| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
University of Aukland The Liggins Institute for Medical Research Auckland, New Zealand
Address all correspondence and requests for reprints to: Peter D. Gluckman, The Liggins Institute for Medical Research, University of Auckland, Private Bag 92019, Auckland, New Zealand.
 |
Introduction |
|---|
 Top Introduction References |
|---|
The fetal origins of adult disease (FOAD) hypothesis developed from such studies. It proposes that environmental influences acting in fetal life are reflected in altered birth size/phenotype and, in some way, permanently affect structure and metabolism leading to a greater risk of developing the metabolic syndrome or elements of it in later life (2). The term "programming" was introduced to describe the processes by which physiological development is altered in utero so as to impact on adult pathophysiology. As yet, programming remains a conceptual term rather than having any precise biological basis.
The FOAD model, also termed the Barker hypothesis, was subject to much criticism. It was suggested that confounders such as socioeconomic status and smoking led to the apparent associations. However, in recent years significant amounts of prospective clinical and experimental data have been amassed to support the general model. For example, Hofman et al. (6) studied a group of carefully matched prepubertal short children either born with IUGR or of normal birth weight and demonstrated insulin resistance to be present in those children with IUGR. Similar findings in children and young adults have been reported by several groups (7, 8).
Barker et al. (9, 10) propose that nutritional influences might be the dominant cause of programming. Such conclusions have been greatly influenced by studies of the Dutch Winter famine of 1944/1945. Fetal growth and nutrient supply are directly linked and the latter is dependent on maternal nutritional status and health, uterine blood flow, placental function and fetal cardiovascular and endocrine status (11). There are thus many potential points at which fetal size can be affected. Relationships between maternal nutrition and subsequent outcome have only recently been identified in the non-famine situation (12), and this has created a shift in focus from the total caloric intake in maternal diet to the balance of diet.
Experimentally there is now overwhelming evidence in animal models to support the concept that an adverse fetal environment, as reflected in birth phenotype, can lead to poor health outcome in adulthood. The original studies in rat used a low (9%) protein isocaloric diet and have repeatedly demonstrated that mothers exposed to this diet across pregnancy give birth to offspring that develop hypertension and glucose intolerance in adult life (13), and exposure for any one week of pregnancy induces similar degrees of hypertension in the offspring (14). Similar data have been obtained in the rat and guinea pig using hypocaloric undernutrition (15, 16) or exposure to glucocorticoids in the rat and sheep (17, 18).
Adverse outcome is greatly amplified in those born small who later develop obesity or an increased ponderal index. Conversely, those born small or thin who stay thin are relatively protected. Recent data have demonstrated an interaction between prenatal exposure and postnatal dietary intake in determining the level of hypertension and insulin resistance (19). Limited data are available that also raise the question as to whether catch-up growth is also an amplification factor (20). Such issues are important as they generate questions as to the optimal growth and nutritional management of a child born small.
Timing of the adverse event during gestation must also be considered. In the Dutch winter famine, early gestational exposure had no effect on the development of postnatal carbohydrate intolerance, whereas third trimester exposure had a marked effect (21). Similarly in sheep, maternal glucocorticoid exposure at day 27 of pregnancy, but not at day 64, induces hypertension in the offspring (18); and in rat, changes in hepatic enzymes are reported after late gestational insults not seen in early pregnancy following the same insult (14). No consistent picture has yet emerged reflecting the difficulties of trans-species comparisons of developmental windows.
That events during critical windows of early development can have permanent effects is not surprising. Teratogenesis due to toxins or infections is well recognized. The sexual differentiation of the hypothalamus and the genitalia are both due to exposure to androgens at critical periods in development independent of chromosomal sex (22). However, what the FOAD model proposes is that normal fetal development is influenced by more subtle factors. The association between birth phenotype and adult disease extends across the normal range of birth size suggesting that even subtle influences, as reflected in the normal range of birth weights, are sufficient to alter the propensity for adult disease.
In humans, the demonstration of adult disease can be viewed as a mismatch between prenatal adaptation and postnatal environment. These observations suggest a teleological explanation of programming. Namely, that there are multiple mechanisms designed to achieve an evolutionary goal of survival of the species; these mechanisms must act via a transient environmental change too brief a period for genetic selection to act. We propose that it is appropriate in evolutionary terms for the fetus to detect its potential postnatal environment as nutritionally disadvantageous and program its development for such an environment. In the short-term, survival to reproduction would be protected despite smaller postnatal size. However, if the postnatal environment is not deprived then postnatal growth may be inappropriate for the settings and obesity and its sequelae could occur. It is therefore physiologically advantageous to maintain a small lean body mass and fat mass if born into a nutritionally deprived environment. Such a model may well explain the high propensity of metabolic syndrome appearing in transitional societies.
Several key questions are the focus of current and needed research: what is the signal(s) that affects the fetus, during what "critical window" of development does this signal act, what mechanisms are involved and what is the role of postnatal amplification? Is there a specific biological process that underpins programming or are there multiple pathways with the same outcome? What are the range of diseases that are programmed and is the programmed state irreversible?
