• arterial structure and compliance;
  • clinical science;
  • developmental biology;
  • endothelium;
  • fetal programming;
  • microcirculation


  1. Top of page
  2. Abstract.
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conflict of interest
  8. Acknowledgements
  9. References

Objective.  Low birth weight is associated with increased prevalence of hypertension and cardiovascular disease in adults. The aim of this study was to evaluate genetic and intrauterine environmental contributions to blood pressure (BP) and vascular functions in twins with discordant growth in utero.

Subjects.  We studied 31 twin pairs (21 monozygous and nine dizygous), mean age 8 years) with large within-pair differences in birth weight. Among the monozygous pairs, nine had suffered from twin-to-twin-transfusion syndrome (TTTS).

Methods.  Apart from BP, we determined diameters and elasticity of the carotid artery and abdominal aorta with ultrasonography, and endothelial function in skin vessels with a laser Doppler technique, before and after transdermal delivery of acetylcholine and nitroglycerin.

Results.  Eight of 62 twin subjects had a systolic BP above the 90th percentile in a North-American reference population. Among these, seven/eight were monozygous with a history of poor fetal growth and/or TTTS. In monozygous twin pairs without TTTS, systolic BP and pulse pressure were higher and vascular endothelial function was impaired in the lower birth weight twin. In the TTTS group, the lighter twin had a narrower carotid artery but there was no within-pair difference in arterial elasticity. Pre-eclampsia during the index pregnancy enhanced within-pair differences in BP but abolished within-pair differences in endothelial function.

Conclusions.  Severe fetal growth retardation contributes to higher BP, arterial narrowing and endothelial dysfunction in childhood. Pre-eclampsia may act both as an effect modifier and confounder of these associations.


abdominal aorta




appropriate for gestational age


blood pressure


common carotid artery


diastolic blood pressure






pulse pressure


perfusion unit


systolic blood pressure


small for gestational age


twin-to-twin transfusion syndrome


  1. Top of page
  2. Abstract.
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conflict of interest
  8. Acknowledgements
  9. References

There is substantial evidence of an inverse relation between birth weight and later blood pressure (BP) [1–4] although recently the strength of the association has been questioned [5]. It is currently a subject of debate whether this association reflects common genetic pathways or results from lasting programming effects because of adverse environmental influences in utero. To assess the various hypotheses, twin studies have been carried out, which focus on birth weight discordances within pairs and subsequent differences in BP [6–14]. However, the findings in these studies have been contradictory. Some ascribe the association with genetic factors [8, 12], whilst others maintain that fetal programming accounts for some of the prevalence of later hypertension [6, 7, 9]. Both or other explanations have also been proposed [10, 11, 13, 14]. As previously pointed out, the differences in results may partly reflect methodological problems, including bias in self-reported birth weights, crossing over of twin identities in various registers, unknown chorionicity and imprecise determinations of BP [13, 15].

Vascular programming has recently been shown in 1-year-old monozygous (MZ) twin infants who had suffered from haemodynamic stress in utero because of twin-to-twin transfusion syndrome (TTTS) [16]. Increased brachial artery stiffness was found in the smaller, but not in the heavier twin. Although no differences in BP were seen shortly after birth, loss of natural arterial elasticity has been proposed to contribute to an increase in BP at an older age [17].

If fetal programming of BP occurs, we hypothesized that differences could be detected in MZ twins discordant in fetal growth and development. In this study, we measured BP in prepubertal twins with large differences in weight at birth. Amongst MZ pairs, we included those with and without a history of TTTS. Because loss of arterial elasticity and endothelial dysfunction have been associated with fetal growth retardation [16, 18–21] and may predispose to adult hypertension and atheroma formation, we also included measurements of these vascular functions.


  1. Top of page
  2. Abstract.
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conflict of interest
  8. Acknowledgements
  9. References

From log-books of all births (n > 60 000) at Karolinska and Danderyd in Stockholm during 1991–1997, we found 35 liveborn same-sexed twin pairs in whom one was appropriate for gestational age (AGA; birth weight = mean ± 2 SD) and the other small for gestational age (SGA; birth weight < mean − 2 SD according to Swedish reference data [22]). Exclusion criteria were very preterm delivery (gestational age: <32 weeks), congenital syndromes and chronic illness in need of medication. Four families were lost to follow up. The remaining 31 families were invited and 22 accepted to participate in the study. The birth weights and gestational ages of those who declined did not differ significantly from those who participated.

