Risk of allergy and nutrition to reduce that risk

Abstract

A method to predict the risk of eczema is provided based on differences in the development of microbiota and its metabolites in healthy infants and infants that develop eczema, and nutritional remedies based on this finding, in the form of lactate utilizing bacteria and fibres stimulating lactate utilizing bacteria.

Claims

1. A nutritional composition, comprising lactate utilizing bacteria and at least one fibre selected from the group consisting of polyfructose with an average degree of polymerisation (DP) of 10 or higher, polydextrose with an average DP of 10 or higher and partially hydrolysed guar gum with an average DP of 10 or higher.

2. The nutritional composition according to claim 1, wherein the lactate utilizing bacteria are selected from the group consisting of Anaerostipes and Eubacterium.

3. The nutritional composition according to claim 1, wherein the lactate utilizing bacteria are selected from the group consisting of Anaerostipes caccae, Anaerostipes butyraticus, Anaerostipes hadrus, Anaerostipes rhamnosivorans, E hallii and E. limosum.

4. The nutritional composition according to claim 1, wherein the lactate utilizing bacteria comprise Anaerostipes caccae.

5. The nutritional composition according to claim 1, wherein the amount of lactate utilizing bacteria is 10.sup.2 to 10.sup.9 cfu per g dry weight.

6. The nutritional composition according to claim 1, wherein the fibre comprises partially hydrolysed guar gum.

7. The nutritional composition according to claim 6, wherein the partially hydrolysed guar gum has a DP between 10 and 300.

8. The nutritional composition according to claim 1, wherein the fibre comprises polyfructose.

9. The nutritional composition according to claim 8, wherein the polyfructose has an average DP between 10 and 300.

10. The nutritional composition according to claim 8, wherein the polyfructose is inulin.

11. The nutritional composition according to claim 1, wherein the fibre comprises polydextrose.

12. The nutritional composition according to claim 8, wherein the polydextrose has a DP of 10 to 50.

13. The nutritional composition according claim 8, wherein the amount of fibre is 0.2 to 8 g per 100 g dry weight.

14. The nutritional composition according to claim 1, which further comprises protein, fat and carbohydrates, wherein the protein provides 5 to 16% of the total calories, the fat provides 35 to 60% of the total calories and the carbohydrates provide 25 to 75% of the total calories.

15. The nutritional composition according to claim 14, wherein the protein provides 5 to 12% of the total calories, the fat provides 39 to 50% of the total calories and the carbohydrates provide 40 to 55% of the total calories.

16. The nutritional composition according to claim 1, which is a follow on formula or young child formula.

17. A method of preventing allergy or reducing the risk of allergy in an infant or young child comprising administering to the infant or young child a nutritional composition.

18. A method of preventing allergy or reducing the risk of allergy in an infant or young child, comprising administering to the infant or young child the nutritional composition according to claim 1.

Description

DESCRIPTION OF THE FIGURE

[0061] The FIGURE shows Principal Response Curves representing the changes in bacterial metabolites, i.c. levels of short chain fatty acids (SCFA) and lactic acids, across time and its interaction with developing eczema.

EXAMPLES

Example 1

Aberrant Temporal Dynamics in Infants Developing Eczema in the First 18 Months of Life; Decreased Microbial Conversion of Lactic Acid in Infants Developing Eczema

[0062] Faecal samples were taken when infants were 4, 12 and 26 of age. Faecal samples were collected by the parents into 10 ml stool containers (Greiner Bio-One, Kremsmünster, Austria), immediately frozen (−12° C. to −20° C.) and transported within three months in a cold storage bag with ice-packs to the hospital. Upon arrival at the hospital and prior to evaluation at the laboratory, samples were kept and transported at ultra-low temperatures (−75° C. to −85° C.).

