• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptNIH Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Am J Hypertens. Author manuscript; available in PMC Aug 20, 2010.
Published in final edited form as:
PMCID: PMC2924638
NIHMSID: NIHMS35729

Ethnic differences in the association of Birth Weight and Blood Pressure -the Georgia Cardiovascular Twin Study

Sara Oberg, MD,1 Dongliang Ge, PhD,2 Sven Cnattingius, MD PhD,1 Anna Svensson, MSc,1 Frank Treiber, PhD,2 Harold Snieder, MD PhD,2,3,4 and Anastasia Iliadou, PhD1

Abstract

Background

African Americans (AA) not only have higher blood pressure levels, but also an increased risk of low weight at birth, compared to European Americans (EA). In light of fetal programming theories it has been suggested that ethnic differences in blood pressure originate in utero. However, most previous studies in bi-ethnic samples have not found a significant inverse association between birth weight and blood pressure in AAs.

Methods

In 562 EA and 465 AA adolescent twins of the Georgia Cardiovascular Twin Study we investigated the potential ethnic difference in blood pressure - birth weight association, with the ability to control for potential confounding by familial factors.

Results

Blood pressure levels were significantly higher in AAs compared to EAs, independent of birth weight (p<0.01). After adjustment for parental factors and BMI, the difference in systolic blood pressure per kg birth weight was -1.1 mm Hg (95% confidence interval -2.7, 0.48, p=0.17) in EAs and -2.5 mm Hg (95% CI -4.7, -0.40, p=0.02) in AAs. A significant ethnic interaction was revealed in paired analysis where the inverse association remained in AAs, but not in EAs. Associations with diastolic blood pressure were generally weaker and non significant.

Conclusion

We could show that low birth weight was associated with an elevated systolic blood pressure in AAs, independent of familial factors. The results also suggest that the association between birth weight and blood pressure may be more pronounced in AAs in adolescence.

Keywords: Birth Weight, Blood Pressure, Fetal Programming, Ethnicity, Twins

Introduction

It is well established that African Americans (AAs) have higher blood pressure levels than European Americans (EAs) and these differences manifest already in childhood and adolescence (1, 2). Less is known about the origins of these discrepancies and the possible influence of other risk factors for elevated blood pressure, including socioeconomic status, body size and birth weight.

Considering birth weight a proxy for fetal growth and blood pressure a mediator of future cardiovascular morbidity, the well documented inverse relationship of birth weight and blood pressure in EAs has been put forward as supporting evidence of fetal programming, suggesting that fetal growth restriction may increase offspring risks of cardiovascular disease later in life (3, 4). Since AAs are at increased risk of delivering low birth weight babies in both singleton and twin pregnancies (5, 6), a ‘fetal programming’ mechanism could explain part of the difference in blood pressure levels between AAs and EAs (7, 8). However, while an inverse association between birth weight and blood pressure has been widely established in EA populations of all ages, findings in AAs (mainly children and adolescents) have been inconsistent (9-16).

Critics of the ‘fetal origins hypothesis’ have mainly focused on methodological objections regarding study designs, adjusting for current body size, and publication bias (17). Birth weight is also a very crude proxy for fetal growth, and influenced by several maternal factors such as age, ethnicity, parity, nutrition and smoking (18). In addition other factors, such as socioeconomic status and genetic factors, may be independently associated with both birth weight and blood pressure and hence confound the association. In the evaluation of the birth weight - blood pressure association, twin studies uniquely enable to control for the influence of familial factors such as shared environment and common genes, since monozygotic twins share all of their common genes and dizygotic on average 50%, while both share intrauterine and early childhood environment. Theoretically, if there is an association among unrelated twins (i.e. between twins) it can be further evaluated within each twin pair, thereby adjusting for factors shared within twins (19). In addition, this approach allows for an approximation of fetal growth that has been controlled for other common factors shared by a twin pair, such as gestational age and maternal characteristics.

Considering the previous bi-ethnic findings, we hypothesized that there might be ethnic differences in the birth weight-blood pressure association, if anything reflecting weaker or non-existing associations in AAs. We investigated this in the adolescent twins of the Georgia Cardiovascular Twin Study, with ability to adjust for important confounders such as maternal factors, socioeconomic status, shared environment and common genes.

