Vitamin D physiology

Optimal vitamin D status

The best measure of an individual's vitamin D status is serum 25-hydroxyvitamin D (25(OH)D)⁽ Reference Holick ^{6 ,} Reference Rosen ^{8 )} which reflects both sunlight exposure and dietary vitamin D intake. The definition of vitamin D deficiency remains controversial. At present there is no target level set for people with Crohn's disease beyond recommendations for the general, healthy population. The US Institute of Medicine define deficiency as <30 nmol/l, and use 40 and 50 nmol/l to define the estimated average requirement and recommended daily allowance respectively with intakes of 15μg (600 IU) vitamin D₃/d recommended for adults and children, a tolerable upper intake level of 100μg (4000 IU)/d and a no observed adverse effect level of 250μg (10 000 IU) vitamin D₃/d^{( 17 )}. The US Endocrine Society's Clinical Practice Guideline suggests 75 nmol/l as a cut-off for adequacy and intakes of 37·5–50μg (1500–2000 IU) vitamin D₃/d to achieve this concentration. Irrespective of the cut-off applied (30, 50 or 75 nmol/l), several studies have reported a high prevalence of vitamin D insufficiency and deficiency in established IBD cases (Table 1) and in 80 % of new Crohn's disease diagnoses⁽ Reference Leslie, Miller and Rogala ^{9 )}. In paediatric cases, 25 % of patients have severe deficient levels⁽ Reference Alkhouri, Hashmi and Baker ^{18 )} (Table 1).

Table 1.

Prevalence of suboptimal vitamin D status in inflammatory bowel disease in patients with active and quiescent disease

CD, Crohn's disease; IBD, inflammatory bowel disease; UC, ulcerative colitis; 25(OH)D, 25-hydroxyvitamin D.

* Paediatric studies.

Factors influencing vitamin D levels in Crohn's disease

Several factors predict vitamin D deficiency in Crohn's disease including; longer disease duration, higher Crohn's disease activity index (CDAI) scores, C-reactive protein levels, poor nutrition status, smoking⁽ Reference Siffledeen, Siminoski and Steinhart ^{28 –} Reference Gilman, Shanahan and Cashman ^{31 )}, small bowel involvement⁽ Reference Gilman, Shanahan and Cashman ^{31 )} and resection⁽ Reference Driscoll, Meredith and Sitrin ^{32 )}, non-Caucasian ethnicity⁽ Reference de Bruyn, van Heeckeren and Ponsioen ^{33 )}, sunlight exposure, impaired conversion of vitamin D to its active metabolite, increased catabolism and increased excretion due to steatorrhoea.

In Crohn's disease dietary intakes and supplemental intakes appear inadequate for achieving sufficient 25(OH)D status. Less than half (43 %) of the patients are consumers of a vitamin D supplement, with multivitamin preparations being the most common form reported providing on average 5.6μg (5–10μg); 225 IU (200–400 IU) vitamin D daily. Moreover, for bone health, present guidelines suggest intakes of 20μg (800 IU)/d⁽ Reference Mowat, Cole and Windsor ^{11 )} which may or may not result in 25(OH)D concentrations ≥75 nmol/l. Studies have indicated intakes of 30, 50 or 125μg (1200, 2000 or 5000 IU)/d may be required to achieve levels ≥75 nmol/l, depending on baseline levels⁽ Reference Jørgensen, Agnholt and Glerup ^{15 ,} Reference Yang, Weaver and Smith ^{26 ,} Reference Raftery, Lee and Cox ^{34 )}.

Diet in Crohn's disease provides approximately 1·0 μg/d (95 % CI 0·6, 1·9)⁽ Reference Suibhne, Cox and Healy ^{35 ,} Reference Vogelsang, Klamert and Resch ^{36 )} with the main food sources being oily fish (38 %), followed by eggs (27 %)⁽ Reference Suibhne, Cox and Healy ^{35 )}. Despite being low, dietary intakes in Crohn's disease are comparable with population intakes⁽ Reference McCarthy, Duggan and O'Brien ^{29 ,} Reference Filippi, Al-Jaouni and Wiroth ^{37 ,} Reference Bin, Flores and Alvares-da-Silva ^{38 )}. Poor dietary intakes may also be hindered by reduced absorption. Vitamin D is absorbed in the proximal small intestine, particularly in the jejunum⁽ Reference Hollander and Truscott ^{39 )}. The effect on vitamin D status due to small bowel involvement is uncertain. In a small study of twelve Crohn's disease patients with a terminal ileum resection a decline in vitamin D absorption correlating with the length of the resection was observed⁽ Reference Leichtmann, Bengoa and Bolt ^{40 )}. Conversely Ulitsky et al.⁽ Reference Ulitsky, Ananthakrishnan and Naik ^{41 )} reported no difference in vitamin D levels between those with a resection v. no resection.