[0063] Frozen stool samples were defrosted on ice and were ten times diluted in PBS buffer (150 mM NaCl, 10 mM Na.sub.2HPO4, 20 mM NaH.sub.2PO.sub.4, pH 7.4) and homogenized by the addition of 5 to 10 glass beads (3 mm in diameter) and vortexing for 3 min. Glass beads and larger particles were removed by centrifugation at 300×g for 1 min. Several 1 ml portions of the homogenized suspension were stored at ultra-low temperatures until further processing and analysis.

[0064] Faecal suspensions were thawed on ice and centrifuged for 10 minutes at 14.000×g. Next, 350 μl supernatant was heated for 10 minutes at 100° C. to inactivate all enzymes and centrifuged again. A portion of the supernatant was used to quantitatively determine the SCFA acetic, propionic, n-butyric, isobutyric, isovaleric and n-valeric acids by gas chromatography. Another portion of the supernatant was used to enzymatically analyze the levels of lactate using a D-/L-lactic acid assay kit (Megazyme, Wicklow, Ireland).

[0065] 16S rRNA-gene sequencing was used to characterize the microbiota composition of faeces collected at 4 and 26 weeks of age in a set of vaginally born infants, including breastfed infants (BF, n=30) and infants receiving the experimental formula (n=51) or control formula (n=57). Diagnosis of eczema at the age of 18 months was based on 2 out of 3 positive scores according the modified Hanifin and Rajka criteria (Kunz et al., 1997, Dermatology 195(1):10-19). Of these 138 infants 52 infants had developed eczema when 18 month of age. Faecal samples for microbial analysis were selected with the following criteria: (I) infants were selected from the breastfed reference group and from the key group of interest (KGI), which consisted of those infants that started formula before 4 weeks of age, (II) when born vaginally and (III) if stool specimens were available at 4, 12 and 26 weeks of age.

[0066] Faecal suspensions were thawed on ice and 200 μl of each suspension was mixed with 450 μl of extraction buffer (100 mM Tris-HCl, 40 mM EDTA, pH 9.0) and 50 μl of 10% sodium dodecyl sulfate. Phenol-chloroform extractions combined with bead-beating were subsequently performed as described by Matsuki, et al., 2004, Appl Environ Microbiol 70:167-173 except that DNA was re-suspended in 0.1 ml of TE (10 mM Tris-HCl, 1 mM EDTA, pH 8.0). The V3-V5 regions of the 16S rRNA gene were amplified using the forward primer 357F, and a ‘bifidobacteria-optimised’ reverse primer 926Rb. The reverse primers included a 12 base-pair error-correcting Golay barcode. PCR was carried out in quadruplicate as previously described in Sim et al., 2012, PLoS One 7: e32543). Replicate amplicons were pooled and purified and three 454 Life Sciences GS FLX (Roche, Branford, Conn., USA) pyrosequencing runs were carried out following the Roche Amplicon Lib-L protocol.

[0067] The ‘Quantitative Insights Into Microbial Ecology’ (QIIME) v1.5.0 package was used to analyse shotgun processed data (Caporaso et al, 2010, Nat Methods 7: 335-336). Sequencing data was first denoised using AmpliconNoise followed by chimera-removal with Perseus (Quince et al, 2011, BMC Bioinformatics 12: 38). Alignment was carried out using the SILVA rRNA database (SSU_REF108) (Pruesse et al, 2007 Nucleic Acids Res 35: 7188-7196) for reference and clustered at 97% sequence identity into operational taxonomic units (OTUs) using the UCLUST algorithm (Edgar, 2010 Bioinformatics 26: 2460-2461). Rarefaction was performed and sequences present only once in the dataset (singletons) were removed.

