Early life and adult disorder: research themes

Early life and adult disorder: research themes M G Marmot International Centre for Health and Society, Department of Epidemiology and Public Health, U...
Author: Linda Sutton
2 downloads 1 Views 397KB Size
Early life and adult disorder: research themes M G Marmot International Centre for Health and Society, Department of Epidemiology and Public Health, University College London Medical School, London, UK

Correspondence fo: Prof. M G Marmot, International Centre for Health and Society, Department of Epidemiology and Public Health, University College London Medical School, 1 - ?9 Torrtngton Place, LondonWClE6BT,UK

This series of papers is about the role factors operating early in life may play in the cause of mental and physical disease in adulthood. Genes are the endowment coming from the earliest moment, when sperm penetrates ovum, that have an obvious influence carrying over into later life. They shape appearance and physiological function and, according to the new findings of behavioural genetics, influence how people think and act. Their influence, relative to environment, over the pattern and occurrence of disease in later life is still the subject of vigorous debate. The debate has, of late, taken a new form. There may be influences from early life that influence adult disease rates that are environmental, not genetic. These may act in utero or during infancy, childhood and adolescence. The debate is, therefore, not only of the relative importance of genes versus environment but of the relative importance of early environment versus later environmental exposures. No student of Freud has any doubt that early psycho-social environment has effects on subsequent mental health. Indeed, much of psychoanalysis is devoted to reversing the effects of early environmental exposures. In neuroscience, the concept of the critical period has had profound effect; i.e there is a period in development of an individual's brain during which it develops certain capacities if exposed to appropriate environmental stimuli; or perhaps fails to develop others if it is not exposed to the appropriates stimuli1. If the capacity is not developed at the appropriate time, the lack carries over into adulthood. Barker has suggested that there may be similar critical periods in other organ systems2. Exposures at this crucial time in development may have profound effect on susceptibility to disease in later life. There may be programming. Rubella, thalidomide..., we are in no doubt that influences on the mother in pregnancy can have dramatic influence on pathology in the offspring that stay with them for the rest of their lives. It is also clear that influences on the mother such as smoking, malnutrition, drug and alcohol abuse can influence outcome of pregnancy. What has been considered to a lesser extent is the influence of these more 'subtle' effects on subsequent health and disease. The chapters gathered in this issue explore aspects of the influence from early life on adult pathology. The purpose of this chapter is to pick

Bnfith Med/col Bulletin 199743 (No l):3-9

©The Bntuh Council 1997

Fetal and early childhood environment: long-term health implications

up some of the themes that emerge in these chapters. It is appropriate to start with two interlinking themes: the recognition of multiple determinants of disease and of interactions.

Gene - environment These two themes can be discussed with reference to debates about genes and environment. The dominant model in the social sciences has been that of social causation. At a time when some medical scientists were 'discovering' the importance of the social environment in the causation of disease, some social scientists have been pointing to the underappreciated importance of genetics. For example, Pinker, critical of previous environmental research, marshals the evidence for the genetic causation of language acquisition3. There is some chance that genetic causation may become the dominant paradigm in sections of both medical and social science. Four comments can be made. First, most risk factors for human disease do not show a discrete distribution in the population. Thus cerebral haemorrhage may be a discrete event-you have it or you do not. But blood pressure, which is the major risk factor that predisposes to it, is continuously distributed in the population. This is inconsistent with causation by a single gene. At the least it suggests the operation of multiple genes. In fact, and this is the second comment, there may be a difference between determinants of the position of the distribution, i.e the population mean and the determinants of where an individual lies within the distribution4. For most characteristics relevant to human disease, the position of the distribution is likely to be related to the environment; the position of individuals within a population will be determined by a combination of genes and environment. The third point is one made by Hales in his chapter5. The relative importance of genetic effects may be overestimated, at least in part, because environmental effects are mistakenly thought to be genetic. He takes the example of non insulin dependent diabetes mellitus (NIDDM) which is widely believed to be genetic in origin. Hales suggests that the estimates of the genetic component of NIDDM that come from twin studies are too large: (1) because monozygotic twins may have been chosen from diabetic clinics; and (ii) monozygotic twins may share similarities of intrauterine environment not shared by others. It may be these intrauterine influences that account for the concordance of NIDDM in monozygotic twins. He goes on to point out that there is not a clear pattern of inheritance of NIDDM and there is widespread variation in prevalence among populations that is unlikely to be genetic. 4

Br.tuf. Medico/BuH.f.n 1997,53 {No 1)

