Fermentation

Fermentation of inulin and oligofructose by lactic acid-producing bacteria results in an increase in bacterial biomass and the production of SCFA (acetate, propionate and butyrate), lactic acid and the gases CO₂ and H₂. The SCFA are utilised rapidly by the microbial flora or are absorbed, mainly through the mucosal surface of the caecum and ascending colon, and used as energy sources by the host. Mucosal cells use butyrate efficiently for their maintenance. In this way, endproducts of inulin and oligofructose fermentation can help to strengthen the intestinal wall by providing valuable sources of energy to the intestinal epithelium. Other SCFA (acetate and propionate) cross the intestinal wall without being metabolised to a significant extent and are delivered to the systemic circulation, where they reach the portal blood and enter the liver. The estimated energy value is about 4·2 kJ (1·0 kcal)/g for inulin and about 6·3 kJ (1·5 kcal)/g for oligofructose (Roberfroid et al. Reference Roberfroid, Gibson and Delzenne1993).

Dietary fibre properties

Through their fermentation and stimulation of bacterial growth, both inulin and oligofructose increase bacterial biomass and therefore have stool-bulking properties. In man, inulin and oligofructose have a bulking index (g stool weight increase per g fibre ingested) of about 1·5–2, which is comparable with soluble dietary fibre components, such as pectins and gums (Den Hond et al. Reference Den Hond, Geypens and Ghoos2000). A description of the dietary fibre properties of inulin and oligofructose is beyond the scope of the present paper and the reader is referred to Cherbut (Reference Cherbut2002) and Nyman (Reference Nyman2002).

Fermentation by bifidobacteria and lactobacilli in cultures

Batch-culture experiments with isolated bacteria in vitro have demonstrated that bifidobacteria (Wang & Gibson, Reference Wang and Gibson1993, Reference Wang and Gibson1994; Gibson & Wang, Reference Gibson and Wang1994a; Hopkins et al. Reference Hopkins, Cummings and Macfarlane1998; Bielecka et al. Reference Bielecka, Biedrzycka and Majkowska2002; Van der Meulen et al. Reference Van der Meulen, Avonts and De Vuyst2004) as well as certain lactobacilli, such as Lactobacillus plantarum (Cebeci & Gürakan, Reference Cebeci and Gürakan2003), are able to grow well in culture medium with inulin or oligofructose as the single source of carbon. The unique characteristic of bifidobacteria to grow on inulin and oligofructose can be ascribed to the presence of an inducible β-fructofuranosidase enzyme able to hydrolyse the β (2–1) glycosidic linkages between the fructose moieties. In addition, it appears that bifidobacteria are equipped with specific transport systems for oligofructose uptake when the DP is below 8 (Janer et al. Reference Janer, Rohr, Pelaez, Laloi, Cleusix, Requena and Meile2004). This suggests that bifidobacteria hydrolyse inulin and oligofructose at the cell surface, degrade the shorter fractions intracellularly, and at the same time metabolise the released fructose moieties. In this way, bifidobacteria prevent the loss of digestion products to nearby microbial competitors. This specific transport system provides bifidobacteria with the ability to compete efficiently with other commensal bacteria in the gut for this growth substrate (Van der Meulen et al. Reference Van der Meulen, Makras, Verbrugge, Adriany and De Vuyst2006). Fructosidases able to hydrolyse fructose moieties have been identified in certain lactobacilli, enabling them to degrade and metabolise inulin-type fructans efficiently (Barrangou et al. Reference Barrangou, Klaenhammer and Altermann2005).

Prebiotic effect in vitro

The selective growth of bifidobacteria on inulin and oligofructose in more complex media was demonstrated first by Wang & Gibson (Reference Wang and Gibson1993). In batch cultures inoculated with faecal slurries, a selective increase in numbers of viable populations of bifidobacteria was seen solely when inulin and oligofructose were added to the medium. This effect could not be obtained when other carbohydrates (pectin, polydextrose or starch) that had a more general effect upon bacterial genera were used as substrates. Further studies in vitro were able to reproduce the positive effects that inulin and oligofructose have on the numbers of bifidobacteria. In addition, a parallel decrease in faecal pathogens, bacteroides, clostridia and coliforms, was observed with inulin and oligofructose (Gibson & Wang, Reference Gibson and Wang1994b). Studies in vitro with more sophisticated, three-stage continuous-culture systems configured so as to reproduce the conditions in the three major regions of the human bowel also demonstrated the bifidogenic effects of oligofructose, inulin, and a combination of both (Synergy1), and this effect was observed even along the different parts of the colon (proximal, transverse and distal). Oligofructose stimulated bifidobacteria growth mostly in the proximal colon, whereas Synergy1 and HP-inulin proved to be particularly efficient for improving bifidobacteria growth towards the more distal regions of the colon (GR Gibson and KM Tuohy, unpublished results).

