Based on its physiological characteristics, resistant starch (RS) has been acknowledged as part of the dietary fibre fraction and, thus, exhibits beneficial health effects. Like other nondigestible carbohydrates, RS contributes, to a considerable extent, to the substrates entering the colon, where it is subject to bacterial degradation. Dietary fibres in general are of importance as an energy source for colonic microorganisms. Moreover, they may selectively stimulate the growth and/or activity of specific colonic bacteria, which are thought to have a beneficial effect on gastrointestinal health. Short-chain fatty acids (SCFA), as main fermentation products, are absorbed and metabolised in various organs. Resistant starch, in particular, is known to lead to comparatively large amounts of butyrate (1,2), which is mainly used as energy source for colonocytes. It is therefore regarded as a key factor in potential protection against colon cancer because it can act at different levels, such as reducing tumour cell growth, inducing cancer cell differentiation, inhibiting protooncogene expression and inducing apoptosis (3,4). In addition, RS fermentation is known to result in significant amounts of acetate and propionate, which enter the systemic circulation to have a positive impact on insulin sensitivity of adipocytes. The latter may be of importance for reducing risks for developing type 2 diabetes and metabolic syndrome (5).
Because of changed eating habits, the mean intake of RS in industrialised countries has dropped drastically (6). Assessments in European countries indicate values less than 10 g/d (7). Although no recommendations exist, some experimental studies suggest that intakes on the order of 20 g/d or more may be needed to obtain some of the intestine-related benefits (8).
Among the generally recognised 4 types (8,9), RS type 2 (digestion-resistant raw starch granules) and RS type 3 (retrograded starch/maltodextrins) are increasingly produced by food ingredient companies to generate new types of dietary fibre–rich foods because of their interesting technological and sensory properties (10).
After birth, the capacity to digest macronutrients is poor because most digestive enzymes of the gastrointestinal tract are present at concentrations lower than in adults, with levels of pancreatic α-amylases activities being extremely low (11). However, human milk contains α-amylase, which may help breast-fed infants to digest starch at early weaning. Moreover, salivary amylase increases its activity rapidly after birth to reach adult levels within the first year. Thus, salivary amylase may play an important role in young infants because of the deficiency of the pancreatic isoenzyme, which reaches mature levels only after 5 to 12 years. The slow maturation of pancreatic α-amylases and the poor chewing ability in young infants may explain the excretion of significant amounts of starch in the faeces of children (12). Moreover, the presence of starch points to the fact that not only the digestion but also the fermentation of starch in infants is incomplete.
At birth, the infant colon is sterile, but it becomes rapidly colonised by bacteria both from the mother and from the environment (13). Early diet has a major influence on the colonisation process. In breast-fed infants, primarily bifidobacteria and some lactobacilli are found, whereas formula-fed infants have a more diverse bacterial profile containing more bacteroides, clostridia and enterobacteria. The introduction of solid food during weaning causes major changes in microbiota composition. The formula-fed infants' microbiota develops more quickly than that of breast-fed infants (14). At approximately 12 months, the microbiota of all infants begin to adjust to that of adults, although a similarity may be reached much later in childhood. The colon of adults harbours a complex and relatively stable bacterial ecosystem composed of bacteroides, bifidobacteria, eubacteria, lactobacilli and smaller numbers of clostridia, enterobacteria and streptococci (15). More than 400 species have been estimated to constitute the dominant microbiota. However, new techniques measuring rRNA or DNA have suggested that up to 80% of the bacterial RNA in faeces is not-accounted-for known microorganisms. Later in life, various physiological changes are commonly observed. It is difficult to determine whether the concomitant alterations in bacterial composition in elderly subjects are the cause or effect of these physiological changes (16). It has been shown in vitro that bifidobacteria strains isolated from elderly subjects adhere less to human intestinal mucus than those of adults, which explains the decrease in the number of bifidobacteria with age (17). In contrast, the number of clostridia rise significantly; the number of lactobacilli, streptococci and enterobacteria also increase (13).
