Dietary Advice for Patients With Inflammatory Bowel Disease
Dietary Advice for Patients With Inflammatory Bowel Disease
Intravenous Feeding and 'Bowel Rest'. Total parenteral (intravenous) nutrition (TPN) with complete 'bowel rest' was shown by Ostro et al. to be effective in the primary management of complicated CD. In a retrospective study of 100 patients who were otherwise refractory to conventional medical management, 90 received complete nutrient replacement and 10 received protein-sparing therapy. In 77 patients, a clinical remission was achieved. The location of the intestinal involvement did not influence the remission rate: 73% in those with small bowel disease only, 78% in those with combined small and large bowel disease and 100% in six with isolated colonic disease.
In UC, however, controlled trials of intravenous feeding and bowel rest showed no influence on the response of severe acute disease to corticosteroids ( Table 1 ). This has strongly influenced clinicians towards the conclusion that diet has little impact on UC, but it needs to be remembered that these were short-term studies in patients with established severe disease. It is reasonable to speculate that diet might still play a subtler, longer term, role in UC pathogenesis.
Enteral Nutrition. The Cochrane review on the use of sole enteral nutrition to induce remission in CD concludes that 'the effectiveness of enteral nutrition for the induction of remission in CD is evident from the remission rates' (up to 84%). The review is, however, usually interpreted negatively by clinicians as it shows better efficacy for corticosteroids than enteral nutrition when compared on an intention-to-treat basis – across the seven included studies, the overall remission rate was 49% for enteral nutrition and 75% for corticosteroids. ( Table 1 ) However, efficacy is equivalent when the comparison is made per protocol (only about 2/3 of adult patients tolerate enteral nutrition sufficiently to complete a 3-week course); moreover, equivalent efficacy was also seen when only the 'highest quality' studies were analysed – showing remission rates of 79% for enteral nutrition and 64% for corticosteroids. Different enteral feeds were used in the various studies and it is unclear whether some of the differences in efficacy may have reflected real differences resulting from the different feed constituents, thus making meta-analysis across the different feeds of uncertain validity. There is also reasonable evidence that maintenance treatment with enteral nutrition, e.g. given as 50% of calorie intake, is effective. This includes a randomised controlled trial in which 51 patients with CD in recently established remission (achieved by various means including corticosteroids, enteral nutrition or surgery) were randomised, 26 to receive half their calorie intake as enteral nutrition and 25 to a free diet. Over a mean follow-up of 11.6 months, the relapse rate was 34.6% in the 'half enteral' group compared with 64.0% in those receiving the free diet (HR 0.40; 95% CI 0.16–0.98).
Evidence taken exclusively from paediatric studies gives even stronger support for the role of enteral nutrition, showing equivalent efficacy with corticosteroids and better improvements in growth and mucosal healing. A recent worldwide survey of 35 paediatric IBD centres demonstrated wide variation in the use of enteral nutrition in CD. The duration of exclusive enteral nutrition was most commonly 6–8 weeks and 90% used polymeric formulas. The reintroduction of food after exclusive enteral nutrition varied greatly: the most common recommendations were for an initial low-fibre diet (26%) or the gradual re-introduction of normal food as the formula volume decreased (52%). Enteral nutrition taken orally, usually with a flavouring such as Nesquick, is as effective as continuous feeding via a naso-gastric tube.
It has been suggested that the efficacy of enteral nutrition is greater for CD involving the small bowel than the colon; however, many of the studies do not differentiate between locations of CD and improvements have also been reported for patients with primarily colonic disease. A major problem that prevents wider use of enteral nutrition as primary therapy for CD is the high relapse rate, approximately 50% within 6 months, when patients return to their normal diet.
The mechanism by which enteral nutrition benefits CD is unclear. Possible mechanisms include low residue resulting in reduced gut microbiota, perhaps particularly in the distal small intestine; avoidance of long-chain fat which, if taken up by macrophages, impairs their function; avoidance of other harmful components of 'normal' food – these might include food additives such as emulsifiers or nano-particles; addition of anti-inflammatory substances such as TGFbeta, present in milk-derived feeds where casein is the main protein source.