Another twin-cohort, consisting of all consecutive and prospectively diagnosed TTTS pairs in which both survived and were born in Stockholm during 1994–1997, was also included (n = 9 pairs). TTTS is a condition that complicates monochorionic twin pregnancies and is due to an unbalanced blood flow through unidirectional vascular anastomoses in the placenta. It is characterized by fetal growth restriction in the smaller, donor twin and hypervolaemia and heart failure in the heavier, recipient twin. In our study, diagnosis was based on serial antenatal ultrasound examinations according to internationally accepted criteria [23]. Most subjects in our study had mild to moderate disease, as evidenced by dual survival. During pregnancy, six mothers of the TTTS children had been treated with amnioreduction alone or in combination with indomethacin, to reduce the risk of rupture of the membranes and preterm labour. None of the women underwent treatment with laser coagulation of the placental anastomoses. All families gave their informed consent before the investigations and the study was approved by the local Ethics Committee.

Altogether 62 twins (34 girls) were studied at an age of 7.8 (range: 4.6–11.8) years. They were divided into three groups: (i) monozygous twins without TTTS (MZ without TTTS), (ii) monozygous with TTTS (MZ with TTTS) and (iii) same-sexed dizygous (DZ) twins. Chorionicity was established on the basis of histology. In cases of dichorionic placentas (n = 12) or unknown chorion type (n =2), the determination of monozygosity relied on parental reporting of mistaken identity. All parents were healthy Caucasians. In vitro fertilization had been performed in three mothers, resulting in one MZ and two DZ twin pairs. Maternal and pregnancy data are summarized in Table 1.

Table 1.  Subject characteristics
 MZ-no TTTS (n = 13 pairs)MZ-TTTS (n = 9 pairs)DZ (n = 9 pairs)
  1. MZ-no TTTS, monozygous twins without twin-to-twin transfusion syndrome; MZ-TTTS, monozygous twins with TTTS; DZ, dizygous twins; BW, birth weight; SDS, standard deviation score (normalized for gestational age and gender in a Swedish reference population of singletons [22]); BL, birth length; AGA, appropriate for gestational age at birth; SGA, small for gestational age at birth.

  2. Intra-pair differences in BW and BL were highly significant whereas current weight or height did not differ within pairs.

  3. Values are expressed as mean (SEM) or proportions.

  4. *P < 0.05 vs. MZ-TTTS.

Maternal data
 Age (years)31 (1.4)30 (1.2)32 (1.2)
Pregnancy data
 Antenatal steroid therapy5/13*9/91/9*
Neonatal data
 Gestation (weeks)35 (0.6)33 (1.2)37 (0.6)*
 BW-SDS for AGA-twin−0.2 (0.3)−0.3 (0.3)−0.4 (0.2)
 BW-SDS for SGA-twin−2.9 (0.3)*−1.8 (0.1)−2.6 (0.1)*
 BL-SDS for AGA-twin−0.3 (0.5)−0.4 (0.4)−0.9 (0.3)
 BL-SDS for SGA-twin−3.1 (0.6)−1.7 (0.2)−2.2 (0.3)
Current age (weight and height)
 Age (years)7.9 (0.7)6.4 (0.3)9.1 (0.5)*
 Weight, AGA-twin (kg)29.1 (2.5)22.3 (1.3)30.1 (2.0)*
 Weight, SGA-twin (kg)27.1 (2.9)21.1 (1.8)29.1 (2.2)
 Height, AGA-twin (cm)130 (4.1)119 (2.4)136 (3.5)*
 Height, SGA-twin (cm)128 (4.6)118 (2.5)135 (4.0)*

Blood pressure determinations

After resting for 20 min, heart rate and BP were recorded with the subject lying down with the arm at heart level. An automated oscillometric sphygmomanometer (Boso MedicusTM, Jungingen, Germany) was used with an appropriately sized cuff (9 × 28 cm) around the right upper arm. The mean of six consecutive determinations, taken at least at 3-min intervals, was regarded as the subject's BP. The coefficient of variation (CV) was 9.5% for systolic (SBP) and 9.9% for diastolic blood pressure (DBP), i.e. the same as in other paediatric studies [10, 24].

