Dietary determinants of child asthma

Abstract

Although morbidity and mortality from asthma decreased in recent years, the burden of the disease remains considerable. The fetal origins of adult disease hypothesis claims that exposure at critical periods in fetal life may alter susceptibility to disease later in life. Numerous studies examined the association between maternal diet and pregnancy outcomes or child atopic disease with often conflicting results. This phenomenon may be due to considerable variation in study populations, sample size, study design, confounding control, and other methodological issues. Dietary influences on disease may also differ for different subgroups of atopic sensitization. This Thesis investigates the later issue in a novel way. Further, no previous study has investigated whether the effect of maternal diet on child asthma is mediated through or independent of the infant anthropometric status or non-milk fluids are associated with asthma. Methods: The 1988-89 Tasmanian Infant Health Survey (initially designed to study Sudden Infant Death Syndrome) provided data on the maternal diet during pregnancy and infant anthropometric measures. A semi-quantitative food frequency questionnaire collected data on maternal intake of protein, fat, carbohydrate, calcium, iron, magnesium, zinc, vitamin A, vitamin C, {u03C9}-3 and {u03C9}-6 polyunsaturated fatty acids, and dietary supplements. The anthropometric variables of interest included birthweight, birth length, head circumference, placental weight and derived indices. Childhood questionnaires administered during the 1995 Childhood Asthma Survey and 1997 Childhood Asthma and Respiratory Health Study collected information on atopic outcomes of wheezing, asthma, eczema, and hay fever. House dust mite (HDM) and ryegrass sensitization were defined by skin prick testing. Adjusted odds ratios (AORs) and their 95% confidence intervals (CIs) were derived from logistic regression models, and adjusted relative risks (ARRs) and their 95% CIs - from generalised linear models with log link function and binomial error structure. Associations between continuous dependent and independent variables were examined by multiple linear regression. The contribution of infant anthropometric status to the change in the magnitude of association between a nutritional risk factor and child asthma was calculated by the proportionate decline in the excess risk. Reliability of maternal recall was measured by kappa (k) and weighted kappa (kw). Results: This study found that an increase of 10g of absolute maternal protein intake per day was associated with a reduction in birthweight of 17.8g (95% CI: -32.7, -3.0, p = 0.02). Protein intake was also associated negatively with ponderal index ({u00DF} = -0.01, 95% CI: -0.02, -0.00, p = 0.01) and positively with relative head circumference ({u00DF} = 0.05, 95% CI: 0.01, 0.10, p = 0.02). A greater risk of asthma was found in infants who were heavier (AOR: 1.34, 95% CI: 1.10, 1.62), longer (AOR: 1.36, 95% CI: 1.12, 1.66) or fatter (AOR: 1.24, 95% CI: 1.02, 1.52) at birth as measured by standard deviation scores of these anthropometric variables. An increased risk of asthma was also associated with intakes of protein (AOR: 1.83, 95% CI: 1.13, 2.95), zinc (AOR: 1.08, 95% CI: 1.01, 1.15), {u03C9}-3 polyunsaturated fatty acids (AOR 1.6, 95% CI: 1.1, 2.4) and nuts (AOR: 2.04, 95% CI: 1.04, 4.00). Higher vitamin C intake decreased the risk of asthma hospitalization (AOR: 0.28; 95% CI: 0.08, 0.98). Birth length, head circumference and placental weight substantially ({u2265}10%) altered the magnitude of associations between selected nutrients and recent wheeze, any atopic sensitization, HDM sensitization, and, to a lesser extent, ryegrass sensitization. The findings suggest that different dietary factors may influence sensitization to different allergens differently. Specifically, a significant association between child fish intake and ryegrass-pure (AOR: 0.37; 95% CI: 0.15, 0.90) but not HDM-pure sensitization (AOR: 0.87, 95% CI: 0.36, 2.13) was found. In contrast, breastfeeding appeared to be protective against HDM (ARR: 0.71, 95% CI: 0.51, 0.99) but not ryegrass (ARR: 0.63, 95% CI: 0.28, 1.46) sensitization. Furthermore, breastfeeding was protective against asthma linked to HDM (ARR: 0.25, 95% CI: 0.07, 0.85) but not ryegrass (ARR: 0.81, 95% CI: 0.46, 1.43) sensitisation. None of the non-milk fluids appeared to be a significant predictor of asthma. When reliability of recalled infant nutrition was analysed, breastfeeding duration was recalled better (kw = 0.86) than the age of solids introduction (kw = 0.36-0.46). Presence of asthma and related conditions altered the reproducibility of recalling solids but not breastfeeding. Conclusions: This study raised the possibility that the effect of high protein in altering infant anthropometry at birth may involve changes in body composition. Examined for the first time, non-milk fluids did not appear to be significant predictors of asthma. Another important contribution from this Thesis was demonstration of the differential effect of diet for sensitization to different aeroallergens. Lastly, the Thesis showed poor agreement between prospective and retrospective reports of the age of solids introduction. The modest effect of maternal nutrition on the pregnancy outcome and child asthma found here is not surprising because associations in nutritional epidemiology are often of small magnitude and are susceptible to confounding bias and misclassification of dietary intake and disease outcome. Difficulty in separating maternal factors from environmental influences during pregnancy and the multifactorial aetiology of asthma may also play a role. Understanding what maternal nutrition constitutes optimal will help refine dietary recommendations. Due to ethical and practical difficulties, observational cohort studies, as reported here, will remain an important source of new knowledge. Greater understanding of molecular and genetic mechanisms involved in asthma should also be a considerable part of future research. Show more