Whereas most of the predictors of low serum 25(OH)D in Crohn's disease are consistent throughout the literature, the effect of Crohn's disease activity on the vitamin D status is not confirmed. Some studies have reported no difference in 25(OH)D-based disease activity,⁽ Reference Gilman, Shanahan and Cashman ^{31 ,} Reference Hassan, Hassan and Seyed-Javad ^{42 )} whereas Jorgensen et al. reported low levels were associated with active disease⁽ Reference Jørgensen, Hvas and Agnholt ^{43 )}. A clear trend of decreasing 25(OH)D from remission (64 nmol/l) to mild disease (49 nmol/l) and moderately active disease (21 nmol/l) (P < 0·01) was reported⁽ Reference Jørgensen, Hvas and Agnholt ^{43 )}. A recent study confirmed these findings insofar as patients with active Crohn's disease had lower 25(OH)D levels than those in clinical remission; this measurement was independent of season or reported supplement use⁽ Reference Ham, Longhi and Lahiff ^{44 )}. There also appears to be wide variation in the absorption of vitamin D in Crohn's disease; for example, Farraye et al.⁽ Reference Farraye, Nimitphong and Stucchi ^{45 )} reported that even in quiescent disease ability to absorb vitamin D is reduced by an average of 30 % in comparison with normal subjects after supplementation with 1250μg (50 000 IU) vitamin D₂. Whether or not the outcome would have been similar had vitamin D₃ been used remains to be seen.

In symptomatic/active disease cholestryamine may also be prescribed to reduce post-resectional diarrhoea. It also reduces bile acids, which are required for vitamin D absorption and may induce vitamin D malabsorption. Protein loosing enteropathy is a condition which can arise in severe disease and can result in the loss of vitamin D-binding protein along with the vitamin D bound to it⁽ Reference Mouli and Ananthakrishnan ^{46 )}. Moreover, genetic variants in vitamin D hydroxylation and transport may also contribute substantially to both the development of vitamin D insufficiency and poor response to supplementation⁽ Reference Wang, Zhang and Richards ^{47 )}.

Beyond diet, sunlight and casual UVB exposure is the main source of vitamin D for most of the population. However, in Crohn's disease immunosuppressive therapy, such as azathioprine and adalimumab, can increase the risk of skin cancer. For this reason patients prescribed such medications are counselled regarding the careful use of sunscreen, which also prevents UVB synthesis of vitamin D. Sun exposure may also have a link to Crohn's disease pre-diagnosis. UVB exposure is often reduced at higher latitudes and coincides with a higher prevalence of autoimmune diseases and colorectal cancer in these regions compared with those more southerly⁽ Reference Simpson, Blizzard and Otahal ^{48 ,} Reference Peyrin-Biroulet, Oussalah and Bigard ^{49 )} suggesting a possible relationship between latitude and Crohn's disease.

Epidemiological evidence: low vitamin D status and Crohn's disease

Environmental triggers for IBD have been difficult to identify⁽ Reference Frolkis, Dieleman and Barkema ^{50 )}. A German twin cohort study confirmed the strong genetic element to IBD, yet concordance rates between monozygotic twins are nonetheless low (35 % for Crohn's disease and 16 % for ulcerative colitis). This suggests important environmental interactions with disease-inducing genes⁽ Reference Spehlmann, Begun and Burghardt ^{51 )}. One potential environmental risk factor is the UVB exposure. Recently a link between latitude and incidence rates of Crohn's disease has been identified in a large prospective study⁽ Reference Khalili, Huang and Ananthakrishnan ^{52 )}. By tracking the location and lifestyle information of approximately 175 000 female American nurses biennially over 20 years, the authors detected a greater increase in the incidence rates of Crohn's disease and ulcerative colitis the farther subjects lived from the equator. At age 30 years, living in southern latitudes was associated with a roughly halved risk of developing Crohn's disease and approximately a 40 % reduced risk of developing ulcerative colitis. Similarly Ananthakrishnan et al.⁽ Reference Ananthakrishnan, Khalili and Higuchi ^{12 )} found that women with a higher serum vitamin D level had a significantly reduced risk of Crohn's disease (hazard ratio: 0·38) suggesting a protective effect of vitamin D sufficiency.

