SEARCH

SEARCH BY CITATION

Keywords:

  • allergy;
  • atopy;
  • IgA;
  • mucosal inflammation;
  • probiotic;
  • synbiotic

Abstract

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Conflict of interest
  8. References

Kukkonen K, Kuitunen M, Haahtela T, Korpela R, Poussa T, Savilahti E. High intestinal IgA associates with reduced risk of IgE-associated allergic diseases. Pediatr Allergy Immunol 2010: 21: 67–73. © 2009 John Wiley & Sons A/S

Development of oral tolerance and its stimulation by probiotics are still incomprehensible. Microbial stimulation of the gut may induce a subtle inflammation and induce secretion of mucosal IgA, which participates in antigen elimination. In a cohort of allergy-prone infants receiving probiotics and prebiotics or placebo we studied intestinal IgA and inflammation in the development of eczema, food allergy, asthma, and rhinitis (allergic diseases). We performed a nested unmatched case–control study of 237 infants participating in a randomized double-blind placebo-controlled allergy-prevention trial using a combination of four probiotic strains pre-natally and during 6 months form birth. We measured faecal IgA, α1-antitrypsin (α1-AT), tumour necrosis factor-alpha (TNF-α), and calprotectin at the age of 3 and 6 months. By age 2 yr, 124 infants had developed allergic disease or IgE-sensitization (cases) and 113 had not (controls). In infants with high faecal IgA concentration at the age of 6 months, the risk of having any allergic disease before the age of 2 yr tended to reduce [odds ratio (OR: 0.52)] and the risk for any IgE-associated (atopic) disease reduced significantly (OR: 0.49). High faecal calprotectin at the age of 6 months associated also with lower risk for IgE-associated diseases up to age 2 yr (OR: 0.49). All faecal inflammation markers (α1-AT, TNF-α, and calprotectin) correlated positively with faecal IgA (p < 0.001). Probiotics tended to augment faecal IgA (p = 0.085) and significantly increased faecal α1-AT (p = 0.001). High intestinal IgA in early life associates with minimal intestinal inflammation and indicates reduced risk for IgE-associated allergic diseases.

In Western developed countries, hygienic environments have altered the commensal gut microbiota, which has weakened their stimulatory effect on the mucosal immune system, a system essential to the development of oral tolerance (1). Gut microbiota is essential for the initiation of an infants’ own immunoglobulin (Ig) A production (2, 3) and influences the development of the entire IgA system (4). IgA is the most plentiful immunoglobulin on mucosal surfaces, where it neutralizes harmless food antigens and prevents them from penetrating the epithelium (5). Therefore, IgA may be important in the development of non-responsiveness to harmless antigens, i.e., induction of oral tolerance. That low IgA and selective IgA deficiency associate with allergic manifestations, supports the important role of IgA in the protection against allergic diseases (6, 7).

To maintain immune homeostasis and tolerance, gut microbiota may stimulate the immune system via induction of a subtle intestinal inflammation (8–10). Intestinal inflammation may be assessed by measuring faecal α1-antitrypsin (α1-AT), tumour necrosis factor alpha (TNF-α), and calprotectin (11–13). α1-AT suppresses local mast cell degranulation and restrains in vitro IgE-dependent histamine release (11, 12). TNF-α is a pro-inflammatory cytokine secreted by intestinal epithelial cells in response to pathogen invasion (13), and calprotectin, which is abundant in the cytosol of neutrophils, lyses phagosomal membranes and inhibits growth of pathogenic bacteria (14).

Probiotics and prebiotics have been immunostimulatory (15–17) among atopy-prone infants (18), but the mechanisms of action are still somewhat obscure. In our cohort of allergy-prone infants, probiotics administered from birth to age 6 months reduced risk of atopic eczema by the age of 2 yr (19). In a nested case–control sample from these same infants, we sought to study whether early intestinal immune markers associate with the development of allergic diseases such as eczema, food allergy, asthma, and rhinitis, and the role of probiotics on them.

