Anencephalus and water hardness in south-west Lancashire.

The well-known regional variation in the incidence of anencephalus in England and Wales, whereby the defect occurs more frequently in Wales and in northem England than in East Anglia and southeast England, has suggested a possible relationship to water hardness (Penrose, 1957; Stocks, 1970; Fedrick, 1970). When the whole area of England and Wales is examined, a strong negative correlation is found between these two variables. In order to establish whether the connexion is causal, further information is needed. First, correlations between water hardness and anencephalus need to be sought within much more circumscribed areas, as has been done in South Wales by Lowe, Roberts, and Lloyd (1971). Second, correlations need to be sought between tSnnual variations in water hardness and anencephaly within one area, especially where water hardness has been deliberately altered by the introduction of water-softening plant. If soft water does indeed contribute towards a higher incidence of anencephalus, it remains a matter of conjecture as to when and how this effect is brought about. Serial observations over a period shorter than a generation can only shed light on the possibility of short-term effects, i.e., the possibility that the hardness of the water consumed by a parent at the time of relevant gametogenesis, or by a pregnant woman in early pregnancy, may affect neural-tube closure in the embryo. The possibility of an effect operative during the mother's prenatal life or early childhood could be investigated only over a much longer period. Congenital malformations have been ascertained systematically in the population of Liverpool since 1960 (Smithells, 1962). From July 1963 until June 1968, the area of ascertainment was extended to include neighbouring parts of south-west Lancashire. This is a predominantly soft-water area but includes Southport, where the water is hard, and St. Helens, where it was hard at the beginning of the study period but was softened after three years (Table I). All data on anencephalic births have oeen checked by reference to maternity department records. These sources were also used to provide information