Background: Childhood allergy and sibship

There is consistent and strong evidence that the prevalence of atopy and hayfever are lower in children from large families or of higher birth orders1. The pattern is less consistent for childhood asthma, probably because its phentoype is less specific and easily confused with infection-related and other forms of non-atopic wheeze.

Sibship size and birth order are inevitably correlated - especially where average families are small - and it is not always easy to examine their independent effects. The apparent protection offered by increasing family size was originally ascribed to variations in the rates of early childhood infection and thus formed the basis for the 'hygiene hypothesis'2. This remains the most plausible and parsimonious explanation for much of the epidemiology of childhood allergy; but it does not preclude additional mechanisms related to birth order. For example, two studies of UK populations, separated in age by about 30 years, were unable to account for birth order effects by examination of recorded infections in early childhood3;4.

Within representative cohorts of families living in Ashford (Kent), Barcelona and Menorca we originally observed, on cross-sectional analysis, that mothers - but not fathers - who have given birth to more children are less often atopic. This relationship was not explained by maternal age5. Subsequently and prospectively, within the Ashford population, we reported that over seven years the loss of maternal atopy and hayfever are associated with a higher number of intervening pregnancies6.

One explanation for these observations is that they reflect changes in maternal and consequently maternal-foetal immunology with successive pregnancies, and that this in turn influences risk of allergic disease in the offspring. If this were the case one might expect to see a relation between parity and cord IgE. The observation that the birth order effect may be stronger if older siblings are male7, could reflect stronger immune responses of the mother to male compared to female foetuses. If maternal-foetal immunology is important, then one might expect a mother's change of partner to influence the birth order effect and for a child with older sibs born to a new father to have the same risk of atopy as a first-born child. This is analogous to the risk of pre-eclampsia which is higher in first than in subsequent pregnancies, but reverts to the risk for primiparous women in later pregnancies if the mother changes partner.