The effect of nutrition of the fetus and neonate on cardiovascular disease in adult life

Early evidence that long-term effects of an adverse environment in fetal life and infancy may be determinants of CVD in humans came from geographical studies. The large geographical differences in death rates from CVD in England and Wales remain largely unexplained. Variations in adult diet and cigarette smoking do not explain why the highest rates are in industrial areas in the north and west of the country, and in some of the less affluent rural areas such as North Wales. Rates are low throughout the south and east, including London. One possible explanation is that the causes of the geographical differences begin to operate not in adult life but during childhood. The existence of detailed records of infant mortality from the beginning of the century allows current death rates in any area of England and Wales to be compared with infant mortality rates 60 years or more ago. This comparison can be made with the country divided into 212 local authority groupings. The correlation between past infant mortality and current mortality from CVD is remarkably strong, the correlation coefficient being 0.73 (Barker & Osmond, 1986). Infant mortality is, of course, no more than a general indicator of an adverse environment. But such a strong relationship is, at the very least, suggestive that some aspects of poor living conditions in childhood determine risk of CVD in adult life. This conclusion was first put forward by Forsdahl (1977), who found a similar geographical relationship between infant and cardiovascular mortality in the twenty counties of Norway. The detailed infant mortality records in England and Wales make it possible to refine this general conclusion. The records distinguish neonatal mortality (deaths before 1 month of age) from post-neonatal mortality (deaths from 1 month to 1 year). They reveal the new, and perhaps surprising, clue that cardiovascular mortality in adults is more closely linked to neonatal mortality than to post-neonatal mortality (Barker et al. 1989b).