It was initially thought that whole proteins or peptides, being potentially antigenic, might need exclusion, but subgroup analyses performed to evaluate the different types of elemental (amino-acid-based) and non-elemental diets (semi-elemental – peptide-based and polymeric – whole protein-based, usually with casein as the protein source) have shown no significant difference in their efficacy. Amino-acid-based feeds have a lower energy density, are more hyperosmolar and may have a lower adherence rate so polymeric feeds are generally now preferred.
Although an earlier meta-analysis suggested an inverse correlation between percentage of energy in enteral nutrition feeds given as long-chain fat and efficacy for inducing remission in CD, the more recent Cochrane analysis covered seven trials including 209 patients treated with enteral nutrition of differing fat content (low fat: <20 g/1000 kcal vs. high fat: >20 g/1000 kcal) and showed no significant difference in efficacy, OR (1.13, 95% CI 0.63–2.01) for low fat vs. high fat. There was also no consistent difference when comparing feeds with high omega-3 vs. high omega-6 fatty acid content.
Conclusion:
(i) CD – Enteral nutrition as sole feed can induce clinical remission and mucosal healing. It possibly works better when the small intestine is involved. Enteral nutrition given as 50% of calories is effective in maintaining remission. Whole protein ('polymeric') feeds work as well as amino-acid-based feeds and are generally less hyperosmolar and easier to flavour. Mechanisms of action are unclear. Use of TPN is associated with higher costs and significant risks including line sepsis unless carefully managed and should be restricted to patients who for some reason, e.g. obstruction or short bowel, cannot take adequate nutrition enterally.
(ii) UC – There is no evidence that bowel rest by either enteral nutrition or intravenous feeding is effective therapy for active UC, although nutritional support is appropriate if the patient is malnourished.
Dietary Supplementation With Omega 3 Fatty Acids. CD: Earlier positive results for supplementation with omega-3 (n-3) fatty acids (fish oil) have not been consistently reproduced and although a recent meta-analysis of six published trials shows a small benefit for maintenance supplementation (RR of relapse 0.77, 95% CI 0.61–0.98), the studies were significantly heterogenous (consistency index I = 58.4%, P for heterogeneity = 0.03) and two well-performed studies with a relatively large sample size showed no benefit. There remains controversy around this with some investigators suggesting that negative studies may have resulted from use of unsatisfactory omega-3 preparations with variable bioactivity, thus impacting on the results.
UC: Meta-analysis of three trials of maintenance with omega-3 fatty acid supplementation in UC showed no significant benefit (RR for relapse 1.02, 95% CI 0.51–2.03). A systematic review has also been performed of studies that assessed omega-3 fatty acid supplementation as treatment for active UC, but there is considerable heterogeneity between the trials and the data are inconclusive. Further studies using enteric coated capsules may be justified.
Conclusion: Despite theoretical evidence that omega-3 fatty acids might be beneficial in CD and UC, current evidence is weak.
Dietary Supplementation With Curcumin. Curcumin is a natural phenol found in turmeric and has had a long history of use in traditional herbal remedies. In vitro studies have demonstrated a variety of potentially beneficial effects including antioxidant activity and suppression of NFkappaB activation. A placebo-controlled maintenance trial of curcumin 2 g/day for 6 months showed a trend towards a reduced relapse rate (4% vs. 18%, RR 0.24, P = 0.06) and a significantly reduced endoscopic score.
Conclusion: Curcumin shows promise as a dietary supplement as adjunctive therapy for UC maintenance, but data are currently inconclusive and a further large-scale trial is needed.
Dietary Component Modification. Sugar and Fibre: Intervention studies do not support avoidance of sugar or increased intake of fibre (nonstarch polysaccharide) in CD. A controlled trial of maintenance with a high-fibre, low refined sugar diet increased CD symptoms and was associated with a high withdrawal rate. A separate trial that assessed reduction in sugar intake without altering fibre was also negative. It should be noted that insoluble fibre, with its potential for causing obstruction at CD strictures, and soluble fibre, which should have little risk for obstruction, might have quite different effects on IBD and further studies are needed to assess this.