The production of SCFA reflects the sequential colonisation of the intestinal tract by its microbiota. The faeces of breast-fed infants contain mainly acetate and lactate, whereas those of formula-fed infants are characterised by large amounts of acetate and some propionate (14). Butyrate, however, can hardly be detected in either group. In contrast, propionate and butyrate account for approximately one third of the SCFA in stool samples of adults (18). No age-related changes in SCFA production seem to occur until approximately 70 years of age.
The aim of the present study was to compare the fermentation patterns of an RS type 3 preparation by faecal microbiota obtained from humans at different ages. Experiments with inocula from breast-fed infants, formula-fed infants, infants at weaning, adults and elderly subjects were performed by applying a well-established in vitro batch fermentation method. Lactulose, known to be easily fermented and to stimulate beneficial bacteria in the colon (19), was used as a positive control in all experiments. To determine the amount of endogenous fermentation capacity of the inocula, blank samples without substrates were included in each experiment.
MATERIALS AND METHODS
A new type of resistant maltodextrins, chemically classified as type 3 retrograded starch (RS-3; Actistar; Cargill-Cerestar, Vilvoorde, Belgium), was used as substrate in this study. It is obtained from a partially hydrolysed tapioca starch by enzymatic debranching using an isoamylase, followed by subsequent retrogradation (20). It consists of linear α-D-glucans, of which more than 50% have a polymerisation degree between 10 and 35 in the resistant fraction. To remove digestible compounds, it was subjected to an in vitro digestion (21). Lactulose (purity, ≥98.0%; Fluka Chemie GmbH, Buchs, Switzerland) was used as an easily fermentable standard.
The inocula were obtained by collecting faecal samples from different population groups, as described in Table 1. The faecal materials of individuals within every age group were mixed and prepared as described earlier (22). The exclusion criteria for all groups were antibiotic treatment during the last 4 weeks (for breast-fed infants, this criterion was also applied to their mothers), other medical treatment affecting the alimentary system, gastrointestinal disorders and special diets. Additional criteria are listed in Table 1. Moreover, all donors were identified as nonmethanogenic.
In Vitro Fermentation
In vitro fermentations were performed by applying a standardised batch technique (22). Incubations were performed under strictly anaerobic conditions with inocula mixtures of each age group as described above. Blank samples (inoculum without substrate) were used as negative controls. Duplicate samples were taken at 0, 4, 6, 8 and 24 hours. Mercuric chloride was used to stop the fermentation. The pH, total gas, hydrogen and SCFA production in an acidified aliquot of the supernatant were determined, as described in the standard procedure. The residues were freeze dried, ground in a ball mill and used to quantify the substrate degradation. Starch disappearance was measured as total starch using the Megazyme total starch assay kit (23). Starch was solubilised using dimethyl sulfoxide and quantitatively hydrolysed to glucose using amyloglucosidase. Subsequently, glucose was quantified enzymatically using the hexokinase/glucose-6-phosphate-dehydrogenase assay (24). Residual lactulose in fermentation residues was determined by high-performance anion exchange chromatography with pulsed amperometric detection on a Dionex BioLC system (Dionex Corporation, Sunnyvale, CA). Lactulose was extracted with water and eluted (flow rate, 0.25 mL/min) on a CarboPac PA1 column with 6 mmol/L NaOH (duration, 0–25 min), followed by a linear gradient to 22 mmol/L NaOH (duration, 25–30 min) and an isocratic elution at 22 mmol/L NaOH. A 0.5-mol/L NaOH (flow rate, 0.05 mL/min) was added postcolumn, and detection took place in an ED50 electrochemical detector (Dionex Corporation) with potentials as follows: E1 = 0.10 V (duration, 0.00–0.40 s); E2 = −2.0 V (duration, 0.41–0.42 s); E3 = 0.60 V (duration, 0.43 s); and E4 = −0.10 V (duration, 0.44–0.50 s). The integration window was from 0.20 to 0.40 seconds. For quantification, 2-D-deoxyribose was used as an internal standard, and calculations were performed using PeakNet chromatography software (Dionex Corporation).