[0068] The statistical analyses of the 16S rRNA-gene markers (frequencies of OTUs) and bacterial metabolite data (levels of SCFAs and lactate) were performed using Canoco 5 software (Šmilauer and Lepš, 2014, Multivariate Analysis of Ecological Data using CANOCO 5, 2nd ed. edn. Cambridge University Press: Cambridge Books Online) and differential abundance testing using the R-package MetagenomeSeq (Paulson et al., 2013, Nat Meth 10:1200-1202). The OTU count data were first aggregated at the genus level, resulting in a total of 142 genera, normalized by total sum scaling and log-transformed. Genera that were present in less than 10 samples were discarded to account for the sparsity of the data. This resulted in 58 taxonomic features that were used as input for the statistical analyses performed. Canoco 5 was used to model the taxonomic data using constrained ordination methods and Monte Carlo permutation tests (MCPT) to evaluate the explanatory power of sample covariates (Ter Braak, 1986, Ecology 67: 1167-1179; Van den Brink and Ter Braak, 1999, Environmental Toxicology and Chemistry 18: 138-148) with significance at 0.05. The sample covariates identified were subsequently used in all models and comparisons. Benjamini-Hochberg false-discovery rate (FDR) was used to account for multiple comparisons (Benjamini and Hochberg, 1995 Journal of the Royal Statistical Society Series B (Methodological) 57: 289-300) with significance for adjusted P-values (P-adj) at 0.05.

[0069] Constrained ordination methods combined with forward selection of variables in Canoco 5 was used to identify the sample covariates that explained most of the variation in microbial taxonomic composition (Legendre and Legendre, 2012 Numerical ecology, vol. 24. Elsevier: Amsterdam; Ter Braak, 1986). The approach of forward selection identified age (in weeks), feeding group (formula 1, 2 or breastfed reference group), ethnicity (Asian, caucasian or other) and having siblings (yes or no) as significantly influencing the bacterial taxonomic composition of the fecal samples (with Benjamini-Hochberg method to control for false-discovery rate with significance at 0.05). The sample covariates identified were subsequently used in all models and comparisons, either as explanatory variables or as covariates with significance for adjusted P-values (P-adj) at 0.05.

[0070] Principal Response Curves (PRC) and the Monte Carlo Permutation test (MCPT with 499 permutaions and significance at 0.05) were used to assess the effects of developing eczema on the bacterial metabolites (SCFA and lactic acids) over time (Van den Brink and Ter Braak, 1999). Differential abundant bacterial genera were evaluated using MetagenomeSeq with Benjamini Hochberg false discovery rates with significance at 0.05.

[0071] Univariate data analysis were performed using GraphPad Prism version 6.02 for Windows (GraphPad Software, La Jolla, Calif., USA) applying Mann-Whitney test for two-group comparisons and Kruskall-Wallis with Dunn's multiple comparisons test for three or more groups with significance at 0.05.

[0072] Results

[0073] The amount of faecal lactate in time is shown in table 1. At 4 and 12 weeks the amount of lactic acid tended to be lower in the infants that developed eczema, but the difference was not statistically different. At 26 weeks however, the amount of D and L-lactic acid was significantly higher in infants that developed eczema. A similar result was observed when the lacic acid was expressed as % of the total organic acids (the sum of D-lactic acid, L-lactid acid, acetic acid butyric acid, propionic acid, valeric acid, isobutyric acid and isovaleric acid), see table 2.

[0074] The effect of developing eczema on the fecal bacterial metabolite composition over time was investigated using Principal Response Curves (PRC, Van den Brink and Ter Beek, 1999), while correcting for all significant covariates identified. Significant temporal differences on the first constrained axis produced were observed for the interaction. The differential dynamics were most pronounced from 12 to 26 weeks of age. Infants developing eczema were characterized by decreased levels of both isomers of lactate at 12 weeks, a pattern which was subsequently reversed at 26 weeks of age. In the FIGURE the horizontal axis represents time and the vertical axis the PRC score values. Infants not developing eczema (NO ECZEMA; dashed line) were used as reference level and has zero PRC values and so its curve lays over the horizontal axis. The change for infants developing eczema in the first 18 months of life (ECZEMA; solid line -) is shown as a response curve relative to this reference. The microbial response scores are shown on the separate vertical (one-dimensional) plot. The multiple of the PRC score with the response score provides a quantitative interpretation as well as the direction of the microbial change at the respective time points (4, 12 and 26 weeks) in infants developing eczema as compared to those that did not. Result of the PRC analysis for faecal SCFA and LA are shown for the first PRC set, which was significant for the interaction (MCPT: P=0.034, explained variation 84.4%, 499 permutations).

