Examining the hygiene hypothesis: the Trial of Infant Probiotic Supplementation


Michael D. Cabana, MD, MPH, Division of General Pediatrics, UCSF Children's Hospital, 3333 California Street, Laurel Heights Bldg #245, San Francisco, CA 94118, USA. E-mail: michael.cabana@ucsf.edu


The hygiene hypothesis suggests that the absence of infectious exposure at a critical point in immune system development leads to a greater risk for the later development of atopic disease. As a result, it may be possible to devise strategies that can block the onset of atopic diseases such as asthma. This paper outlines the rationale, background and design for the Trial of Infant Probiotic Supplementation study, which is designed to test the effectiveness of a daily infant probiotic supplement in the first 6 months of life in preventing the development of early markers of asthma.


It is hypothesised that environmental factors during early infancy can affect immune system development and subsequent risk of allergic disease.1 Within the mature immune system, T-helper (Th) cells are part of a system that help recognise foreign antigens and secrete cytokines to help activate other components of the immune system. The two subtypes of Th cells, Th-1 and Th-2 cells, are defined for the most part by the specific cytokines they produce. The hygiene hypothesis suggests that the absence of infectious exposure at a critical point in the development of the immune system leads to an unfavourable Th1/Th2 balance and, subsequently, a greater risk for the later development of atopic disease and asthma.2

There are considerable arguments in favour and against this hypothesis.3,4 However, if the hygiene hypothesis is correct, it may be possible to devise strategies that can establish a Th1/Th2 balance that blocks the onset of asthma or slows the progression of disease. The Trial of Infant Probiotic Supplementation (TIPS) study, funded by the National Institutes of Health, is designed to test the effectiveness of a daily infant probiotic supplement in the first 6 months of life in preventing the development of early markers of asthma. This paper outlines the rationale, background and design for the TIPS study. In addition, we outline specific challenges that we have encountered regarding study design and implementation.

Selecting an exposure: potential, feasibility and safety

An appropriate exposure to utilise in an intervention study should fulfil three criteria: feasibility of incorporating the exposure into routine care, a strong safety record and the presence of promising data for the use of such an exposure. There are many potential exposures linked to the hygiene hypothesis, but few fulfil all three criteria listed above.

First, the exposure should be feasible and practical. Specific environments (e.g. farms) are associated with decreased risk of developing asthma. Although trials of drastic environmental change to decrease asthma morbidity have been attempted,5 it would not be practical or cost-effective to expose all children at risk for asthma to these environments for sustained periods.6 Likewise, it would also be impractical for many families to incorporate specific exposures (e.g. livestock, farm animals) that are associated with a decreased likelihood of the development of asthma.

Second, the exposure should present a justifiable risk to the patient. For example, a history of infections, such as hepatitis A infection, has been associated with decreased risk for atopic disease.7 However, such infections carry serious risks that easily outweigh any theoretical benefits, and it would be unethical to encourage exposure to such pathogens. Although decreased antibiotic exposure has also been associated with a decreased likelihood of developing asthma,8 trials that limit the use of antibiotics would also present ethical challenges. An exposure should present low risk of harm, with some possibility of benefit.

Finally, the exposure should be based on published observational and in vitro studies that support the hypothesis that the exposure promotes a Th1 phenotype. For example, randomised controlled trials of a probiotic exposure, Lactobacillus, suggest benefit in decreasing the risk of eczema. In a randomised, controlled, double-blind study of 159 newborns, Kalliomaki et al. found that early Lactobacillus exposure as a probiotic supplement leads to decreased risk of atopic disease.9 In addition, a follow-up study found that such effects are sustained past infancy. Although the benefits of Lactobacillus exposure are only associated with the prevention of atopic dermatitis, early development of eczema is associated with later development of asthma.10 For example, the likelihood for developing persistent wheezing, after presenting with atopic dermatitis, is significantly higher, compared with the baseline likelihood of 12% in most populations. Furthermore, Lactobacillus is a promising exposure, based on in vitro studies. Several studies have reported results consistent with the hygiene hypothesis with Lactobacilli exposure leading to patterns consistent with a Th-1 phenotype.11,12

Rationale for probiotics

Based on these criteria, probiotic bacteria are a potentially promising, practical and justifiable antigen exposure. Probiotics are live micro-organisms, usually found as supplements or in fermented foods that benefit the host by stabilising the intestinal microflora.13 Recent studies have demonstrated the variety of ways they can be used in the management of infectious disease and prevention of diarrhoea.14

Lactobacillus is a probiotic bacterium that is commonly found in the childhood diet and found in many foods, such as yogurt.15 In addition, Lactobacillus is used as a supplement to prevent the development of diarrhoea.16 The efficacy of one probiotic strain does not imply that other strains will be equally efficacious. As a result, when reviewing the literature, it is important to note the organisms and preparations being used in a particular study, as other strains may not have equal effectiveness. For the TIPS study, based on the criteria above, we have selected Lactobacillus casei sps. Rhamnosus (Lactobacillus GG).

