By definition, the term probiotic is limited to viable microorganisms (1). However, treatments with their inactivated counterparts have also been demonstrated to have beneficial health effects, and in such cases, their usage may be favorable because of lower risks (2). Safety is a particular concern in infant therapy because infants' immature gut defense barrier readily allows translocation for the luminal bacteria with a potentially heightened risk for infections (3). The use of inactivated instead of viable microorganisms would also result in technical benefits in the form of longer shelf-life and reduced requirements for refrigerated storage (2).
Recent studies draw attention to the potential use of probiotic supplementation of infant nutrition to control atopic eczema and local and systemic inflammatory responses, and in early prevention of onset of atopic disease (4–6). The administration of exogenous bacteria has been suggested to counterbalance the aberrance in the infant gut microbiota associated with the development of atopy and manifestation of atopic eczema (7–11). Without the counterbalance, such microbiota may provide means for initiating or perpetuating allergic inflammation (10–12). An aberrant gut microbial composition, such as inadequate bifidobacterial biota, may also deprive the developing immune system from counterregulatory signals against the Th2-mediated allergic responses (8,13,14).
Our primary aim in this study was to test whether viability is an essential property for the probiotic activity of Lactobacillus rhamnosus strain GG (LGG) for the management of atopic eczema. We also determined the effects of viable and heat-inactivated LGG on the presence of some predominant gut bacterial genera to elucidate whether the endogenous microbiota may be a target of these treatments.
Subjects and study design
The study cohort comprised 43 infants who had been referred to the Turku University Hospital Pediatric Department for suspected cow's milk allergy. Their symptoms included pruritus, atopic eczema, gastrointestinal symptoms (such as loose stools, gas, and vomiting), and subjective symptoms reported by the parents (e.g., crying and restlessness). These infants were randomly assigned into placebo, viable LGG, or heat-inactivated LGG groups and accordingly given, in a double-blind manner, either unsupplemented extensively hydrolyzed whey formula (EHF, Valio Ltd., Helsinki, Finland; n = 10) or the same formula supplemented with viable (n = 17) or heat-inactivated (n = 16) LGG (ATCC 53103, Valio Ltd., Helsinki, Finland). The inclusion criteria for the infants were 1) tolerance to EHF, and 2) the infants could not suffer from acute diarrhea during the sample collection. These criteria were fulfilled by 35 of 43 infants (8 in the placebo group, 14 in the viable LGG group, and 13 in the heat-inactivated LGG group), and they formed the study population.
At enrollment, the mean age of the study population was 5.5 months (interquartile range [IQR], 3.5–6.8), the serum total IgE concentration (Phadebas IgE Prist, Pharmacia, Uppsala, Sweden), which describes the extent of atopic sensitization, was 19 kU/L (IQR, 2–32), and the SCORAD score, which describes the severity of atopic eczema, was 16 (IQR, 9–20). The SCORAD scores were assessed as previously described (15). Skin prick test reactivity (16) to egg, cow's milk, and wheat were observed in 12, 4, and 2 infants, respectively. Before or after the treatment period, 25 infants were assigned to open or double-blind placebo-controlled cow's milk challenge, and 15 exhibited an allergic response to cow's milk. The severity of atopic eczema and the presence of selected fecal microbial groups were assessed before the introduction of the randomly assigned formula and after the intervention period, which had a mean length of 7.5 weeks (range, 0.4–45.3; IQR, 3.4–8.0).
The Turku University Ethical Committee approved the study protocol, and written informed consent to participate was obtained from the children's parents.
Extensively hydrolyzed whey formula was supplemented with viable LGG in concentration equivalent to 1 × 109 colony-forming unit (cfu)/g, which we have previously shown to result in an approximate mean intake of 3 × 1010 cfu/kg body weight (5). The inactivation by heat treatment was performed by an independent microbiologist at Valio Ltd. Liquid LGG concentrate of pharmaceutical quality was heated in a flask in small aliquots (approximately 250 mL) in a boiling water bath. When the concentrate had reached the temperature of 80°C, the flask was heated for further 15 seconds, after which it was transferred to ice, and the concentrate was cooled to 5°C. Viable and inactivated concentrates were freeze-dried before added to EHF. The efficacy of the heat treatment and the bacterial concentration in the formulas was controlled using a standard plate count method.