Methods

Study population

The Georgia Cardiovascular (CV) Twin Study is a longitudinal study of over 500 AA and EA twin pairs recruited from schools within 120 miles of the study location of Augusta, Georgia, USA. 308 EA and 223 AA pairs attended a baseline visit in 1997-2000, and during their scheduled follow-up-visit during 2001-2004, an additional 53 EA and 51 AA pairs were recruited for their baseline visit. In order to enhance power, baseline measurements for all individuals in the cohort (N=1270, 635 pairs) were selected for analyses. However, due to missing birth weight information (N=243) our final study population comprised of 1027 individuals. Of these, 562 were EA (279 pairs and 4 singletons) and 465 AA (232 pairs and 1 singleton). Distributions of zygosity and gender were in EA: 63 pairs and 1 singleton monozygotic (MZ) males, 73 pairs and 2 singleton MZ females, 36 pairs dizygotic (DZ) males, 34 pairs DZ females and 73 pairs and 1 singleton opposite-sexed (OS), and in AA: 61 pairs MZ males, 58 pairs MZ females, 31 pairs DZ males, 41 pairs DZ females and 51 pairs and 1 singleton OS. A detailed description of the demographics and overall study design has been previously published elsewhere (20, 21). The Georgia CV Twin Study has been approved by the institutional review board and by written consent from all subjects (and parents if subject under 18 years).

Measurements

Tests were performed in a laboratory setting and the measurements and procedures used were the same in 1997-2000 and 2001-2004. Blood pressures were examined using a Dinamap Vital Signs Monitor (model 1864 SX, Criticon Inc, Tampa, FL) after 11, 13 and 15 minutes in supine position (where subjects were instructed to relax on a hospital bed). An average of the three measurements was calculated and used as the resting value. Because of equipment failure there were missing blood pressure values on one subject. Based on parental report on previous medical history, all subjects were healthy when entering the study and were not using any anti-hypertensive drug.

Determination of ethnic designation and zygosity has been described and published previously (22, 23). Measurements of height and weight enabled calculation of body mass index (BMI) as weight (in kilograms) divided by height (in square meters). For two individuals BMI values were considered as outliers (BMI 54.9 kg/m2 and 58.6 kg/m2) and set to missing. Information on birth weight was self reported by the mother of the twins. In a subset of twins born in the state of Georgia, birth weight information was also available from Birth Registry data of the state of Georgia. This register data was used when available (n=617, 53% EA and 47% AA) except in one individual where it was considered unreliable. In this and remaining cases self-reports were used (n=410, 58% EA and 42% AA). The ethnic distributions did not differ significantly between the two groups (p=0.23). In the 399 with both types of information available, self-reports was found in good agreement with registers (r=0.92, p<0.001). Among the missing birth weight cases EA were more frequent (64% compared to 54%) as were MZs (56% compared to 48%). After adjustment for gender and ethnicity, missing cases were not different from the study population in any of the characteristics presented in Table 1.

Table 1
General characteristics (Mean, ± Standard Deviation or Numbers and Frequencies) of 1027 twins, stratified by sex and ethnicity.

Other self reported information used in this study included family social status, mothers’ age at birth (categorized as <20, 20-30 and >30) and mothers’ smoking habits at the time of the visit/interview (yes/no). To control for confounding by socioeconomic status we used the parental education level, defined as the highest level of either parent if married, and in any other case the mothers level of education, as it has been proposed to be the most stable socioeconomic variable over time (24, 25).

Statistical analysis

Comparisons between ethnic groups (and missing and non missing cases) were investigated at a significance level of 5%. Comparisons and regression analyses were performed with the GENMOD procedure, using generalized estimating equations (GEEs)(26) and specifying a covariance matrix to correct for the dependence within twin pairs. All statistical analyses were performed in SAS version 9.

In multivariate analyses birth weight was assessed as independent predictor of the two separate outcomes systolic and diastolic blood pressure. The multivariate models tested and the covariates included in them were as follows; Model 1 included age, sex and ethnicity; Model 2 included model 1 covariates plus parental factors (such as mothers’ age at birth, mothers’ smoking and family education); Model 3 included all covariates in model 2 plus current BMI categorized. An interaction term between birth weight and ethnicity was tested, and all models were also run and presented stratified by ethnicity.