In Europe an evident north–south gradient of incidence and prevalence also exists⁽ Reference Shivananda, Lennard-Jones and Logan ^{53 –} Reference Schultz and Butt ^{55 )}. For example, low sunlight exposure was associated with an increased incidence of Crohn's disease in France and no association with ulcerative colitis⁽ Reference Burisch, Pedersen and Cukovic-Cavka ^{56 )}. Migration of populations who live near the equator to countries of greater latitude also increases the rate of Crohn's disease⁽ Reference Pinsk, Lemberg and Grewal ^{57 ,} Reference Sewell, Yee and Inadomi ^{58 )}. More recently, Limketkai et al.⁽ Reference Limketkai, Bayless and Brant ^{59 )} reported that lower UV exposure is associated with greater rates of hospitalisation, prolonged hospitalisation and the need for bowel surgery in IBD. Further studies are needed to determine if this association is causal and also the role of other environmental factors that might explain these findings such as pollutants, diet and commensal or pathogenic microorganisms.

Vitamin D and immune function in Crohn's disease: experimental data

Vitamin D appears to have an important role in innate immunity⁽ Reference Cantorna, Zhu and Froicu ^{14 ,} Reference Cantorna and Mahon ^{60 )}. For example, human cathelicidin antimicrobial peptide and beta defensins are antimicrobial peptides of the innate immune system, which are expressed by the gastrointestinal epithelium⁽ Reference Jäger, Stange and Wehkamp ^{61 )}. Antimicrobial peptides protect against bacterial invasion⁽ Reference Tollin, Bergman and Svenberg ^{62 )} and human cathelicidin antimicrobial peptide is important in maintaining and re-establishing intestinal barrier integrity⁽ Reference Otte, Zdebik and Brand ^{63 )} and in the healing of human intestinal epithelial cells⁽ Reference Otte, Zdebik and Brand ^{63 )}. Moreover in vitro studies have shown that 1,25(OH)₂D can induce the expression of the gene encoding human cathelicidin antimicrobial peptide⁽ Reference Gombart, Borregaard and Koeffler ^{64 )}. However, the largest body of experimental evidence for an immunoregulatory role for vitamin D in IBD concerns the adaptive T-cell response. Several types of T-cells are important for the regulation of homeostasis in the gastrointestinal tract and either induce or suppress IBD. The VDR and 1,25(OH)₂D inhibit Th1 and Th17 functions by suppressing the production of particular cytokines⁽ Reference Cantorna and Mahon ^{13 ,} Reference Daniel, Sartory and Zahn ^{65 ,} Reference Tang, Zhou and Luger ^{66 )} which restores gastrointestinal homeostasis post infection or chemical injury. In addition, 1,25(OH)₂D stimulates dendritic cell production of IL-10, and T-cell levels of CTLA-4 (an inhibitory co-stimulatory signal), which further enhances its anti-inflammatory effect⁽ Reference Jeffery, Burke and Mura ^{67 )}.

Vitamin D and intestinal permeability in Crohn's disease: experimental data

Animal studies have shown that vitamin D may be linked to Crohn's disease severity and the function of the epithelial barrier. Vitamin D deficiency increased symptoms of several experimental models of IBD⁽ Reference Cantorna ^{68 )} and VDR deficiency increased susceptibility of mice to colitis⁽ Reference Froicu and Cantorna ^{69 ,} Reference Froicu, Weaver and Wynn ^{70 )}. Conversely treatment with 1,25(OH)₂D improves IBD symptoms and blocks the progression of colitis in mice⁽ Reference Daniel, Sartory and Zahn ^{65 ,} Reference Froicu, Weaver and Wynn ^{70 ,} Reference Cantorna, Munsick and Bemiss ^{71 )}

