primary prevention of coeliac disease - Associazione Italiana ...

OR (95%CI)
INFANT FEEDING PRACTICES AND COELIAC DISEASE
47
study was population-based, with a high participation rate, and the results should thus
be representative for Sweden at large. Since our findings are consequences of
biological phenomena, they should be valid for infants in general. Only incident cases,
i.e. newly diagnosed cases, were included, which reduced the recall period. A
comprehensive questionnaire concerning children's diet and health in general was
mailed to the families; it did not reveal our special interest in coeliac disease. A semiquantitative
food frequency questionnaire was used to assess both the consumption of
gluten-containing cereals at the time when these cereals were introduced into the diet
and the total amount consumed by seven months of age. The amount of glutencontaining
flour in homemade foods was calculated based on standard Swedish recipes
and the amount in industrially produced foods was obtained from the manufacturers.
Multivariate analyses were used to adjust risk estimates for confounding and to suggest
57
causal relationships .
Our main finding was that the risk for coeliac disease was reduced if the child was
breast-fed during the time period when gluten-containing foods were introduced [Odds
57
Ratio (OR) = 0.59, 95% Confidence Interval (CI) 0.42-0.83] . This protective effect
was even more pronounced if the child was also breast-fed beyond the period of gluten
introduction (OR=0.36, 95% CI 0.26-0.51) (Fig. 2). These risk estimates were adjusted
for the age of the infant when gluten was introduced into the diet and the amount of
gluten that was given.
4
2
1
½
¼
Discontinued Continued Continued beyond
Breast feeding status at introduction of gluten
Fig. 2. Breast-feeding status (BF) at introduction of gluten-containing flour into the diet
and risk (OR, 95% CI) for coeliac disease before two years of age.
A protective effect of breast-feeding was supported by our ecological study of the
Swedish epidemic, i.e. using aggregated data to explore any temporal relationship
between the changes in incidence rate and changes in infant dietary patterns. Both the
rise and later fall in incidence had a temporal relationship to a change in the proportion
36
of infants introduced to gluten while still being breast-fed .
It is important to note that at the time of these studies the majority of Swedish