Search in the Abstract Database

Abstracts Search 2017

* = Presenting author

P182 Assessing the effect of ethnicity on urinary metabolic profiles in inflammatory bowel disease

Hicks L.*1, Powles S.1, Swann J.2, Chong L.1, Holmes E.2, Williams H.1, Orchard T.1

1Imperial College Healthcare NHS Trust St. Mary's Hospital, Department of Gastroenterology, London, United Kingdom 2Imperial College London, Computational and Systems Medicine, London, United Kingdom


Urinary metabolic profiling can distinguish patients with inflammatory bowel disease (IBD) from healthy controls (HC) and can also separate ulcerative colitis (UC) from Crohn's disease (CD). Clinical phenotype varies between Caucasians (Cau) and South Asians (SA) however discriminatory metabolites have not been studied in different ethnic populations. The aim of this study was to compare the urinary metabolic profiles of IBD patients and HC in Cau and SA backgrounds.


Samples from 405 IBD patients (283 Cau and 122 SA) and 137 HC (98 Cau and 48 SA) were analysed by H1NMR spectroscopy. Clinical and dietary data were collected. Orthogonal partial least squares discriminant analysis (OPLSDA) was performed to examine whether there were differences in metabolic data between Cau and SA. R2 (variance), Q2 (quality assessment) and p values (validity) for each model were described. Targeted profiling was performed using integral regions of 19 metabolites relating to the gut microbiota, TCA energy cycles and amino acids, as identified in published literature on IBD metabonomics in Cau.


The phenotype of SA CD was not significantly different to Cau CD in this cohort. In the SA UC group there was more pancolitis (p=0.051) and less proctitis (p=0.008). There were more vegetarians in the SA group. OPLSDA was able to separate patients with IBD from HC, and also UC from CD, in the Cau cohort, but this separation could not be replicated in SA (negative Q2 values).

Table 1. OPLSDA models examining HC vs CD, HC vs UC, and UC vs CD in Cau and SA cohorts. Implicated discriminatory metabolites identified from loadings plot of each model

R2Q2p value (1000 permutation testing)Implicated metabolitesR2Q2
HC vs CD0.630.630.003Beta-aminoisobutyrate (3-aminoisobutyrate), Hippurate, Glycine, Creatine0.60−0.21
HC vs UC0.820.800.012Fatty acyl chain protons, Hippurate, Citrate, Cholate0.34−0.15
UC vs CD0.880.460.008N-Acetylglutamate, Hippurate, Alpha-hydroxyisobutyrate (2-hydroxyisobutyrate), Citrate0.22−0.07

OPLSDA models also separated SA HC from Cau, and SA CD from Cau CD, but in UC no robust model could be made. In targeted analysis 4-cresol and hippurate were lower in SA compared to Cau regardless of IBD or HC. In IBD patients trimethylamine-N-oxide and succinate were different between ethnicities in both subtypes of IBD, alanine and methanol were different between Cau and SA controls.

Figure 1. Targeted analysis of 4-cresol sulphate.


SA manifest IBD phenotypically differently and display a different metabolic profile to Cau with IBD. Differences are also present between SA and Cau HC, and a combination of genetics, diet and microbiome may account for these differences. IBD has been sparsely studied in populations other than Cau, but with an incidence in SA migrants superseding that of the native UK population, it is important to better understand and interpret IBD metabolic profiles of different ethnicities.