Nanoparticles: It has been suggested that very small insoluble particles consumed in the diet may have a harmful effect. A typical example might be titanium oxide – an insoluble white powder used not only in white paint but also in substantial quantities in foods such as mayonnaise as a whitener. Because of their very high ratio of surface area to weight, such 'nanoparticles' can act as haptens – for example enhancing by several orders of magnitude the cytokine response of lymphocytes to bacterial lipopolysaccharide. They are also taken up by M (microfold) cells overlying Peyer's patches, accumulate in gut macrophages, and can impair macrophage phagocytic activity. Western diets regularly expose the gastrointestinal tract to large quantities (>10/day) of man-made, submicron-sized, particles derived from food additives and excipients. An initial small trial showed benefit of a particle exclusion diet in CD, but a larger controlled trial performed by the same investigators was negative. Further larger scale trials are probably still warranted.
Milk and Dairy Products: An early clinical trial (77 patients) showed that about one in five patients with UC benefited from removal of milk and cheese from the diet, but this has never been repeated and further trials are needed.
Lactose: Hypolactasia is commonly present in IBD, but double-blind challenge with 240 mL of milk in non-IBD patients with proven hypolactasia did not cause significant symptoms and therefore, strict lactose exclusion, even though widely practised, is not usually necessary.
Avoidance of Various Specific Dietary Components. In an extended study of staged cumulative re-introduction of dietary components in a patient with stricturing small bowel CD, various foods (bread, lamb, potatoes, oranges, sugar, meats, dairy products, flour, rice) were tolerated without problems until relapse, accompanied by elevation of serum C-reactive protein (CRP), occurred when green vegetables were introduced. Following re-induction of remission with a 3-week period of enteral feeding, a return to the same low-residue diet taken prior to introduction of green vegetables repeatedly (×2) resulted in immediate relapse accompanied by elevation of serum CRP. This implied that the low-residue diet that had been well tolerated when the patient was in remission was no longer tolerated following recent relapse. It seems plausible that the initial relapse had resulted in mucosal ulcers and that once ulcers had developed, intolerance set in to foods that had previously been well tolerated.
Promising results have been reported with an approach in which CD patients re-introduce specific foods one by one and then leave out those that they find problematic, thus generating an individually tailored exclusion diet. A controlled trial was performed in which patients who had entered remission with elemental enteral nutrition were randomised to either tapering corticosteroids or exclusion diet and showed better long-term maintenance with the exclusion diet. Foods which most commonly caused problems and were therefore omitted included wheat, dairy products and yeasts.
Conclusions: Strong statistical associations exist between diet and IBD, particularly CD and a high intake of refined carbohydrate (see later); however, interventional studies of sugar exclusion and a combination of high-fibre, low-sugar diet have been negative. Because of the increased symptoms seen in the trial of high fibre, low refined sugar, coupled with the knowledge that most patients with small intestinal CD are likely to develop stricturing at some time in their course, it seems appropriate to recommend a low intake of foods that are high in insoluble fibre. Evidence from a single detailed case study suggests that a diet that is well tolerated when a patient is in long-standing remission may not be tolerated following recent relapse. A single controlled trial suggested good results for maintenance of remission in CD with an exclusion diet with wheat, dairy product and yeasts, the most common foods excluded. There is insufficient evidence at present to recommend a particle-free diet.
Early studies showing possible benefit in UC from milk and dairy product exclusion need repeating. Hypolactasia, although common, is probably overestimated as a cause of symptoms.
Vitamin and Mineral Supplementation. Vitamin D and Calcium: There is a resurgence of interest in the possible role of vitamin D deficiency in various conditions, including CD, that are commoner or more severe in northern Europe where diminished sunlight exposure is associated with vitamin D deficiency. Vitamin D is important for the function of the innate immune system and vitamin D deficiency is linked with increased susceptibility to tuberculosis. The actions of Vitamin D are relevant to all the potential pathogenic mechanisms of CD that have been highlighted by the genome-wide association studies, i.e. mucosal barrier function, innate immunity and immune regulation.