Within twin pair analysis uniquely enable to assess potential confounding by shared familial factors (shared environment and common genes), by analyzing the association of a within component of the exposure (birth weight) on outcome (blood pressure)(19). All twins share intrauterine and early childhood environment, MZ twins share all of their common genes and DZ twins share on average half of their segregating genes. The within component, defined as the individual deviation from the twin pair mean (xij-xi where i=twin pair number and j=individual twin number) reflects the degree of likeness between twin siblings(19). If an association is due to familial effects (common genes and/or shared environment), the association of the within component and the outcome is expected to decrease or even diminish with degree of likeness. If unaltered, we conclude that the association cannot be explained by familial factors. Stratifying the analysis according to zygosity enables even further insight, as the potential difference of the within component in MZ and DZ twins allows us to investigate the influence of genes and environment. Thus, an attenuated effect across both zygotic groups suggests that both environment and genes can confound the association. However, a larger attenuation in the MZ group reflects additional genetic confounding. In our analysis the within coefficient expresses the change in blood pressure by 1 kg deviance in birth weight from the twin mean.

Results

Descriptive statistics of the study population, stratified by sex and ethnicity, are presented in Table 1. In males, there was virtually no ethnic difference in birth weight (3 grams), while corresponding difference among females was 202 grams. Males were also similar in current body constitutions while AA females, though lighter than their EA counterparts at birth, grew up to be heavier in adolescence (BMI 23.1 kg/m2 vs. 21.4 kg/m2 respectively). In both genders, AAs had significantly higher systolic and diastolic blood pressure levels compared to EAs (all p<0.01). There was no significant interaction between sex and ethnicity predicting blood pressure (pSBP=0.74, pDBP=0.65), neither in models including main effects (pmodel1=0.59, pmodel2=0.91, pmodel3=0.55 for systolic blood pressure and pmodel1=0.51, pmodel2=0.68, pmodel3=0.57 for diastolic blood pressure).

In the whole cohort there was a significant inverse association between birth weight and systolic blood pressure, further strengthened by adjustments to -1.8 mmHg/kg (95% CI - 3.0, -0.46, p<0.01) as presented in Table 2; model 3. Ethnicity remained an independent predictor of blood pressure (p<0.01) in all models, while an interaction between birth weight and ethnicity was not significant (pmodel1= 0.20 pmodel2=0.34 pmodel3=0.38). However, stratifying analyses by ethnicity revealed a more pronounced inverse association between birth weight and systolic blood pressure in AAs. Although adjustments seemed to increase EAs estimates more, they never reached significance (pmodel1= 0.49, pmodel2=0.34, pmodel3=0.17). With full adjustments, the difference in systolic blood pressure per kg birth weight was -2.5 mm Hg (95% CI -4.7, -0.40, p=0.02) in AAs and -1.1 mm Hg (95% CI - 2.7, 0.48, p=0.17) in EAs. Associations to diastolic blood pressure were weaker, non significant and generally attenuated by adjustments (data available on request).

Table 2
Regression coefficients (βC), confidence intervals (95% CI) and p-values of systolic and diastolic blood pressure (SBP and DBP) for 1 kg increase in birth weight.

Paired investigation of the inverse association between birth weight and systolic blood pressure identified a significant interaction between ethnicity and deviance from the twin pair mean birth weight (pcrude=0.02 padjusted=0.01), and results are therefore presented stratified by ethnicity (Table 3). After adjustment for BMI, the association in AAs was — 3.8 mmHg/kg (95% CI -7.2, -0.34, p=0.03). The association remained (although no longer significant) in both DZs and MZs, when stratifying by zygosity. In contrast, the generally weaker cohort effects in EAs were found reversed, i.e. positive, in paired analysis (regardless of zygosity). Adjusting for BMI seemed to slightly attenuate these positive associations. Since initial cohort effects for diastolic blood pressure were weak and non significant, paired analysis was of little interest (although data are available on request).

Table 3
Regression coefficients (βw) and confidence intervals (95% CI) of systolic blood pressure for 1 kg deviance from the twin pair mean birth weight (mmHg/kg) stratified by ethnicity.

Lastly, adjustment for current body size has been a rather controversial issue in fetal programming. Hence, we also investigated the influence of BMI on the birth weight — blood pressure association. In short, BMI was a positive predictor of systolic blood pressure in both ethnic groups, also in models adjusted for main effects (p<0.05 in all). Interestingly, BMI was found positively correlated with birth weight in EAs (r=0.17 p<0.01) but not in AAs (r=0.02 p=0.68). Consequently, adjusting for BMI in the birth weight-systolic blood pressure association increased estimates, mostly in EAs. There was no significant interaction between birth weight and BMI in the whole twin sample, nor in ethnic groups separate (pALL=0.54, pEA=0.92, pAA=0.34 in model 3).