Analysis of the bacteriology of fecal samples using genetic probes
The parents collected fecal samples by scraping a specimen from diapers after defecation. The specimens were immediately cooled at 6°C to 8°C and delivered to researchers within 24 hours. All the samples were frozen at −75°C directly on delivery. Some samples were not provided or contained inadequate amount of biomass for quantitative microbiologic analysis; the bacteriology was analyzed from the samples of 4 infants in the placebo group, 7 in the viable LGG group, and 10 in the heat-inactivated LGG group.
The fecal samples were homogenized, bacterial cells were fixed with paraformaldehyde, and fluorescence in situ hybridization (FISH) was performed as previously described (17,18). Bifidobacteria were enumerated with probe BIF164 (´5-CATCCGGCATTACCACCC) (17), bacteroides with BAC303 (´5-CCAATGTGGGGGACCTT) (19), lactobacilli and enterococci with LAB158 (´5-GGTATTAGCA(T/C)CTGTTTCCA) (20), and Clostridium histolyticum group bacteria (from here onward referred to as clostridia) with HIS150 (´5-TTAT GCGGTATTAATCT(C/T)CCTTT) (21). The probe HIS150 detects a representative proportion of the known clostridia species commonly present in the infant gut, e.g., Clostridium paraputrificum, Clostridium butyricum, and Clostridium perfringens (but not Clostridium difficile) (22). Total cell numbers were counted using a nucleic acid-stain 4´, 6-diamidino-2-phenylindole (DAPI). Cells were enumerated visually using a Leica Laborlux D epifluorescence microscope (Wetzlar, Germany). Fifteen microscopic fields were counted per assay.
Wilcoxon signed rank test was used to analyze whether the SCORAD scores or bacterial numbers within a treatment group were different before the intervention compared with the respective values after intervention. Kruskal-Wallis test was used to analyze whether the changes during intervention differed between the treatment groups. Mann-Whitney test was used for post hoc analysis. All these analyses were performed using StatView for Windows version 4.57 (Abacus Concepts, Inc.). Fisher test for 2 × 3 tables was used to analyze whether a diarrhetic episode that parents associated with the beginning of treatment was reported significantly more frequently within any of the treatment groups. This, the retrospective sample size, and power analysis were performed using SISA online software (23). Exact unconditional test for 2 × 2 tables was used to analyze whether high numbers of clostridia or bacteroides were more or less common among infants who had adverse gastrointestinal symptoms in response to heat-inactivated LGG than among the rest of the study population (24).
For ethical reasons, the recruitment of the study cohort was prematurely terminated, and the intervention assignment code was broken after complaints from children's parents concerning adverse gastrointestinal symptoms. These complaints were limited to the heat-inactivated LGG group; the parents of 5 of 13 infants reported that their child had diarrhea from several days to weeks after the introduction of the supplemented formula, whereas no adverse reactions were reported in the placebo or the viable LGG groups (P = 0.05).
The pilot study population was large enough to demonstrate that atopic eczema and subjective symptoms improved significantly in all three treatment groups during the study period (Fig. 1). However, the Kruskal-Wallis test indicated that some differences between the treatment efficacies exist (P = 0.08). According to the post hoc analysis, the mean decrease in the SCORAD scores tended to be greater in the viable LGG group than in the placebo group (P = 0.02).
The bacteriologic analysis showed that before the intervention, high numbers (defined as >1 SD greater than the mean) of potential lactobacilli antagonists (i.e., clostridia and bacteroides) were present significantly more frequently (P = 0.04) among the infants who had adverse symptoms after the intervention (3 of 5) than within the rest of the study population (2 of 16). None of the three treatments was observed to cause significant changes in the bacterial numbers (Table 1). The retrospective power analysis indicated that, depending on the bacterial genera, the power of this study to detect >1 log changes in the bacterial numbers during the treatment period at a double-sided 5% significance level was variable: 20% to 37% in the placebo group, 21% to 94% in the viable LGG group, and 51% to 91% in the inactivated LGG group.