Low serum concentrations of Vitamin D have been reported in 63% of patients with CD, but caution is needed in interpreting this as 25-OH vitamin D acts as a negative acute phase reactant and serum concentrations may be an unreliable guide to vitamin D deficiency. Prediction of vitamin D deficiency from diet and lifestyle in a prospective cohort study of 72 719 women (age, 40–73 years) enrolled in the Nurses' Health Study showed that those predicted to have vitamin D levels in the top quartile compared with those in the bottom quartile had a significantly reduced risk for CD [Hazard Ratio 0.54 (95% CI 0.30–0.99)] and a nonsignificantly reduced risk for UC HR 0.65 (0.34–1.25). A randomised placebo-controlled trial of maintenance supplementation with 1200 IU/day of vitamin D in 94 patients with CD, regardless of vitamin D status on entry, showed a reduction in relapse rate (from 29% to 13%) that very nearly reached significance (P = 0.06). The vitamin D supplement dose of 1200 IU was relatively low and 2000 IU/day has been shown to be safe i.e. without risk of hypercalcaemia.
Patients with IBD have up to 40% increased risk of fractures compared with the general population and osteoporosis is common, particularly in patients who have received corticosteroids. Routine vitamin D supplementation is therefore not unreasonable even though clear evidence of its efficacy is lacking for prevention of disease relapse or osteoporosis.
It is common for people with IBD to believe that they are intolerant to dairy foods and this may lead to an inadequate intake of calcium. The daily recommendation for calcium intake in IBD is 1000 mg. However, calcium supplementation has come under scrutiny recently with some studies linking it to arterial calcification and a meta-analysis of data from 11 921 participants in 11 randomised controlled trials (not including IBD) showed that calcium supplements (≥500 mg/day) were associated with increased risk for myocardial infarction (RR 1.27, 95% CI 1.01–1.59) and smaller, nonsignificant, increases in the risk of stroke and mortality.
Vitamins C and E. Oxidative stress has been suggested to play a role in tissue damage in IBD. The nutritional antioxidants, including vitamins C and E, are attractive therapeutics because they are inexpensive and have relatively little toxicity. However, vitamin C can be pro-oxidant under certain conditions, and systemically altering the redox state may have unexpectedly negative effects on the inflammatory response. Clinical studies of vitamin C and E supplementation in CD have shown effects on biomarkers of oxidative stress, but have not so far been shown to have significant clinical efficacy.
Conclusions: Vitamin D deficiency may contribute to the cause and progression of IBD, particularly CD. Further trials of vitamin D supplementation are under way in both UC and CD, but meanwhile, low-dose vitamin D supplementation seems reasonable in all patients with CD.
Evidence From Interventional Studies
Intravenous Feeding and 'Bowel Rest'. Total parenteral (intravenous) nutrition (TPN) with complete 'bowel rest' was shown by Ostro et al. to be effective in the primary management of complicated CD. In a retrospective study of 100 patients who were otherwise refractory to conventional medical management, 90 received complete nutrient replacement and 10 received protein-sparing therapy. In 77 patients, a clinical remission was achieved. The location of the intestinal involvement did not influence the remission rate: 73% in those with small bowel disease only, 78% in those with combined small and large bowel disease and 100% in six with isolated colonic disease.
In UC, however, controlled trials of intravenous feeding and bowel rest showed no influence on the response of severe acute disease to corticosteroids ( Table 1 ). This has strongly influenced clinicians towards the conclusion that diet has little impact on UC, but it needs to be remembered that these were short-term studies in patients with established severe disease. It is reasonable to speculate that diet might still play a subtler, longer term, role in UC pathogenesis.
Enteral Nutrition. The Cochrane review on the use of sole enteral nutrition to induce remission in CD concludes that 'the effectiveness of enteral nutrition for the induction of remission in CD is evident from the remission rates' (up to 84%). The review is, however, usually interpreted negatively by clinicians as it shows better efficacy for corticosteroids than enteral nutrition when compared on an intention-to-treat basis – across the seven included studies, the overall remission rate was 49% for enteral nutrition and 75% for corticosteroids. ( Table 1 ) However, efficacy is equivalent when the comparison is made per protocol (only about 2/3 of adult patients tolerate enteral nutrition sufficiently to complete a 3-week course); moreover, equivalent efficacy was also seen when only the 'highest quality' studies were analysed – showing remission rates of 79% for enteral nutrition and 64% for corticosteroids. Different enteral feeds were used in the various studies and it is unclear whether some of the differences in efficacy may have reflected real differences resulting from the different feed constituents, thus making meta-analysis across the different feeds of uncertain validity. There is also reasonable evidence that maintenance treatment with enteral nutrition, e.g. given as 50% of calorie intake, is effective. This includes a randomised controlled trial in which 51 patients with CD in recently established remission (achieved by various means including corticosteroids, enteral nutrition or surgery) were randomised, 26 to receive half their calorie intake as enteral nutrition and 25 to a free diet. Over a mean follow-up of 11.6 months, the relapse rate was 34.6% in the 'half enteral' group compared with 64.0% in those receiving the free diet (HR 0.40; 95% CI 0.16–0.98).