Vitamin D may also function on the epithelial barrier. Epithelial cells are connected by intercellular junctions, comprising tight junctions and adherens junctions⁽ Reference Henderson and van Limbergen ^{72 )}. Patients with Crohn's disease have increased small intestine permeability⁽ Reference Marchiando, Graham and Turner ^{73 )} resulting in part from defects in these junctions. Compromised barrier function in Crohn's disease has been associated with inflammation, dysbiosis⁽ Reference Cantorna, McDaniel and Bora ^{74 )}, disease pathogenesis and as a predictor of clinical relapse⁽ Reference D'Incà, Di Leo and Corrao ^{75 ,} Reference Wyatt, Vogelsang and Hübl ^{76 )}. Evidence suggests that vitamin D increases tight junction proteins and enhances gut mucosal healing post-injury⁽ Reference Kong, Zhang and Musch ^{77 )}. For example, following exposure to dextran sulphate sodium, a chemical which induces colitis, the VDR knock out mice were unable to maintain the integrity of the epithelial barrier⁽ Reference Froicu and Cantorna ^{69 ,} Reference Ooi, Li and Rogers ^{78 )} and had lower expression of tight junction proteins than in wild-type mice⁽ Reference Kong, Zhang and Musch ^{77 –} Reference Zhang, Leung and Richers ^{79 )}. As a result of reduced tight junction proteins, vitamin D-deficient and VDR knock out mice had increased gut permeability compared with vitamin D-sufficient wild-type mice⁽ Reference Ooi, Li and Rogers ^{78 )}. Whilst the basic science supports a role for vitamin D in Crohn's disease as reviewed elsewhere⁽ Reference Hewison ^{80 )}, further work is required to establish if this translates to human studies.

Observational studies: association between vitamin D levels, disease activity and surgery in Crohn's disease

Whilst epidemiological, animal and experimental data are promising the full possible range of effects of vitamin D in Crohn's disease are unknown, as are the optimal level(s) for inducing them. Observational studies which have focused on vitamin D and its effect on clinical markers such as CDAI (a research tool used to quantify the symptoms of patients with Crohn's disease) and inflammatory markers have been inconclusive. In cross-sectional IBD cohort studies El-Matary et al.⁽ Reference El-Matary, Sikora and Spady ^{81 )} and Hassan et al.⁽ Reference Hassan, Hassan and Seyed-Javad ^{42 )} reported no association between 25(OH)D and CDAI. CDAI levels <150 are indicative of remission, whereas levels above that suggest active disease. The mean 25(OH)D in these two studies were 66·7 (sd 27·3) nmol/l and 32·7 (sd 28·3) nmol/l, respectively. The cohorts included both Crohn's disease and ulcerative colitis patients and the sample sizes were small. Another cross-sectional study exclusive to Crohn's disease (n 34) reported a significant inverse association between 25(OH)D and CDAI with mean concentrations of 53·5 (sd 27) nmol/l⁽ Reference Joseph, George and Pulimood ^{82 )}. Similar findings were reported by Ulitsky et al.⁽ Reference Ulitsky, Ananthakrishnan and Naik ^{41 )} who observed greater disease activity in those with lower 25(OH)D levels (Table 2).

Table 2.

Observational studies of the association between 25-hyroxyvitamin D (25(OH)D) status and disease related outcomes in inflammatory bowel disease (IBD)

CD, Crohn's disease; UC, ulcerative colitis; IBD, inflammatory bowel disease; HBI, Harvey Bradshaw index; CDAI, Crohn's disease activity index; CDI, Clostridium difficile infection; CRP, C-reactive protein; UCAI, ulcerative colitis activity index.

Almost two-thirds of patients with Crohn's disease will eventually require surgery as part of their clinical course. Ananthakrishnan et al.⁽ Reference Ananthakrishnan, Cagan and Gainer ^{83 )} reported that 25(OH)D levels >50 nmol/l in Crohn's disease were associated with fewer surgeries and hospitalisations compared with those with levels below this threshold. A more aggressive disease course and need for surgery among those with vitamin D deficiency was also seen in a South Asian cohort⁽ Reference Boyd and Limdi ^{84 )}. Overall, despite limitations inherent in cross-sectional studies, such as mixed cohorts of ulcerative colitis and Crohn's disease, a reduced spread of vitamin D levels, different methods of data analysis, role of causality and various primary outcome measures, most of these studies suggest positive correlations between vitamin D and Crohn's disease-related outcomes (Table 2).

Clinical studies; association between 25-hyroxyvitamin D levels, disease activity and relapse in Crohn's disease