Prebiotic effect in human subjects

Antagonistic activities in vitro

Data from batch cultures have shown that when oligofructose is added to pure cultures of Bifidobacterium infantis and two potential harmfully organisms (Escherichia coli and C. perfringens), a rapid decline in counts of both pathogens can be obtained as compared with conditions where glucose is present in the fermentation medium (Wang & Gibson, Reference Wang and Gibson1993). The fermentation of oligofructose by bifidobacteria led to an increase in the concentration of acetate and lactate in the medium, which lowered the pH of the culture. At a pH of 4·5–5·0, the numbers of E. coli and Clostridium fell to zero whereas the growth of bifidobacteria was unaffected (Wang & Gibson, Reference Wang and Gibson1993). Subsequent co-culture experiments done by the same researchers (Gibson & Wang, Reference Gibson and Wang1994b) showed that in the presence of oligofructose, several species of bifidobacteria are able to excrete antimicrobial substances with a broad spectrum of activity. In this way, the growth of several species belonging to the genera Salmonella, Listeria, Campylobacter and Shigella as well as Vibrio cholerae was affected. Fig. 1 illustrates the degree of inhibition of pathogenic micro-organisms by bifidobacteria (Gibson & Wang, Reference Gibson and Wang1994b). This antagonistic activity of bifidobacteria against the growth of Salmonella enterica ser. Typhimurium and E. coli was confirmed recently by Makras & De Vuyst (Reference Makras and De Vuyst2006). According to these authors, the antagonistic activity of bifidobacteria against pathogenic Gram-negative bacteria appears to be widespread among the bifidobacteria. Other co-culture experiments performed by Fooks & Gibson (Reference Fooks and Gibson2002) with bifidobacteria or lactobacilli in the presence of enteropathogens such as E. coli, Campylobacter jejuni and S. enteritidis again showed an inhibition of the growth of these pathogens when either oligofructose or inulin was added to the medium. Particularly, the combination of B. bifidum Bb12 or L. plantarum with oligofructose was very effective, with the latter causing a 6-log decrease in the numbers of E. coli and compromising the growth of C. jejuni and S. enteritidis to undetectable levels (Fooks & Gibson, Reference Fooks and Gibson2002). Fig. 2 shows the inhibition of enteropathogens by L. plantarum in co-culture experiments in the presence of oligofructose or starch (data from the latter are added for comparison) (Fooks & Gibson, Reference Fooks and Gibson2002).

Fig. 1 Antimicrobial effect of bifidobacteria (Bifidobacterium infantis) measured by the clear areas on agar plates (inhibition zone) inoculated with the test bacteria and surrounding discs containing methanol–acetone extracts of B. infantis. Values are means of triplicate determinations, with their standard errors represented by horizontal bars (Gibson & Wang, Reference Gibson and Wang1994b).

Fig. 2 Changes in the numbers of the pathogens Escherichia coli, Campylobacter jejuni and Salmonella enteritidis (colony-forming units (cfu)/ml) in co-culture experiments with Lactobacillus plantarum with the addition of oligofructose () or starch () after 24 h incubation. Values are means of three determinations, with standard errors represented by vertical bars. There was a significant change in pathogen numbers from the baseline (start of the fermentation): * P < 0·01, ** P < 0·001 (Fooks & Gibson, Reference Fooks and Gibson2002).

