The clustering of respiratory diseases in early childhood.

BACKGROUND

It has been suggested that early childhood respiratory morbidity is related to asthma and chronic bronchitis in later life. To understand these relations, it is necessary to know how early childhood respiratory diseases are distributed, interrelated, and clustered. This study analyzed the distribution, interrelations, and the clustering of early childhood respiratory morbidity.

METHODS

We prospectively collected data of episodes of respiratory illness among 1140 children in the first five years of life, in family practice. To determine if the occurrence of respiratory illness was randomly distributed, we compared the frequency of respiratory illness to random Poisson distribution. We used factor analysis to determine if respiratory illnesses occurred in clusters or patterns according to type of illness.

RESULTS

Five percent of subjects never presented a respiratory disease, and 53% presented at least six episodes. The episodes of each illness were not randomly distributed, and most conditions were associated with each other. Factor analysis showed that asthma, acute bronchitis, and pneumonia formed one cluster, and tonsillitis, otitis media, and common cold another cluster.

CONCLUSIONS

Early childhood respiratory morbidity is unevenly distributed, but a clear-cut group of children with a much higher rate of more severe respiratory morbidity than other children could not be distinguished. Both upper and lower respiratory tract illnesses show clustering. In research on the long-term prognosis of early childhood respiratory morbidity, this morbidity may best be grouped into the categories identified in this research.