Vascular studies

Arterial dimension and stiffness.  The mechanical properties of the abdominal aorta (AA) and the left common carotid artery (CCA) were studied with ultrasonography. A computer-generated pair of electronic echo-trackers (DiamoveTM, Teltec AB, Lund, Sweden) was used to measure the end-diastolic diameter (Dd; mm), pulse amplitude of the diameter (ΔD; mm) and relative strain (ratio: ΔD/Dd, %). The minimum diameter possible to measure with the Diamove equipment is 7.8 μm [25]. These diameter data and those of simultaneously measured BPs were computed to yield the stiffness index (β) using the equation:

  • image

The use of noninvasively measured BPs in the brachial artery might lead to an underestimation of the true aortic and carotid BPs. However, the consistent use of arm BP in the calculations should not influence the comparisons of the results between and within the twin pairs in our study. The mean value of three recordings, each consisting of about 6–10 consecutive heart cycles, was taken as the subject's reading.

Endothelial function.  A laser Doppler (LD) instrument and a micropharmacology system (Perimed AB, Kista, Sweden) were used to measure changes in perfusion during vascular drug provocations in the skin of the dorsum of the hand. The LD signal is proportional to the number and velocity of moving blood cells in the illuminated superficial skin microvessels and expressed in perfusion units (PU) of output voltage (1 PU = 10 mV). The temperature of the LD probe facing the skin was standardized to 32 °C. After adjustment to the room temperature (22 °C) for 20 min, the vascular studies were performed with the child lying supine with both arms beside the body. To study endothelium-dependent vasodilation, perfusion was recorded after transferal of acetylcholine (ACh) across the skin by iontophoresis (anodal current of 0.1 mA for 20 s repeated six times at 60-s intervals). To study endothelium-independent vasodilation, the effect of an exogenous nitric oxide (NO) donor (1% nitroglycerine) was tested on the contralateral hand. Details concerning vascular methods have been described elsewhere [18].

Statistical analyses

Data are presented as mean (SEM) values or proportions. Chi-square and anova were used to test for group differences. Mean within-pair differences [95% confidence interval (CI)] were calculated (=value in the lighter twin at birth − value in the heavier twin at birth) and two-tailed paired t-test was used for cotwin comparisons. A P-value of <0.05 was considered significant and a P-value between 0.05 and 0.10 as borderline significant. Within-pair differences in BP, arterial diameter and pulsatile diameter changes and maximum skin perfusion response to ACh (endothelial function) were considered as primary outcome variables. Regression analyses were used to evaluate contributions to within-pair differences from the following risk factors or confounders: group (MZ twins without TTTS, MZ with TTTS and DZ), pre-eclampsia, use of antenatal steroids, gender, current age, weight and height. Initially, stepwise regression was performed. Independent variables with a P-value of <0.20 in this model were entered into a multiple regression model.


  1. Top of page
  2. Abstract.
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conflict of interest
  8. Acknowledgements
  9. References

The mean gestational age was 35 (range: 28–40) weeks as determined by early ultrasound. The heavier twin had a mean birth weight of 2567 (1305–3705) g and a mean birth length of 46 (40–52) cm. The lighter twin weighed on average 1855 (922–2630) g and was 43 (36–48) cm long at birth, corresponding to an inter-twin difference in birth weight of 28 (95% CI: 24–31)%. At the time of the study, there were no statistically significant within-pair differences in weight or height (Table 1).

Blood pressure

The overall SBP was 101 (1.3) and DBP 59 (0.7) mmHg. After taking age, gender and height into account, eight of 62 twin subjects had a SBP in the high-normal range, i.e. above the 90th centile, according to normative data from 61 206 North American children (Swedish reference data is lacking [24]). Amongst those with high SBP, four were MZ twins with TTTS subjects (two donors and two recipients), three were MZ twins without TTTS and small at birth and one was DZ with a normal birth weight. The gestational age, current age and gender distribution were similar in those with and those without a high SBP. The SBP and DBP in children with a history of maternal pre-eclampsia or antenatal steroid therapy did not differ significantly from those without such a history (P-value: 0.67–0.83).