Total gas production and hydrogen accumulation during the 24-hour fermentation period are shown in Figure 1. All inocula led to a fast total gas production from lactulose and blank samples, which was almost completed within the first few hours. From RS-3, however, it differed considerably between the 5 inocula. Whereas the production in breast-fed and formula-fed infants' faecal slurries was rather similar to blank values, the amounts reached end values comparable with those obtained from lactulose. Moreover, the elderly subjects and the infants at weaning showed a delayed increase in total gas production compared with adults. Hydrogen accumulation from lactulose differed substantially between the inocula. The slurries obtained from infants at weaning, from adults and from elderly subjects showed the typical pattern of hydrogen being an intermediate product and reaching a maximum value after approximately 4 hours. The breast-fed and formula-fed infants' microbiota, however, led to a continuing increase in hydrogen. The values obtained for RS-3 were nearly identical to those of blank samples. Despite the low accumulation, it can be observed that hydrogen produced from RS-3 by breast-fed and formula-fed infants' inocula did not decline, thus leading to somewhat higher values after 24 hours compared with blank samples.
Figure 2 gives an overview of SCFA production, patterns and changes in pH value. The values of both lactulose and blank samples obtained from adults' faecal samples are well within the variation limits based on more than 50 experiments conducted with the same fermentation system. In breast-fed infants, however, the total SCFA production from lactulose was found to be slightly lower. Most remarkably, their pattern was clearly dominated by acetate, which correlates well with the exceptionally low pH value reached in this sample. A clear tendency of decreasing total SCFA production from inocula blank samples with increasing age was observed. Concerning RS-3, both SCFA production and pH value decrease exhibited clearly visible differences between the inocula. In breast-fed and formula-fed infants, the amounts hardly deviated from those of blank samples. However, in the inocula obtained from infants at weaning, the adults and the elderly subjects, a steady increase of total SCFA during the whole fermentation period was observed, and the end values were found to be 3 to 4 times higher than those of the blank samples. Moreover, a tendency of pH values to approach those of lactulose was observed. Both effects point clearly to an increasing ability to degrade RS-3 with increasing age. However, slight differences in SCFA distribution were observed. The fermentation of RS-3 by microbiota from infants at weaning led to a proportional increase of the 3 main SCFAs during the whole fermentation period. A similar but less pronounced effect was found in the experiment with the elderly subjects. Rather surprisingly, no further increase in butyrate concentrations after 8 hours was observed in the experiment with adults, leading to unexpectedly low end values of this metabolite.
Substrate degradation can be seen in Figure 3. Lactulose was easily and completely fermented by all population groups. However, there was a somewhat delayed decline detected in the elderly subjects. As expected from the results shown in Figures 1 and 2, the degradation patterns of RS-3 were found to be distinctly different. The most rapid metabolisation, compared with all of the other population groups, was observed in adults, whereas a clearly delayed decrease observed in the slurry obtained from the elderly subjects, comparable with that of infants at weaning, was found. In the slurries obtained from breast-fed and formula-fed infants, however, most of the substrate resisted metabolisation within 24 hours.