Discussion

We have found inverse associations between birth weight and blood pressure in a cohort of bi-ethnic adolescent twins. Associations were generally stronger for systolic blood pressure and more pronounced in AAs compared to EAs. Paired analysis indicated that the association between birth weight and systolic blood pressure in AAs was not confounded by familial factors.

Ethnic differences in blood pressure have been suggested to partly originate in utero (7, 8). The increased risk of low birth weight in AAs could, according to fetal programming, contribute to AAs having higher blood pressure compared to EAs. AA participants of the Georgia CV Twin Study presented higher levels of both systolic and diastolic blood pressure compared to EAs, and these differences remained in models where birth weight was included. However, we found that an inverse association between birth weight and systolic blood pressure was more pronounced in AAs. Hence, we could not exclude that fetal programming may contribute to some of the ethnic differences in blood pressure.

An inverse association between birth weight and systolic blood pressure has been extensively reported for all ages in EA populations, and mostly so after controlling for current body size. In the only studies of AA adults to date, no correlation between size at birth and blood pressure has been found (9, 14), and findings in bi-ethnic children and adolescents have been inconsistent between studies. In one longitudinal study an inverse association between birth weight and blood pressure developed over time, and appeared among older children and adolescents in both ethnic groups (10-12). In contrast, another longitudinal study found an inverse association only in children (most pronounced in EA), but the effect was attenuated in late adolescence and significance was lost (15). There are also studies showing positive associations in AAs, while associations in EAs were absent (16) or inverse (13).

The inconsistency of results from previous bi-ethnic samples could be due to small sample sizes, differing age distributions when measuring blood pressure, and inadequate control for confounding factors. While the participants of the Georgia CV Twin Study where recruited from the same geographical area, they represent a socio-demographically heterogeneous sample, with significant ethnic differences across all socioeconomic measures. Attempts to adjust for some of these differences could not explain the ethnic discrepancy in the association of birth weight and blood pressure in the cohort analyses, although residual confounding by other socioeconomic factors could not be excluded. In the paired analyses, the association between birth weight and systolic blood pressure in AAs remained without attenuation in both MZ and DZ twins. These findings indicate that the association between birth weight and systolic blood pressure in AAs was not confounded by shared environment or common genes, and lends support to the ‘fetal programming effect’ of intrauterine growth impairment in this ethnic group. In contrast, associations were reversed in EAs. Although these results suggest confounding by familial factors in this ethnic group, interpretation should be careful since the cohort effects were not significant to begin with, nor were the estimates of the paired analysis.

The association of birth weight and blood pressure has been reported to attenuate during adolescence, possibly due to hormonal changes. Potential ethnic differences, in for instance sexual maturation, could thus in part account for disparities in the birth weight — blood pressure association. Irrespective of this, we would in general expect weaker associations in this adolescent cohort compared to findings in young children and/or adults. Further, catch-up growth, and especially centile crossing catch-up in childhood by babies small at birth, is often discussed as an important component in the pathway towards future morbidity (27). Without knowledge of the early postnatal growth trajectories in this cohort, we cannot exclude that ethnic differences with respect to catch-up growth could explain some of our findings. Interestingly, there were some indications of ethnic differences in the influence of current body size on the association between birth weight and blood pressure. Current body size could be considered in the pathway between exposure (birth weight) and outcome (blood pressure); associated with both, but also an effect of the exposure. Positive correlations to both exposure and outcome would tend to obscure an underlying inverse association. However, it has also been argued that attempts to estimate a more direct effect by controlling for current body size could introduce a spurious association between exposure and outcome (28). In this study, BMI positively predicted systolic blood pressure in both ethnic groups, but was only significantly (positively) correlated to birth weight in EAs. A more potent masking effect of BMI could have explained the generally weaker effects in EAs. However, even with BMI inclusion EA estimates did not reach AA levels and remained non significant. In conclusion, there was little evidence of a BMI ‘interference’ explaining the ethnic differences in the association of birth weight to blood pressure in our study.