TABLE-US-00001 TABLE 1 D-lactic acid and L-lactic acid in mmol per kg of wet weight faeces Age Lactic acid Statistics NO ECZEMA ECZEMA Summary  4 weeks D-lactic n 83 48 ns acid Mean 2.36 2.78 Median (Q1-Q3) 0.020 (0.020-2.92) 0.020 (0.020-4.34) L-lactic n 83 48 ns acid Mean 11.1 7.30 Median (Q1-Q3) 0.020 (0.020-9.88) 2.30 (0.020-9.62) 12 weeks D-lactic n 81 49 ns acid Mean 3.83 3.35 Median (Q1-Q3) 0.020 (0.020-4.17) 0.020 (0.020-4.27) L-lactic n 81 49 ns acid Mean 10.2 6.75 Median (Q1-Q3) 0.020 (0.020-14.4) 0.020 (0.020-9.64) 26 weeks D-lactic n 86 52 ** acid Mean 2.33 5.39 Median (Q1-Q3) 0.020 (0.020-2.11) 0.020 (0.020-9.28) L-lactic n 86 52 * acid Mean 7.74 14.3 Median (Q1-Q3) 0.020 (0.020-7.88) 4.37 (0.020-24.2) n = number of non-missing subjects, Q1 = 25% quartile and Q3 = 75% quartile. Statistical summary is based on Kruskall-Wallis with Dunn's multiple comparisons test comparing infants with eczema versus infants without at the same age (ns = P > 0.05, * = P ≤ 0.05, ** = P ≤ 0.01)

[0075] The effect of developing eczema on the fecal bacterial metabolite composition over time was investigated using Principal Response Curves (PRC, Van den Brink and Ter Beek, 1999), while correcting for all significant covariates identified. Significant temporal differences on the first constrained axis produced were observed for the interaction. The differential dynamics were most pronounced from 12 to 26 weeks of age. Infants developing eczema were characterized by decreased levels of both isomers of lactate at 12 weeks, a pattern which was subsequently reversed at 26 weeks of age. In the FIGURE the horizontal axis represents time and the vertical axis the PRC score values. Infants not developing eczema (NO ECZEMA; dashed line - - - ) were used as reference level and has zero PRC values and so its curve lays over the horizontal axis. The change for infants developing eczema in the first 18 months of life (ECZEMA; solid line -) is shown as a response curve relative to this reference. The microbial response scores are shown on the separate vertical (one-dimensional) plot. The multiple of the PRC score with the response score provides a quantitative interpretation as well as the direction of the microbial change at the respective time points (4, 12 and 26 weeks) in infants developing eczema as compared to those that did not. Result of the PRC analysis for faecal SCFA and LA are shown for the first PRC set, which was significant for the interaction (MCPT: P=0.034, explained variation 84.4%, 499 permutations).