The influence of early life

The most general point is that the dichotomy between genes and environment is a false one. Adult height can be predicted from height in childhood. Is this a reflection of genes or environment? Karl Pearson showed that the height of sons is correlated with the height of fathers presumably largely a genetic effect. But the mean height of populations can change rapidly as a result of improved nutrition, i.e. environment. The height of any individual will be a reflection of genes and environment. This can be illustrated further with an example relevant to our theme. One animal model, much used for studying hypertension, is that of the spontaneously hypertensive rat (SHR). An experiment in cross fostering, however, showed the importance of early environment in expressing this characteristic6. When pups of SHR rats were cross-fostered with KyotoWistar mothers, they did not develop hypertension as they matured. This 'pure' genetic characteristic could not manifest itself as the phenotype of hypertension without the appropriate environmental stimulus. The chapter by Suomi illustrates a similar phenomenon (see below). This is an obvious example of gene-environment interaction. Sylva, writing as a psychologist in this issue7, presents 5 models of how nurture can modify nature in the development of a characteristic as an individual matures: maturation-emergence of the characteristic is programmed genetically and there is no environmental effect maintenance-the form of a skill or behaviour is innate, but it is maintained by environmental stimuli facilitation-a trait that would appear anyway but the environment influences the timing of its occurrence attunement- environment leads to lasting higher or lower performance of a genetically determined capability induction-skill or behaviour that is shaped completely by the child's experience. If these represent the variety of gene-environment interactions for the development of psychological traits, an obvious question is the extent to which they apply to development of predisposition to physical disease. Such evidence as we have, suggests that we cannot understand genetic effects without taking into account gene-environment interactions. In their chapter, for example, Doull and Holgate point out that genetic factors are important in asthma and atopy but there has been an increase in prevalence of asthma that must have some other explanation8. British M.d>co/ Buf/*hn 1997,53 (No 1)


Fetal and early childhood environment: long-term health implications

Early and later effects Only an extreme position on the effects of early life would suggest that later environmental effects are unimportant. The bulk of the evidence suggests that, as with genes and environment, there are interactions between effects from early life and influences operating later in life. Wadsworth illustrates with the example that poor fetal growth and adult obesity lead to increased risk of NIDDM. Similarly, hypertension is influenced by low birth weight, educational attainment and adult levels of obesity9. In discussing early and later life effects, Power and Hertzman consider a variant to interaction10. Rather they consider two types of effect of early life on adult health: latency and pathway models. The latency model is the effect that Barker describes. Exposure in utero during a critical period can programme the individual to susceptibility to subsequent disease risk. This, of course, does not exclude the possibility of interaction with later life influences. In the pathway model, people who are exposed to adverse circumstances early in life are exposed to adverse circumstances later. It may be these circumstances, or the cumulation of exposures that increase disease risk in adulthood. Power and Hertzman say that the evidence is insufficient to distinguish between the two models in general. There are, however, two clear examples in this collection in the field of mental health that illustrate the combined importance of early and later life effects. Maughan and McCarthy review the evidence for development of psychosocial disorders in adulthood11. They show that adverse childhood experiences are associated with elevated rates of psycho-social disorder in adult life. Important mediators of the relationship are: variations in severity or pervasiveness of early risk; individual differences in susceptibility; and interaction with later stressors. Although Maughan and McCarthy observe that knowledge of intervening factors is limited, their conclusions seem to fit with a pathway, or interactive, model. They state that 'longer term sequelae will depend on a series of shorter term links, each acting to maintain, amplify or reduce previous difficulties'. Specifically, women who were institutionalised as young children have poor mental health outcomes in adulthood. Positive school experiences and marriage to a non-deviant partner were associated with better outcomes. Pathways to poor marital support could be tracked through a series of intervening stages-lack of family support in adolescence, association with deviant peers, and early pregnancy. At each link in the chain earlier trajectories could be reinforced or at times redirected. It could perhaps be argued that this careful accumulation of evidence is not in fact evidence of a pathway model but that these intervening, reinforcing or protective steps are all the result of individual BnW,M«/.ca/Bu//«hn 1997^3 (No 1)

The influence of early life

characteristics determined by early exposure. This is far from a theoretical discussion. If risks are set early in life and apparently modifying factors are also set early in life, prospects for prevention given that early exposure has taken place would be limited. If indeed early risks can be modified by later experience this implies prospects for prevention.

Prospects for breaking the links Animal evidence suggests that interventions may make a difference to pre-set risks. Suomi's chapter12 reviews his long term elegant studies among rhesus monkeys Macaca mulatto. About 20% of any troop are high reactors. They are more likely than others to exhibit depressive responses to maternal separation with greater and more prolonged activation of the hypothalamic-pituitary-adrenal axis, more dramatic sympathetic arousal, more rapid central norepinephrine turnover and greater selective immunosuppression. These differential responses remain quite stable throughout development. Evidence suggests that the pattern of high reaction is genetically determined but it can be reproduced in non-genetically pre-disposed animals by raising them without their mothers. Interestingly these high reactors tend to end up at the bottom of the social hierarchy. There is also potential intergenerational transmission, non-genetic, of this tendency. Females raised without their mothers are likely to be abusive or neglectful of their first born offspring. This genetic high reactor destiny can be interrupted by changing the environment. When animals genetically predisposed to become high reactors were cross fostered with especially nurturant mothers they showed no signs of the behavioural disorders usually associated with being a high reactor. Rather they showed signs of precocious behavioural development and rose to the top of the hierarchy as adults. When they became mothers they showed the maternal style typical of their especially nurturant foster mothers. There is evidence that intervention can disrupt the effects of deprivation in humans. Sylva refers to the Head Start programmes in the USA7. These were aimed at low income deprived children before they started school. Initial reviews of the Head Start programme suggested that there were immediate effects on self-esteem, scholastic achievement and social behaviour but these effects did not last beyond the first year of school. The long term evaluation of the Perry pre-school project, following participants and controls to age 27, suggested that there were long term benefits. Participants had higher monthly earnings, higher BnhW. M^dxalBulletm 1997;53 (No 1)


Fetal and early childhood environment: long-term health implications

home ownership, higher level of completed schooling, a lower per cent receiving social services, and fewer arrests. A cost benefit analysis suggested that for $1000 invested in a pre-school programme at least $7160, after adjustment for inflation, was returned to society.

