![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
Department of Anthropology, Northwestern University, Evanston, IL 60208
2To whom correspondence should be addressed. E-mail: kuzawa{at}northwestern.edu.
 |
ABSTRACT |
|---|
 TOP ABSTRACT MATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
KEY WORDS: • fetal nutrition • cardiovascular disease • maternal nutrition • birth weight
The fetal origins hypothesis (FOH)2 proposes that intrauterine nutrition influences the development of various hormonal systems and organs, with lasting effects on adult risk for cardiovascular disease (CVD) (1). The hypothesis has generated considerable interest for its potential insights into health trends in populations experiencing the nutrition transition, where more common problems of poor maternal pregnancy nutrition and low birth weight may be contributing to the emerging CVD epidemic (2).
The FOH hypothesis gains strongest support from animal models showing that maternal nutritional restriction during pregnancy (3,4) and direct modification of fetal nutrition through restricted uterine or placental blood flow (5–7) result in elevated blood pressure, reduced insulin sensitivity and abnormal cholesterol profiles after birth. However, in humans, fetal nutritional sufficiency must be inferred through the use of proxy measures, most typically birth weight (BW). Although nutrition is a key factor influencing fetal growth rate, the common assumption that BW-CVD risk factor relationships reflect an effect of fetal nutrition is problematic, because BW is also influenced by a wide range of hormonal, genetic and epigenetic (imprinting) factors that vary by individual and across populations. The most important non-growth-related influence on birth size is prematurity, and only some of the many studies using retrospectively collected birth weight have had access to information on gestational age at birth. But even in a hypothetical population of optimally nourished newborns carried to term, birth weight would follow a normal distribution (8). If fetal nutritional sufficiency is indeed a common pathway through which prenatal factors influence later CVD risk, babies who are "constitutionally" small should not be at increased risk for CVD as adults. While any given small baby is more likely to have experienced growth restriction than a large baby, this interpretation is more defensible if there is evidence that the small baby had a higher growth potential—as indicated, for instance, by being born to a mother who is tall or who was large herself as a baby (9,10)—and thus might have been expected to be larger at birth.
The supply-demand model as a refined test of the fetal origins hypothesis
This logic, which uses a combination of offspring and maternal characteristics to define a de facto conditional probability defining fetal growth sufficiency, provides a basis for a refined test of the FOH, illustrated by the supply-demand model outlined in Figure 1(11). The model incorporates three functionally distinct markers—measures of maternal supply, fetal demand and birth outcome—to infer fetal nutritional sufficiency. Although not a comprehensive list of hypotheses, the model predicts an elevated risk of CVD among (Hypothesis 1) small babies born to tall mothers (or that otherwise show evidence of a higher growth potential) and (Hypothesis 2) small babies born to poorly nourished mothers (or otherwise born under conditions that suggest poor nutritional supply to the fetus). Finally, confidence in the interpretation that small birth size results from poor nutrition is greatest if (Hypothesis 3) the mother is both tall and poorly nourished during pregnancy.
|
Of course, the converse hypotheses also follow from the model; e.g., large babies born to short and well-nourished mothers are the least likely to have experienced intrauterine nutritional shortfall and growth restriction, and are thus predicted to have the lowest CVD risk. These hypotheses use BW and different combinations of maternal traits to test, in a more targeted fashion, the central premise of the FOH: that fetal nutritional insufficiency both reduces birth size and elevates CVD risk. Expanding our previous analysis of Hypothesis 4 (11), this paper tests the above hypotheses with data from the Philippines, using systolic blood pressure (SBP) and low density lipoprotein cholesterol (LDL-C) in adolescence as measures of CVD risk.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Predictor variables: supply, demand and birth outcome
The supply-demand model incorporates measures of maternal supply, fetal demand and birth outcome. Building from the bulk of prior FOH research, birth weight was chosen as a measure of birth outcome. In selecting a measure of maternal nutritional supply, maternal weight and BMI were excluded because they in part reflect the weight of the feto-placental unit. Maternal arm fat area (MAFA) measured during the third trimester of pregnancy (30 ± 4 wk of gestation) was deemed the most appropriate index of maternal nutritional supply available for this population because it is a marker of energy balance that is correlated with maternal energy intake (r = 0.2, P < 0.00001) and relates positively to offspring birth weight in this sample (17). Maternal third-trimester energy intake was also used as a complimentary measure of maternal energetics, under the assumption that a mother with low fat stores is even less likely to meet the energy demands of pregnancy when her energy intake is also constrained.