Within all pairs there was a trend towards higher SBP (mean + 2.6 mmHg, P = 0.08) and pulse pressure (PP; mean + 3.0 mmHg, P = 0.09) in the lighter than in the at birth heavier twin. The differences were more marked amongst MZ twins without TTTS; SBP was +5.2 (95% CI: +0.1 to +10, P = 0.04) and PP was +5.8 (95% CI: −0.4 to +12, P = 0.07) mmHg higher in the lighter than in the at birth heavier twin sibling. Within-pair SBP differences did not show any significant relation to the degree of birth weight discordancy (regression coefficient: +4.8 mmHg kg−1 lower birth weight in the lighter twin, P = 0.39). Moreover, we found no significant within-pair differences in DBP or heart rate (Table 2).

Table 2.  Within-pair blood pressure differences in 8-year-old twins with discordant birth weights
Difference in:MZ-no TTTS (n = 13)MZ-TTTS (n = 9)DZ (n = 9)
  1. MZ-no TTTS, monozygous twins without twin-to-twin transfusion syndrome; MZ-TTTS, monozygous twins with TTTS; DZ, dizygous twins; SBP, systolic blood pressure; DBP, diastolic blood pressure; PP, pulse pressure; bpm, beats per minute.

  2. Difference: (Value in the lighter − Value in the heavier twin at birth)/Value in the heavier twin at birth × 100. Adjusting for differences in current weight and height did not alter the results.

  3. Values are expressed as mean (lower to upper limits for 95% confidence interval) values; bold indicates P < 0.05.

SBP (mmHg)+5.2 (+0.1 to +10)−0.1 (−3.2 to +3.1)+1.6 (−6.3 to +9.4)
DBP (mmHg)+1.4 (−2.0 to +4.8)−1.4 (−7.4 to +4.5)+1.1 (−3.5 to +5.7)
PP (mmHg)+5.8 (−0.4 to +12)+1.1 (−3.6 to +5.9)+0.9 (−7.5 to +9.3)
heart rate (bpm)−3.3 (−7.7 to +1.0)−1.3 (−6.1 to +3.5)+4.3 (−1.2 to +9.8)

According to multivariate regression analyses, contributors to the within-pair difference in SBP were pregnancy history of pre-eclampsia, antenatal steroid treatment and monozygosity (Table 3). Age (P = 0.18), gender (P = 0.28), current weight (P = 0.91) and height (P = 0.51) differences did not significantly affect SBP discordance. Pulse pressure differences did not show any significant relations to the independent variables.

Table 3.  Factors independently related to within-pair blood pressure differences in twins with discordant birth weights (n = 31 pairs)
Dependent variableRegression factorRegression coefficient P-value
  1. SBP, systolic blood pressure; DBP, diastolic blood pressure.

SBP differencePre-eclampsia+3.9<0.01
Antenatal steroid treatment+2.70.07
DBP differenceAntenatal steroid treatment+2.50.09

The within-pair difference in DBP tended to be associated with age (regression coefficient: +1.1 mmHg, zero intercept 7.4 years, P = 0.06; Table 3). No other significant associations were noted between risk factors and within-pair differences in DBP or PP.

Arterial dimension and stiffness

The overall mean end-Dd of the CCA was 5.9 (0.1) mm, pulsatile increase in diameter was (ΔD) 0.90 (0.02) mm, strain (ratio: ΔD/Dd) was 15 (0.4)% and CCA stiffness index 3.7 (0.12). These findings are comparable with reference data for healthy school-aged singletons with normal birth weights [18].

Amongst MZ without TTTS and DZ twins, no intra-pair differences were detected in CCA dimensions, changes in pulsatile diameter or stiffness. In the MZ with TTTS group, the lighter (donor) twin had a CCA diameter that was −0.3 (95% CI: −0.6 to 0) mm or 5% narrower than that of the heavier (recipient) twin. The lighter TTTS twin also showed +3 (95% CI: +1 to +5)% higher carotid strain than the heavier sibling (Table 4). According to multivariate regression analyses, there were no significant associations between other risk factors than MZ-TTTS birth weight discordance and within-pair differences in CCA dimensions or pulsatile diameter changes.

Table 4.  Intra-pair differences in arterial dimensions, strain and stiffness in twins with discordant birth weights
 MZ-no TTTS (n = 13)MZ-TTTS (n = 9)DZ (n = 9)
  1. MZ-no TTTS, monozygous twins without twin-to-twin transfusion syndrome; MZ-TTTS, monozygous twins with TTTS; DZ, dizygous twins.