Social or physical effects The Barker hypothesis relates to nutrition. The evidence reviewed above relates to deprivation which presumably is, in part, the result of psychosocial factors. In fact the two may be related in a variety of ways. First, psychological and physical disorders may be related in adult life. People with depression, for example, are at higher risk of coronary heart disease. Hence the evidence showing that a pattern of disturbance, acting through the life course, leads to psychological disorder may, by analogy, or by direct influence have implications for physical disorder in adult Life. Second, there may be interaction between nutrition and social effects. Morley and Lucas report13 that underfed rats have poorer performance on behavioural measures which are worse if they are fed artificially rather than by their mothers. This raises the possibility that maternal stimulation may be important. Against this, they showed in trials in pre-term infants those receiving breast milk, via a nasogastric tube, had developmental advantage whether or not their mothers went on to breast feed. Third, there may be simple additive effects. McGregor's trial in Jamaica in stunted children showed that nutritional supplements and psychosocial stimulation had additive effects on children's mental development. Fourth, social factors may have strong impact on processes of programming. Wadsworth points out that social factors influence both poor fetal growth and adult obesity which together relate to the development of raised blood pressure9. Illustrations of how nutrition and psycho-social factors can interact in subtle ways comes from a review of studies of malnutrition in Guatemala14. Brown and Pollitt argue that malnutrition may have effects on cognitive performance not because of permanent brain effects but because malnutrition leads to less participation of children in activities, and to social withdrawal; and these lead to poorer mental and social development which in turn lead to social deprivation. They found that in poorly nourished children there was a marked effect of poverty on cognitive performance, that was less in evidence in well-nourished children. In an experiment to test the effects of nutritional supplementation among children living in poverty, they found that supplementation led to significant improvement in cognitive performance but did not bring test scores up to the level observed among middle class children living in the city. Bnhifi Medical Bul/.hn 1997,53 (No 1)

The influence of early life

They suggest that the positive effects of nutritional supplementation can be explained by the differences in the children's motor skills, physical growth and social and emotional development. A better nourished child may be better able to explore the environment and hence have faster cognitive growth.

An introduction not a conclusion Some of us working on the social, environmental, and life style determinants of ill-health in adults do so on the assumption that changing circumstances of adults will lead to changing disease rates. There is much evidence to support this position, but this is not the place to review it. The evidence gathered together in these papers suggests we can not ignore how people got to be where they are. Both the circumstances impinging on them in utero, the 'Barker hypothesis', and in early childhood, and their subsequent experiences will affect their predisposition to disease and the circumstances in which they find themselves in adulthood. These papers also hold out the prospect that the pathways linking early life effects to adult disease may be interrupted. This is clearly a fertile area for research and policy for which this issue serves as a timely introduction.

References 1 Cynader MS. Mechanisms of brain development and their role in health and well-being. Daedalus 1994, 123- 155-65 2 Barker DJP. Fetal nutrition and cardiovascular disease in later life. Br Med Bull 1997; 53: 96108 3 Pinker S. The Language Instinct. New York: William Murrow, 1994 4 Rose G. The Strategy of Preventive Medicine. Oxford: Oxford University Press, 1992 5 Hales CN. Non-insulin-dependent diabetes melhtus. Br Med Bull 1997; 53: 109-122 6 Cierpial MA, McCarthy R Hypertension in SHR rats: contribution of maternal environment. Am } Physiol 1987; 253: H980-4 7 Sylva K. Critical periods in childhood learning. Br Med Bull 1997, 53. 185-197 8 Doull IJM, Holgate ST. Asthma: early predisposing factors. Br Med Bull 1997; 53: 71-80 9 Wadsworth M£J. Changing social factors and their long-term implications for health. Br Med Bull 1997; 53: 198-209 10 Power C, Hertzman C. Social and biological pathways linking early life and adult disease. Br Med Bull 1997; 53 210-221 11 Maughan B, McCarthy G. Childhood adversities and psychosocial disorders. Br Med Bull 1997, 53: 156-169 12 Suorru SJ. Early determinants of behaviour: evidence from primate studies. Br Med Bull 1997, 53: 170-184 13 Morley R, Lucas A. Nutrition and cognitive development. Br Med Bull 1997; 53: 123—134 14 Brown JL, Polhtt E Malnutrition, poverty and intellectual development. Set Am 1996; 38-43

Bnhsh Mmdical Bu//«fm 1997;53 (No. 1)

Suggest Documents