Methods

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Conflict of interest
  8. References

Study design

Our nested unmatched case–control study took place at the Helsinki University Central Hospital, and comprised 237 (23%) of the 1018 intention-to-treat infants participating in an allergy-prevention trial of probiotics. Pregnant mothers of risk families (mother or father with a doctor-diagnosed allergic disease) were randomized to receive a mixture of probiotic bacteria (Lactobacillus rhamnosus GG, L. rhamnosus LC705, Bifidobacterium breve Bb99, and Propionibacterium freudenreichii ssp. Shermanii, total amount of viable bacteria minimum 1 × 109) or a placebo for 4 wk before delivery. Their infants received the same probiotics along with 0.8 g of galacto-oligosaccharides daily for 6 months from birth, or a placebo. The study paediatrician examined the infants when they were 3, 6, and 24 months old, and if showing any symptoms of allergic diseases. Food allergy was diagnosed if the infant showed symptoms (urticaria, eczema, wheeze, vomiting, diarrhoea) in an open food challenge after improvement during a 2-wk elimination diet (20). Eczema was defined as an itchy skin condition and ≥3 of the following symptoms: familial history of atopic disease, dry skin during the last year, history of eczema and visible eczema involving typical sites (21). Asthma was defined as ≥2 physician-diagnosed wheezing episodes, accompanied by persistent cough or exercise-induced symptoms (22), and rhinitis according to ARIA (23). At the 2-yr visit, we performed skin prick tests against cow’s milk, egg white, wheat, fish, cat, dog, birch, and timothy, (24) and measured serum-specific IgE (ImmunoCAP Phadia, Uppsala, Sweden) against cow’s milk, egg white, cat, dog, birch, and timothy. Infants with one or more positive skin prick test reactions (wheal size ≥3 mm), or any serum-specific IgE concentration of ≥0.7 kU/l were considered sensitized (25). Allergic diseases associated with sensitization were defined as IgE-associated (atopic).

Objectives

Primarily, we studied the association of faecal IgA, faecal inflammatory markers and development of allergic diseases, sensitization, and IgE-associated (atopic) diseases from birth until the age of 2 yr. Secondarily, we studied the effect of probiotics on faecal immune markers.

Assessment of cases and controls

Cases (n = 124) were defined as infants who were sensitized at the age of 2 yr or were diagnosed eczema, food allergy, asthma, or rhinitis (any allergic disease) at any time-point between birth and the age of 2 yr (cumulative incidence). Non-sensitized infants who were not diagnosed any allergic disease at any time-point between birth and the age of 2 yr were controls (n = 113).

Faecal IgA and faecal inflammatory markers

Parents took faecal samples from the diapers of their 3- and 6-month-old infants, and froze (at −18°C) the samples within 15 min. The samples were brought frozen in screw-capped containers to the hospital, where they were stored at −40°C until analysed. These samples were homogenized with phosphate-buffered saline (w/v ratio 1:4) on a shaker for 30 min at +4°C and then centrifuged for 15 min at 10 000 g at +4°C. The supernatants were stored at −70°C for subsequent analysis. Faecal IgA concentrations were measured by enzyme-linked immunosorbent assay (details in 15, 26). The detection limit was 5 μg IgA/l. Faecal α1-AT was determined by a single radial immunodiffusion method. The antiserum (Orion Diagnostica, Espoo, Finland) was diluted to 1:37.5 in the agar. Diffusion was allowed to occur for 1 wk at +4°C. The detection limit was 80 μg/g. TNF-α was measured by the Quantikine HS TNF-α Immunoassay kit (R&D Systems, Inc., Minneapolis, MN, USA) with a detection limit of 5 pg/g. Faecal calprotectin was measured with an enzyme immunoassay (Phical Test; Valpro AS, Oslo, produced by NovaTec Immunodiagnostica, Dietzenbach, GmBH, Germany) with a detection limit of 9.75 μg/g. All samples were analysed in duplicate. For undetectable inflammatory marker concentrations, we used the value of the detection limits divided by two.

Statistical methods

Sample size

The sample size of the nested case–control study was based on the number of available samples of cases and controls.

Statistical analyses

Faecal immune markers between cases and controls were compared after logarithmic transformation of the skewed distributions with the t-test for independent samples (1-anova). The results are given as geometric means and 95% confidence intervals (CI). anova for repeated measures was used to analyse the effect of time on faecal markers. Immune marker concentrations, which were possible predictive variables for allergic diseases by the age of 2 yr, were divided into tertiles (T) and then the association of faecal markers with allergic diseases was analysed using logistic regression. Results are given as odds ratios (OR) and 95% CI.

Confounding variables

Breast feeding, gender and probiotic intervention may affect the primary outcome and were therefore assessed as possible confounding factors. As IgA, α1-AT, TNF-α, and calprotectin may be detected in breast milk, breast feeding was included as a covariate. Any antibiotic use during the first 6 months was considered a possible confounding factor for the probiotic treatment (2-anova) because less infants in the probiotic group (22%) used antibiotics than in the placebo group (32%) (19). Association between faecal IgA and inflammation markers at 3 and 6 months were analysed by Pearson correlation coefficient and by two repeated observations (27). Data were analysed with spss statistical software (version 15.0; SPSS Inc, Chicago, IL, USA).