These results show that lactic acid-producing bacteria have antibacterial activity against pathogens. The production of organic acids (acetic acid and lactic acid) and antimicrobial substances that are active against virulent micro-organisms might be two of the underlying inhibitory mechanisms. The presence of increasing concentrations of organic acids acidifies the medium. The lower pH is of benefit for their growth, since lactic acid-producing bacteria are very acid-tolerant and, hence, their growth is less affected. This characteristic provides bifidobacteria and lactobacilli with the advantage to survive acidic environments that inhibit the growth of certain noxious bacteria. Bacteriocins are bactericidal proteinaceous molecules produced by bacteria. The bacteriocin family includes a wide variety of peptides and proteins in terms of their size, microbial targets and mechanisms of action and immunity. More research is needed to characterise these molecules and to elucidate their mechanisms of action with regard to their inhibitory potential against pathogen viability (Servin, Reference Servin2004).

Data from epithelial layer experiments in vitro

An in vitro model of the small-intestinal tissue of pigs was used by Naughton et al. (Reference Naughton, Mikkelsen and Jensen2001) to study the effect of non-digestible oligosaccharides on intestinal colonisation by S. enterica ser. Typhimurium and non-pathogenic E. coli. Since the small intestine is the main site of invasion for both of these bacteria, the model is very suitable. Pigs were fed a diet containing inulin (4 %) for 4 weeks. Sections of jejunum and ileum were taken from the pigs and challenged with Salmonella and E. coli. Fig. 3 illustrates the numbers of E. coli associated with intestinal piglet tissue after challenge (Naughton et al. Reference Naughton, Mikkelsen and Jensen2001). The numbers of Salmonella sp. associated with the ileal sections from animals fed the inulin diet were lower (although not significantly) compared with the control diet. Jejunal cultures from animals fed inulin showed a significant reduction (P < 0·05) in the E. coli numbers associated with the tissue v. the controls (Naughton et al. Reference Naughton, Mikkelsen and Jensen2001). Other in vitro models are confluent T-84 intestinal epithelial monolayers and intestinal epithelial–mesenchymal cell co-culture models that mimic spatial and functional interactions of these cell types in vivo. It appeared that SCFA, which are products of oligosaccharide fermentation, improved barrier function and had mucoprotective effects (Van Tol, Reference Van Tol and Cavaillon2005).

Fig. 3 Numbers of Escherichia coli (log₁₀ colony-forming units(cfu)/g wet weight) in the jejunal tissue from pigs fed a diet with or without (control) inulin (4 %) after 1 h incubation (n 20 organ cultures). * Mean value was significantly different from that in the control pigs (P < 0·05) (adapted from Naughton et al. Reference Naughton, Mikkelsen and Jensen2001).

Studies in vitro show that inulin and oligofructose inhibit pathogen colonisation of the intestinal epithelial tissues and that endproducts of oligosaccharide fermentation can support the intestinal barrier function.