The blank samples from breast-fed infants were characterised by clearly higher hydrogen accumulation and slightly elevated amounts of total gas compared with those of all of the other groups. Moreover, there was a pattern of decreasing SCFA production from inocula blank samples with increasing age. This trend was also reflected in decreasing amounts of residual dry matter in blank inocula (results not shown), pointing to a metabolisation of inoculum-derived unfermented compounds. The fact that higher amounts of SCFA were produced from microbiota of breast-fed infants, compared with formula fed, has been reported earlier (25). The clearly higher fermentation capacity is assumed to be due to the presence of human milk oligosaccharides in the inoculum, which may have been fermented during the in vitro experiment. Human milk contains considerable amounts of complex oligosaccharides (26), which have been shown to be only minimally digested in vitro (27). They seem to play a role as growth factors for bifidobacteria and as factors influencing the local immune system for the intestine in breast-fed infants (26). It has been reported previously that they are not completely fermented in the colon (28). Moreover, lactose partly escapes into the colon, where it is available for fermentation (15). To test the hypothesis that milk oligosaccharides were still present in notable amounts in the inoculum of breast-fed infants, blank samples of all age groups at time zero were analysed by high-performance anion exchange chromatography with pulsed amperometric detection. The chromatograms indeed showed considerable amounts of unfermented oligosaccharides, disaccharides and, possibly, monosaccharides to be present in the slurry of breast-fed infants (results not shown). However, more work would be needed to identify these carbohydrates.
According to published data, acetate and lactate dominate the SCFA profile in stool samples of breast-fed infants; only little propionate and no butyrate could be detected (29). In the present study, lactate concentrations in the samples were not measured. However, acetate was found to be clearly dominant in blank samples at fermentation time zero; very low amounts of propionate and no butyrate were present. Accordingly, the SCFA pattern obtained from formula-fed infants' faecal materials in this study corresponds well with literature values (29), both showing acetate-proprionate ratios of approximately 4:1 and only negligible amounts of butyrate. In adults, the major faecal SCFA are acetate, propionate and butyrate. The molar proportions determined in this study are in accordance with those reported for freshly collected stool samples (18) and for distal parts of sudden-death victims (30). In addition, agreement was found between the elderly subjects' stool samples and the baseline stool samples reported earlier (31).
Lactulose was found to be rather quickly and completely fermented by the microbiota of all age groups. However, distinct qualitative differences in fermentation metabolite production were observed. Hydrogen is known to be an intermediate in human colonic fermentation. Its accumulation in both infant groups before weaning leads to the hypothesis that hydrogen-reducing bacteria are only being colonised with the introduction of solid food.
With lactulose as a substrate for breast-fed infants' microbiota, acetate constituted 97 mol% of total SCFA, but no butyrate at all was found. This clearly points to the presence of specific acetate-producing bacteria. Moreover, it is known that butyrate production, which is thought important in adults, is at very low levels in breast-fed infants (18). The physiological significance of its virtual absence remains to be uncovered. However, milk-derived or locally synthesized growth factors control infection at this stage (32). The microbiota of formula-fed infants seems to have already diversified considerably because a shift from acetate to propionate in particular and partly to butyrate can be observed.
Taking all fermentation parameters into account, it can be stated that the ability to degrade RS-3 is only being established during weaning. Very likely, microbial enzymes that degrade α-glucans are lacking in the microbiota of infants before weaning. However, the bacteria of formula-fed infants may be slightly adapted to starch, which is often used in infant formulas, according to the producers' information. Yet, in the present study, no such effect could be observed. The bacteria collected from the faecal slurry of infants at weaning was able to ferment RS-3 slowly but completely, whereas those of adults were clearly better adapted to this substrate. For the elderly subjects, a slightly diminished activity was found.
The amounts of butyrate produced from RS-3 by microbiota of adults turned out to be exceptionally low. Although this result is unexpected because RS-3 is known to be butyrogenic (1,2) and former experiments led to higher butyrate proportions (6), ongoing studies have revealed that the donor-related effects on butyrate accumulation are great. Five experiments performed with the identical batch of RS-3 but varying donor collectives led to a mean butyrate production of 158 ± 95 μmol/100 mg substrate, whereas the variation for lactulose was considerably lower (218 ± 32 μmol/100 mg). Thus, it may be speculated that the differences in microbiota involved in butyrate metabolism may be responsible for this effect.