Results

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Conflict of interest
  8. References

Of the 237 infants, one refused to participate in allergy tests. Of the 124 cases, 117 were IgE-sensitized, 99 had an allergic disease (95% eczema), and 92 had an IgE-associated allergic disease. The control group comprised 113 healthy non-sensitized infants (Table 1). Of the infants 120 received probiotics and 117 were on placebo. Of the cases, 94 (76%) had eczema, 30 (24%) had food allergy, 15 (12%) had asthma, and nine (7%) had rhinitis.

Table 1.   Characteristics of infants who by age of 2 yr went on to develop eczema, food allergy, asthma or IgE-sensitization (cases) and infants without an allergic disease or IgE-sensitization (controls)
 Cases n = 124Controls n = 113
  1. None of the characteristics in Table 1 were significantly associated with allergic diseases or sensitization by age 2 (univariate logistic regression analysis).

Male sex (%)5243
Birth weight (g) (s.d.)3674 (535)3577 (393)
Birth length (cm) (s.d.)51.0 (2.0)50.5 (1.5)
Vaginal delivery (%)7881
Partially breast-fed ≥6 months (%)7174
Solid foods initiated ≥4 months of age (%)6463
Age (in months) of initiating solid foods (s.d.)4 (0.7)4 (0.7)
Antibiotic treatment at age 0–6 months (%)2727
Maternal smoking (%)1013
Maternal smoking during pregnancy (%)03
Maternal allergy (%)8079
Both parents allergic (%)4441
Probiotic group (%)4755

Faecal IgA

The geometric mean faecal IgA among infants who by the age of 2 yr went on to develop IgE-sensitization, eczema, food allergy or asthma compared with the healthy, non-sensitized controls did not differ significantly (Table 2). However, in a logistic regression model to study the association of faecal IgA on later development of allergic diseases, infants with higher (T2 and T3) faecal IgA concentrations at 6 months of age were at lower risk of developing any allergic disease, more pronouncedly so for any IgE-associated (atopic) disease (Table 3). Compared with infants in the lowest tertile, infants with higher faecal IgA (T2 and T3) were also at lower risk for eczema (adjusted OR 0.44; 95% CI: 0.22–0.91 and OR: 0.52; 95% CI: 0.26–1.04, respectively) and for atopic eczema (adjusted OR: 0.45; 95% CI: 0.22–0.93 and OR: 0.51; 95%CI 0.25–1.03, respectively) (data not shown).

Table 2.   Faecal IgA and inflammatory markers (geometric mean, 95% CI) at 3 and 6 months of age in infants who by age 2 went on to develop an allergic disease or sensitization (cases) and in infants without an allergic disease or sensitization (controls)
Faecal markerAge (mo)Cases n = 124Controls n = 113*p-value
nG-mean95% CInG-mean95% CI
  1. *Significance between cases and controls by t-test for independent samples using logarithmically (ln) transformed values.

IgA (μg/ml)311016561503–18249918181665–19860.161
6119898813–993105980882–10880.239
α1-antitrypsin (μg/ml)3111410353–47799444386–5110.450
6119183155–217105200166–2410.482
TNF-α (pg/ml)311110.78.6–13.49912.810.0–16.40.288
61197.96.4–9.61057.86.1–9.80.929
Calprotectin (μg/g)3122180154–212111152127–1830.167
61243125–381133629–430.357
Table 3.   The incidence (%) of allergic disease and IgE-associated allergic disease up to age 2 yr in tertiles of faecal IgA and inflammatory markers measured at the age of 6 months
Faecal markerEczema, food allergy, or asthmaIgE-associated eczema, food allergy, or asthma
%CrudeAdjusted †%CrudeAdjusted †
  1. The results of comparisons between tertiles are given as odds ratios (OR) with 95% confidence intervals (CI).

  2. †Adjusted for probiotic intervention, gender and duration of breastfeeding (≥ 6 months); *p < 0.05; **p< 0.10.