  2. Difference: (Value in the lighter − Value in the heavier twin at birth)/Value in the heavier twin at birth × 100.

  3. Data in bold type indicate significant intra-pair differences.

  4. Strain: pulsatile diameter increase/end-diastolic diameter (%).

  5. Values are expressed as mean (lower to upper limits for 95% confidence interval) values.

Common carotid artery, difference in:
 end-diastolic diameter (mm)−0.1 (−0.4 to +0.3)0.3 (−0.6 to 0)+0.3 (−0.2 to +0.8)
 strain (%)+0.9 (−0.7 to +2.5)+3.1 (+0.9 to +5.3)0 (−3.5 to +3.6)
 stiffness index0.03 (−0.6 to +0.7)−0.5 (−1.2 to +0.2)+0.5 (−0.4 to +1.4)
Abdominal aorta, difference in:
 end-diastolic diameter (mm)−0.3 (−1.4 to +0.7)−0.3 (−1.3 to +0.7)+0.5 (−0.5 to +1.5)
 strain (%)+2.8 (0 to +5.6)+1.5 (−4.2 to +7.1)−3.0 (−7.6 to +1.7)
 stiffness index−0.1 (−0.7 to +0.4)−0.3 (−1.5 to +0.8)+0.5 (−0.1 to +1.2)

The overall mean Dd of the AA was 8.2 (0.2) mm, the ΔD 1.7 (0.05) mm, strain was 21 (0.6)% and AA stiffness index 2.8 (0.11). These findings do not differ significantly from reference data for healthy school-aged singletons with normal birth weights [18].

Irrespective of group, the AA Dd was the same in the lighter and heavier twins. In MZ twins without TTTS, the strain of the AA was +2.8 (95% CI: 0 to +5.6)% higher in the lighter than in the heavier twin. By calculating the arterial stiffness index, we could adjust for differences in strain because of those in BP. We found no intra-pair differences in AA stiffness indices (Table 4) and no association between arterial dimensions and BP.

There were no significant associations between pre-eclampsia, use of antenatal steroids, gender, current age, weight and height and within-pair differences in AA dimensions or pulsatile diameter changes.

Endothelial function

The overall mean basal skin perfusion was 11 (1.5) PU and the maximum increase in perfusion induced by the endothelium-dependent vasodilator ACh was 110 (5.7) PU. These findings do no differ significantly from reference data for healthy school-aged singletons with normal birth weights [18]. Basal skin perfusion did not differ between the lighter and heavier twins. In the MZ group without TTTS, the lighter twin had a 40 PU lower maximum response to ACh than the heavier twin (P < 0.05, Fig. 1). In the MZ with TTTS and DZ groups, the within-pair differences in responses to ACh were not significant.


Figure 1. Skin perfusion in response to repeated transdermal (iontophoresis) doses of acetylcholine (ACh; an endothelium-dependent vasodilator) in healthy 8-year-old twins (n = 62) discordant in birth weight. Amongst monozygous twins without TTTS (MZ-no TTTS, n = 13 pairs), the lighter twin at birth responded significantly less to ACh than the heavier sibling (P < 0.05, paired t-test for maximum perfusion response). In twins suffering from the twin-to-twin transfusion syndrome (MZ-TTTS, n = 9 pairs) during pregnancy and in dizygous (DZ) twins (n = 9 pairs), no significant within-pair differences in ACh response were found (mean values and SEM bars are shown).

Download figure to PowerPoint

Pre-eclampsia reduced the within-pair difference in maximum skin perfusion response to ACh (6.2 PU, n.s.). By contrast, in pregnancies with discordant twin growth but without pre-eclampsia the ACh skin perfusion response was 34 PU lower in the at birth lighter twin (P < 0.05).

The mean basal skin perfusion before testing endothelium-independent vasodilation was 11 (1.0) PU. The maximum increase in skin perfusion after topical application of nitroglycerin was 35 (2.4) PU.

The mean difference in response to nitroglycerin between lighter and heavier twins was −4.8 PU (95% CI: −14 to +4.4). There were no significant within-pair differences between groups of twins. According to multivariate regression analyses, gender was the only factor contributing to the within-pair difference in endothelium-independent vasodilation. The mean difference in skin blood flow response to nitroglycerin between lighter and heavier twins was statistically significant amongst girls (−13 PU, P < 0.05) but not amongst boys (−4.5, n.s.).