There has been considerable focus in the experimental literature on what might be the trigger to programming, and it is in this context of this question that the paper of Lesage et al. (23) is of interest. Many of the experimental paradigms used to replicate programming have been based on severe maternal undernutrition. As demonstrated in this paper such studies may have been confounded by activation of the maternal HPA axis. Lesage and colleagues present compelling evidence that at least acute undernutrition in pregnancy profoundly activates the maternal HPA axis sufficient to affect the fetal HPA axis secondary to transplacental corticosterone transfer. This is of relevance in that other experimental paradigms in both rats and sheep have demonstrated that maternal exposure to betamethasone or dexamathasone can induce hypertension in the offspring (17, 18). Their paper, however, suggests that fetal growth retardation is directly a result of undernutrition, although it is established that maternal glucocorticoid exposure can induce intrauterine growth retardation both in animals and man.
It is less clear whether this is an issue of import. Undernutrition itself can down-regulate placental 11ß-hydroxysteroid dehydrogenase type 2 (11ß-HSD2) (24). 11ß-HSD2 creates the barrier between the maternal HPA system and the fetus, converting cortisol to cortisone. Thus, undernutrition itself may cause more subtle elevations in fetal glucocorticoid exposure sufficient to induce fetal growth retardation without impacting on the development of the fetal HPA axis. On the other hand, glucocorticoids can impact on glucose transporters and thus impact on nutrient availability (25). Moreover, the human situation is less likely to relate to acute nutritional insults, but more probably either to chronic nutritional imbalance or, perhaps, to specific acute illness. The latter has been suggested in controversial studies of the relationship between influenza in pregnancy and subsequent psychiatric disorder (26). Furthermore, programming can be demonstrated even when the insult is applied in the preimplantation period (27).
It seems unlikely that a singular integrative mechanism can explain programming given that insults occurring at different stages in development can have similar outcomes. Many of the insults are broad and extend over long periods of development, thus spanning multiple critical periods. Indeed, it seems more likely that both macronutrient and micronutrient supply may influence fetal development directly and indirectly, and that additionally glucocorticoid exposure can also impact on fetal development. The adult phenotype presumably is the net sum of genomic factors, and fetal and postnatal environmental influences. In any event, it is clear that in interpreting adult metabolic, endocrine, and cardiovascular phenotypes greater consideration must be given to the fetal experience.
Only a limited number of basic processes can explain programming: altered gene expression or altered patterns of cell proliferation, death, or differentiation. Data to support each of these processes are available in multiple organ systems. For example, in the protein undernutrition model, increased apoptosis has been reported in the fetal pancreas that might explain the altered ß cell mass observed (28); in the kidney, reduced nephron number has been reported (29); in the brain, there is reduced cerebral and mesenteric vascularity (30); and in the liver a change in pattern of hepatic zonation (31, 32), as reflected in altered expression patterns of gluconeogenic enzymes. In and of themselves, none of these are sufficient explanations for the programming of metabolic syndrome. At a more synthetic level it has been suggested that multiple hormone insensitivity including GH, IGF-1, insulin, and leptin might be involved (33). Hypothalamic dysfunction may also be involved: there being evidence of central disturbance of the HPA axis and of vegetative functions, including appetite (19). Other authors have suggested endothelial dysfunction, or alterations in the autonomic nervous system as core elements (34, 35).
In human studies subtle disturbances of the HPA axis have been reported in those with evidence of being born small (36). These disturbances are compatible with defective negative feedback, and hypothalamic and hippocampal glucocorticoid receptor abnormalities have been described in the experimental animal model following either nutritional or glucocorticoid administration to the mother. McCormick et al. (37) have demonstrated in rat that programming induced by glucocorticoids alters the utilization of alternative exon 1 sequences coding for promotor regions on the glucocorticoid receptor gene. In addition, postnatal handling, which is a form of behavioral programming that induces alterations in the HPA axis, influences preferential usage of particular exon 1 subtypes. This is some of the strongest evidence that programming can cause alterations in gene expression.
Maternal cortisol is an important way in which maternal experience can be signaled to the fetus. The paper by Lesage et al. (23) provides further evidence of the interaction between the maternal and fetal HPA axis. This study, as others, has shown changes in neuroendocrine control in the neonate that if persistent would lead to defective negative feedback within the HPA axis. In other models of programming, such persistence has been shown (38).
More importantly, this perspective is relevant to the increasing use of betamethasone in human pregnancy. Since the original studies of Liggins and Howie (39), it has become commonplace to offer glucocorticoids to pregnant women in premature labor to promote lung maturation. However, endocrine changes are observed in human newborns following betamethasone treatment to induce lung maturation (40). There is no doubt that a single dose of glucocorticoid treatment is well-validated as an essential therapy in the prophylactic management of the premature infant, but there has been an increasing tendency – not supported by clinical trials—toward repeated glucocorticoid administration. Clearly excessive fetal exposure to glucocorticoids can cause IUGR in experimental animal models with supportive data in humans, and the complications of IUGR and prematurity increase morbidity and mortality. Furthermore, excess glucocorticoid exposure in utero has been implicated experimentally to impact on neural development (41). The paper by Lesage et al. (23) is part of the growing evidence that should inspire caution. Repeated exposure of the fetus to glucocorticoids should be avoided in the absence of well-validated clinical trials.
Received February 23, 2001.
|
References |
|---|
|
TopIntroductionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  R. Ain, L. N Canham, and M. J Soares Dexamethasone-induced intrauterine growth restriction impacts the placental prolactin family, insulin-like growth factor-II and the Akt signaling pathway J. Endocrinol., May 1, 2005; 185(2): 253 - 263. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
