METHODS

Study population

The study population consisted of those children from the original BOKAAL study, the parents of whom had consented to being approached for follow up at age 5. The protocol for the study was approved by the Institutional Review Board of the Academic Medical Center of the University of Amsterdam, and written informed consent from the parents was again obtained for each child. A questionnaire was completed, assessing demographic variables, family characteristics and risk of atopy, indoor and outdoor risk factors such as pets and smoking, and an assessment of the current health status of the children and the medical history of the past years. All children (except for two living abroad) were seen by two investigators (MdJ and VS) and were examined for the presence of atopic eczema according to the ISAAC protocol,⁸ for (rhino)conjunctivitis, wheezing, and other clinical signs of allergy. A venous blood sample was taken for the assessment of specific IgE by radioallergosorbent test (RAST) against cows' milk, hen's egg, house dust mite (Dermatophagoides pteronyssinus), cat dander and dog dander, which were all tested at age 1; additionally grass pollen (a mix of Phleum pratense and Dactylis glomerata), tree pollen (a mix of birch, alder, hazel, and willow) and moulds (a mix of Aspergillus fumigatus, Alternaria alternata, Cladosporium herbarum, and Penicillium notatum) were tested. The results of the laboratory tests were expressed as negative (normal) or ranged from 1+ (dubious) to 5+ (strongly positive).

Outcome definitions

The primary outcome at age 5 was the presence of atopic disease according to the following categories: obvious, possible, and no atopic disease. We classified the children as having “obvious atopic disease” if any of the following symptoms occurred: (1) wheeze in the past year; (2) sneezing with itchy–watery eyes in the past year; or (3) itchy rash in the past year with any flexural involvement (around the eyes, around the sides or front of the neck, fronts of the elbows, behind the knees, or on the fronts of the ankles). Children who fulfilled at least one of the following criteria were classified as having “possible atopic disease”: (1) an affirmative answer to at least one of the questions: Has your child ever had asthma? Has your child ever had hay fever? Has your child ever had eczema?; (2) any flexural involvement ever, but no itchy rash in the past year; (3) itchy rash in the past year but no flexural involvement; (4) sneezing during April to September, but not accompanied by itchy–watery eyes; (5) a positive answer to the question: Is your child allergic for house dust, for pets, or for pollen? The remaining children were classified as having “no atopic disease”.

For the prevalence measures the following core questions on asthma, allergy, and eczema of the International Study on Childhood Asthma and Allergy (ISAAC) were incorporated in the questionnaire.⁹

Wheeze at age 5 was defined as a positive answer to the question: Has your child had wheezing or whistling in the chest in the past 12 months? In addition, the questions: How many attacks of wheezing has your child had in the past 12 months? and How often, on average, has your child's sleep been disturbed because of wheezing in the past 12 months? were used to get insight in prevalence and severity.

Sneezing with itchy–watery eyes, as an indicator for hay fever, was defined as an affirmative answer to the following two questions: In the past 12 months, has your child had a problem with sneezing or a runny or blocked nose when he/she did not have a cold or the flu? In the past 12 months, has this nose problem been accompanied by itchy–watery eyes?

Eczema was defined as a positive response to the question: Has your child had an itchy rash in the past 12 months? To gain insight in the severity of symptoms, we received categorical answers to the question: In the past 12 months, on average, has your child been kept awake at night by this itchy rash? Flexural dermatitis was defined as the presence of dermatitis at physical examination on at least one of the following five areas: around the eyes, around the sides or front of the neck, fronts of the elbows, behind the knees, or on the fronts of the ankles.

We defined a positive family history for atopy at age 5 if at least one of the parents reported to have ever suffered from asthma or hay fever (ISAAC questionnaire). Before birth of the child and at age 1 we included questions about shortness of breath in several circumstances and affirmative answers on sensitisation to house dust mite, pets, and pollen. We excluded questions on eczema from the definition of family history, as earlier experience showed that eczema in an adult population was very poorly associated with sensitisation.¹⁰ Siblings were excluded from the definition of family risk of atopy to make the association more pure. Maternal smoking during pregnancy was assessed separately from exposure to environmental tobacco smoke after birth, defined by the presence of any smoking inside the home. For exposure to pets, we included all pets during the first years and only furred pets kept inside the home at age 5.

Maternal education was used as a measure for socioeconomic status, in three categories of increasing educational achievement.

Sensitisation at age 5 was defined by any of the eight above mentioned RASTs being clearly positive (2+ or more).

We also combined the results of the clinical and laboratory assessments at age 5 by classifying children into three groups, as at age 1: (1) obvious atopy was defined as obvious atopic disease, regardless of the sensitisation by RAST, or as possible atopic disease with any positive RAST outcome; (2) possible atopy was defined as possible atopic disease without any positive RAST outcome, or as no atopic disease with any RAST positive outcome; (3) no atopy was defined as both negative clinical and negative laboratory outcome.

The cumulative incidence of atopic disease was analysed as the most severe clinical outcome at age 1, 2, or 5.

Data were analysed using SPSS 9. We calculated simple odds ratios (OR) or relative risks (RR) with corresponding 95% confidence intervals (CI) from cross tabulations. Multivariate analyses, adjusting for confounding factors, were conducted with binary logistic regression.

