Epidemiology of Congenital Malformations, by E G Knox and R J Lancashire, University of Birmingham. Pp 221 (HMSO 1991). 29.00

This is an analysis of the splendid Malformation Register which was kept in Birmingham for 21 years from 1964 to 1984. It had to be disbanded in 1985 owing to the restructuring of the local health authorities so that the single Birmingham Area Health Authority became five separate Health Districts, each with their individual and independent methods for recording health problems, thus providing an example of administrative procedures ruining research into the causes of ill-health.

Multiple sources were used for ascertainment of malformations with the aim that the Register would be accurate rather than an immediate source of trends. Data for a single year were not considered complete until ascertainment had continued for six years after birth, by which time the malformation rate had risen from 19.1 per thousand to 26.7 per thousand. As well as type and date of malformation, the following information was collected: maternal age, social class, ethnic group, parity, pregnancy interval preceding the birth of the malformed child, incidence of similar or other malformations in sibs and twins of patients, birth weight, gestation, address of mother, housing of mother, and type and place of delivery and labour.

The database recorded information on 9965 infants with one or more malformations. These included Down syndrome, neural tube defects, genito-urinary malformations, intestinal obstructions (omitting pyloric stenosis), cleft lip and/or palate, heart disease, dislocated hips and talipes, cataract, syndactyly and polydactyly and a few others. The data were analysed for long-term trends in temporal or geographic incidence.

The following were some of the observations made. Hypospadias showed a systematic increase in incidence; the decline of neural tube defects could not be entirely explained by selective terminations of pregnancies; certain specific heart malformations exhibited varying prevalences in different calendar years: neural tube defects were the only malformations to show a seasonal variation; there was a slight fall in the incidence of Down syndrome attributed mainly to decreasing maternal age; and there was a fall in the incidence of dizygotic twinning which was loosely related to their mothers' dates of birth. Looking at twins in more detail, it was noted that there was a deficiency of twins amongst spina bifida babies, but an excess for hydrocephalus, anal atresia and congenital heart disease; these last three malformations were associated with an unexpectedly high proportion of discordant pairs. The following malformations: exomphalos (106 babies), tracheo-oesophageal fistula (122), anal atresa (195), limb reduction deformities (168), cataract (62) and accessory auricle (222) showed no recurrence in sibs, and congenital heart disease (1317 patients) showed an unusually low recurrence in sibs.

The authors searched for any effects of proximity to motorways, factories, overhead cables, tall chimneys etc and found little association with malformations except that a greater proportion of children with cataracts were both near gasworks and incinerators compared to children with other malformations. Space-time interactions were sought and were most striking with hydrocephalus, where in addition, a relation with young maternal age was observed. Both these features demonstrate that infective factors are important in its aetiology. Investigations of communicable disorders showed an association between Coxsackie A infections and hydrocephalus, and between Coxsackie A9 and anophthalmos and atrial septal defect.

The data from the Birmingham Malformation Register show that environmental teratogenic effects must be involved in the aetiology of many malformations, and that amongst the viruses, several strains of Coxsackie A and B, and the Echoviruses may be involved in a chronic or recurring manner. The discrepancy between recurrence in twins and sibs, with the latter being greater than the former, suggests that genetic factors play a small part and/or that the mechanism of twinning is inextricably linked to the development of certain malformations.

This monograph has not received the attention it deserves. It is an outstanding example of how to collect data carefully and how to analyse it in the fullest ways possible. At 29 it is an undoubted bargain for those who wish to learn about the causes of malformations and about the applications of epidemiological methods of analysis.

Sarah Bundey FRCP, Senior Lecturer in Clinical Genetics