Data from animal studies

Evidence on the role of bifidobacteria against the development of disease came from inoculation of germ-free quails with a flora from patients with necrotising enterocolitis (which contained C. butyricum and C. perfringens). Lesions occurred rapidly after inoculation with the necrotising enterocolitis flora (thickening of the caecal wall with gas cysts, haemorrhagic ulcerations, and necrotic areas), whereas they did not in the presence of B. infantis and B. longum. Colonisation with bifidobacteria suppressed the growth of C. butyricum and led to the complete disappearance of C. perfingens (Butel et al. Reference Butel, Roland, Hibert, Popot, Favre, Tessedre, Bensaada, Rimbault and Szylit1998). Additional effects of supplementation with oligofructose upon those seen with bifidobacteria alone were demonstrated in subsequent investigations in similar models. Quails were harvested in either a gnotobiotic environment or an ordinary environment permitting post-colonisation by exogenous bacteria. In both environments, oligofructose (3 %) in the diet significantly increased the level of bifidobacteria and this increase was associated with a significant decrease (P < 0·05) in pathogen numbers (E. coli, C. perfringens, C. difficile, and C. ramosum) v. the control diet (lactose) (Catala et al. Reference Catala, Butel, Bensaada, Popot, Tessedre, Rimbault and Szylit1999). The research group confirmed their findings in a subsequent study (Butel et al. Reference Butel, Catala, Waligora-Dupriet, Taper, Tessedre, Durao and Szylit2001). According to the authors, oligofructose inhibited bacterial adhesion by acting as a cell-receptor analogue, promoting the clearance of bacteria (or their toxins). Pathogen numbers of several clostridia strains in the caecal contents of the birds fed the oligofructose diet (3 %) v. the control (lactose) diet are given in Fig. 4 (Butel et al. Reference Butel, Catala, Waligora-Dupriet, Taper, Tessedre, Durao and Szylit2001). Further evidence on the efficacy of inulin and oligofructose at enhancing enteric protection against challenge by an intestinal pathogen was provided by Buddington et al. (Reference Buddington, Donahoo and Buddington2002a,Reference Buddington, Kelly-Quagliana, Buddington and Kimurab). Mice were administered orally with Candida albicans, and their ability to clear the pathogen from the mid-small intestine was measured. The yeast densities in the contents of the small intestine were similar in mice fed oligofructose or inulin diets (10 %) but both were significantly lower (P < 0·05) than values in the control mice (cellulose). Bomba et al. (Reference Bomba, Nemcova, Gancarcikova, Herich, Guba and Mudronova2002) studied the administration of oligofructose together with L. paracasei in weaning pigs. Pigs were selected because they provide a model relevant to man for studies on gastrointestinal physiology. The administration of the synbiotic resulted in a significant increase in both lactobacilli (P < 0·01) and bifidobacteria (P < 0·05) in faeces compared with the controls. These changes were associated with a significant decrease in clostridia (P < 0·05), enterobacteria (P < 0·01) and enterococci counts (P < 0·001), as well as with a decrease in coliforms (NS) v. the control group (Bomba et al. Reference Bomba, Nemcova, Gancarcikova, Herich, Guba and Mudronova2002). Further studies in pigs were performed by Oli et al. (Reference Oli, Petschow and Buddington1998), who added oligofructose to an oral electrolyte solution (5 g/100 ml) to study changes in the microbiota in pigs with acute secretory diarrhoea due to the administration of a cholera toxin. This toxin was chosen because it can induce secretory diarrhoea in man as well. The pigs receiving the oligofructose solution had significantly more lactobacilli in their small intestine (P < 0·05) and caecum (P < 0·05), whereas densities of Enterobacteriaceae in the mucosa of the small intestine (P < 0·05), caecum (P < 0·05) and colon (P < 0·05) were significantly lower compared with the non-supplemented pigs. Bacterial densities of lactobacilli and enterobacteria at different sites of the pig intestine (small-intestinal lumen and mucosa, caecum and colon) in the oligofructose v. the non-supplemented group (after 24 h) are given in Fig. 5 (Oli et al. Reference Oli, Petschow and Buddington1998).

Fig. 4 Influence of oligofructose (3 %; ■) on the bacterial composition (Clostridium perfringens, C. paraputrificum and C. difficile; log₁₀ colony-forming units(cfu)/g) in the caecum of gnotobiotic quails after inoculation with a faeces from preterm infants suffering from necrotising enterocolitis. Values are means, with standard deviations represented by vertical bars. * There was a significant change in pathogen numbers compared with the control diet (□) (P < 0·05) (adapted from Butel et al. Reference Butel, Catala, Waligora-Dupriet, Taper, Tessedre, Durao and Szylit2001.)

Fig. 5 Bacterial densities (log₁₀ colony-forming units(cfu)/g wet weight) at different intestinal sites from the oligofructose-supplemented pigs (5 g/100 ml; ■) v. non-supplemented pigs (□). Values are means, with standard deviations represented by vertical bars. * There was a significant change in bacterial numbers compared with the control diet (P < 0·05) (Oli et al. Reference Oli, Petschow and Buddington1998).

The role of inulin and oligofructose in the defence against systemic invaders was assessed by recording survival in rats 2 weeks after intra-peritoneal infection with the pathogens Listeria monocytogenes and S. typhimurium. Systemic challenge with virulent strains of the pathogen L. monocytogenes resulted in 28 % mortality in the control group (cellulose), whereas none of the mice fed the inulin diet (10 %) died (P < 0·05). Mortality was intermediate for mice fed the oligofructose diet (10 %), but this was still significantly lower (P < 0·05) than that of the control mice. After Salmonella challenge, mice fed the inulin diet had significantly lower (P < 0·05) mortality rates compared with the controls, whereas mortality was intermediate with oligofructose, but this difference was not significant compared with the control group (Buddington et al. Reference Buddington, Kelly-Quagliana, Buddington and Kimura2002a). In the study reported by Bovee-Oudenhoven et al. (Reference Bovee-Oudenhoven, Ten Bruggencate, Lettink-Wissink and van der Meer2003), oligofructose (4 %) was shown to inhibit the colonisation of Salmonella in rats but to increase translocation to spleen and liver compared with the control diet (cellulose). One major drawback of this study was that rats were given diets very low in Ca, which itself can damage barrier function and enhance translocation. After increasing the Ca content of the diet, the rats recovered their resistance against translocation (Ten Bruggencate et al. Reference Ten Bruggencate, Bovee-Oudenhoven, Lettink-Wissink, Katan and Van der Meer2004).

