Malt was obtained free of charge by Carlsberg AB (Falkenberg, Sweden), and milled to a particle size less than 0.5 mm before inclusion in the diets. Five lyophilised probiotics, Paracasei, HEAL 9, HEAL 19, CURE 21 and Lr 271 were obtained from Probi AB (Lund, Sweden).

The six diets used in the study contained a basal diet mixture (329 g/kg, dry weight) and were prepared according to Henningsson et al. (23). Malt was added at a level of 68 g dietary fibre per kg dry feed (corresponding to 481 g malt per kg feed, dry weight basis, dwb), and wheat starch (189.7 g/kg, dwb) was used to adjust the dry matter weight. The starch used makes no contribution to the hindgut formation of SCFAs due to complete digestion in the upper gastrointestinal tract (24). The malt diet without probiotics is referred to as the control diet, and the other five diets containing one of the five probiotic strains are referred to as test diets. The daily amount of probiotics was 1×10⁹ colony forming units (CFU) for each rat, and was added at feeding time.

Male Wistar rats (n=42) obtained from Taconic (Ejby, Denmark) were used in the study. The rats were randomly divided into six groups and housed individually in metabolic cages in a room maintained at 22°C and 60% relative humidity in a 12-h light and night circle. After 1 week of acclimatisation to the environment, the rats (131.5±1.1 g) were allowed to adapt to the diet for 7 days before a 5-day long experimental period. Food intake was restricted to 12 g (dwb) per day, and water was given ad libitum. During the experimental period, faeces and feed residues were collected. Faecal material was freeze-dried and milled before analysis of the content of dietary fibre monomers. At the end of the experiment, the rats were anaesthetised by injecting Hypnorm-Dormicum in sterile water (1:1:2) at a dose of 0.15 ml per 100 g body weight. Portal blood was taken for analysis of SCFAs, amino acids and ammonia. Caecal content and proximal and distal colon content were collected for analysis of SCFAs. Caecal pH and the weight of caecal content and tissue were measured directly. The animal experiment was approved by the Ethics Committee for Animal Studies at Lund University (M 30-09).

One rat fed malt +CURE 21 suffered a weight loss (6.4 g) during the experiment and showed very low concentrations of SCFAs both in serum and caecal content. The data from this rat were therefore excluded from the statistical analysis as this animal was considered an outlier.

The caecal pool was calculated as the level of each SCFA obtained in the analysis multiplied by the amount of caecal content, which was extrapolated to the complete intake of dietary fibre (4.08 g). Body weight gain during the experimental period was calculated as g/g feed intake. Bulking capacity was calculated as faecal dry weight in g/g fibre ingested.

The results are presented as means and their standard error means (SEM). Effects of probiotics were evaluated by one-way ANOVA followed by Dunnett's procedure, where the test diets were compared (P<0.05) with the control diet. Box-Cox transformation was carried out when data was not normally distributed. Minitab version 14 was used for statistical analysis.

The content of dietary fibre polysaccharides in malt was 110 g/kg (dwb), of which only a small part was soluble (7%) (Table 1). The main components of the dietary fibre polysaccharides were xylose (37%), glucose (34%) and arabinose (19%). Considerable amounts of uronic acids (5%), galactose and mannose (5%) were also detected. The amount of Klason lignin was 25 g/kg (dwb).

Dietary fibre polysaccharides in malt were rather resistant to hindgut degradation in rats and only 30% of the ingested fibre was fermented. Among the dietary fibre components, glucose was most resistant to fermentation, and only 14% was utilised by the hindgut microbiota. The other main components, xylose and arabinose were utilised to 32 and 52%, respectively, by the hindgut microbiota. The minor components were fermented to similar degrees: 37–58%. None of the probiotic strains affected the fermentability of the dietary fibre to any greater extent. The only significance seen was the higher degree of fermentation with malt and Paracasei compared with malt and CURE 21 or Lr 271 when suing another statistical test, that is, student's t-test, than one-way ANOVA.

The rats appeared to be healthy throughout the study and ate most of the food provided during the experimental period (11.2–11.8 g/day), which corresponded to 93–98% of the feed intake (Table 2). All the rats included in the analysis gained weight, and no significances were detected between the groups.

Caecal content and tissue weight were similar in rats fed different diets (2.2±0.1 and 0.53±0.01 g, respectively). The caecal pH was between 6.5 and 6.9, with no differences between groups. No significant difference was observed in the bulking capacity between groups (1.38–1.49 g/g fibre).

The main SCFA formed in the caecum of all rats was acetic acid (104–131 µmol), followed by butyric acid (22–36 µmol) and propionic acid (16–21 µmol) (data not shown), which corresponded to 67–70%, 14–18%, and 10–12% of the total amount of SCFAs formed, respectively.

The caecal levels of total SCFAs were lower in rats fed malt together with HEAL 9 (P<0.01), HEAL 19 (P<0.01) and Lr 271 (P<0.05) than without these strains, mainly due to a decrease in the levels of acetic acid (P<0.01, 0.01 and 0.05, respectively), but also due to decreased levels of butyric acid (P<0.01 for HEAL 9 and P<0.05 for HEAL 19). The probiotic groups had lower levels of SCFAs in the distal colon than the group fed diets containing no probiotics, although this was not significant (Table 3).

Acetic acid (820–1,090 µmol/l), propionic acid (45–77 µmol/l) and butyric acid (55–130 µmol/l) were the major SCFAs presented in portal serum (Table 3). Considerable amounts of isobutyric acid (12–14 µmol/l), isovaleric acid (13–15 µmol/l) and valeric acid (4–10 µmol/l) were also detected (data not shown).

Levels of total SCFAs in the portal serum of rats were lower with malt together with HEAL 9 and Lr 271 than in rats fed only malt (P<0.05), which was mainly due to the lower levels of acetic acid (P<0.05). There was also a lower level of butyric acid in the portal serum of rats together with malt and HEAL 9 than with malt only (P<0.05) (Table 3).

The main amino acids in the portal plasma were glutamine (16%), alanine (14%), lysine (13%) and glycine (10%). Nineteen α-amino acids and ammonia were analysed in the portal blood, but only ammonia, glutamine and those showing significant differences between rats fed malt alone and malt with probiotics are shown in Table 4, that is, asparagine, proline, phenylalanine and glycine. Plasma levels of glycine from rats given malt together with HEAL 9 (P<0.001) or 19 (P<0.05) were higher than with only malt, while the levels of asparagine (P<0.05) and proline (P<0.01) were higher when CURE 21 was added to the malt diet than without this strain. Lr 271 contributed to higher levels of proline (P<0.05) and phenylalanine (P<0.01) in the rats. None of the probiotic strains affected the glutamine levels (Table 4).

The viable counts of Lactobacillus, Bifidobacterium and Enterobacteriaceae in the caecal content were not affected by any of the probiotic strains added (Fig. 1), although all the strains given to the rats, according to the supplier, survived and reached the caecum. In rats fed Paracasei, HEAL 9 or 19 or Lr 271, the given probiotic strain was also the dominant Lactobacillus in the caecum content (Table 5). Apart from the administered probiotic strains, two other Lactobacilli were isolated and found in all groups, an exception was in rats fed HEAL 19 where only one of these strains was found. One of the strains belonged to species of Lactobacillus animalis and the other belonged to species of Lactobacillus reuteri. CURE 21 was not detected in the caecum of rats, which may be due to that the strain did not survive in the passage through the hindgut or to the limited culture technique.