  3. The lowest tertile group (T1) is the reference group in all logistic regression analyses. Tertiles (T1, T2, T3): IgA μg/ml at 3 months: <1420.68, 1420.68–2196.57, >2196.57; at 6 months: <711.85, 711.85-1235.13, >1235.13; α 1-antitrypsin μg/ml at 3 months: <375.48, 375.48–629.57, >629.57; at 6 months: <155.79, 155.79-311.66, >311.66; TNF-α pg/ml at 3 months: <6.38, 6.38–25.83, >25.83; at 6 months: <2.50, 2.50-12.84, >12.84; Calprotectin μg/g at 3 months: <123.76, 123.76–259.34, >259.34; at 6 months: <20.46, 20.46-51.00, >51.00.

IgAT1591.001.00581.001.00
T2410.48 0.24–0.96*p = 0.0380.43 0.21–0.90*p = 0.025390.47 0.23–0.94*p = 0.0340.42 0.20–0.89*p = 0.023
T3430.52 0.26–1.03**0.48 0.23–1.00**400.49 0.24–0.99*p = 0.0390.45 0.21–0.97*p = 0.041
α 1-antitrypsinT1501.001.00481.001.00
T2480.91 0.46–1.810.98 0.48–1.99440.85 0.42–1.720.92 0.44–1.90
T3450.82 0.41–1.61 0.88 0.43–1.77440.84 0.42–1.680.93 0.45–1.90
TNF-αT1441.001.00411.001.00
T2521.38 0.66–2.891.35 0.64–2.83501.44 0.67–3.081.33 0.62–2.84
T3491.19 0.63–2.231.16 0.61–2.18481.36 0.72–2.581.26 0.66–2.39
CalprotectinT1541.001.00521.001.00
T2480.80 0.41–1.540.80 0.41–1.55470.82 0.42–1.590.83 0.42–1.63
T3390.55 0.28–1.07**0.57 0.29–1.13340.49 0.25–0.98*p = 0.0440.51 0.25–1.04**

Faecal inflammation markers

Geometric means of faecal inflammation markers at 3 and 6 months of age were comparable between all cases and healthy non-sensitized controls (Table 2). Neither α1-AT, nor TNF-α, associated significantly with any allergic or any IgE-associated disease (Table 3). However, the highest level (T3) of faecal calprotectin at the age of 6 months associated to reduced risk for IgE-associated diseases (unadjusted OR: 0.49) and tended to parallel lower risk for IgE-associated eczema (unadjusted OR: 0.52; 95% CI: 0.26–1.04, p = 0.066). Faecal IgA at 3 months and notably at 6 months correlated positively with faecal inflammation markers. The correlation coefficients were statistically significant (p < 0.001) for IgA vs. α1-AT (r = 0.68), IgA vs. TNF-α (r = 0.30), and IgA vs. calprotectin (r = 0.66). In all infants, IgA, α1-AT, TNF-α, and calprotectin levels decreased significantly from 3 to 6 months of age (time effect p < 0.001).

Effect of probiotics and prebiotics on faecal immune markers

Probiotic and prebiotic treatment, compared with placebo, tended to elevate faecal IgA (p = 0.085) and TNF-α (p = 0.099), and significantly elevated α1-AT (p = 0.001) at the age of 3 months, but these differences disappeared with age (Table 4). Antibiotic treatment abolished the probiotic effect on IgA, α1-AT, and TNF-α, as the probiotic effect was evident only among infants without any antibiotic use during the intervention (data not shown). When adjusted for antibiotic use, faecal calprotectin at the age of 6 months was significantly higher in the probiotic group (p = 0.045).

Table 4.   Faecal immune markers (geometric mean, 95% CI) at 3 and 6 months of age in infants who received probiotics and prebiotics or placebo from birth to 6 months
Faecal markerAgeProbiotic n = 120Placebo n = 117
(months)G-meanMinMaxG-meanMinMax*p-Value**p-Value
  1. *Independent samples t-test.

  2. **Analysis of variance (2-anova); adjusted for antibiotic use during the 6-month intervention.

IgA (μg/ml )31832166120201633149617820.0850.107
6938842104693384710270.9360.838
α1-antitrypsi (μg/ml)35064495703573044200.0010.192
62011682401821532160.4380.926
TNF-α (pg/ml)313.410.616.810.18.012.90.0990.874
67.45.99.28.36.710.30.4180.428
Calprotectin (μg/g)31701442021631371930.7000.874
63629433125380.3220.045

Discussion

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Conflict of interest
  8. References

Having a high IgA level (above the median) in the faeces in early life associated with fewer IgE-associated (atopic) diseases at age 2 yr. Faecal IgA level correlated positively with intestinal inflammation markers. Probiotics, which initially induced a subtle intestinal inflammation, tended to augment faecal IgA.