Two studies suggest that early exposure to Lactobacillus GG is associated with decreased risk of developing atopic dermatitis.9,10 Other studies that have utilised other Lactobacillus strains have not yielded similar results. The results from a randomised controlled trial of a 6-month exposure of Lactobacillus acidophilus to 231 infants showed that early probiotic supplementation did not reduce the risk of atopic dermatitis.17 A randomised controlled trial of a prenatal and a postnatal 6-month exposure of a combination of four probiotic strains and probiotics showed no effect on the incidence of allergic diseases by 2 years of age.18

Lactobacillus is a practical and accessible exposure, as it can be incorporated into the infant diet. Infants are commonly exposed to Lactobacillus already in foods commonly initiated during the first year of life. In many countries, probiotics are available as a supplement for infant formulas. This experience has been characterised by minimal side-effects or issues with the use of probiotics for infants.19

Finally, Lactobacillus is an exposure with justifiable risk with a long safety record and well-documented benefits. To date, only two cases of bacteraemia attributable to Lactobacillus supplementation, with identical molecular clinical and supplement isolates, have been reported in children. Both patients had complicated and prolonged hospitalisations, in which they received multiple courses of antibiotics via intravenous access. Furthermore, probiotic supplementation was initiated after both were symptomatic with profuse diarrhoea, possibly representing mucosal epithelial compromise.20 Although the potential risks of probiotic supplementation are low, care should be taken when probiotics are used with infants who are immunosuppresed or have complicated medical histories.

Study design

The TIPS study is a randomised, controlled trial designed to evaluate the effectiveness of Lactobacillus GG exposure in decreasing the likelihood of the development of early markers for asthma. Subjects are randomly assigned after enrolment in the nursery and the exposure occurs during the first 6 months of life. Figure 1 describes the study design and sequence of events. Initial recruitment occurs during the third trimester, with final screening, consent and randomisation performed in the nursery after the infant is born.

Figure 1.

TIPS study: study design.

Description of the intervention

The timing of the intervention differs from previous work, as the supplement is initiated in the postnatal phase without prenatal supplementation to the mother in the third trimester.9,10 The intervention group will receive a daily, 6-month course of Lactobacillus GG after the infant is born. The control group will receive a daily 6-month course of placebo, which will contain 325 mg of inulin. The intervention will start within the first 4 days of life, as infants in the intervention arm will receive 10 billion colony-forming units of Lactobacillus GG and 325 mg of inulin from each dose. This dosage level was selected as it is commercially available and has been demonstrated to achieve effective colonisation in children in previous studies.9,10 The supplement (active or control) can be dissolved in 2 mL of pumped breast milk, infant formula or sterile water and fed to the infant. To promote breast feeding, all participating families are encouraged to use breast feeding and are provided with a breast pump.

Subjects, parents and investigators are blinded to group assignment. The placebo does not differ from the contents of the active capsule in terms of appearance, taste, smell or texture. However, after the end of the intervention period at 6 months, parents are surveyed to determine whether they suspect that they have received the active or control capsules.

Compliance with the protocol will be assessed using three methods. Pill counts will be conducted at 1, 3 and 6 months, in addition to a parent survey during the same follow-up time points that will include questions about adherence to the protocol. Stool samples will also be obtained and cultured for Lactobacillus GG to determine whether infants in the control group were exposed to the probiotic, as well as to determine whether infants in the intervention group received the probiotic supplement.

Study site and population

The TIPS study is based at several institutions in San Francisco, CA, USA. The city has a diverse population that has an international character, as one-third of the population is from outside the US. Although the percentage of the population that is self-described as African American or white is less than the overall percentages in the US, the large size of the city overall (population 776 733 in 2000) offers a large population from which to recruit (Table 1) with a potential demographic mix similar to the US. The age distribution of the city offers a potential challenge, as the city has a low proportion of children (15%) compared with other large American cities.

Table 1.  Comparative demographics, based on 2000 US census data
LocationSan FranciscoCaliforniaUnited States
Population776 73333 871 648273 643 269
White (%)49.759.577.4
African American (%)7.86.711.8
Hispanic (%)14.132.412.5
Asian (%)30.810.90.8

Potential families are being recruited from several sites in San Francisco, as the study involves the collaboration of both private and public medical institutions. Specific enrolment criteria include infants who (i) have at least one parent who has a history of asthma and (ii) have parents who are willing to add the supplement/placebo to the baby's breast milk or formula for at least one feed each day. Exclusion criteria include patients with any major congenital birth deformities, acute illness at enrolment, or chronic conditions affecting food intake or metabolism.