We acknowledge that the present study has several limitations. The small sample size limited power, especially for stratifications but also in detecting interactions. Forty percent of the birth weight information was self-reported. Although the correlation between records and self-reports was high, we cannot exclude some degree of misclassification in the self-reported group. An unequal ethnic distribution in this group could have introduced a bias. However, the ethnic distribution among self-reports was not significantly different from those with birth records. The cohort analyses were also limited by the lack of information on maternal morbidity and gestational age. Low birth weight may in some cases be an effect of prematurity rather than impaired fetal growth, and such non-differential misclassification of exposure could, if anything, lead to under-estimation of an association between fetal growth and blood pressure. On the other hand, the within component of the paired analyses can be considered an estimate of fetal growth by default controlled for factors shared by a twin pair, such as gestational age, maternal body size and maternal morbidity.

Lastly, there is the issue of generalisability in twin studies. Fetal conditions differ between single and twin pregnancies and twins have lower mean birth weights compared to singletons. Thus, twins could be expected to have higher blood pressure according to fetal programming. However, several studies have shown blood pressure levels and future cardiovascular morbidity to be no different in twins from that of the general population (29, 30). Apart from supporting generalisability of twins, this implies that mechanisms of the ‘fetal programming’ pathway, though still unknown, may act independently of the fetal environmental factors that separate single and twin pregnancies(29). That is, although growth impairment solely due to twinning does not seem to invoke fetal programming, twins may still be susceptible to fetal programming effects. For instance, maternal constraint due to poor nutrition or maternal morbidity, may affect a twin pregnancy similar to a single pregnancy. Also, the unequal sharing of the supply line that can make twin pairs differ in size, may potentially invoke fetal programming. The increasing reports of inverse associations between birth weight and blood pressure in twin samples (both between and within twin pairs) can be considered supporting evidence of this, and suggests that information from twin studies may help further understanding of the fetal programming hypothesis (31).

In conclusion, we found a significant association between birth weight and systolic blood pressure in AAs, independent of familial factors such as shared environment and common genes. Findings indicate that the generally less favorable birth weight outcome in AAs may be of consequence for future blood pressure, and support further tracking of the association to cardiovascular morbidity/mortality in adulthood in this ethnic group.

Acknowledgements

The study was supported by grants from the National Institute of Health (grant: NIH HL56622) and from the Swedish council for working life and social research (grant nr: 2004-1654).

Supported by grants from the National Institute of Health (grant: NIH HL56622) and the Swedish council for working life and social research (grant nr: 2004-1654)