More than 70% of infants in this study received breast milk during the 6-month intervention. Breast milk contains prebiotics, is rich in IgA, and may also contain α1-AT, TNF-α, and calprotectin. Faecal IgA includes the amount received from breast milk. It can be argued, whether the protective effect of high IgA on atopic diseases reflects the effect of breast feeding. However, considering breast feeding as a possible confounding factor had no significant effect on the main results.

Previously low salivary IgA has associated with sensitization, allergic rhinitis, and atopic eczema (28) and high salivary IgA has protected sensitized infants from allergic symptoms (18). Whether local IgA really protects from development of IgE-associated diseases or is merely a marker of immune maturation may be argued. That infants receiving colostrums with low concentration of IgA have increased risk for cow’s milk allergy, and those receiving colostrums with little casein-specific IgA have raised risk for allergy at age 4 (29, 30) suggest an important role for local IgA in the protection against allergy. Selective IgA deficiency is a risk factor for bronchial hyper responsiveness, however, the risk was associated with sensitization to mites (7).

The effects of probiotics on immune markers should be interpreted in the light that the study was designed to primarily learn about mucosal immune markers in the development of allergies. The tendency of infants in the probiotic group to produce more IgA is in line with previous knowledge. In mice L. johnsonii administered during weaning has increased faecal IgA and prevented atopic dermatitis from developing (31) and B. bifidum has enhanced intestinal IgA secretion (2). In bottle-fed infants L. rhamnosus GG along with B. lactis has increased the number of circulating cow’s milk specific IgA-secreting cells (16). In infants with eczema, L. rhamnosus GG and the same mixture of four probiotic strains as used in this study tended to increase faecal IgA (15). Furthermore, in the same infants L. rhamnosus GG increased interleukin (IL)-6 and the mixture increased IL-10, both of which are involved in IgA secretion. However, L. acidophilus administered to prevent allergies lacked any beneficial clinical effect or effect on mucosal IgA (32). That also prebiotics have stimulated the secretory IgA system may be due to their capability to enhance bifidobacteria in the gut (17).

Intestinal inflammation correlated with high faecal IgA levels. Also probiotics, which tended to augment faecal IgA levels, induced a subtle intestinal inflammation. Stimulation of toll-like receptors in antigen-presenting cells by indigenous intestinal microflora is a prerequisite for intestinal epithelial homeostasis and for oral tolerance (1, 33). Following antigen stimulation, mucosal dendritic cells secrete IL-6, which is a proinflammatory cytokine, and induce IgA secretion of naïve B-cells (34). Also regulatory cytokines, IL-10 and transforming growth factor-β, induce secretion of faecal antigen-specific IgA, which relates to the establishment of tolerance (35).

The initially high levels of inflammation markers and fading of the inflammation over time may be a physiological response to the establishment of commensal bacteria (9). The phenomenon was more pronounced in the probiotic group. However, high α1-AT concentrations observed in our probiotic group were 100-fold lower than reported for infants who manifest food allergy, eczema, or inflammatory bowel disease (13). In infants receiving probiotics in this same study, we observed increased concentrations of circulating inflammatory marker C-reactive protein and IL-10 (10). Probiotic bacteria seem thus to induce more a subtle than a clinical mucosal inflammation. In vitro studies have shown that several Lactobacillus strains elicit proinflammatory activities through NFκB pathway and elevate concentrations of proinflammatory cytokines such as TNF-α (8, 36). However, the inflammation is much weaker than that induced by pathogens (37).

In conclusion, high faecal IgA is a marker of immune maturation which associates with reduced risk for atopic diseases during the first 2 yr of life. Induction of inflammation may be among the mechanisms by which probiotic bacteria stimulate immune maturation.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Conflict of interest
  8. References

We thank Sanna Liuhanen for laboratory analyses and nurses Taina Koskikare and Anne Nikkonen for assistance during the study.

Conflict of interest

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Conflict of interest
  8. References

Financial support came from the Helsinki University Central Hospital Research Funds and Valio Ltd, Helsinki, Finland. K.K. has received salary, and M.K. has received part-time salary from the Clinical Research Institute of Helsinki University Central Hospital Ltd, partly funded by Valio Ltd. R.K. is employed by Valio Ltd Research Centre. The statistician (T.P.) received consulting fees from Valio Ltd. The researchers were independent of the company funding, and the clinical trial follows guidelines on acceptable publication practice.

References

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Conflict of interest
  8. References