The results may indeed reflect the changes in bacterial composition during life. The breast-fed and formula-fed infants' microbiota were unable to ferment RS-3, very likely because of their low diversity. Bifidobacteria, dominating in breast-fed infants, are suggested to have a limited capacity to use RS as a substrate (32). Consequently, it can be stated that no beneficial effect from RS consumption may be expected for infants before weaning. Along with the introduction of solid food and, thus, complex carbohydrates, a change in colonisation occurs, resulting in a more diverse microbiota. At this stage of growth, increased consumption of RS-3 may be of as much interest as for adults. As for the elderly subjects, the information concerning diet-related effects on colon health is limited. However, the changes in microbiota composition could be of significance for health-promoting capabilities, such as relief from constipation or some immunomodulation (16). Whether RS-3 may contribute to such effects remains to be elucidated.
The assessment of bacterial composition of faecal samples and their changes during in vitro fermentations are considered a prerequisite for further studies. Moreover, more emphasis needs to be given to the investigation of the influence of complex carbohydrates, including RS, on human colonic microbiota in general (32). New molecular identification methods (33) will help elucidate such effects. A greater understanding of the colonisation process in early life is required before the changes in bacterial composition among infants in general may be taken into account. In vitro fermentations offer an appropriate screening tool to follow such alterations. After all, in vivo studies in this population group need to be conducted to investigate whether RS consumption may result in conditions that are assumed to be beneficial to health, similar to those observed in adults and in elderly subjects.
The authors thank Dr Christian P. Braegger (Department of Gastroenterology and Nutrition, University Children's Hospital, Zurich, Switzerland) and Dr Miriam Thumshirn and Bernadette Stutz (Department of Internal Medicine, Gastroenterology and Hepatology, University Hospital, Zurich, Switzerland) for their support in obtaining ethical approval and in organising the faecal sample collection. The authors also thank Caroline Fässler, Alessandra Frazzoli, Marianna Gulfi, Sandro Janett and Dorothea Kemmler for technical assistance.
1. Champ MMJ. Physiological aspects of resistant starch and in vivo measurements. J AOAC Int 2004; 87:749–755.
2. Asp NG, van Amelsvoort JMM, Hautvast JGAJ. Nutritional implications of resistant starch. Nutr Res Rev 1996; 9:1–31.
3. Velázquez OC, Rombeau JL. Butyrate—potential role in colon cancer prevention and treatment. Adv Exp Med Biol 1997; 427:169–181.
4. Smith JG, Yokoyama WH, German JB. Butyric acid from the diet: actions at the level of gene expression. Crit Rev Food Sci Nutr 1998; 38:257–297.
5. Robertson MD, Bickerton AS, Dennis AL, et al
. Insulin-sensitizing effects of dietary resistant starch and effects on skeletal muscle and adipose tissue metabolism. Am J Clin Nutr 2005; 82:559–567.
6. Brouns F, Kettlitz B, Arrigoni E. Resistant starch and “the butyrate revolution”. Trends Food Sci Technol 2002; 13:251–261.
7. Champ M, Langkilde AM, Brouns F, et al
. Advances in dietary fibre characterisation. 2. Consumption, chemistry, physiology and measurement of resistant starch; implications for health and food labelling. Nutr Res Rev 2003; 16:143–161.
8. Baghurst PA, Baghurst KI, Record SJ. Dietary fibre, non-starch polysaccharides and resistant starch: a review. Food Aust 1996; 48:3–35.
9. Englyst HN, Kingman SM, Cummings JH. Classification and measurement of nutritionally important starch fractions. Eur J Clin Nutr 1992; 46:33–50.
10. McCleary BV, Prosky L. Advanced Dietary Fibre Technology
. Oxford, England: Blackwell Science Ltd, 2001. Chapters 34–37.
11. Christian M, Edwards C, Weaver LT. Starch digestion in infancy. J Pediatr Gastroenterol Nutr 1999; 29:116–124.
12. Verity K, Edwards CA. Resistant starch in young children. Proc Nutr Soc 1994; 53:105A.
13. Mitsuoka T. Intestinal flora and health. Asia Pac J Clin Nutr 1996; 5:2–9.
14. Edwards CA, Parrett AM. Intestinal flora during the first months of life: new perspectives. Br J Nutr 2002; 88:S11–S18.
