A prospective study of proneness to acute respiratory illness in the first two years of life.

R. Douglas,A. Woodward,H. Miles,S. Buetow,D. Morris

DOI: https://doi.org/10.1093/IJE/23.4.818

IF: 9.685

1994-08-01

International Journal of Epidemiology

Abstract:BACKGROUND This study sought explanations for the proneness to respiratory events in young Australian children. METHODS Prospective respiratory symptom diaries on 836 children collected data on respiratory symptoms and episodes. Questionnaires to mothers and birth and pregnancy records provided 56 known and possible predictors which were tested against two summary respiratory outcomes in each of the first and second years of life. RESULTS The two summary respiratory variables recorded for first and second year of life give four outcome variables. In fitting multivariate regression models to predict outcomes, use of child care in early childhood and mothers' experience of respiratory illness in the 12 months before birth were significant predictors for all four outcomes. Number of siblings was a predictor for three of the four outcomes. Sleep difficulty during pregnancy in the mother, and respiratory hospitalization of the infant in the first year, were significant predictors for both first-year outcomes. Unexpected and unexplained findings emerged for alcohol intake during pregnancy, passive smoking and breastfeeding in relation to the second year respiratory outcomes. Less than 9% of variance in outcome scores was explained in any of the four multiple regression models but this rose to between 24% and 31% when a corresponding score from the other year was added to the model. CONCLUSIONS Proneness to respiratory illness is an important entity; its determinants are largely unknown and events in pregnancy or the perinatal period explain only a small proportion of the between-infant variability.