Evidence taken exclusively from paediatric studies gives even stronger support for the role of enteral nutrition, showing equivalent efficacy with corticosteroids and better improvements in growth and mucosal healing. A recent worldwide survey of 35 paediatric IBD centres demonstrated wide variation in the use of enteral nutrition in CD. The duration of exclusive enteral nutrition was most commonly 6–8 weeks and 90% used polymeric formulas. The reintroduction of food after exclusive enteral nutrition varied greatly: the most common recommendations were for an initial low-fibre diet (26%) or the gradual re-introduction of normal food as the formula volume decreased (52%). Enteral nutrition taken orally, usually with a flavouring such as Nesquick, is as effective as continuous feeding via a naso-gastric tube.
It has been suggested that the efficacy of enteral nutrition is greater for CD involving the small bowel than the colon; however, many of the studies do not differentiate between locations of CD and improvements have also been reported for patients with primarily colonic disease. A major problem that prevents wider use of enteral nutrition as primary therapy for CD is the high relapse rate, approximately 50% within 6 months, when patients return to their normal diet.
The mechanism by which enteral nutrition benefits CD is unclear. Possible mechanisms include low residue resulting in reduced gut microbiota, perhaps particularly in the distal small intestine; avoidance of long-chain fat which, if taken up by macrophages, impairs their function; avoidance of other harmful components of 'normal' food – these might include food additives such as emulsifiers or nano-particles; addition of anti-inflammatory substances such as TGFbeta, present in milk-derived feeds where casein is the main protein source.
It was initially thought that whole proteins or peptides, being potentially antigenic, might need exclusion, but subgroup analyses performed to evaluate the different types of elemental (amino-acid-based) and non-elemental diets (semi-elemental – peptide-based and polymeric – whole protein-based, usually with casein as the protein source) have shown no significant difference in their efficacy. Amino-acid-based feeds have a lower energy density, are more hyperosmolar and may have a lower adherence rate so polymeric feeds are generally now preferred.
Although an earlier meta-analysis suggested an inverse correlation between percentage of energy in enteral nutrition feeds given as long-chain fat and efficacy for inducing remission in CD, the more recent Cochrane analysis covered seven trials including 209 patients treated with enteral nutrition of differing fat content (low fat: <20 g/1000 kcal vs. high fat: >20 g/1000 kcal) and showed no significant difference in efficacy, OR (1.13, 95% CI 0.63–2.01) for low fat vs. high fat. There was also no consistent difference when comparing feeds with high omega-3 vs. high omega-6 fatty acid content.
Conclusion:
(i) CD – Enteral nutrition as sole feed can induce clinical remission and mucosal healing. It possibly works better when the small intestine is involved. Enteral nutrition given as 50% of calories is effective in maintaining remission. Whole protein ('polymeric') feeds work as well as amino-acid-based feeds and are generally less hyperosmolar and easier to flavour. Mechanisms of action are unclear. Use of TPN is associated with higher costs and significant risks including line sepsis unless carefully managed and should be restricted to patients who for some reason, e.g. obstruction or short bowel, cannot take adequate nutrition enterally.
(ii) UC – There is no evidence that bowel rest by either enteral nutrition or intravenous feeding is effective therapy for active UC, although nutritional support is appropriate if the patient is malnourished.
Dietary Supplementation With Omega 3 Fatty Acids. CD: Earlier positive results for supplementation with omega-3 (n-3) fatty acids (fish oil) have not been consistently reproduced and although a recent meta-analysis of six published trials shows a small benefit for maintenance supplementation (RR of relapse 0.77, 95% CI 0.61–0.98), the studies were significantly heterogenous (consistency index I = 58.4%, P for heterogeneity = 0.03) and two well-performed studies with a relatively large sample size showed no benefit. There remains controversy around this with some investigators suggesting that negative studies may have resulted from use of unsatisfactory omega-3 preparations with variable bioactivity, thus impacting on the results.