These observations in various animal models of disease (quails, pigs, mice and rats) indicate that inulin and oligofructose contribute in accelerating the recovery of beneficial bacteria while slowing the growth of pathogens. Moreover, the results available provide evidence for a protective effect of inulin and oligofructose against pathogen colonisation (either endogenous or from the external environment) and systemic translocation, suggesting that inulin and oligofructose could contribute to improve the barrier function of the intestinal epithelium.

Human data

The ability of inulin and oligofructose to protect against pathogen colonisation and invasion is somewhat difficult to study in human subjects, since challenge studies are ethically not approved. However, indirect proof can be obtained either by investigation of the changes in microbial composition of the faeces or biopsies in several disease states or by the relief of disease and/or accompanying symptoms. Otherwise, intestinal barrier function can be assessed by measuring intestinal permeability in human subjects. This represents the passive crossing of the intestinal epithelium by water and water-soluble inert small molecules through intercellular spaces. In the study by Den Hond et al. (Reference Den Hond, Geypens and Ghoos2000), constipated subjects were administered HP-inulin (15 g/d) or placebo (sucrose) for 2 weeks. Inulin did not change intestinal permeability (determined by the measurement of urinary [⁵¹Cr]EDTA excretion) compared with the control group. This means that neither inulin nor endproducts of prebiotic fermentation exerted any negative effect on intestinal permeability. However, all subjects fermented the inulin well, since it affected stool frequency positively and increased faecal bulk compared with the controls (P < 0·05) (Den Hond et al. Reference Den Hond, Geypens and Ghoos2000).

In critically ill patients, the balance of the gut microbial ecology is altered in such a way that the numbers of beneficial bacteria are lower and the healthy interaction between the host and microbes is disturbed. Some of the resident bacteria are potential pathogens that can be a source of illness when the gut barrier is functionally or physically breached. In this way, patients with critical illnesses are generally known to have a high risk of bacterial translocation resulting in a worsening of the disease state. In the study by Jain et al. (Reference Jain, McNaught, Anderson, MacFie and Mitchell2004), patients at high risk for developing sepsis and admitted to an intensive care unit were randomised to receive either oligofructose (15 g/d) in a synbiotic preparation or a placebo (sucrose) for 10 d. Intestinal permeability was high at baseline and lowered after therapy but the difference between the synbiotic group and the placebo group was not significant. The synbiotic changed the bacterial composition in nasogastric aspirates in such a way that significantly lower (P < 0·05) potential pathogenic bacteria were found compared with the placebo. In another study, patients who had undergone major abdominal surgery and who were supplemented with oligofructose (32 g/d) and probiotics in a 2-week pre-operative course did not display any adverse effect on the barrier function when compared with the control group (sucrose). Bacterial translocation did not differ between these two groups (Anderson et al. Reference Anderson, McNaught, Jain and MacFie2004).

From the results of these studies, it appears that treatment with oligofructose during critical illnesses can restore the balance in microbial communities in a beneficial way without adverse effects on intestinal permeability or bacterial translocation. The way in which this positively affects the outcome of disease is investigated further.

Proof of the positive effect of oligofructose on the outcome of disease was obtained by Saavedra & Tschernia (Reference Saavedra and Tschernia2002). In their study they enrolled 140 infants (age 4–24 months) attending day-care centres, which implies that they are of high risk for infectious disease. Infants supplemented with oligofructose (0·55 g/15 g dry cereal) showed significant lower events of fever, medical attention and day-care absenteeism compared with the non-supplemented group (P < 0·05). In addition, the frequency of vomiting, regurgitation and abdominal discomfort was also significantly lower in the oligofructose group compared with the control group (P < 0·05).