RESULTS

The BOKAAL study has a high response rate. About 90% of mothers asked by the midwives before birth agreed to participate (n = 1693). After birth, 160 children were not randomised (see de Jong and colleagues⁸ for reasons for exclusion). The drop outs occurred mostly during the second year, mainly because of logistic circumstances, and were equally distributed over the intervention groups. We analysed baseline characteristics and outcome at age 1 and 2 for responders and non-responders at age 5 and found no statistical differences, although the responders tended to have somewhat higher educated mothers who smoked less during pregnancy.

From 1992 to 1994, 1533 children were randomised for the first part of the BOKAAL study. At age 2, 1180 parents (in 1220 fulfilled questionnaires) agreed to follow up at age 5. Eventually, 1108 children (72.2% of the original cohort) participated in this second part of the study by questionnaire, 1058 of them underwent a physical examination, and a venous blood sample was taken from 943 children. As shown in table 1, no substantial differences were found in the baseline characterics of the study population between the intervention groups. Mean age at physical examination was 5 years 10 months (range 5.5–7.5 years).

Obvious atopic disease was found in 26.3% of the children exposed to cows' milk and in 25% of the placebo exposed children. The relative risk was 1.05 (95% CI 0.86 to 1.29).

Wheezing during the past 12 months at age 5 was reported for 62 children (11.5%) in the cows' milk intervention group (n = 541) and in 58 (10.3%) of the placebo group (n = 564). The associated odds ratio was 1.11 (95% CI 0.76 to 1.62). Furthermore, all the other analysed clinical outcome measures showed no differences between the cows' milk and placebo groups (table 2). When adjusted for duration of exclusive breast feeding, sex, parental history of atopy, smoking during pregnancy and smoking in the household at age 1, having older sibs, day care attendance at age 1, the introduction of solid foods before the age of 5 months, and pets at age 1, no substantial differences were found.

At age 1, we found 9.4% children in the cows' milk group positive for any RAST (2+ or more) versus 7.9% in the placebo group. At age 5, these figures were 17.3% versus 17.7%. As was to be expected, positivity for hen's egg was less frequent at age 5 than at age 1, while positivity for aeroallergens was much more frequent, mainly house dust mite and grass pollen (table 3). Sensitisation to cows' milk was found in 5.8% in the cows' milk group versus 3.0% in the placebo group, with a relative risk of 1.77 (95% CI 0.93 to 3.37).

A per protocol analysis was performed for all children who received at least three times the randomised supplementation of cows' milk or placebo (intervention feeding) and at least six weeks exclusively breast feeding (table 4). Again we found no differences between the intervention groups.

Parental history of atopy was assessed before birth and at age 5. The proportion of positive parental risk at birth (in the remaining group at age 5) was 37.6%, and 51% assessed at age 5 (560 children). This shows that parents of 181 children switched from negative at birth to positive at age 5. However, in this “switch group” there was no association with the outcome of obvious atopic disease at age 5 in a trend analysis, neither in the negative parental history group (Mantel–Haentzel with 1 df, value = 2.298, p = 0.13), nor in the positive group (Mantel–Haentzel with 1 df, value = 1.108, p = 0.29).

DISCUSSION

This study indicates that breast fed children who received cows' milk supplementation during the first three days of life showed the same risk of clinical or serological expression of atopic disease at age 5 as those who received a placebo formula. Again, the “dangerous bottle” concept could not be confirmed in our study group.⁶ It is unlikely that the amount of cows' milk was not sufficient to sensitise the children. As reported earlier, the mean volume of formula feeding in our study (120 ml) was considerably more than the estimated minimum amount required to induce sensitisation. No differences were found by Gustafsson et al, who reported a cumulative incidence of atopic disease in Swedish children at 7 and 11 years who received supplementary human donor milk or a cows' milk based formula in maternity wards.¹¹ Neither could they find a relation between the dose of cows' milk formula and the subsequent incidence of atopic disease, despite literature suggesting that small doses of foreign proteins may be more likely to provoke sensitisation than larger ones.¹²

Saarinen et al found, in her prospective study, that supplementation with cows' milk in maternity hospitals increased the risk of cows' milk allergy at 18 months, just like obvious parental atopy, in comparison with feeding of other supplements. Exclusive breast feeding did not eliminate the risk.⁷

As Hånson expressed, the positive effects of breast feeding are beyond dispute, as immunomodulatory, anti-inflammatory, nutritional, and other components are provided in human milk.¹³

We previously reported the outcome at age 1 and 2 and found no differences between the intervention groups regarding atopic dermatitis and gastrointestinal disease caused by food allergy. Now, at age 5, the incidence of atopic disease increased in both groups to over 25%, mainly because of wheeze and eczema. With regard to our whole cohort, 9% of the children at age 5 still showed flexural dermatitis at physical examination and 18% reported itchy rash in the past year. This is considerably more than the 9.6% prevalence of itchy rash Wieringa et al found in a cross sectional study in 6–7 year old Belgian children.¹⁴ It is also higher than the 13.5% flexural dermatitis found in Dutch school children, although in that group, as in our study, 25% also reported one or more atopic symptoms.¹⁵

However, our data indicate that if there is any effect of brief neonatal exposure to cows' milk, the effect is not as harmful as might be thought, even in high risk children, who are exposed to many risk factors.

Conclusion

Early and brief exposure to cows' milk in breast fed children is not associated with atopic disease or allergic symptoms up to age 5.