UC: Meta-analysis of three trials of maintenance with omega-3 fatty acid supplementation in UC showed no significant benefit (RR for relapse 1.02, 95% CI 0.51–2.03). A systematic review has also been performed of studies that assessed omega-3 fatty acid supplementation as treatment for active UC, but there is considerable heterogeneity between the trials and the data are inconclusive. Further studies using enteric coated capsules may be justified.
Conclusion: Despite theoretical evidence that omega-3 fatty acids might be beneficial in CD and UC, current evidence is weak.
Dietary Supplementation With Curcumin. Curcumin is a natural phenol found in turmeric and has had a long history of use in traditional herbal remedies. In vitro studies have demonstrated a variety of potentially beneficial effects including antioxidant activity and suppression of NFkappaB activation. A placebo-controlled maintenance trial of curcumin 2 g/day for 6 months showed a trend towards a reduced relapse rate (4% vs. 18%, RR 0.24, P = 0.06) and a significantly reduced endoscopic score.
Conclusion: Curcumin shows promise as a dietary supplement as adjunctive therapy for UC maintenance, but data are currently inconclusive and a further large-scale trial is needed.
Dietary Component Modification. Sugar and Fibre: Intervention studies do not support avoidance of sugar or increased intake of fibre (nonstarch polysaccharide) in CD. A controlled trial of maintenance with a high-fibre, low refined sugar diet increased CD symptoms and was associated with a high withdrawal rate. A separate trial that assessed reduction in sugar intake without altering fibre was also negative. It should be noted that insoluble fibre, with its potential for causing obstruction at CD strictures, and soluble fibre, which should have little risk for obstruction, might have quite different effects on IBD and further studies are needed to assess this.
Nanoparticles: It has been suggested that very small insoluble particles consumed in the diet may have a harmful effect. A typical example might be titanium oxide – an insoluble white powder used not only in white paint but also in substantial quantities in foods such as mayonnaise as a whitener. Because of their very high ratio of surface area to weight, such 'nanoparticles' can act as haptens – for example enhancing by several orders of magnitude the cytokine response of lymphocytes to bacterial lipopolysaccharide. They are also taken up by M (microfold) cells overlying Peyer's patches, accumulate in gut macrophages, and can impair macrophage phagocytic activity. Western diets regularly expose the gastrointestinal tract to large quantities (>10/day) of man-made, submicron-sized, particles derived from food additives and excipients. An initial small trial showed benefit of a particle exclusion diet in CD, but a larger controlled trial performed by the same investigators was negative. Further larger scale trials are probably still warranted.
Milk and Dairy Products: An early clinical trial (77 patients) showed that about one in five patients with UC benefited from removal of milk and cheese from the diet, but this has never been repeated and further trials are needed.
Lactose: Hypolactasia is commonly present in IBD, but double-blind challenge with 240 mL of milk in non-IBD patients with proven hypolactasia did not cause significant symptoms and therefore, strict lactose exclusion, even though widely practised, is not usually necessary.
Avoidance of Various Specific Dietary Components. In an extended study of staged cumulative re-introduction of dietary components in a patient with stricturing small bowel CD, various foods (bread, lamb, potatoes, oranges, sugar, meats, dairy products, flour, rice) were tolerated without problems until relapse, accompanied by elevation of serum C-reactive protein (CRP), occurred when green vegetables were introduced. Following re-induction of remission with a 3-week period of enteral feeding, a return to the same low-residue diet taken prior to introduction of green vegetables repeatedly (×2) resulted in immediate relapse accompanied by elevation of serum CRP. This implied that the low-residue diet that had been well tolerated when the patient was in remission was no longer tolerated following recent relapse. It seems plausible that the initial relapse had resulted in mucosal ulcers and that once ulcers had developed, intolerance set in to foods that had previously been well tolerated.
Promising results have been reported with an approach in which CD patients re-introduce specific foods one by one and then leave out those that they find problematic, thus generating an individually tailored exclusion diet. A controlled trial was performed in which patients who had entered remission with elemental enteral nutrition were randomised to either tapering corticosteroids or exclusion diet and showed better long-term maintenance with the exclusion diet. Foods which most commonly caused problems and were therefore omitted included wheat, dairy products and yeasts.