Treatment with antimicrobials changes the gut microflora in a dramatic way and disrupts the normal ecological balance. This results in an altered gut barrier function with increased susceptibility to pathogens. The antibiotic cefpodoxime is often used to treat upper and lower respiratory tract infections. Treatment is associated commonly with an overgrowth of enterococci and yeast in the intestines, and colonisation with C. difficile also often occurs. Orrhage et al. (Reference Orrhage, Sjöstedt and Nord2000) gave volunteers an antibiotic treatment together with oligofructose (15 g/d) and probiotics or a placebo for 3 weeks. Administration of the antimicrobial agent induced a marked decrease in the anaerobic microflora, mainly with a loss of bifidobacteria and an overgrowth in enterococci. Numbers of lactobacilli were significantly higher (P < 0·05) in the synbiotic group compared with the placebo group. In addition, colonisation with C. difficile was significantly lower (P < 0·05) in the synbiotic group compared with the placebo group. In patients who developed C. difficile-associated diarrhoea triggered by antibiotic therapy, the administration of oligofructose was efficient. Administration of oligofructose (12 g/d) indeed increased numbers of bifidobacteria significantly (P < 0·0001) compared with their baseline levels. In addition, relapse of diarrhoea was much less common (P < 0·001) in the patients taking oligofructose compared with the placebo (sucrose) group (Lewis et al. Reference Lewis, Burmeister and Brazier2005).

Chronic gastrointestinal diseases such as inflammatory bowel disease (ulcerative colitis and Crohn's disease), pouchitis and colonic cancer have been linked to the composition of the colonic microbial community and its activities. The pathogenesis of inflammatory bowel disease involves multiple factors (genetic and physiological) of the host in relationship to a magnitude of complex interactions with the microbiota in the gut and the emergence of pathogens. These ‘silent’ pathogens are repressed when the gut microbiotia is functioning normally. Studies in patients with ulcerative colitis have revealed that their bifidobacterial populations are about 30-fold lower compared with that of healthy individuals (Macfarlane et al. Reference Macfarlane, Furrie, Cummings and Macfarlane2005). Supplementation of the diet with Synergy1 (12 g/d) together with a probiotic for 4 weeks in ulcerative colitis patients was able to restore these bacterial levels and resulted in a 42-fold increase (P < 0·05) in bifidobacterial colonisation in mucosal biopsies in the synbiotic group compared with the control ulcerative colitis group (maltodextrin) (Macfarlane et al. Reference Macfarlane, Furrie, Cummings and Macfarlane2005). The restoration of the microbial populations paralleled a decrease in the severity of the symptoms when the synbiotic was administered compared with the control patients (maltodextrin), indicating modulating effects on the progression of the disease. Indeed, sigmoidoscopy scores decreased (P < 0·05) in the Synergy1 group after the intervention compared with the placebo. Biopsies in the test group showed reduced inflammation and regeneration of the epithelial tissue, which did not appear in the controls. Synbiotic treatment of ulcerative colitis patients improved the full clinical appearance of chronic inflammation (Furrie et al. Reference Furrie, Macfarlane, Kennedy, Cummings, Walsh, O'Neil and Macfarlane2005). A reduction in inflammation and associated factors was observed also in patients with an ileal pouch–anal anastomosis after inulin therapy (24 g/d). Concentrations of butyric acid were increased significantly (by 62 %; P < 0·01) in patients receiving inulin compared with the non-supplemented group (Welters et al. Reference Welters, Heineman, Thunnissen, van den Bogaard, DipBact, Soeters and Baeten2002).

Studies in treated cancer patients and patients at high risk for developing the disease revealed beneficial changes in the composition of the microbial flora after administration of Synergy1. Data from the SYNCAN project demonstrated recently that addition of Synergy1 (12 g/d) in the diet of human patients increased the numbers of bifidobacteria and lactobacilli (P < 0·05), while at the same time lowering (P < 0·05) the number of coliforms and clostridia compared with cancer patients receiving a placebo (maltodextrin) (Van Loo et al. Reference Van Loo, Clune, Bennett and Collins2005). This was paralleled by changes in several biomarkers of colorectal cancer risk, indicating preventive effects on the development of the disease.

Data from well-controlled intervention trials on the use of inulin and oligofructose in patients with intestinal disorders or disease, or prone to critical illnesses, demonstrate that inulin and oligofructose restore balance when the gut microbial community is altered, which may inhibit the progression of the disease or may even prevent the disease from developing.