TABLE-US-00002 TABLE 2 Relative abundance of D-lactic acid and L-lactic acid as mol % of total organic acids measured Age Lactic acid Statistics NO ECZEMA ECZEMA Summary  4 weeks D-lactic n 83 48 ns acid Mean 2.88 4.19 Median (Q1-Q3) 0.080 (0.030-0.58) 0.145 (0.040-8.24) L-lactic n 83 48 ns acid Mean 9.93 9.35 Median (Q1-Q3) 0.100 (0.030-14.7) 4.63 (0.0525-12.9) 12 weeks D-lactic n 81 49 ns acid Mean 4.26 3.39 Median (Q1-Q3) 0.060 (0.030-5.89) 0.060 (0.030-6.51) L-lactic n 81 49 ns acid Mean 10.7 7.41 Median (Q1-Q3) 0.200 (0.030-15.7) 0.120 (0.030-12.2) 26 weeks D-lactic n 86 52 *** acid Mean 1.96 4.61 Median (Q1-Q3) 0.025 (0.020-2.52) 1.90 (0.030-8.58) L-lactic n 86 52 ** acid Mean 7.97 11.8 Median (Q1-Q3) 0.030 (0.020-8.03) 5.51 (0.030-18.6) n = number of non-missing subjects, Q1 = first quartile and Q3 = third quartile. Statistical summary is based on Kruskall-Wallis with Dunn's multiple comparisons test comparing infants with eczema versus infants without eczema at the same age (ns = P > 0.05, * = P ≤ 0.05, ** = P ≤ 0.01, *** = P ≤ 0.001)

[0076] So, at 26 weeks of age the level and percentage of lactic acid was higher in the faeces of infants that had developed eczema at 18 months compared to infants that had not developed eczema at 18 months, and also the level and percentage of faecal lactic acid was increased when compared to the earlier time points of 4 and 12 weeks. In infants that did not develop eczema on the other hand the faecal lactate levels were lower at 26 weeks when compared with the earlier time points.

TABLE-US-00003 TABLE 3 Relative abundance (%) of Eubacterium spp Age Statistics NO ECZEMA ECZEMA Summary  4 weeks n 82 48 ns Mean 0.00780 0.599 26 weeks n 86 52 ns Mean  0.275 (1.15) 0.0996 Difference n (sample pairs) 82 48 * 26 weeks − Mean 0.280* (1.17) −0.507 4 weeks n = number of non-missing subjects; Statistical summary is based on Kruskall-Wallis with Dunn's multiple comparisons test comparing infants with eczema versus infants without eczema at the same age (ns = P > 0.05, * = P ≤ 0.05, ** = P ≤ 0.01, *** = P ≤ 0.001)

TABLE-US-00004 TABLE 4 Relative abundance (%) of Anaerostipes spp Age Statistics NO ECZEMA ECZEMA Summary  4 weeks n 82 48 ** Mean 0.00171 0.245 26 weeks n 86 52 ns Mean 0.434 0.314 26 weeks − n (sample pairs) 82 48 # 4 weeks Mean 0.350 0.0942 n = number of non-missing subjects; Statistical summary is based on Kruskall-Wallis with Dunn's multiple comparisons test comparing infants with eczema versus infants without eczema at the same age (ns = P > 0.05, # = P ≤ 0.1, ** = P ≤ 0.01)

[0077] MetagenomeSeq was used to assess differential abundances of bacterial taxa over time in infants developing and not developing eczema, as well as the taxa being differential over time comparing the two groups, while correcting for the covariates identified. The increases over time observed for Eubacterium and Anaerostipes spp. were more pronounced for infants not developing eczema as compared to infants developing eczema, see tables 3 and 4. Both genera are associated with a specialist group of microbes known to convert lactate together with acetate into mainly butyrate, hence referred to as lactate-utilizing bacteria (LUB). The levels of lactic acid and presence lactic acid utilizing bacteria were indeed inversely correlated.

[0078] The increased levels of acetate and lactate, and decreased amounts of propionate and butyrate, early in life, at 4 and 12 weeks, before starting weaning, may be crucial for the establishment of LUB like Eubacterium and Anaerostipes spp. around 6 months of age, as observed in this study for subsequent infants not developing eczema in contrast to infants developing eczema.

[0079] In conclusion, this study indicates for the first time a link between the functionality of the microbiota and the expression of allergic phenotypes in early life. It emphasizes the importance of the early life microbial succession of species and metabolite cross-feeding to develop a gut physiology that supports gut development, but also the development of normal immune responses towards environmental triggers. These observations could aid the development of optimal nutritional strategies to support timely gut colonization of keystone species in the gradually diversifying infant gut.