Conclusions: Strong statistical associations exist between diet and IBD, particularly CD and a high intake of refined carbohydrate (see later); however, interventional studies of sugar exclusion and a combination of high-fibre, low-sugar diet have been negative. Because of the increased symptoms seen in the trial of high fibre, low refined sugar, coupled with the knowledge that most patients with small intestinal CD are likely to develop stricturing at some time in their course, it seems appropriate to recommend a low intake of foods that are high in insoluble fibre. Evidence from a single detailed case study suggests that a diet that is well tolerated when a patient is in long-standing remission may not be tolerated following recent relapse. A single controlled trial suggested good results for maintenance of remission in CD with an exclusion diet with wheat, dairy product and yeasts, the most common foods excluded. There is insufficient evidence at present to recommend a particle-free diet.
Early studies showing possible benefit in UC from milk and dairy product exclusion need repeating. Hypolactasia, although common, is probably overestimated as a cause of symptoms.
Vitamin and Mineral Supplementation. Vitamin D and Calcium: There is a resurgence of interest in the possible role of vitamin D deficiency in various conditions, including CD, that are commoner or more severe in northern Europe where diminished sunlight exposure is associated with vitamin D deficiency. Vitamin D is important for the function of the innate immune system and vitamin D deficiency is linked with increased susceptibility to tuberculosis. The actions of Vitamin D are relevant to all the potential pathogenic mechanisms of CD that have been highlighted by the genome-wide association studies, i.e. mucosal barrier function, innate immunity and immune regulation.
Low serum concentrations of Vitamin D have been reported in 63% of patients with CD, but caution is needed in interpreting this as 25-OH vitamin D acts as a negative acute phase reactant and serum concentrations may be an unreliable guide to vitamin D deficiency. Prediction of vitamin D deficiency from diet and lifestyle in a prospective cohort study of 72 719 women (age, 40–73 years) enrolled in the Nurses' Health Study showed that those predicted to have vitamin D levels in the top quartile compared with those in the bottom quartile had a significantly reduced risk for CD [Hazard Ratio 0.54 (95% CI 0.30–0.99)] and a nonsignificantly reduced risk for UC HR 0.65 (0.34–1.25). A randomised placebo-controlled trial of maintenance supplementation with 1200 IU/day of vitamin D in 94 patients with CD, regardless of vitamin D status on entry, showed a reduction in relapse rate (from 29% to 13%) that very nearly reached significance (P = 0.06). The vitamin D supplement dose of 1200 IU was relatively low and 2000 IU/day has been shown to be safe i.e. without risk of hypercalcaemia.
Patients with IBD have up to 40% increased risk of fractures compared with the general population and osteoporosis is common, particularly in patients who have received corticosteroids. Routine vitamin D supplementation is therefore not unreasonable even though clear evidence of its efficacy is lacking for prevention of disease relapse or osteoporosis.
It is common for people with IBD to believe that they are intolerant to dairy foods and this may lead to an inadequate intake of calcium. The daily recommendation for calcium intake in IBD is 1000 mg. However, calcium supplementation has come under scrutiny recently with some studies linking it to arterial calcification and a meta-analysis of data from 11 921 participants in 11 randomised controlled trials (not including IBD) showed that calcium supplements (≥500 mg/day) were associated with increased risk for myocardial infarction (RR 1.27, 95% CI 1.01–1.59) and smaller, nonsignificant, increases in the risk of stroke and mortality.
Vitamins C and E. Oxidative stress has been suggested to play a role in tissue damage in IBD. The nutritional antioxidants, including vitamins C and E, are attractive therapeutics because they are inexpensive and have relatively little toxicity. However, vitamin C can be pro-oxidant under certain conditions, and systemically altering the redox state may have unexpectedly negative effects on the inflammatory response. Clinical studies of vitamin C and E supplementation in CD have shown effects on biomarkers of oxidative stress, but have not so far been shown to have significant clinical efficacy.
Conclusions: Vitamin D deficiency may contribute to the cause and progression of IBD, particularly CD. Further trials of vitamin D supplementation are under way in both UC and CD, but meanwhile, low-dose vitamin D supplementation seems reasonable in all patients with CD.