Selection of endpoints

It is difficult to establish the diagnosis of asthma at a young age, as a significant proportion of children who wheeze have transient symptoms and never wheeze again.21 As a result, during the initial phase, the TIPS study will focus on the effect of the intervention on early markers of asthma. We will assess the infants for each of these markers monthly during the first year of life, and then every 6 months until the children are 3 years of age. Physical examinations will be performed at 1, 3, 6, 12, 24 and 36 months of age. Serum markers will be obtained at 12, 24 and 36 months of age.

Based on a review of the literature, we have selected published early markers of asthma that have a high predictive value. These early clinical markers of asthma include a history of frequent wheezing, wheezing without colds, atopic dermatitis and rhinitis using a standard history assessment developed by Castro-Rodriguez et al.22 Additional laboratory markers include elevated serum immunoglobulin E (IgE) and eosinophilia.

Elevated IgE level has been associated with severe asthma across different populations.23–25 Martinez et al. reported that children with wheezing at 6 years of age were more likely to have elevated IgE levels.21 Serum eosinophilia is non-specific but has been associated with hospitalisation for asthma,26 and asthma severity in some cases.27 Karakoc et al. reported that the development of asthma was associated with persistent eosinophilia.28

Wheezing without colds may be a more specific indicator of asthma because bronchiolitis, which commonly presents with wheezing in infancy, will also be characterised by upper respiratory symptoms or cold-like symptoms and fever.29 Allergic rhinitis and atopic dermatitis are other common childhood conditions that are associated with paediatric asthma.30

Confounding variables

Many factors are associated with the later development of asthma. Multiple-treatment interference will be addressed by measuring whether subjects have had other exposures based on published risk factors associated with the development of childhood asthma listed in Table 2.31

Table 2.  Variables that may affect the development of asthma
Genetic factors
 Maternal history of asthma
 Paternal history of asthma
 In utero
  Maternal stress
  Tobacco exposure
  Antibiotic exposure
  Maternal nutrition
  Maternal use of probiotics
  Maternal age
 Household tobacco exposure
 Socio-economic status
 Birth order/sibling exposure
 Farm exposure
 Pet exposure
 Travel exposure
 Daycare enrolment
 History and duration of breast feeding
 Timing of introduction to cows' milk
 Timing of introduction to solid foods
Medical history
 Mode of delivery (Caesarean section vs. vaginal delivery)
 Gestational period
 History of bacterial infection
 History of viral infection
  Respiratory syncytial virus infection
  Rhinovirus infection
 Frequency and duration of antibiotic use

For example, there are many factors in the gestational period alone that may potentially affect the development of asthma in childhood. The naïve immune system can be exposed to new antigens prenatally, through the placenta, or postnatally via the newborn diet.32 Because the prenatal period may be a crucial period to prime the Th phenotype, we will systematically survey the mothers of all subjects to assess for other factors that may affect the likelihood of developing asthma (e.g. prenatal infections, probiotic/yogurt intake during pregnancy).

As we will randomise subjects into the control and intervention groups, it is likely that such exposures will be evenly distributed between such groups. However, in our analysis, we will systematically assess whether these exposures are asymmetric.

Overview of analysis plan

The analysis will be designed to address two principal study questions: (i) the effect of Lactobacillus GG exposure on the proportion of infants with early clinical and immunological markers for the development of asthma; and (ii) the effect of Lactobacillus GG exposure on the time to presentation of early clinical and immunological markers for infants at risk for the development of asthma.

The intention-to-treat analyses will assume that the subjects who are lost to follow-up are random. For the first study question, multivariable logistic regression analysis for each of the early markers of asthma will be used to estimate the treatment effect, adjusted for the covariates, such as family history, dietary history, environmental exposures and adherence to the study protocol. Analysis of the second question will be similar, except that a proportional hazards model will be used.


The TIPS study is a randomised controlled trial designed to determine whether a daily probiotic exposure of Lactobacillus GG for the first 6 months of life can prevent the development of early markers of asthma. The causes of asthma are assumed to be multifactorial, and the analysis will take into account a number of familial and environmental exposures. This study will refine current understanding about the development and progression of asthma, as it will be one of the few studies that will examine the hygiene hypothesis using a randomised control design. The study will collect additional data to add insight to the mechanisms of how such an antigen exposure can affect the developing immune system.

This intervention is a potentially novel method for the primary prevention of asthma with enormous potential to have public health impact. Currently, there are no known methods to prevent the onset of asthma. In addition, the intervention we propose to study represents a potentially passive method to prevent asthma with large-scale public benefits, similar to folic acid fortification of grains or universal salt iodisation. Given the prevalence of asthma, the impact of such an intervention could potentially be tremendous.

Conflicts of interest

The authors have declared no conflicts of interest.


Funded in part by the National Institutes of Health (HL 080074) and the Clinical and Translational Science Institute (UL1 RR024131) at UCSF.