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

1. Dekkers JC, Snieder H, Van Den Oord EJ, Treiber FA. Moderators of blood pressure development from childhood to adulthood: a 10-year longitudinal study. J Pediatr. 2002;141:770–9. [PubMed]
2. Bao W, Threefoot SA, Srinivasan SR, Berenson GS. Essential hypertension predicted by tracking of elevated blood pressure from childhood to adulthood: the Bogalusa Heart Study. Am J Hypertens. 1995;8:657–65. [PubMed]
3. Barker DJ, Osmond C, Golding J, Kuh D, Wadsworth ME. Growth in utero, blood pressure in childhood and adult life, and mortality from cardiovascular disease. Bmj. 1989;298:564–7. [PMC free article] [PubMed]
4. Lucas A. Programming by early nutrition in man. Ciba Found Symp. 1991;156:38–50. discussion 50-5. [PubMed]
5. Salihu HM, Fitzpatrick L, Aliyu MH. Racial disparity in fetal growth inhibition among singletons and multiples. Am J Obstet Gynecol. 2005;193:467–74. [PubMed]
6. Martin JA, Kochanek KD, Strobino DM, Guyer B, MacDorman MF. Annual summary of vital statistics--2003. Pediatrics. 2005;115:619–34. [PubMed]
7. Lopes AA, Port FK. The low birth weight hypothesis as a plausible explanation for the black/white differences in hypertension, non-insulin-dependent diabetes, and end-stage renal disease. Am J Kidney Dis. 1995;25:350–6. [PubMed]
8. Forrester T. Historic and early life origins of hypertension in Africans. J Nutr. 2004;134:211–6. [PubMed]
9. Falkner B, Hulman S, Kushner H. Birth weight versus childhood growth as determinants of adult blood pressure. Hypertension. 1998;31:145–50. [PubMed]
10. Donker GA, Labarthe DR, Harrist RB, Selwyn BJ, Wattigney W, Berenson GS. Low birth weight and blood pressure at age 7-11 years in a biracial sample. Am J Epidemiol. 1997;145:387–97. [PubMed]
11. Mzayek F, Sherwin R, Fonseca V, et al. Differential association of birth weight with cardiovascular risk variables in African-Americans and Whites: the Bogalusa heart study. Ann Epidemiol. 2004;14:258–64. [PubMed]
12. Cruickshank JK, Mzayek F, Liu L, et al. Origins of the “black/white” difference in blood pressure: roles of birth weight, postnatal growth, early blood pressure, and adolescent body size: the Bogalusa heart study. Circulation. 2005;111:1932–7. [PubMed]
13. Rostand SG, Cliver SP, Goldenberg RL. Racial disparities in the association of foetal growth retardation to childhood blood pressure. Nephrol Dial Transplant. 2005;20:1592–7. [PubMed]
14. Hughson MD, Douglas-Denton R, Bertram JF, Hoy WE. Hypertension, glomerular number, and birth weight in African Americans and white subjects in the southeastern United States. Kidney Int. 2006;69:671–8. [PubMed]
15. Li C, Huang TK, Cruz ML, Goran MI. Birth weight, puberty, and systolic blood pressure in children and adolescents: a longitudinal analysis. J Hum Hypertens. 2006;20:444–50. [PubMed]
16. Hemachandra AH, Klebanoff MA, Furth SL. Racial disparities in the association between birth weight in the term infant and blood pressure at age 7 years: results from the collaborative perinatal project. J Am Soc Nephrol. 2006;17:2576–81. [PubMed]
17. Huxley R, Neil A, Collins R. Unravelling the fetal origins hypothesis: is there really an inverse association between birthweight and subsequent blood pressure? Lancet. 2002;360:659–65. [PubMed]
18. Kramer MS. Intrauterine growth and gestational duration determinants. Pediatrics. 1987;80:502–11. [PubMed]
19. Carlin JB, Gurrin LC, Sterne JA, Morley R, Dwyer T. Regression models for twin studies: a critical review. Int J Epidemiol. 2005;34:1089–99. [PubMed]
20. Snieder H, Treiber FA. The Georgia Cardiovascular Twin Study. Twin Res. 2002;5:497–8. [PubMed]
21. Snieder H, Harshfield GA, Treiber FA. Heritability of blood pressure and hemodynamics in African- and European-American youth. Hypertension. 2003;41:1196–201. [PubMed]
22. Jackson RW, Snieder H, Davis H, Treiber FA. Determination of twin zygosity: a comparison of DNA with various questionnaire indices. Twin Res. 2001;4:12–8. [PubMed]
23. Snieder H, Dong Y, Barbeau P, et al. Beta2-adrenergic receptor gene and resting hemodynamics in European and African American youth. Am J Hypertens. 2002;15:973–9. [PubMed]
24. Liberatos P, Link BG, Kelsey JL. The measurement of social class in epidemiology. Epidemiol Rev. 1988;10:87–121. [PubMed]
25. Krieger N, Williams DR, Moss NE. Measuring social class in US public health research: concepts, methodologies, and guidelines. Annu Rev Public Health. 1997;18:341–78. [PubMed]
26. Burton P, Gurrin L, Sly P. Extending the simple linear regression model to account for correlated responses: an introduction to generalized estimating equations and multi-level mixed modelling. Stat Med. 1998;17:1261–91. [PubMed]
27. Lucas A, Fewtrell MS, Cole TJ. Fetal origins of adult disease-the hypothesis revisited. Bmj. 1999;319:245–9. [PMC free article] [PubMed]
28. Hernan MA, Hernandez-Diaz S, Werler MM, Mitchell AA. Causal knowledge as a prerequisite for confounding evaluation: an application to birth defects epidemiology. Am J Epidemiol. 2002;155:176–84. [PubMed]
29. de Geus EJ, Posthuma D, Ijzerman RG, Boomsma DI. Comparing blood pressure of twins and their singleton siblings: being a twin does not affect adult blood pressure. Twin Res. 2001;4:385–91. [PubMed]
30. Christensen K, Wienke A, Skytthe A, Holm NV, Vaupel JW, Yashin AI. Cardiovascular mortality in twins and the fetal origins hypothesis. Twin Res. 2001;4:344–9. [PubMed]
31. Morley R. Can we generalise from findings in twins? Paediatr Perinat Epidemiol. 2005;19(Suppl 1):54–9. [PubMed]
PubReader format: click here to try

Formats:

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...

Links

  • MedGen
    MedGen
    Related information in MedGen
  • PubMed
    PubMed
    PubMed citations for these articles

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...