Fetal growth and adult disease

Recent findings suggest that ischaemic heart disease and stroke, and the associated conditions hypertension and noninsulin dependent diabetes, originate through impaired growth and development during fetal life and infancy. These diseases and certain others, including obstructive lung disease, may be consequences of ‘programming’, whereby a stimulus or insult at a critical, sensitive period of early life results in long term changes in physiology or metabolism (Lucas 1991). Animal experiments provide many examples of programming. For example, the insulin response to glucose is permanently impaired if rats are weaned onto a low protein diet for only 3 weeks (Swenne et al. 1987). This is likely to reflect persisting damage to the pancreatic [3 cells, which develop during late fetal life and infancy. Programming occurs because organs and systems mature during periods of rapid growth in fetal life and infancy. There are critical windows of time during which maturation must be achieved; and failure of maturation is to some extent irrecoverable. Nutritional deprivation in early life can programme the size and DNA content of many organs and systems. Which is affected depends on the precise time at which undernutrition occurs (Winick & Noble 1966). Suspicion that programming plays a major role in the development of cardiovascular disease arose from a series of geographical studies in England and Wales. These showed that the large differences in cardiovascular mortality (deaths from ischaemic heart disease and stroke) between different areas of the country are paralleled by similar differences in neonatal mortality around 70 years ago (Barker & Osmond 1986). In those days high neonatal mortality in a population indicated a high incidence of low birthweight and poor maternal nutrition. The programming hypothesis is now being systematically explored by studying adults in middle and old age whose birth measurements and infant growth were recorded. From 1911 onwards every baby born in the county of Hertfordshire was weighed at birth, visited periodically by a health visitor throughout the first year, and weighed again at one year of age. Records of these visits have survived so that it is possible to trace men and women born 60 and more years ago and to relate these measurements to the later occurrence of illness and death and to the levels of known risk factors for cardiovascular disease. Similar long term follow-up studies are being carried out in Preston and a number of other places in Britain where in the past maternity hospitals made unusually detailed measurements on all newborn babies. The first study in Hertfordshire was of 5600 men born in the eastern part of the county between 19 1 1 and 1930. Among men who weighed 18 Ib (8.2 kg) or less at one year of age, death rates from ischaemic heart disease were almost three times greater than among those who weighed 27 Ib (12.3 kg) or more (Barker et al. 1989). Death rates fell progressively with increasing weight at one year. There were similar, though less strong, trends with birthweight. Death rates from stroke showed similar associations. Examination of samples of men and women still living in Hertfordshire and Preston has shown that birthweight and infant weight are associated with adult blood pressure, glucose tolerance, plasma concentrations of fibrinogen, factor VII and apolipoprotein B, and with a tendency to store fat abdominally rather than peripherally (Barker 1992). These associations parallel those with death rates from cardiovascular disease in that higher early weight is associated with lower levels of each risk factor. The associations are remarkably strong and graded. For example, the percentage of men with impaired glucose tolerance, or non-insulin dependent diabetes, falls progressively from 40 among men with birthweights of 5.5 Ib or less to 14 among those with birthweights of 9.5 Ib or more (Hales et al. 1991). The associations between early growth and adult measurements are independent of social class, either at birth or currently, and of other confounding variables such as smoking. A striking feature of these findings is that different risk factors are each related to different patterns of early growth. For example, blood pressure is related to birthweight but not independently to weight at one year, whereas plasma fibrinogen concentrations are strongly related to weight at one year but are unrelated to birthweight. This suggests that the critical period when blood pressure is sensitive to programming is during fetal life not infancy. Examination of birth measurements additional to weight identifies two groups of babies who develop high blood pressure as adults. One group, who also develop impaired glucose tolerance, have below average birthweight, head circumference and length, taking account of gestational age. Babies in this group would be recognized clinically as symmetrically growth retarded, the result of sub-optimal growth beginning in early pregnancy. An additional characteristic is that they have an above average ratio of placental weight to birthweight. A second group of babies who develop high blood pressure have above average birthweight, head circumference and placental weightshirthweight ratio, but below average length. These babies, who are not usually recognized clinically, have ‘asymmetrical’ growth retardation, thought to result from slowing of growth near term. The relations between early growth and adult disease rates and levels of risk factors are continuous. They fall progressively up to the highest values of birthweight and weight at one year. If the criteria for successful fetal growth include adult health and longevity these findings reinforce the view that infants with significant intrauterine growth retardation need not necessarily be ‘light-for-gestational age’ (Altman & Hytten 1989). Intrauterine growth retardation seems to be widespread. It affects many babies whose birthweights are within the normal range, not just