  1. Top of page
  2. Abstract.
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conflict of interest
  8. Acknowledgements
  9. References

In this 8-year follow-up study, we found that amongst MZ twins without TTTS with discordant fetal growth, the in utero growth-retarded (SGA) twin had significantly higher SBP and impaired endothelial function in childhood, when compared to the cotwin with appropriate birth weight. In addition, amongst MZ twins suffering from TTTS in fetal life, the smaller donor twin showed narrower carotid artery at school-age. Given the fact that familial effects (genetic and postnatal environmental factors) were controlled for, different fetal exposures are likely to be involved in the explanation of these results. Longitudinal studies will be needed to fully understand the significance of these findings in adult life. They may, independently or in combination, contribute to an increased risk for future coronary heart disease and stroke in the affected twin.

Although our focus was on intra-pair differences, we note that in the subgroup of twins with an adjusted SBP above the 90th percentile, seven of eight had suffered from adverse intrauterine influences in terms of severe growth restriction or TTTS. These findings indicate that fetal programming may contribute to later BP in humans, in accordance with some previous studies [6, 7, 9].

The twins of pre-eclamptic mothers did not show higher BPs than those without such a history. However, in pregnancies complicated by pre-eclampsia, the growth-restricted twin had a higher SBP than its cotwin at follow up. Although several twin studies have suggested that the fetoplacental unit is involved in fetal programming [6, 7, 9], there are no other twin-data on the role of pre-eclampsia. Pre-eclampsia is a hypertensive disorder of the pregnant woman, triggered by vascular lesions and endothelial dysfunction of the placenta. It is two to three times commoner in women with twin than in those with singleton pregnancies [26] and even more common in twin gestations complicated by severe discordancy in fetal growth [27]. Its interaction with twin discordances in birth weight and later differences in BP is intriguing as it decreased rather than increased covariance. Although a definite explanation cannot be given, the effect modification by pre-eclampsia can be interpreted in one of two ways: either it represents an additional environmental disadvantage to unequal placental sharing or it represents a genetic susceptibility triggered by fetal growth restriction.

Fetal exposure to antenatal steroids has been associated with a rise in BP in adolescence [28]. We found no difference in BP between twins who had received such treatment and those who did not. In MZ twins with TTTS, all had been exposed to antenatal steroids and amongst those with high SBP at follow up, there was an equal distribution between donors and recipients. As TTTS and/or antenatal steroids affected both twins in utero, both could have been programmed. In TTTS, humoral factors compensating for anaemia, hypoxia and hypovolaemia in the donor twin fetus reasonably also reaches the recipient twin fetus. Such an interpretation would help to explain why no within-pair differences in BP were found in the subgroup of MZ twins suffering from TTTS.

An interesting hypothesis of the inverse association between birth weight and BP concerns early reduction in arterial compliance. A combination of abnormal haemodynamics and malnutrition accompanying growth restriction in utero has been proposed as a cause of impaired elastin incorporation in the walls of the aorta and large arteries [17]. Neither the present nor other studies in children have shown a clear link between fetal growth restriction, premature stiffening of the aorta and a subsequent rise in BP [18]. However, aortic narrowing has been described amongst schoolchildren born SGA [29]. We have previously found a stiffer carotid artery in schoolchildren with low-birth weight. Present data from TTTS pairs show that poor fetal growth is also associated with narrowing of the carotid artery [18]. From our data and those of others [19], it seems that disturbed carotid haemodynamics in the fetus could be an important determinant of long-term carotid development and growth.

Systemic impairment of endothelial function has been found in newborn infants, children and young adults with low-birth weight [18, 20, 21]. Endothelial dysfunction predicts later hypertension, accelerated atherosclerosis and coronary heart disease. The present study provides evidence of an intrauterine contribution to later endothelial dysfunction in humans. The lack of a significant difference in endothelial function amongst MZ twins with TTTS may be due to their shorter gestation (four pairs delivered before 32 weeks of gestation) because preterm delivery attenuates developmental programming of the endothelium [30].

Formula feeding of infants has been associated with a higher BP in children and adolescents [31, 32]. We cannot exclude confounding from different within-pair infant feeding practices amongst the subjects of this study. The routine in our units, however, is to give the smaller twin sibling priority if the amount of breast milk is insufficient for both.