Maternal height was chosen as a marker of fetal growth potential and thus of fetal demand for nutrients. Of the candidate variables at our disposal, maternal height is among the strongest established predictors of birth weight (18) and, compared to other measures of maternal body size, is only weakly correlated with third-trimester MAFA (r = 0.18). Paternal height measurements were not available for this analysis. We assume that a fetus born to a tall mother has, on average, a higher fetal growth potential, thus requiring a greater supply of nutrients to avoid nutritional insufficiency, growth restriction and the suite of adaptations that persist to elevate risk for CVD.
Statistical analyses
All analyses were performed with the Stata Statistical Package, Version 8 (Stata, College Station, TX). Means and standard deviations were calculated for each CVD risk factor, predictor and control variable. Because descriptive statistics for the lipid analysis subsample were published previously (11,15,16), descriptive statistics are provided for the total (blood pressure) sample. Mean CVD risk factor levels were adjusted for potential confounding factors with multivariate regression and stratified on two or more of the variables in the supply-demand model, following from the hypotheses outlined above. In stratified analyses, P-values for the difference between the highest and lowest tertiles of the exposure variable (e.g., tertiles of maternal height) were assessed using dummy variables in regression models stratified on the level of the other predictor variable of interest (e.g., high and low birth weight). Hypothesized interactions were also tested using interaction terms in full regression models.
|
RESULTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
|
|
The LDL-C data for males were most consistent with predictions. The general pattern is that LDL-C levels were highest among lower-birth-weight males who likely had a higher fetal growth potential, as indicated by being born to a mother who was not in the shortest tertile of height (Hypothesis 1). Also consistent with expectations was the finding that higher-birth-weight babies born to shorter mothers had the lowest levels of LDL-C. Stratifying on third-trimester MAFA rather than maternal height revealed additional, although still modest, support for Hypothesis 2. The highest LDL-C levels were found among individuals who had below-median birth weights and were born to lean mothers, whereas higher-birth-weight babies born to better-nourished mothers had the lowest LDL-C levels as adolescents. Also consistent with predictions, MAFA and LDL-C were associated only among individuals who had below-median birth weights.
Although these relationships were not particularly clean, the general trends support the premise that maternal height and nutritional status during pregnancy may be used to refine the interpretation of birth weight when predicting LDL-C levels in males in this population. Individuals born small despite evidence of higher growth potential, or born to a mother who was relatively poorly nourished during pregnancy, are those most likely today to have elevated LDL-C levels. These findings are as expected if fetal nutritional sufficiency, reflected in birth outcome, programs CVD risk. However, the hypothesis that gains the strongest support is Hypothesis 4, which stratifies LDL-C on a measure of growth potential (maternal height) and maternal supply (MAFA) alone. There is a clear dose-response increase in LDL-C with declining MAFA, but only among individuals with higher growth potential, as indicated by being born to a tall mother. This finding suggests that poor maternal nutritional status (low supply) during pregnancy may have implications for lipid profiles in males, but only when there is evidence for higher fetal growth potential (high demand).
The most specific hypothesis, Hypothesis 3, predicts that the risk of CVD is highest in individuals who had a high fetal growth potential, were born to relatively poorly nourished mothers, and were small at birth. Figure 2A explores this hypothesis for LDL-C levels in the males. Adjusted mean LDL-C level is presented for the total population (dark bars), limited to individuals born to mothers with below-median MAFA (light bars), and further limited to the subset of these mothers who also had below-median energy intake (white bars). This figure incorporates all three elements of the model—supply, demand and birth outcome—simultaneously.
|
In addition to confirming the expectations of the supply-demand model, Figure 2 also illustrates our previously reported finding that maternal nutritional status predicts LDL-C level independent of both fetal growth potential (maternal height) and birth outcome (11). Although the association between elevated LDL-C level and poor maternal nutrition is most apparent among individuals who were small at birth but born to taller mothers, LDL-C level is relatively consistently elevated among offspring of poorly nourished mothers, and this is present among individuals of both above- and below-average birth weight and irrespective of maternal height.
