 5) history of cardiovascular disease (CVD) or thrombotic events (TE); and 6) smoking habits. Chart review was undertaken to confirm medical histories and to determine disease phenotype IBD or history of surgery. The disease activity of the subjects IBD was determined by using the Harvey–Bradshaw index (for CD) and the Powell–Tuck index (for ulcerative colitis, UC).14, 15
Only patients who were over the age of 18, not pregnant, not receiving enteral or parenteral nutrition, and had no history or presence of renal disease, liver disease, or diabetes were included in this study. All patients agreed to attend at least three appointments with the Research Dietitian over the course of 1 year following enrollment into the study.
At each follow-up appointment, use of medication and vitamin/mineral supplementation was documented. Changes in fistula or stricturing disease were recorded. Disease activity was assessed at each appointment. Serum was drawn at each visit. If patients were in remission on recruitment, they were advised to contact the research team should a flare-up occur, at which point the subject was asked to come in for a visit and to complete the biochemical tests. If no disease flare occurred, then subjects had sequential visits while maintaining clinical remission. Disease was considered to have flared when either the Harvey–Bradshaw or Powell–Tuck indices achieved a score of ≥5. Disease flare was also corroborated by reviewing the physician's (C.N.B.) clinical assessment. For those who did not present with a flare, serial measurements of homocysteine and B vitamins were undertaken at ≈6-month intervals from the time of enrollment.
After an overnight fast of 12 hours, blood samples were collected at each visit to measure: serum homocysteine (Hcy), vitamin B12, vitamin B6, folate, creatinine, and C-reactive protein (CRP). For vitamin B6 and folate testing, two aliquots of serum (1 mL per aliquot) per subject were frozen at −70°C and later shipped to the Hospitals In Common Laboratory (Toronto, Ontario, Canada) for analysis.
The intraassay coefficient of variance (CV) for measurements of homocysteine ranges from 3.2%–5.1% (interassay CV is not available), and for folate the intaassay CV is 3.2%–6.2% and the interassay CV is 7.7%–8.6%. For vitamin B6 the interassay CV is 3.2%–5.9% (the intraassay CV is not available) and for vitamin B12 the intraassay CV is 2.0%–3.9% and the interassay CV is 1.2%–2.9%.
At the initial visit the research dietitian completed a food-frequency checklist with each subject. The questionnaire was adapted from a previously validated questionnaire.7 At each appointment, subjects were then given a set of forms to record all food consumed for 3 consecutive days. Verbal and written guidelines were provided for completing the food records. For vitamin, mineral, or other nutrition supplements, subjects were asked to provide the dietitian with all containers/information. Nutrient analysis was performed using the Food Smart Nutrition Management Solutions v. 6.0 (Sasquatch software, Vancouver, British Columbia, Canada) nutrient analysis program. Nutrition intake from supplements was added to the final analysis.
Results are expressed as mean and standard deviations or as stated. Repeated measures analysis was done to determine whether there was any significant difference by IBD type between visits for blood levels of vitamins (B12, B6, and folate), blood markers (homocysteine, CRP, creatinine), and intake of B vitamins. Multivariate analysis of variance (MANOVA) was used to determine if there was a significant difference in blood parameters by disease activity. A Pearson correlation test was performed to analyze the association between B vitamins and homocysteine levels, and serum B vitamins and dietary intake. Significance was set at P < 0.05. Multiple logistic regressions and multiple linear regressions were used to model the effects of all the groups adjusted for one another and possible confounders.