15. Edwards CA, Parrett AM. Dietary fibre in infancy and childhood. Proc Nutr Soc 2003; 62:17–23.
16. Tuohy KM, Likotrafiti E, Manderson K, et al. Improving gut health in the elderly. In: Remacle C, Reusens B, eds. Functional Foods, Ageing and Degenerative Disease
. Cambridge, England: Woodhead Publishing Ltd, 2004:394–415.
17. He F, Ouwehand AC, Isolauri E, et al
. Differences in composition and mucosal adhesion of bifidobacteria isolated from healthy adults and healthy seniors. Curr Microbiol 2001; 43:351–354.
18. Szylit O, Andrieux C. Physiological and pathophysiological effect of carbohydrate fermentation. World Rev Nutr Diet 1993; 74:88–122.
19. Havenaar R, van Dokkum W. Lactulose: a review on effects, clinical results, and safety aspects in relation to its influence on the colonic environment. In: Sungsoo Cho S, Dreher ML, eds. Handbook of Dietary Fibre
. New York, NY: Marcel Dekker, 2001:179–93.
20. Kettlitz B, Coppin J, Röper H, et al. Highly fermentable resistant starch. US Patent 2000: No. 6043229.
21. Lebet V, Arrigoni E, Amadò R. Digestion procedure using mammalian enzymes to obtain substrates for in vitro fermentation studies. Lebensm Wiss Technol 1998; 31:509–513.
22. Lebet V, Arrigoni E, Amadò R. Measurement of fermentation products and substrate disappearance during incubation of dietary fibre sources with human faecal flora. Lebensm Wiss Technol 1998; 31:473–479.
23. Megazyme International Ireland Ltd. Total starch assay procedure, 2001.
24. Boehringer GmbH, Mannheim, Germany. Methoden der enzymatischen Lebensmittelanalytik mit Einzelreagentien: Bestimmung von Glucose, 1984.
25. Parrett AM, Edwards CA. In vitro fermentation of carbohydrate by breast-fed and formula-fed infants. Arch Dis Child 1997; 76:249–253.
26. Kunz C, Rudloff S, Baier W, et al
. Oligosaccharides in human milk: structural, functional, and metabolic aspects. Annu Rev Nutr 2000; 20:699–722.
27. Gnoth MJ, Kunz C, Kinne-Saffran E, et al
. Human milk oligosaccharides are minimally digested in vitro. J Nutr 2000; 130:3014–3020.
28. Coppa GV, Gabrielli O, Pierani P, et al. Oligosaccharides from milk and their role in bacterial adhesion. In: Renner B, Sawatzki G, eds. New Perspectives in Infant Nutrition
. Stuttgart, Germany: Georg Thieme, 1993:43–9.
29. Edwards CA, Parrett AM, Balmer SE, et al
. Faecal short chain fatty acids in breast-fed and formula-fed babies. Acta Paediatr 1994; 83:459–462.
30. Macfarlane GT, Macfarlane S, Gibson GR. Validation of a three-stage compound continuous culture system for investigating the effect of retention time on the ecology and metabolism of bacteria in the human colon. Microb Ecol 1998; 35:180–187.
31. Kleesen B, Sykura B, Zunft HJ, et al
. Effects of inulin and lactose on faecal microflora, microbial activity and bowel habit in elderly constipated persons. Am J Clin Nutr 1997; 65:1397–1402.
32. Topping DL, Fukushima M, Bird AR. Resistant starch as a prebiotic and synbiotic: state of the art. Proc Nutr Soc 2003; 62:171–176.
33. Harmsen HJM, Wildeboer-Veloo ACM, Raangs GC, et al
. Analysis of intestinal flora development in breast-fed and formula-fed infants by using molecular identification and detection methods. J Pediatr Gastroenterol Nutr 2000; 30:61–67.
© 2006 Lippincott Williams & Wilkins, Inc.