The findings for SBP in males were less consistent with the predictions of the model. Instead, they suggest that maternal energy status or adiposity, as indicated by MAFA and energy intake, are relatively consistent, if modest, predictors of SBP in males, with effects that are independent of size at birth or maternal stature (Fig. 3A). As such, there is no evidence that fetal nutritional restriction reflected in reduced fetal growth programs SBP in the CLHNS males.
|
We reported previously that the inverse associations between MAFA and lipids and BP in the males are absent in the females in this adolescent sample (13,15). None of the relationships present in the female data were consistent with the predictions of the supply-demand model (Table 2). For instance, among females, LDL-C levels were highest among small babies but only among those born to well-nourished mothers. Levels of LDL-C were also elevated among babies born to mothers who were well-nourished and short, a group predicted by the model to have the lowest CVD risk. The highest LDL-C levels were found among small babies born to tall mothers, but also among large babies born to short mothers (both nonsignificant). For SBP, the only significant relationship was a modest increase in SBP with declining MAFA, but only among individuals who had above-median birth weights. None of these relationships support the hypothesis that fetal nutritional sufficiency has long-term effects on CVD risk among the CLHNS females. The absence of any clear patterns in Figures 2B and 3B and of interactions in the regression models (not shown) further support this interpretation.
|
DISCUSSION |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
In contrast, there was only a modest association between MAFA and SBP in males that appeared to be independent of birth weight and maternal stature. Because this effect was present at all levels of birth weight and maternal stature, it is probably not mediated through fetal nutritional sufficiency or growth. Findings that do not conform to model expectations may hint at the importance of alternate programming pathways. The most defensible interpretation of the SBP data is that some nonnutritional correlate of maternal adiposity, perhaps hormonal, programs SBP in males (19). Thus, although both the LDL-C and SBP findings support the hypothesis that the prenatal milieu has long-term implications for CVD risk in males, only the patterns observed for LDL-C support the specific prediction that fetal nutritional sufficiency both reduces birth weight and programs CVD risk. Consistent with our prior work at Cebu (13,15), fetal nutritional sufficiency as defined by the model failed to predict either elevated LDL-C level or SBP among the CLHNS females (discussed below).
The present approach differs from most tests of the FOH that incorporate BW, MNS or maternal diet, which typically treat all variables as independent proxies for fetal nutrition (13,15,20,21). Instead, the model interprets birth weight in light of these other measures. What the model loses in simplicity it makes up in specificity. An association between low birth weight and later hypertension is difficult to interpret, as it could reflect a wide range of factors, including genetics, or programming by maternal hormones, stress, or other correlates of birth size. The supply-demand model uses maternal traits in conjunction with BW to isolate individuals who likely were small at birth as a result of compromised nutrition and growth restriction.
An additional strength of the model is its flexibility, because it can be tested using a wide range of maternal and offspring characteristics (Fig. 2). In theory, variables representing any two or more of the supply-demand model’s categories—supply, demand or birth outcome—should provide a more refined criteria of growth sufficiency than any single variable used in isolation. This flexibility also warrants caution, however, as it is likely that in any population there will be some combination of predictors that yields significant results. For this reason, analyses should be limited to tests of a priori and biologically informed hypotheses. Analytically, there are many possible approaches to testing the model, which will vary by the type of data available, the goals of analysis, and the sample size. In addition to formal testing of interaction terms, the consistency, biological plausibility and dose dependence of any set of interactions should be assessed. Visual inspection of the stratified male LDL-C data revealed a relatively clear set of relationships that support the model and its predictions. The regression model confirmed the significance of the interactions, but would have been difficult to interpret on its own owing to the need to consider multiple interaction terms.
What sets fetal growth potential? A central assumption of the supply-demand model is that the "sufficiency" of BW or of a given level of nutritional supply is best evaluated against the benchmark of that fetus’s unique growth potential, indexed here by maternal stature. Previously, Leon et al. (22) and Hennessy and Alberman (23) used the child’s adult stature to index fetal growth potential, an approach that is arguably less capable of distinguishing an effect of fetal growth restriction from that of postnatal catch-up growth. In a study predicting perinatal mortality, Skjaerven (9) evaluated BW against the yardstick of the mother’s birth weight, a measure that shows promise.