In this study, adverse gastrointestinal side effects were associated with the supplementation of infant nutrition with heat-inactivated probiotics. This result presents a problem that should be addressed if further studies are conducted to evaluate the efficacy of using nonviable probiotics for the management or prevention of infant diseases. The bacteriologic data in this study indicated that high numbers of bacteroides and clostridia might have been a predisposing factor for these side effects. It is possible that some members of these genera respond antagonistically to the intrusion of the administered lactobacilli, e.g., by secreting diarrhetic toxins. The adverse symptoms would then be confined to the inactivated preparation because viable LGG could counterrespond and restrain the antagonistic activity. Another possibility is that the heat-inactivation process may cause denaturation of surface peptides and expression of heat-shock proteins, thus modifying the immunostimulatory properties of LGG in such way that the heat-inactivated form would induce inflammatory responses and consequently increase gut permeability. However, the production of significant amounts of heat-shock proteins during the relatively fast inactivation process seems unlikely.
The administration of viable LGG did not appear to cause changes in the bacterial numbers in the gut. Accordingly, a previous study of preterm infants showed that enteral feeding with this strain did not reduce the numbers of nosocomial pathogens (25), and another study of healthy adults demonstrated that oral administration of this strain did not increase the numbers of lactobacilli (26). However, in other studies with healthy adults, the consumption of viable LGG has been shown to result in an increase in the numbers of bifidobacteria, lactobacilli, bacteroides, and clostridia, whereas in milk-hypersensitive adults, the numbers of bifidobacteria remained unaltered (19,27). Whether the numbers of bifidobacteria would similarly increase in healthy infants remains to be assessed. To our knowledge, this is the first study to assess the effects of heat-inactivated LGG on the microbiota, and we found no indication that the genera studied would be quantitatively affected by this intake (Table 1). Due to the low power of the study and heterogeneous diets, the lack of changes in the microbiota must be interpreted with caution.
In conclusion, the results of this study support the previous reports demonstrating the efficacy of viable LGG for the management of atopic disease, whereas there was no indication of such efficacy with heat-inactivated LGG. The relatively small pilot study population does not allow an indisputable conclusion for the question whether viability is essential for this probiotic activity of LGG. However, the adverse gastrointestinal symptoms in heat-inactivated LGG group question the use of nonviable probiotics for infant therapy in general. The preliminary microbiologic assessments showed no indication that the beneficial effects of LGG treatment would be mediated by changes in the composition of the intestinal microbiota.
The authors thank Ms. Ulla-Maija Eriksson for her invaluable help in coordinating the study and in data recording, and The Academy of Finland for financial support. The infant formulas for the study were provided by Valio Ltd., Helsinki, Finland.
1. Fuller R. Probiotics in man and animals. J Appl Bacteriol 1989; 66:365–78.
2. Ouwehand AC, Salminen SJ. The effects of cultured milk products with viable and non-viable bacteria. Int Dairy J 1998; 8:749–58.
3. Duffy LC. Interactions mediating bacterial translocation in the immature intestine. J Nutr 2000; 130:432S–6.
4. Majamaa H, Isolauri E. Probiotics: a novel approach in the management of food allergy. J Allergy Clin Immunol 1997; 99:179–85.
5. Isolauri E, Arvola T, Sütas Y, et al. Probiotics in the management of atopic eczema. Clin Exp Allergy 2000; 30:1604–10.
6. Kalliomäki M, Salminen S, Arvilommi H, et al. Probiotics in primary prevention of atopic disease: a randomised placebo-controlled trial. Lancet 2001; 357:1076–9.
7. Bjorksten B, Naaber P, Sepp E. The intestinal microflora in allergic Estonian and Swedish 2-year-old children. Clin Exp Allergy 1999; 29:342–6.
8. Kalliomäki M, Kirjavainen P, Eerola E, et al. Distinct patterns of neonatal gut microflora in infants developing or not developing atopy. J Allergy Clin Immunol 2001; 107:129–34.