Only a small number of intervention studies have examined the effects of vitamin D supplementation in a clinical trial setting in IBD (Table 3). Two studies have reported positive associations with disease activity. The first, a prospective open label study compared supplementation with active vitamin D (alfacalcidiol) to 25μg (1000 IU) vitamin D₃ (cholecalciferol) in Crohn's disease⁽ Reference Miheller, Muzes and Hritz ^{85 )}. After 6 weeks alfacalcidiol treatment resulted in a significant decrease in CDAI scores and C-reactive protein levels, as well as improvement in quality of life (QoL) scores. In spite of this at 12 months there were no significant differences between the groups with respect to these variables⁽ Reference Miheller, Muzes and Hritz ^{85 )}. The primary aim of the present study was to examine the effects on bone metabolism and not disease activity; moreover, the paper did not report the 25(OH)D levels obtained by the groups, which may have not been in the therapeutic range at 12 months. Yang et al.⁽ Reference Yang, Weaver and Smith ^{26 )} titrated vitamin D₃ intake until such a point serum levels were ≥100 nmol/l (commonly requiring 125μg (5000 IU)/d) and reported significant improvements in CDAI with a mean reduction in CDAI from 230 (sd 74) to 118 (sd 66; P < 0·0001). This may suggest a minimum level of 100 nmol/l is required to exert a significant effect on disease severity but further research is warranted. Whilst promising these studies were open label trials and therefore have their inherent limitations. A double-blind randomised placebo-controlled study assessed the effectiveness of vitamin D₃ supplementation in preventing clinical relapse. In comparison with the placebo group, oral vitamin D₃ supplementation of 30μg (1200 IU)/d for 12 months reduced the risk of relapse from 29 to 13 % at 1 year (P = 0·056)⁽ Reference Jørgensen, Agnholt and Glerup ^{15 )}. This difference in relapse was not statistically significant and merits further work (Table 3).

Table 3.

Intervention studies; relationship between 25-hyroxyvitamin D (25(OH)D) and outcomes in Crohn's disease

CD, Crohn's disease, CDAI, Crohn's disease activity index; QoL, quality of life.

The current authors previously examined the effects of vitamin D supplementation on intestinal permeability as measured by the lactulose : mannitol ratio and sucrose excretion which indicates small bowel permeability⁽ Reference Raftery, Martineau and Greiller ^{86 )}. In a double-blind placebo-controlled study 27 Crohn's disease patients were randomised to 50μg (2000 IU)/d vitamin D₃ or placebo. At follow-up (3 months) mean (95 % CI) 25(OH)D levels were as expected significantly higher in the vitamin D group 91·6 (75·5–107·6) nmol/l than in the placebo group 40·4 (30·4–50·4) nmol/l (P < 0·001). At 3 months, there was a significant increase in lactulose : mannitol ratio (P = 0·010) and sucrose excretion (P = 0·030) in the controls, but these parameters were unchanged in the vitamin D group, suggesting that 25(OH)D levels ≥75 nmol/l may preserve intestinal integrity.

Clinical studies; association between 25-hyroxyvitamin D levels and muscle function

Compared with healthy controls in Crohn's disease skeletal muscle mass and strength are reduced⁽ Reference Schneider, Al-Jaouni and Filippi ^{87 –} Reference Geerling, Badart-Smook and Stockbrügger ^{89 )} and muscle fatigue is increased⁽ Reference van Langenberg, Della Gatta and Warmington ^{90 )}. Fatigue is a major concern in Crohn's disease⁽ Reference Jelsness-Jorgensen, Bernklev and Henriksen ^{91 –} Reference Maunder, de Rooy and Toner ^{94 )} with two of five patients reporting that it negatively impacts their QoL, even in remission⁽ Reference Minderhoud, Oldenburg and van Dam ^{95 )}. Reasons for this may include elevated pro-inflammatory factors such as TNFα and IL-6⁽ Reference Reimund, Wittersheim and Dumont ^{96 ,} Reference Cominelli ^{97 )} which are associated with lower muscle mass and strength in elderly populations⁽ Reference Visser, Pahor and Taaffe ^{98 ,} Reference Schaap, Pluijm and Deeg ^{99 )}, poor nutrition, physical inactivity and prolonged corticosteroid therapy⁽ Reference Geerling, Badart-Smook and Stockbrügger ^{89 )}. The underlying mechanisms of how vitamin D might improve muscle are poorly understood; however, several lines of evidence support a role of vitamin D in muscle health. First, proximal muscle weakness is a prominent feature of vitamin D deficiency⁽ Reference Al-Shoha, Qiu and Palnitkar ^{100 )} in addition to diffuse muscle pain and gait impairments such as a waddling way of walking⁽ Reference Schott and Wills ^{101 )}. Secondly, skeletal muscle is a major reservoir of 25(OH)D⁽ Reference Mawer, Backhouse and Holman ^{102 )}; however, whilst it was previously thought that VDR were abundantly expressed in muscle cells with roles myogenesis and contractility this is currently under debate⁽ Reference Boillon, Gielen and Vanderschueren ^{103 )}. Although vitamin D supplementation increases muscle strength and balance in some populations for example in the elderly⁽ Reference Pfeifer, Begerow and Minne ^{104 )} data in Crohn's disease are not as widely available. In a cross-sectional study, van Landenberg et al. ⁽ Reference van Langenberg, Gatta and Hill ^{105 )} reported that high 25(OH)D and physical activity may protect against reduced muscle mass⁽ Reference van Langenberg, Della Gatta and Warmington ^{90 )}. Conversely Salacinski et al.⁽ Reference Salacinski, Regueiro and Broeder ^{106 )} were unable to show a relationship between 25(OH)D levels and muscle strength in Crohn's disease. Although they did show that those with higher 25(OH)D levels (≥100 nmol/l) exhibited greater muscle strength (normalised to body weight) than those with lower levels (≤80 nmol/l) suggesting perhaps optimal effects on muscle function with levels ≥100 nmol/l; however, this is tentative data.

