Contents - Previous
Current state of knowledge
Specific hypotheses and possible mechanisms
Implications for future research
Implications for public health
Working Group members: L Grivetti¹, D Leon², K Rasmussen³, PS Shetty4, R Steckel5 and J Villar6
Correspondence: IDECG Secretariat, c/o Nestlé Foundation, POB 581, 1010 Lausanne, Switzerland
¹Department of Nutrition,
University of California Davis, Davis, CA;
²Epidemiology Unit, Department of Epidemiology and Population Sciences, London School of Hygiene & Tropical Medicine, Keppel Street, London;
³Division of Nutritional Sciences, Savage Hall, Cornell University, Ithaca, NY;
4Human Nutrition Unit, London School of Hygiene & Tropical Medicine, Taviton Street, London;
5Economics & Anthropology Departments, Ohio State University, Columbus, OH 43210;
6UNDP/UNFPA/WHO World Bank Special Programme of Research, Development and Research Training in Human Reproduction, World Health Organization, Geneva, Switzerland
There is an established body of literature that size at birth is related to morbidity and mortality in infancy and to various neurodevelopmental outcomes and physical growth in childhood. The question of whether there are long-term effects on blood pressure and insulin mediated glucose uptake in childhood and disease in adulthood remains more controversial. On this question the working group reached a consensus that there is evidence for associations between fetal growth (as measured by birth weight, BW) and adult disease as follows:
- There is a negative association among both children and adults between BW within the normal range and blood pressure.
- There is a J-shaped negative association between BW and death from cardiovascular disease (CVD), and in particular from ischemic heart disease. This association is most securely established in men.
- There is a progressive negative association between BW or ponderal index at birth and non-insulin dependent diabetes mellitus (NIDDM).
- Birth weight is not associated with the aggregate of causes of mortality other than CVD, although there is some evidence for effects on other conditions including a negative association between BW and chronic respiratory disease and a direct association between BW and breast cancer.
- Nearly all these associations have been reported from studies in developed countries.
There are a number of uncertainties that surround these associations; they include:
- Whether they are confounded by factors that were present at birth and continue to act throughout the lifespan ('continuity of circumstance'), although recent data and research in progress suggest that this kind of confounding may not be a major issue.
- Whether the association between BW and CVD is as significant for women as for men.
- Whether the observed association between BW and CVD is dependent on established adult risk factors for CVD (e.g. blood pressure, serum cholesterol).
- Whether it is BW and/or other characteristics of size at birth (e.g. degree of adiposity, birth length, head circumference) that are important for these associations.
- Whether these associations occur to a greater or lesser degree in developing-country settings, where the distribution of BW and the predictors, of variation in fetal growth differ from those characteristic of developed-country settings.
The working group identified several issues that need to be taken into account when interpreting the observed associations between variation in fetal growth and adult disease; these include:
- Potential confounding factors that were not measured at all or that were measured imprecisely.
- The generalizability of findings from any single study within a selected population to the whole population and to other populations.
- The apparently opportunistic use of a range of different obstetric and perinatal measures as explanatory variables, and the over-interpretation of associations with these variables as distinct causal mechanisms.
- How the associations between variation in fetal growth and adult disease are to be reconciled with the secular trends and ecologic differences in BW and the incidence of CVD.
Common to the studies evaluated was a single, general hypothesis that
Impaired fetal growth is a determinant of adult-onset, degenerative disease.
The working group regretted the limited evidence from data in human subjects to demonstrate a mechanism or mechanisms by which variation in fetal growth could cause adult-onset chronic disease.
A number of uncertainties surround this hypothesis, including:
- Whether the association is with metabolic or nutritional aspects of the fetal environment.
- Whether the same intrauterine factor(s) are responsible for the association of fetal growth with outcomes as diverse as raised blood pressure and non-insulin-dependent diabetes mellitus (NIDDM).
- Whether variation in fetal growth is, in itself, sufficient to cause adult-onset chronic disease or whether it induces a susceptibility to developing such disease that is expressed only in conjunction with other, later-acting factors (e.g. adult obesity as suggested by some studies).
- Although this hypothesis is specific to fetal growth, it is difficult to distinguish effects that result from variation in prenatal growth from those that result from variation in postnatal growth.
- Impaired fetal growth may not simply be a consequence of maternal malnutrition or dietary deficiency; therefore, investigation of the hypothesis should ensure that adequate account is taken of other possible factors.
The ultimate proof of causality will require evidence from experimental designs. There may be grounds for including a follow-up dimension to selected future intervention trials in pregnancy that are set up principally to study shorter term outcomes (e.g. birth weight or term-IUGR). In addition follow-up studies of individuals involved in trials of nutritional or non-nutritional interventions that produced variation in fetal growth and trials using ultrasound routinely in pregnancy to document fetal growth could also potentially be useful.
The working group does consider that intervention trials involving adult subjects could usefully investigate whether variation in size at birth modifies the effectiveness of interventions to treat or prevent conditions such as hypertension or NIDDM. The priority here should be to use existing data from trials already completed that may involve retrospective collection of data on size at birth. Depending on the outcomes of such secondary analyses, new CVD intervention trials can be designed.
Although it appears that socio-economic confounding may not be a major issue within individual studies, it is essential that future observational studies collect adequate data to allow adjustment for these factors. There is an urgent need for studies in developing countries in which a larger proportion of babies are growth-retarded at birth and in which the causes of variability in fetal growth differ from populations that have previously been investigated. As work in this area develops, it will become increasingly important to establish what proportion of these adult-onset diseases could be accounted for by variation in size at birth. At present, it is unknown how important such variability might be if it proved to be a causal determinant of adult-onset disease.
Studies in appropriate animal models
continue to be useful. These will be particularly important in helping to define more
specifically the nature and extent of in utero programming mechanisms. Identification of a
more limited set of possible mechanisms would help to focus future mechanistic studies in
This hypothesis potentially has major implications for public health. These might be even more important in developing countries that are experiencing the epidemiologic transition than in developed countries. This is because in developing countries a high proportion of births occur in the range of BW with the highest risk of developing adult disease. Furthermore, the prevalence of factors, such as obesity, that might potentiate the development of high blood pressure, NIDDM and CVD is increasing rapidly in some developing countries, including those where the rates of childhood malnutrition remain alarmingly high. This issue highlights once more the importance of conducting studies in developing countries that will help to address this particular concern.
Funding for this I/D/E/C/G workshop was provided by the United Nations University and the Nestlé Foundation.
This publication is available free of charge from:
The Secretariat of IDECG
c/o Nestlé Foundation
P. O. Box 581
Thanks to a grant from the Nestle
Contents - Previous