Subject characteristics are shown in Table 1. Of the 100 subjects, two consented to participating but did not have any measures complete (lost to follow-up). Of the 98 subjects who had blood work completed, 84 had three measures (61 CD, 23 UC), 11 had two measures (eight CD, three UC), and three had only one measure (one CD, two UC). As shown in Table 1, 33% of CD patients and 46% of patients with UC had a flare-up during the course of the study. We found that among CD and UC, the body mass index (BMI) was 24.9 and 25.6, respectively. Overall, the majority (75%) of our patients were 100% of their ideal body weight as well (110 ± 19% and 113 ± 24% for CD and UC, respectively).
|Crohn's Disease (n=70)||Ulcerative Colitis (n=28)|
|Age (year, mean ± SD)||42.5 ± 12.5||39.4 ± 13.8|
|Gender (M,F)||22, 48||5, 23|
|Individuals with active disease at every visit, n (%)||3 (4.3)||2 (7.1)|
|Remission at every visit, n (%)||44 (62.9)||13 (46.4)|
|Active and remitted disease at different times, n (%)||23 (32.9)||13 (46.4)|
|Duration of disease, years, n (%)|
|0–5||11 (15.7)||9 (32.1)|
|6–10||24 (34.3)||9 (32.1)|
|11–20||17 (24.3)||6 (21.4)|
|>20||18 (25.7)||4 (14.3)|
|Small bowel only, n (%)||28 (40)||0|
|Small and large bowel, n (%)||27 (38.6)||0|
|Large bowel only, n (%)||15 (21.4)||26 (92.9)|
|Rectum only n (%)||0||2 (7.1)|
|None, n (%)||31 (44.3)||27 (96)|
|Small bowel only, n (%)||19 (27.1)||0|
|Small and large bowel, n (%)||12 (17.1)||0|
|Large bowel only, n (%)||8 (11.4)||1 (3.6)|
|History of cardiovascular disease|
|No history, n (%)||43 (61.4)||22 (78.6)|
|Historya, n (%)||2 (2.9)||0|
|Family history, n (%)||19 (27.1)||5 (17.8)|
|Otherbn (%)||6 (8.6)||1 (3.6)|
|Use of medications at baseline|
|No medications or 5-ASA use, n (%)||36 (51.4)||19 (67.9)|
|Immunosuppresivescn (%)||5 (7.1)||3 (10.7)|
|Steroids, n (%)||25 (35.7)||6 (21.4)|
|Immunosuppressives & steroids, n (%)||4 (5.7)||0 (0)|
|Use of vitamin/mineral supplementationd|
|No supplements||9 (12.9)||6 (21.4)|
|Yes, n (%)||61 (87.1)||22 (78.7)|
|Oral B12e, n (%)||33 (47.1)||12 (42.9)|
|Intramuscular B12, n (%)||17 (24.3)||1 (3.6)|
|Vitamin B6e, n (%)||40 (57.1)||14 (50)|
|Folic acide, n (%)||45 (64.3)||1 (3.6)|
|Yes, n (%)||6 (8.6)||1 (3.6)|
|Quit, n (%)||28 (40%)||9 (32)|
|BMI (kg/m2)||24.9 ± 4.5||25.6 ± 5.3|
|Ideal body weight (%)f||110.5 ± 19.6||113.8 ± 23.7|
Biochemical and Dietary Parameters
Table 2 shows the mean level of all biochemical parameters measured in the IBD subjects. CRP is greater among patients with active disease compared to subjects in remission (11.4 ± 27 vs. 5.81 ± 9.8 mg/L, P < 0.05 respectively). Table 3 shows the prevalence of IBD subjects who demonstrated elevated homocysteine levels during the study. Nine patients with CD had changes in homocysteine levels during the study, ranging from within normal range to elevated levels. Of these nine patients, six were in remission for all measurements, one had elevated homocysteine (17.1 μmol/L) while in remission at visit 1, a normal level when disease was in a flare (9.6 μmol/L), and a normal level at visit 3 when disease was in remission (10.5 μmol/L). One patient had elevated homocysteine in remission at baseline (16.5 μmol/L), normal level at visit 2 (11.36 μmol/L) while in remission, and normal at visit 3 while disease was active (9.2 μmol/L). Only one patient with CD with fluctuating Hcy levels above and within the normal range showed normal homocysteine while in remission (11 μmol/L and 10.6 μmol/L) and elevated homocysteine levels in a flare-up (13.6 μmol/L). Only one patient with UC had changes in homocysteine levels during the study. It was normal during remission (8.9 μmol/L and 8.0 μmol/L) and elevated Hcy during active disease (13.1 μmol/L). Among the 10 patients with elevated homocysteine levels at baseline, only 4 (40%) had elevated CRP levels.