Example 2

L-Lactate Accumulation in Faeces of Infants that Develop Allergy

[0080] To gain insights into the development of gut microbiota in initially healthy infants who develop allergy in early life and identify plausible microbiota biomarkers of allergic disease, a nested case-control study of Chinese infants from a large Singaporean birth cohort was performed. The maturation of intestinal microbiota and its metabolism was measured in 20 pair-matched allergic cases and non-allergic controls during the first 6 months of life using 16S rRNA sequencing. The allergic infants were assessed by means of cumulative incidence of clinical allergy symptoms (eczema episode/allergic rhinitis/food allergy) and SCORAD values; according to the study eczema workflow up to 12 months of age. At age of 6 months higher levels of faecal L-lactate (12.25 mmol/kg wet weight faeces) were detected in infants that were allergic at 12 months of age, compared to the L-lactic acid level in infants that did not develop allergy (3.95 mmol/kg wet weight faeces), P<0.05. L-Lactic acid levels in the group of allergic infants were higher at 6 months when compared to week 3 and month 3, which is indicative for lactate accumulation in this group. For non allergic infants it was the other way around and the amount of faecal L-lactic acid was higher at week 3 and month 3.

TABLE-US-00005 TABLE 5 L-lactic acid in mmol per kg of wet weight faeces Faecal L-lactic acid NO ECZEMA ECZEMA (mmol/kg ww faeces) Month 3 Month 6 Month 3 Month 6 L-lactic acid 9.50 3.95 6.55 12.25

Example 3

Selection of Fibres that Stimulate Lactate Utilizing Bacteria

[0081] Fresh faecal samples were collected from four healthy adults, pooled, and divided in smaller portions mixed with glycerol (10%) in an anaerobic cabinet and stored at −80° C. Before the experiment the faecal samples were defrosted and mixed with the fermentation medium in 1:5 in a falcon tube and a t=0 sample was taken, 6 ml of this suspension was added to falcon tube with the substrate of interest and mixed. This mixed suspension was put in a dialysis tube and the dialysis tube was added in a 100 ml bottle filled with 100 ml dialysis medium. Bottles were closed and incubated at 37° C. A fermentation with no added carbohydrates acted as a negative control (blanc), whereas fermentation with glucose served as a positive control. Fibres and glucose were added at concentrations of 200 mg per 6 ml of faeces suspension.

[0082] 19 single fibres and 3 fibre mixtures were tested. Fermentation was at 37° C. at 48 h under anaerobic conditions. Starting pH of the buffer was 6.3.

[0083] Preservative medium: buffered peptone 20.0 g/l, L-cysteine-HCl 0.5 g/l, sodium thioglycollate 0.5 g/l, resazurine tablet 1 per litre, adjusted to pH 6.7±0.1 with 1 M NaOH or HCl. Boiled in microwave. Filled into 30 ml serum bottles with 25 ml medium. Sterilised 15 minutes at 121° C.

[0084] McBain & MacFarlane medium: buffered peptone water 3.0 g/l, yeast extract 2.5 g/l, mucin (brush borders) 0.8 g/l, tryptone 3.0 g/l, L-cysteine-HCl 0.4 g/l, bile salts 0.05 g/l, K.sub.2HPO.sub.4.3H.sub.2O 2.6 g/l, NaHCO.sub.3 0.2 g/l, NaCl 4.5 g/l, MgSO.sub.4.7H.sub.2O 0.5 g/l, CaCl.sub.2 0.228 g/l, FeSO.sub.4.7H.sub.2O 0.005 g/l. Filled into 500 ml Scott bottles with the medium and sterilised 15 minutes at 121° C.