We did not perform an exercise test on these children, which could have unmasked further within-pair differences in BP and heart rate responses. On average, we found 2–3 mm higher resting SBP in the MZ twin sibling suffering from intrauterine growth failure. This is a larger effect than that reported in epidemiological studies [13, 14]. The validity of our findings may be limited to the subgroup of twins with the largest differences in birth weight. In addition, the generalizability of twin studies to the general population is sometimes a concern, especially as twin growth in utero, even under normal conditions, differs from that of singletons. Because prospectively collected reference data on intrauterine twin growth are lacking, our definitions of being small (SGA) or appropriately sized at birth (AGA) were based on reference data for Swedish singletons.

This study has limitations in statistical power and therefore, the results must not be over-interpreted. The total number of twins possible to recruit, permitted detection of a within-pair mean difference of about 0.5 SD or more. Nevertheless, the choice of a well-characterized twin-cohort with large birth weight discordances provided an opportunity to obtain more detailed insights into perinatal contributions to later BP and vascular structure and function.

The birth weight discordance was lower in MZ twins with TTTS, but was greater in MZ without TTTS and DZ twin pairs. This is most likely reflected that all MZ with TTTS, according to the diagnostic criteria, were diagnosed in utero, and that action (delivery with C-section before term) was taken before the condition deteriorated further.

In conclusion, our findings suggest the occurrence of fetal programming of BP, endothelial function and carotid growth. Although twins overall have similar prevalence of hypertension and cardiovascular mortality as singletons, severe complications during pregnancy such as IUGR (intrauterine growth retardation) and TTTS, are much commoner than in singletons. As shown here, this may also affect health after the perinatal period.


  1. Top of page
  2. Abstract.
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conflict of interest
  8. Acknowledgements
  9. References

This study was supported by the Swedish Research Council (project no. 71P-14158), Swedish Heart Lung Foundation, Swedish Society of Medicine, Karolinska Institutet's Research Foundations, Stiftelsen Samariten, Stiftelsen Frimurare Barnhuset and Sällskapet Barnavård, Karolinska Hospital.