|Hcy (μmol/L)||CRP (mg/L)||Vitamin B12 (pmol/L)||Vitamin B6 (nmol/L)||Folate (nmol/L)|
|Normal ranges||4.5–13 μmol/L||<8 mg/L||140–600 pmol/L||20–96 nmol/L||>11.8 nmol/L|
|CD||9.0 ± 2.4||7.7 ± 11.4||416.5 ± 195.2||73.8 ± 92.2||49.1 ± 49.2|
|UC||8.5 ± 2.0||4.5 ± 4.3||442.3 ± 255.5||68.8 ± 60.2||48.4 ± 47.9|
|Active diseaseb||8.9 ± 3.0||11.4 ± 26.5||455.1 ± 247.2||63 ± 73.5||38.3 ± 21.2|
|Inactive diseasec||9.0 ± 2.5||5.8 ± 9.8||420.9 ± 245.0||71.0 ± 95.9||48.2 ± 71.9|
|Crohn's Disease||Ulcerative Colitis|
|% Patients who never had elevated Hcy, n (%)||59 (84.2)||26 (92.9)|
|% Patients with at least one elevated Hcy level, n (%)||11 (15.7)||2 (7.1)|
|% Patients with at least one elevated Hcy level that had consistently elevated Hcy levels, n (%)||2 (2.9)||1 (3.6)|
|% Patients with fluctuating Hcy levels, n (%)b||9 (12.9)||1 (3.6)|
A total of 80 patients completed baseline food records (60 CD, 20 UC). Eighteen patients (18%) did not complete the food records. Of the 80 diet assessments, 37 patients (46%) did not consume adequate amounts of folate (27 CD, 10 UC). Perhaps the normal levels of folate are accounted for by the high rates of vitamin supplementation. There was a lower prevalence of inadequate intake of vitamin B6 (6%) and vitamin B12 (2.5%). Table 4 shows the prevalence of IBD subjects who demonstrated subnormal values of vitamin B status at baseline. Serum vitamin B6 deficiency was the more prevalent deficiency among all IBD subjects (30%) compared to folate (1%) and B12 (11%). Among the 29 IBD patients with low serum vitamin B6 levels, seven had active disease (24%). Among the 29 patients who presented with vitamin B6 deficiency at baseline, 17 had no history of bowel surgery (59%), while only 10 patients had small bowel resections (34%). All B vitamins showed a significant correlation between intake and subsequent serum levels dietary intake of (r = 0.42, P < 0.001 for vitamin B12, r = 0.46, P < 0.001 for vitamin B6, and r = 0.26, P = 0.008 for folate). There was a significant correlation between serum vitamin B12 and homocysteine (r = −0.241, P = 0.008). Serum vitamin B6 and folate were not significantly correlated with serum homocysteine levels (P = 0.30 and P = 0.43, respectively). Finally, data were analyzed to compare those who used vitamin supplementation (87%) versus those who did not use vitamin supplementation (13%). There were no significant differences in homocysteine levels when the two groups were compared (data not shown).
|Serum vitamin B6 (20–96 nmol/L)||n=21 (30%)||n=8 (28.6%)|
|Serum folate (>11.8 nmol/L)||n=1 (1.4%)||n=0 (0)|
|Serum vitamin B12 (>180 pmol/L)||n=8 (11.4%)||n=3 (10.7%)|
We set out to study serum homocysteine levels longitudinally in patients with IBD and to determine whether changes in homocysteine levels were associated with changes in disease activity or B vitamin status, or both. There were three key findings from our work. First, serum homocysteine was mostly normal in patients with IBD and changed minimally over time. The prevalence of having an abnormal level was higher than reported in studies of healthy Canadian controls12, 16 but it was a small minority who had elevated levels. Second, there was no association between a flare-up of IBD and an elevation of serum homocysteine. Finally, we confirmed that serum vitamin B6 levels were low in ≈30% of adult patients with IBD. However, B6 levels did not correlate with serum homocysteine levels.