9. Bjorksten B, Sepp E, Julge K, et al. Allergy development and the intestinal microflora during the first year of life. J Allergy Clin Immunol 2001; 108:516–20.
10. Kirjavainen PV, Apostolou E, Arvola T, et al. Characterizing the composition of intestinal microflora as a prospective treatment target in infant allergic disease. FEMS Immunol Med Microbiol 2001; 32:1–7.
11. Kirjavainen PV, Arvola T, Salminen SJ, et al. Aberrant composition of gut microbiota of allergic infants: a target of bifidobacterial therapy at weaning? Gut 2002; 51:51–5.
12. Kirjavainen PV, Apostolou E, Salminen SJ, et al. New aspects of probiotics—a novel approach in the management of food allergy. Allergy 1999; 54:909–15.
13. Sudo N, Sawamura S, Tanaka K, et al. The requirement of intestinal bacterial flora for the development of an IgE production system fully susceptible to oral tolerance induction. J Immunol 1997; 159:1739–45.
14. Hessle C, Andersson B, Wold AE. Gram-positive bacteria are potent inducers of monocytic interleukin-12 (IL-12) while gram-negative bacteria preferentially stimulate IL-10 production. Infect Immun 2000; 68:3581–6.
15. European Task Force on Atopic Dermatitis. Severity scoring of atopic dermatitis: The SCORAD Index. Dermatology 1993; 186:23–31.
16. Isolauri E, Turjanmaa K. Combined skin prick and patch testing enhances identification of food allergy in infants with atopic dermatitis. J Allergy Clin Immunol 1996; 97:9–15.
17. Langendijk PS, Schut F, Jansen GJ, et al. Quantitative fluorescence in situ hybridization of Bifidobacterium
spp. with genus-specific 16S rRNA-targeted probes and its application in faecal samples. Appl Environ Microbiol 1995; 61:3069–75.
18. Apostolou E, Pelto L, Kirjavainen PV, et al. Differences in the gut bacterial flora of healthy and milk-hypersensitive adults, as measured by fluorescence in situ hybridization. FEMS Immunol Med Microbiol 2001; 30:217–21.
19. Manz W, Amann R, Ludwig W, et al. Application of a suite of 16S rRNA-specific oligonucleotide probes designed to investigate bacteria of the phylum cytophaga-flavobacter-bacteroides in the natural environment. Microbiology 1996; 142:1097–106.
20. Harmsen HJM, Elfferich P, Schut F, et al. A 16S rRNA-targeted probe for detection of lactobacilli and enterococci in faecal samples by fluorescent in situ hybridization. Microbiol Ecol Health Dis 1999; 11:3–12.
21. Harmsen HJM, Gibson GR, Elfferich P, et al. Comparison of the plate culture technique and fluorescence in situ hybridization using specific rRNA-based probes for the quantification of human faecal bacteria. FEMS Microbiol Lett 2000; 183:125–9.
22. Stark PL, Lee A. Clostridia isolated from the feces of infants during the first year of life. J Pediatr 1982; 100:362–5.
23. Uitenbroek DG. SISA-Fisher 2*5, SISA-sample size and SISA-power [SISA web site]. 1997. Available at: http://home.clara.net/sisa
. Accessed September 6, 2001.
24. Berger RL. Exact unconditional homogeneity/independence tests for 2X2 tables. Available at: http://www.stat.ncsu.edu/
∼berger/tables.html. Accessed November 4, 2001.
25. Millar MR, Bacon C, Smith SL, et al. Enteral feeding of premature infants with Lactobacillus
GG. Arch Dis Child 1993; 69:483–7.
26. Saxelin M, Elo S, Salminen S, et al. Dose response colonization of feces after oral administration of Lactobacillus casei
strain GG. Microbiol Ecol Health Dis 1991; 4:209–14.
27. Benno Y, He F, Hosoda M, et al. Effects of Lactobacillus
GG yogurt on human intestinal microecology in Japanese subjects. Nutr Today 1996; 31(suppl 6):9–11.