  1. Top of page
  2. Abstract.
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conflict of interest
  8. Acknowledgements
  9. References
  • 1
    Curhan GC, Willett WC, Rimm EB, Spiegelman D, Ascherio AL, Stampfer MJ. Birth weight and adult hypertension, diabetes mellitus, and obesity in US men. Circulation 1996; 94: 324650.
  • 2
    Law CM, Shiell AW. Is blood pressure inversely related to birth weight? The strength of evidence from a systematic review of the literature. J Hypertens 1996; 14: 93541.
  • 3
    Nilsson PM, Östergren PO, Nyberg P, Söderström M, Allebeck P. Low birth weight is associated with elevated systolic blood pressure in adolescence: a prospective study of a birth cohort of 149378 Swedish boys. J Hypertens 1997; 15: 162731.
  • 4
    Huxley RR, Shiell AW, Law CM. The role of size at birth and postnatal catch-up growth in determining systolic blood pressure: a systematic review of the literature. J Hypertens 2000; 18: 81531.
  • 5
    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: 65965.
  • 6
    Dwyer T, Blizzard L, Morley R, Ponsonby AL. Within-pair association between birth weight and blood pressure at age 8 in twins from a cohort study. BMJ 1999; 319: 13259.
  • 7
    Poulter NR, Chang CL, MacGregor AJ, Snieder H, Spector TD. Association between birth weight and adult blood pressure in twins: historical cohort study. BMJ 1999; 319: 13303.
  • 8
    Ijzerman RG, Stehouwer CD, Boomsma DI. Evidence for genetic factors explaining the birth weight–blood pressure relation. Analysis in twins. Hypertension 2000; 36: 100812.
  • 9
    Loos RJ, Fagard R, Beunen G, Derom C, Vlietinck R. Birth weight and blood pressure in young adults. A prospective twin study. Circulation 2001; 104: 16338.
  • 10
    Zhang J, Brenner RA, Klebanoff MA. Differences in birth weight and blood pressure at age 7 years among twins. Am J Epidemiol 2001; 153: 77982.
  • 11
    Poulsen P, Vaag A, Kyvik K, Beck-Nielsen H. Genetic versus environmental aetiology of the metabolic syndrome among male and female twins. Diabetologia 2001; 44: 53743.
  • 12
    Baird J, Osmond C, MacGregor A, Snieder H, Hales CN, Phillips DI. Testing the fetal origins hypothesis in twins: the Birmingham twin study. Diabetologia 2001; 44: 339.
  • 13
    Johansson-Kark M, Rasmussen F, De Stavola B, Leon DA. Fetal growth and systolic blood pressure in young adulthood: the Swedish young male twins study. Paediatr Perinat Epidemiol 2002; 16: 2009.
  • 14
    Christensen K, Stovring H, McGue M. Do genetic factors contribute to the association between birth weight and blood pressure? J Epidemiol Community Health 2001; 55: 5837.
  • 15
    Leon DA. The foetal origins of adult disease. Interpreting the evidence from twin studies. Twin Res 2001; 4: 3216.
  • 16
    Gardiner HM, Taylor MJO, Karatza A et al. Twin-twin transfusion syndrome. The influence of intrauterine laser photocoagulation on arterial distensibility in childhood. Circulation 2003; 107: 190611.
  • 17
    Martyn CN, Greenwald SE. A hypothesis about a mechanism for the programming of blood pressure and vascular disease in early life. Clin Exp Pharmacol Physiol 2001; 28: 94851.
  • 18
    Martin H, Hu J, Gennser G, Norman M. Impaired endothelial function and increased carotid stiffness in 9-year-old children with low birth weight. Circulation 2000; 102: 273944.
  • 19
    Martyn CN, Barker DJP, Jespersen S, Greenwald S, Osmond C, Berry C. Growth in utero, adult blood pressure and arterial compliance. Br Heart J 1995; 73: 11621.
  • 20
    Leeson CP, Kattenhorn M, Morley R, Lucas A, Deanfield JE. Impact of low birth weight and cardiovascular risk factors on endothelial function in early adult life. Circulation 2001; 103: 12648.
  • 21
    Goodfellow J, Bellamy MF, Gorman ST et al. Endothelial function is impaired in fit young adults of low birth weight. Cardiovasc Res 1998; 40: 6006.
  • 22
    Marsal K, Persson PH, Larsen T, Lilja H, Selbing A, Sultan B. Intrauterine growth curves based on ultrasonically estimated foetal weights. Acta Paediatr 1996; 85: 8438.
  • 23
    Quintero RA, Morales WJ, Allen MH, Bornick PW, Johnson PK, Kruger M. Staging of twin-twin transfusion syndrome. J Perinatol 1999; 19: 5505.
  • 24
    Anonymous. Update on the 1987 task force report on high blood pressure in children and adolescents: a working group report from the National High Blood Pressure Education Program. National High Blood Pressure Education Program Working Group on Hypertension Control in Children and Adolescents. Pediatrics 1996; 98(4 Pt 1): 64958.
  • 25
    Lindstrom K, Gennser G, Sindberg Eriksen P, Benthin M, Dahl P. An improved echo-tracker for studies on pulse waves in the fetal aorta. In: RolfeP, ed. Fetal Physiological Measurements. London, UK: Butterworths, 1987; 21726.
  • 26
    Sibai BM, Hauth J, Caritis S et al. Hypertensive disorders in twin versus singleton gestations. Am J Obstet Gynecol 2000; 182: 93842.
  • 27
    Gonzalez-Quintero VH, Luke B, O'Sullivan MJ et al. Antenatal factors associated with significant birth weight discordancy in twin gestations. Am J Obstet Gynecol 2003; 189: 8137.
  • 28
    Doyle LW, Ford GW, Davis NM, Callanan C. Antenatal corticosteroid therapy and blood pressure at 14 years of age in preterm children. Clin Sci (Lond) 2000; 98: 1278.
  • 29
    Ley D, Stale H, Marsal K. Aortic vessel wall characteristics and blood pressure in children with intrauterine growth retardation and abnormal foetal aortic blood flow. Acta Paediatr 1997; 86: 299305.
  • 30
    Norman M, Martin H. Preterm birth attenuates association between low birth weight and endothelial dysfunction. Circulation 2003; 108: 9961001.
  • 31
    Singhal A, Cole TJ, Lucas A. Early nutrition in preterm infants and later blood pressure: two cohorts after randomised trials. Lancet 2001; 357: 4139.
  • 32
    Wilson AC, Forsyth JS, Greene SA, Irvine L, Hau C, Howie PW. Relation of infant diet to childhood health: seven-year follow-up of cohort of children in Dundee infant feeding study. BMJ 1998; 316: 215.