Although genetics must play an important role in setting growth potential, the fetal growth trajectory may be altered by the mother’s nutritional experiences prior to conception, or even by the nutritional experiences of the grandmother. The Dutch Famine study is one commonly cited example: the offspring of women exposed to famine as fetuses were relatively small, even though the birth weights of the women themselves were not reduced (24). Although the mechanisms underlying such transgenerational effects are unclear, the key issue for the FOH is arguably how they influence growth trajectories—by reducing supply, demand or both. If supply-demand imbalance is key to fetal programming, the latter two scenarios would not necessarily elevate CVD risk. Indeed, the model predicts that certain forms of growth restriction—specifically, early lowering of growth trajectory resulting in reduced fetal demand—would protect the fetus against supply-demand imbalance and thus long-term CVD sequelae.
Sex differences in the long-term effects of the prenatal environment
Sex differences in the relationships between BP or LDL-C and markers of fetal nutritional sufficiency were evident in the CLHNS sample, with few relationships present in females. Lucas (25) hypothesizes that programming effects could be stronger in males owing to their greater size and demand for nutrients (reviewed in 11). A recent study in Boston found that mothers bearing males consumed 10% more energy during pregnancy than mothers bearing females (26). Consistent with this, CLHNS mothers who gave birth to males had increased energy intake late in pregnancy, when energetic requirements are likely near their peak (Fig. 4). Such findings suggest that the nutritional plane below which a fetus is forced to adapt might be higher in males. It is important to note that in the CLHNS data, the excess in energy intake associated with a male fetus was most pronounced among mothers from households with above-median income, perhaps indicating that they had greater opportunities to meet the excess requirements of a male fetus. Sex differences in CVD programming might be greatest when fetuses of poorly nourished mothers are limited in their opportunities to adapt their energy requirements, such as by reducing expenditure.
|
|
FOOTNOTES |
|---|
3 Abbreviations used: BW, birth weight; CLHNS, Cibu Longitudinal Health and Nutrition Survey; CVD, cardiovascular disease; FOH, fetal origins hypothesis; LDL-C, low density lipoprotein cholesterol; MAFA, maternal arm fat area; SBP, systolic blood pressure.
|
LITERATURE CITED |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
1. Barker, D. (1994) Mothers, Babies, and Disease in Later Life 1994 BMJ Publishing London, U.K.
2. Popkin, B., Horton, S. & Kim, S. (2001) The nutrition transition and prevention of diet-related diseases in Asia and the Pacific. Food Nutr. Bull. 22:1-58.
3. Langley-Evans, S. C. (2001) Fetal programming of cardiovascular function through exposure to maternal undernutrition. Proc. Nutr. Soc. 60:505-513.[Medline]
4. Joshi, S., Garole, V., Daware, M., Girigosavi, S. & Rao, S. (2003) Maternal protein restriction before pregnancy affects vital organs of offspring in Wistar rats. Metabolism 52:13-18.
5. Bassan, H., Trejo, L. L., Kariv, N., Bassan, M., Berger, E., Fattal, A., Gozes, I. & Harel, S. (2000) Experimental intrauterine growth retardation alters renal development. Pediatr. Nephrol. 15:192-195.[Medline]
6. Lane, R. H., Kelley, D. E., Gruetzmacher, E. M. & Devaskar, S. U. (2001) Uteroplacental insufficiency alters hepatic fatty acid-metabolizing enzymes in juvenile and adult rats. Am. J. Physiol. Regul. Integr. Comp. Physiol. 280:R183-R190.
7. Wolf, G. (2003) Adult type 2 diabetes induced by intrauterine growth retardation. Nutr. Rev. 61:176-179.[Medline]
8. Chard, T., Yoong, A. & Macintosh, M. (1993) The myth of fetal growth retardation at term. Br. J. Obstet. Gyn. 100:1076-1081.
9. Skjaerven, R., Wilcox, A. J., Oyen, N. & Magnus, P. (1997) Mothers’ birth weight and survival of their offspring: population based study. Br. Med. J. 314:1376-1380.
10. Clausson, B., Gardosi, J., Francis, A. & Cnattingius, S. (2001) Perinatal outcome in SGA births defined by customised versus population-based birthweight standards. Br. J. Obstet. Gyn. 108:830-834.
11. Kuzawa, C. & Adair, L. (2004) A supply-demand model of fetal energy sufficiency predicts lipid profiles in male but not female adolescent Filipinos. Eur. J. Clin. Nutr. (in press).