[0085] Buffered medium: K.sub.2HPO.sub.4.3H.sub.2O 2.6 g/l, NaHCO.sub.3 0.2 g/l, NaCl 4.5 g/l, MgSO.sub.4.7H.sub.2O, 0.5 g/l, CaCl.sub.2 0.228 g/l, FeSO.sub.4.7H.sub.2O 0.005 g/l. Adjusted to pH 6.3±0.1 with K.sub.2HPO.sub.4 or NaHCO.sub.3. Filled into 500 ml Scott bottles with the medium and sterilised 15 minutes at 121° C.

[0086] Faecal suspension: the preserved solution of faeces was centrifuged at 13,000 rpm for 15 minutes. The supernatant was removed and the faeces was mixed with the McBain & Mac Farlane medium in a weight ratio of 1:5.

[0087] Faecal sample was stored in an RNAlater solution (Ambion, Courtaboef, France) for subsequent bacteria DNA or RNA isolation. The faecal sample was homogenized and the volume of RNAlater was adjusted to achieve a fmal faecal solution of 1:10. 200 μl of this 10-fold dilution was added to 1 ml of PBS buffer and centrifuged for 5 min at 5000 g. The supernatant was discarded and the pellet stored at −80° C. Faecal DNA and RNA was extracted. Determination of bacteria abundance was determined based on the quantification of RNA molecules using primers targeting mainly 16S rRNA sequences. In order to have a common scale of comparison with other bacterial quantification methods, the number of detected molecules (RNA or DNA) was given as cell equivalents (CE), for which a bacteria culture of a reference strain was used as a standard curve.

[0088] For determination of the target bacteria present in the faecal samples, three serial 10-fold dilutions of the extracted RNA or DNA sample were applied to qPCR or RT-qPCR, and CT values in linear range of the assay were applied to the standard curve generated in the same experiment to obtain the corresponding bacterial count in each nucleic acid sample, and then converted to the count per sample. Different species of bacteria were analysed amongst which there was one lactate utilizing bacteria, Anaerostipes caccae. Statistical analysis was performed using the Mann Whitney method.

[0089] Results

[0090] At t=0 the amount of Anaerostipes caccae was equivalent to 2*10.sup.8 cfu/ml. Upon 48 h of fermentation with glucose the amount had doubled to 4*10.sup.8 cfu/ml.

[0091] Of the 19 different fibres and 3 fibre mixes tested, a substantial increase in Anaerostipes caccae compared with the controls T=48 h (glucose) was observed with the following fibres: Orafti® GR (Beneo), an inulin with a DP>10 (5-fold), Orafti® HP (Beneo), a long chain inulin with an average DP of >23 (5 fold), Benefiber® (Novartis) a partially hydrolysed guar gum (PHGG) (4 fold), Sta-Lite® (Tate and Lyle) a polydextrose with DP>10 (5 fold) and Novelose® 330 (national starch) a resistant starch (3 fold). One fibre mixture, rich in inulin, also stimulated the growth of A. caccae.

[0092] No increase or even a decrease compared to the glucose control was observed with the other fibres tested, e.g the short chain FOS Actilight® (Meiji) and Frutalose® (Orafti both having an average DP<5. Table 6 shows the psotive results for polyfructose with an average DP of 10 or higher, polydextrose with an average DP of 10 or higher and partially hydrolysed guar gum with an average DP of 10 or higher. For comparison the result of non-positive short chain fructooligosdacchrides in included.

TABLE-US-00006 TABLE 6 Increase of A. caccae upon faecal fermentation of several fibres. Log increase relative t Fibre Log to t = 48 glucose T = 0 blanc 8.3 — T = 48 glucose 8.6 0 T = 48 Orafti ® HP 9.3 0.7 T = 48 Orafti ® GR 9.3 0.7 T = 48 Actilight ® 8.7 0.1 T = 48 Novelose ® 330 9.1 0.5 T = 48 Sta-Lite ® 9.3 0.7 T = 48 Benefiber ® 9.2 0.6