Normal Homocysteine Levels Among IBD Patients
We reported that 10% of our IBD subjects had elevated homocysteine levels at baseline and 13% of our patients had elevated homocysteine levels at least once during the study. The prevalence of hyperhomocysteinemia among IBD patients is higher than healthy controls and has been well documented in the literature.10, 17, 18 Although we did not have a healthy control group in our study, our results support the existing literature that the prevalence of hyperhomocysteinemia among IBD may be higher than the healthy population prevalence, which ranges between 2 and 5%.8 However, the prevalence of hyperhomocysteinemia observed in our patient group is lower than other studies among IBD patients.9, 19 Erzin et al9 reported that among 105 IBD patients, 56% had hyperhomocysteinemia. In a study by Peyrin-Biroulet et al19 among 140 CD patients, 24% of the patients had elevated homocysteine levels (greater than 15 μmol/L). In contrast, we found hyperhomocysteinemia to be uncommon in IBD and only rarely present.
In addition, the mean serum homocysteine level of our IBD patients was lower than previously published studies (9.03 μmol/L and 8.5 μmol/ L for CD and UC, respectively). Dagli et al20 reported a mean homocysteine level of 20.9 ± 8 μmol/L among 40 IBD patients. Erzin et al9 reported a mean homocysteine level of 16.4 ± 8.7 μmol/L among 105 IBD patients.
It is unclear why our IBD patients have a lower prevalence and lower mean serum homocysteine levels compared to other published work. It could be argued that this may be a reflection of an overall “healthier” IBD group compared to other studies. In our recent work that studied the nutritional status of 126 adults with IBD, our data suggested most of our patients reporting to clinic are presenting as well-nourished as assessed by anthropometric measures.6 Using BMI in our present study as a general marker of nutritional status, we found that among CD and UC the BMI is 24.9 and 25.6, respectively. Further, the majority (75%) of our patients were 100% of their ideal body weight as well (110 ± 19% and 113 ± 24% for CD and UC, respectively). While we could not find a correlation between serum homocysteine and serum vitamin B6 and folate, it may be that such a sizable portion of our subjects used vitamin supplements that they helped to stabilize homocysteine levels. Our cohort may also be a “healthier” IBD population as indicated by the number of flare-ups. Almost 60% of our patients were in remission for the duration of the study.
An alternative hypothesis for the lower prevalence of hyperhomocysteinemia and lower values of serum concentrations of homocysteine in our study compared to others may be secondary to our study being conducted in Canada after the implementation of mandatory folic acid supplementation. Both in Canada and the United States, the fortification of the food supply with synthetic folic acid was deemed to be an acceptable public health strategy to increase the intake of folic acid for the general population.16 Both the Canadian and the US programs were introduced in 1998. In the study by Shuaibi et al16 among 95 healthy Manitoban women, there was 0% prevalence of folate deficiency as measured by serum and RBC folate. As serum folate levels inversely correlate with serum homocysteine, it may be that if the population is replete in folate that homocysteine levels will generally be lower. Since the introduction of the program, there has been a population-wide reduction in blood homocysteine levels.21
Because our study was drawn from an outpatient clinic, it may be worth pursuing a similar study among inpatients with IBD who are admitted to hospital with primarily active disease to observe if their homocysteine levels (and B vitamins) vary from that of our outpatient cohort. Venous thromboembolism is a greater risk for sedentary, more ill IBD patients and it is unknown whether hyperhomocysteinemia would be more prevalent in this population.