12. Godfrey, K. M. (1998) Maternal regulation of fetal development and health in adult life. Eur. J. Obstet. Gynecol. Reprod. Biol. 78:141-150.[Medline]
13. Adair, L. S., Kuzawa, C. W. & Borja, J. (2001) Maternal energy stores and diet composition during pregnancy program adolescent blood pressure. Circulation 104:1034-1039.
14. Adair, L. S. & Cole, T. J. (2003) Rapid child growth raises blood pressure in adolescent boys who were thin at birth. Hypertension 41:451-456.
15. Kuzawa, C. W. & Adair, L. S. (2003) Lipid profiles in adolescent Filipinos: relation to birth weight and maternal energy status during pregnancy. Am. J. Clin. Nutr. 77:960-966.
16. Kuzawa, C. W., Adair, L. S., Avila, J. L., Cadungog, J. H. & Le, N. A. (2003) Atherogenic lipid profiles in Filipino adolescents with low body mass index and low dietary fat intake. Am. J. Hum. Biol. 15:688-696.[Medline]
17. Adair, L. & Popkin, B. (1988) Birth weight, maturity and proportionality in Filipino infants. Hum. Biol. 60:319-339.[Medline]
18. Institute of Medicine (1990) Nutrition During Pregnancy. Part I. Weight Gain 1990 National Academy Press Washington, D.C.
19. Lucas, A. (1998) Programming by early nutrition: an experimental approach. J. Nutr. 128:401S-406S.
20. Mi, J., Law, C., Zhang, K. L., Osmond, C., Stein, C. & Barker, D. (2000) Effects of infant birthweight and maternal body mass index in pregnancy on components of the insulin resistance syndrome in China. Ann. Intern. Med. 132:253-260.
21. Roseboom, T. J., van der Meulen, J. H., Osmond, C., Barker, D. J., Ravelli, A. C. & Bleker, O. P. (2000) Plasma lipid profiles in adults after prenatal exposure to the Dutch famine. Am. J. Clin. Nutr. 72:1101-1106.
22. Leon, D., Koupilova, I., Lithell, H., Berglund, L., Mohsen, R., Vagero, D., Lithell, U. & McKeigue, P. (1996) Failure to realise growth potential in utero and adult obesity in relation to blood pressure in 50 year old Swedish men. Br. Med. J. 312:401-406.
23. Hennessy, E. & Alberman, E. (1997) The effects of own fetal growth on reported hypertension in parous women aged 33. Int. J. Epidemiol. 26:562-570.
24. Lumey, L. H. (1992) Decreased birthweights in infants after maternal in utero exposure to the Dutch famine of 1944–1945. Paediatr. Perinat. Epidemiol. 6:240-253.[Medline]
25. Lucas, A., Baker, B. A., Desai, M. & Hales, C. N. (1996) Nutrition in pregnant or lactating rats programs lipid metabolism in the offspring. Br. J. Nutr. 76:605-612.[Medline]
26. Tamimi, R. M., Lagiou, P., Mucci, L. A., Hsieh, C. C., Adami, H. O. & Trichopoulos, D. (2003) Average energy intake among pregnant women carrying a boy compared with a girl. Br. Med. J. 326:1245-1246.
27. Tell, G. (1985) Cardiovascular disease risk factors related to sexual maturation: The Oslo Youth Study. J. Chron. Dis. 38:633-642.[Medline]
28. Dodic, M., Hantzis, V., Duncan, J., Rees, S., Koukoulas, I., Johnson, K., Wintour, E. M. & Moritz, K. (2002) Programming effects of short prenatal exposure to cortisol. FASEB J. 16:1017-1026.
This article has been cited by other articles:
|
|
 |
|
  M. Du, M. J. Zhu, W. J. Means, B. W. Hess, and S. P. Ford Effect of nutrient restriction on calpain and calpastatin content of skeletal muscle from cows and fetuses J Anim Sci, September 1, 2004; 82(9): 2541 - 2547. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. J. Conlisk, H. X. Barnhart, R. Martorell, R. Grajeda, and A. D. Stein Maternal and Child Nutritional Supplementation Are Inversely Associated with Fasting Plasma Glucose Concentration in Young Guatemalan Adults J. Nutr., April 1, 2004; 134(4): 890 - 897. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. S. Adair and A. M. Prentice A Critical Evaluation of the Fetal Origins Hypothesis and Its Implications for Developing Countries J. Nutr., January 1, 2004; 134(1): 191 - 193. [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