The current authors previously reported the results of a 3-month randomised, double-blind intervention study in quiescent Crohn's disease (n 27)⁽ Reference Raftery, Lee and Cox ^{34 )}. Patients were randomised to either 50μg (2000 IU)/d vitamin D₃ or placebo and the primary outcome measures included changes in hand-grip strength, a proxy measure for muscle strength. Post-intervention, both dominant and non-dominant hand-grip strength were significantly higher in the vitamin D-treated group compared with the controls. In the same study group, we also assessed changes in fatigue and QoL⁽ Reference Raftery, Lee and Cox ^{34 )}. At 3 months, patients who achieved 25(OH)D levels ≥75 nmol/l had significantly higher QoL compared with patients below this cut-off (P = <0·0001)]. In line with this, significantly less fatigue was experienced in those with 25(OH)D levels ≥75 nmol/l compared with those below this cut-off, as assessed by question 2 of the IBD questionnaire.

In a cross-sectional study of 504 IBD patients (403 Crohn's disease patients and 101 ulcerative colitis patients) vitamin D deficiency (<50 nmol/l) was associated with lower QoL in Crohn's disease but not ulcerative colitis;⁽ Reference Ulitsky, Ananthakrishnan and Naik ^{41 )} however, muscle function and fatigue were not measured in the present study. Another intervention study⁽ Reference Yang, Weaver and Smith ^{26 )} also showed improved QoL scores following vitamin D supplementation (P < 0·0004), particularly when serum concentrations were ≥100 nmol/l⁽ Reference Yang, Weaver and Smith ^{26 )}. This was paralleled with significant improvements in CDAI scores; however, muscle strength was not measured in this study.

Vitamin D and cancer in Crohn's disease

More recently, associations between vitamin D status and cancer have been examined. Epidemiological studies suggest an increased risk of and mortality from cancer in northern latitudes with reduced UVB exposure, an association possibly mediated by vitamin D⁽ Reference Wacker and Holick ^{107 )}. Furthermore, prospective cohorts have demonstrated an inverse association between 25(OH)D and cancers of the colon, breast and prostate⁽ Reference Feskanich, Ma and Fuchs ^{108 –} Reference Giovannucci, Liu and Rimm ^{111 )} with one intervention study reporting a reduced risk of cancer by 60 %⁽ Reference Lappe, Travers-Gustafson and Davies ^{112 )} with levels >80 nmol/l. Ananthakrishnan et al. ⁽ Reference Ananthakrishnan, Khalili and Higuchi ^{12 )} looked at data from 2809 patients with IBD and a median plasma 25(OH)D level of 65 nmol/l. During a median follow-up period of 11 years, 196 patients (7 %) developed cancer, excluding nonmelanoma skin cancer (forty-one cases of colorectal cancer). Patients with vitamin D deficiency had an increased risk of cancer (adjusted OR, 1·82; 95 % CI 1·25, 2·65) compared with those with sufficient levels. Each 1–2·5 nmol/l increase in plasma 25(OH)D was associated with an 8 % reduction in risk of colorectal cancer (OR, 0·92; 95 % CI 0·88, 0·96). The mean plasma 25(OH)D in patients who subsequently developed cancer was 12·5 nmol/l lower than in those who did not develop cancer (57 v. 69 nmol/l; P < 0·0001). They also reported a statistically significant inverse association for lung cancer (OR, 0·95; 95 % CI 0·90, 0·99). However, the study has its limitations; for example, the confounding impact of inflammation and low BMI on low 25(OH)D status was not reported and there was a lack of information on screening practices and on smoking.