Zepeda-Gomez et al22 argued that fasting serum homocysteine levels underestimate the prevalence of hyperhomocysteinemia and that a methionine loading test is a more accurate test. In their study among 82 patients with IBD, 56% of IBD patients were diagnosed with hyperhomocysteinemia using methionine loading compared to 22% using only fasting serum homocysteine levels. Oldenburg et al23 also suggested that substantial hyperhomocysteinemia can be detected only after methionine loading. The authors demonstrated that homocysteine levels after methionine loading was significantly higher in IBD patients with a history of arterial thrombosis versus those with no history.23 It has been argued, however, that the prognostic value of homocysteine levels that are elevated after methionine challenge (and are otherwise normal in the basal state) is uncertain,12 and therefore the role of loading remains controversial. Since we had few patients with evidence of arterial or venous thrombosis and since we did not undertake methionine loading, it remains possible that even among our ambulatory population there may be abnormalities in homocysteine metabolism that we have not uncovered.
Only two of the patients in our study had a diagnosis of past arterial or venous thromboembolic disease. Interestingly, both of these patients had normal serum homocysteine levels at baseline and throughout the study. Because of this small sample size we cannot conclude if there is an association between serum homocysteine and thrombosis in our IBD population. Among 19 subjects with a family history of first-degree relatives with cardiovascular disease, only three had elevated serum homocysteine levels. Of six with a self-reported history of elevated lipids or hypertension, none had elevated homocysteine. Although these results provide insight on the (lack of) association between homocysteine and past thrombotic events / cardiovascular disease, our sample sizes were too small to make any firm conclusions. There were other potential limitations of our study. One was that we did not record data on coffee and alcohol intake, both of which may impact homocysteine levels. Another critical factor that contributes to hyperhomocysteinemia is the C677 MTHFR genotype; while one study found a significantly higher prevalence of homozygosity for this genotype variant in patients with IBD compared to healthy controls,24 another population-based case control study by our group found a reduction if anything in mutations for this gene in IBD.25 While we did not measure genetic variants that may contribute to hyperhomocysteinemia, ultimately with relatively low prevalence rates of hyperhomocysteinemia, these data may not have greatly impacted our results.
There were two main strengths of our study. These include that it was the first prospective evaluation of serum homocysteine levels in IBD patients, with serial measurements and a correlation of levels not just with vitamin B levels but also with disease activity.
In summary, the results of our present study suggest that the “routine” measurement of serum homocysteine in IBD is not warranted. While homocysteine may have a role in arterial or venous thromboembolism in patients, including in those with IBD, our study does not suggest that it is a common problem among an unselected group of IBD patients presenting to a referral clinic.
In the past year Dr. Bernstein has consulted for Abbott Canada, Astra Zeneca Canada, Janssen Canada, and received an unrestricted educational grant from Axcan Pharma and a research grant from Abbott Canada. Ms. Vagianos has no conflict of interest to declare.
- 3Homocysteine in inflammatory bowel disease: a risk factor for thromboembolic complications? Am J Gastroenterol. 2000; 95: 2825–2830., , , et al.Direct Link:
- 8Low vitamin B6 plasma levels, a risk factor for thrombosis in inflammatory bowel disease: role of inflammation and correlation with acute phase reactants. Am J Gastroenterol. 2003; 98: 112–117., , , et al.Direct Link:
- 13Increased levels of homocysteine in patients with Crohn's disease are related to folate levels. Am J Gastroenterol. 2000; 95: 3498–3502., , , et al.Direct Link:
- 17Homocysteine triggers mucosal microvascular activation in inflammatory bowel disease. Am J Gastroenterol. 2005; 100: 886–895., , , et al.Direct Link:
- 19Association of MTRR 66A>G polymorphism with superoxide dismutase and disease activity in patients with Crohn's disease. Am J Gastroenterol. 2008; 103: 399–406., , , et al.Direct Link:
- 25Mutations in clotting factors and inflammatory bowel disease. Am J Gastroenterol. 2007; 102: 338–343., , , et al.Direct Link: