Correspondence: Professor B. Trudinger, Department of Obstetrics and Gynaecology, University of Sydney at Westmead Hospital, Westmead Hospital, Westmead, NSW 2145, Australia.
Objective To measure the blood apolipoprotein A-1 and apolipoprotein B in the fetal circulation in normal pregnancy and in pregnancy with evidence of vascular disease in the fetal umbilical placental circulation defined in the antenatal period by Doppler ultrasound study.
Design An observational study to compare fetal plasma apolipoprotein levels in normal and complicated pregnancy.
Setting A university hospital tertiary referral obstetric unit.
Samples Umbilical vein blood was collected at delivery from 22 normal fetuses delivered by elective caesarean section for non fetal reasons and 30 fetuses with evidence of umbilical placental vascular disease identified antenatally by Doppler ultrasound study.
Methods Plasma apolipoprotein A-1 and B were determined using an enzyme-linked immunosorbent assay (ELISA) methods.
Main outcome measures Fetal plasma levels of apolipoprotein A-1 and B were measured.
Results There was a significantly lower level of fetal plasma apolipoprotein A-1 in placental insufficiency [placental insufficiency vs normal pregnancy, median 0.30 g/L (interquartile range 0.24, 0.39 g/L) vs 0.35 g/L (0.31, 0.42 g/L), P= 0.045]. In contrast, the levels of fetal plasma apolipoprotein B in placental insufficiency [0.20 g/L (0.17, 0.26 g/L)] were significantly increased compared with normal pregnancy [0.16 g/L (0.14, 0.20 g/L), P= 0.03]. The ratio of fetal plasma apolipoprotein B to A-1 was also substantially higher in placental insufficiency [0.68 (0.55, 0.83)] than in normal pregnancy [0.45 (0.36, 0.60), P= 0.0003].
Conclusions Our study has demonstrated that levels of fetal plasma apolipoprotein A-1, apolipoprotein B and the ratio of apolipoprotein B to A-1 were altered in the fetuses who are victims of umbilical placental insufficiency in the same direction as in adults associated with a high risk of atherogenesis.
A relationship between blood lipid profile and atherosclerotic vascular disease is well established in adults. Elevated plasma levels of low density lipoprotein cholesterol are associated with increased risk1,2. In contrast, high density lipoprotein cholesterol is anti-atherogenic3,4. The carrier protein for low density lipoprotein cholesterol is apolipoprotein B. The principal protein moiety of high density lipoprotein cholesterol is apolipoprotein A-1. These carriers are stronger predictors of atherosclerotic risk than their corresponding lipoprotein cholesterols5. The ratio of apolipoprotein B to A-1 is a better indicator for atherosclerotic risk than either plasma concentration alone6 and a high ratio predicts an increased risk of atherosclerosis7.
The term placental insufficiency has long existed in obstetric language to reflect a circumstance of inadequate blood flow to the placenta. It is associated with the maternal syndrome of pre-eclampsia8 and the fetal syndrome of intrauterine growth retardation9,10. Vascular pathology in both the uterine placental vascular bed and the umbilical placental circulation have been described. On the maternal side, reduced perfusion is an important feature. Acute atherosis in the spiral arteries of the placental bed has been associated with endothelial cell injury and aggregates of fibrin, platelets and lipid loaded macrophages11. Thrombosis is common. On the fetal side thrombosis12, vessel obliteration13 and platelet activation and consumption14 have been demonstrated. Doppler ultrasound study of the umbilical artery blood flow velocity waveform can identify pregnancy associated with this placental pathology in the fetal circulation to the placenta8,9. There has been debate as to whether the primary pathology is in the fetal placental vasculature which determines the uteroplacental response or whether the uterine circulation constrains the placenta and fetus8–15.
There is considerable contemporary interest in the fact that low birthweight (i.e. a small fetus who is possibly the victim of placental insufficiency) predicts those in later life predisposed to atherosclerotic cardiovascular disease16. In what has been referred to as the ‘Barker Hypothesis’ it has been proposed that adult cardiovascular disease is programmed during the period of rapid growth in fetal life. According to this hypothesis somehow an adverse in utero environment changes the structure, physiology and metabolism of the fetus before birth and therewith determines the development of cardiovascular disease in later life. Evidence advanced to support this includes the observations that risk factors for cardiovascular disease including increased blood pressure17 and higher plasma concentrations of glucose, insulin18,19, fibrinogen, factor VII20,21 and apolipoprotein B22 measured in adult life have been retrospectively associated with low birthweight and inadequate fetal nutrition. These associations exist with babies who are born small for gestational age rather than those born prematurely.
We hypothesised that the vascular disease in the fetal umbilical placental circulation in placental insufficiency is similar to the disease of the coronary and cerebral circulations in later adult life and depends upon vessel injury and an atherogenic lipoprotein profile. In this study we investigated changes in the levels of apolipoprotein B and apolipoprotein A-1 in fetal blood when vascular disease was known to be present in the fetal umbilical circulation in placental insufficiency.
Fifty-two women were sequentially recruited from the Department of Obstetrics and Gynaecology, Westmead Hospital. We studied 22 fetuses from normal pregnancy and 30 fetuses from pregnancies with evidence of umbilical placental vascular disease. In the normal group we selected pregnancies which were uncomplicated with no maternal hypertension. All delivered at term (> 37 weeks) by elective caesarean section for non fetal reasons since we anticipated a high incidence of elective caesarean delivery in our fetal compromise group. The birthweight was greater than the tenth centile. The placental insufficiency group was selected because of clinical concern about fetal welfare and the presence of umbilical placental vascular disease as signalled by an abnormal result in the Doppler flow velocity waveform pattern recorded from the fetal umbilical artery. These cases were identified when referred to our Fetal Welfare Laboratory for fetal ultrasound studies. They were a sequential group over the time period of the study. The umbilical artery Doppler systolic to diastolic ratio was greater than the 95th centile using our reported normal range23. In this group 17 women had hypertension (diastolic blood pressure > 90 mmHg on at least two occasions).
Blood was collected from the umbilical vein at delivery into a citrate-containing tube and the plasma was separated within one hour. Plasma was stored at −70°C until analysis.
Plasma levels of apolipoprotein A-1 and apolipoprotein B were measured using enzyme-linked immunosorbent assay (ELISA) methods with the following details, developed in the Department of Cardiovascular Medicine, the Prince Henry/Prince of Wales Hospitals, Sydney:
1Apolipoprotein A-1 assay24: Sheep polyclonal anti-human-apolipoprotein A-1 antiserum was used as a coating antibody to capture the sample apolipoprotein A-1. Horseradish peroxidase-conjugated anti-human-apolipoprotein A-1 antiserum was used to quantify the apolipoprotein A-1 levels.
2Apolipoprotein B assay25: The method for apolipoprotein B assay is similar to apolipoprotein A-1 ELISA method but requires different antibodies and some minor alternations for the incubation time. In the ELISA, anti-apolipoprotein B and anti-apolipoprotein B-horseradish peroxidase conjugate were used instead of anti-apolipoprotein A-1 and anti-apolipoprotein A-1-horseradish peroxidase.
The inter-assay and intra-assay coefficients of variations were 5.1% and 8.2% for the apolipoprotein A-1 assay, and 4.0% and 6.6% for the apolipoprotein B assay.
The clinical data from the two groups of infants and their mothers were compared with the Student's unpaired t test. Because the apolipoprotein levels were not normally distributed, nonparametric statistic analysis (Mann-Whitney U test) was used to compare the differences between the two groups. Fisher's exact test was used for comparison between categorical variables. Correlations were performed using Pearson's correlation coefficient after logarithmic transformation of the apolipoprotein levels. Spearman's ρ correlation coefficient was used for categorical variables. Apolipoprotein data were expressed as median value and interquartile range. A two-tailed P value less than 0.05 was considered significant.
These studies were performed with approval of Western Sydney Area Health Service Ethics Committee.
We studied pregnancy in the third trimester. The normal pregnancy group was by definition delivered after 37 weeks. The pregnancy characteristics are shown in Table 1.
Table 1. Subject characteristics. Values are given as mean [SEM] or n(%).
Normal Pregnancy (n= 22)
Placental insufficiency (n= 30)
*No. of primiparae.
Maternal age (years)
Gestational age at delivery (weeks)
Infant birth weight (g)
There was a significantly lower level of fetal plasma apolipoprotein A-1 in placental insufficiency [median 0.30 g/L (interquartile range 0.24, 0.39 g/L)] than in normal pregnancy group [0.35 g/L (0.31, 0.42 g/L), P= 0.045] (Fig. 1 a). In contrast, the levels of fetal plasma apolipoprotein B in placental insufficiency [0.20 g/L (0.17, 0.26 g/L)] were significantly increased compared with normal pregnancy [0.16 g/L (0.14, 0.20 g/L), P= 0.03] (Fig. 1 b). The ratio of fetal plasma apolipoprotein B to apolipoprotein A-1, a marker of potential atherogenesis, was also substantially higher in placental insufficiency [0.68 (0.55, 0.83)] compared with normal pregnancy [0.45 (0.36, 0.60), P= 0.0003] (Fig. 1 c).
To assess whether the apolipoprotein levels were associated with maternal age, gestational age, actual infant birthweight, birthweight centile and placenta weight, the correlation coefficient was assessed. As the distribution of apolipoproteins was not normal, these data were logarithmically transformed. As shown in Table 2, neither log apolipoprotein B nor log apolipoprotein A-1 were significantly correlated with maternal age, gestational age, infant birthweight and placenta weight in either the placental insufficiency or the normal pregnancy group. The parity was not significantly associated with apolipoprotein A-1 in normal pregnancy (r= 0.18, P= 0.08) and in placental insufficiency (r= 0.16, P= 0.39). Parity was also not associated with apolipoprotein B in normal pregnancy (r=−0.1, P= 0.31) and in placental insufficiency (r= 0.24, P= 0.21). The birthweight centile was significantly correlated with the logarithm of apolipoprotein A-1 concentration (r= 0.53, P= 0.02) in normal pregnancy, but not in placental insufficiency population.
Table 2. Correlation coefficients between Log apo B and Log apo A-1, and maternal age, gestational age at delivery, infant birthweight and placenta weight.
Log apo B
Log apo A-1
In normal pregnancy
Maternal age(n= 22)
Getational age at delivery (n= 20)
Infant birthweight (n= 22)
Birthweight centile (n= 19)
Placenta weight (n= 22)
In placental insufficiency
Maternal age (n= 30)
Getational age at delivery (n= 30)
Infant birthweight (n= 30)
Birthweight centile (n= 30)
Placenta weight (n= 30)
To evaluate whether maternal hypertension had an effect on the fetal lipoprotein profile changes in placental insufficiency, we compared the fetal plasma apolipoprotein levels in the group with placental insufficiency when maternal hypertension was present (n= 17) and absent (n= 13). No significant differences were observed between these two groups in apolipoprotein A-1 [no hypertension vs with hypertension, median 0.34 g/L (interquartile range 0.22, 0.44 g/L) vs 0.29 g/L (0.25, 0.36 g/L), P= 0.59] and apolipoprotein B [0.17 g/L (0.14, 0.29 g/L) vs 0.21 g/L (0.19, 0.26 g/L), P= 0.27]. The ratio of apolipoprotein A-1 to B was also not significantly different between placental insufficiency with hypertension [0.78 (0.57, 0.83)] and without hypertension [0.67 (0.44, 0.72), P= 0.15].
Our study has demonstrated that compared with normal pregnancy umbilical placental insufficiency is characterised by a low fetal plasma apolipoprotein A-1, high apolipoprotein B and high ratio of apolipoprotein B to A-1. Our findings contrast to the report of Spencer et al.26 in that apolipoprotein A-1, apolipoprotein B and the ratio of apolipoprotein A-1 to B were similar in pregnancies with normal or retarded fetal growth during the third trimester. In that study cases were not selected for study on the basis of the antenatal identification of umbilical placental vascular disease by Doppler results. All babies were delivered after 36 weeks of gestation. In our study placental insufficiency led to delivery at an earlier stages. This does suggest that we have studied a more severely affected group of fetuses with definite evidence of placental vascular disease.
The findings in the present study are consistent with our hypothesis that the decreased level of the antiatherogenic apolipoprotein A-1 and increased level of the atherogenic apolipoprotein B in fetal blood may contribute to the vascular disease in the fetal umbilical placental circulation in placental insufficiency. The alterations of plasma apolipoprotein A-1, apolipoprotein B and the ratio of apolipoprotein B to A-1 in the fetuses who are victims of placental insufficiency is the same as in the adults with high risk of atherogenesis. They are widely regarded as risk factors present before disease has occurred rather than the consequences of disease development27. In the fetus we cannot be certain whether the alteration of apolipoprotein A-1 and apolipoprotein B causes vascular disease or accelerates progression after an initiating injury. We speculate that in the present study increased levels of apolipoprotein B and decreased levels of apolipoprotein A-1 could cause ‘acute atherosis’ in umbilical placental circulation similar to atherosclerosis in adult life. An alternative possibility is that the alteration of apolipoprotein A-1, apolipoprotein B and the ratio of apolipoprotein B/A-1 is the consequence of placental insufficiency. As this report is an observational study we can not distinguish cause and effect.
The placental insufficiency group consisted of fetuses who were the victims of clinically identified umbilical placental insufficiency, a fact confirmed by umbilical Doppler flow velocity measurements. These fetuses may be born small for gestational age. The pathology and pathophysiology of umbilical placental vascular disease associated with in utero fetal growth restriction has been studied. Abnormal umbilical artery Doppler flow velocity pattern have been associated with obliteration of the small arteries and arterioles of the umbilical villus vascular tree and degenerative changes in surviving vessels13. Sixty to seventy percent of small for gestational age infants show evidence of umbilical placental vascular disease recognised by an abnormal umbilical artery Doppler flow velocity waveform study9. Recognition of this placental vascular disease is important in clinical practice. Meta-analysis of the randomised controlled trials evaluating the use of this umbilical Doppler flow velocity measurements in high risk pregnancy has shown a 32% reduction in perinatal mortality28. It has furthermore been suggested that fetuses ‘small for dates with normal umbilical artery Doppler (have) a benign disease’29.
In this study we assayed fetal plasma collected at delivery in the third trimester of pregnancy. There were differences in gestational age and maternal age in our placental insufficiency group compared with the normal pregnancy group. Maternal age and gestational period did not significantly correlate with apolipoprotein A-1 and apolipoprotein B in either the normal pregnancy or placental insufficiency groups. It is therefore unlikely that these differences could explain the decreased apolipoprotein A-1 and increased apolipoprotein B levels in placental insufficiency. A weakness of our study design is the difference in gestational age. In our study the gestational age at delivery was different in our two groups and the overlap was small. We elected not to study spontaneous preterm delivery fetuses because of the possibility of unrecognised vascular disease and fetal compromise. There exists no data to show gestational age is associated with decreased levels of apolipoprotein A-1 and increased levels of apolipoprotein B in the third trimester of pregnancy.
It is unlikely that maternal hypertension had a significant effect on the lipoprotein profile changes in placental insufficiency. No significant differences were found in apolipoprotein A-1, apolipoprotein B and the ratio of apolipoprotein B to A-1 between placental insufficiency with and without hypertension.
The birthweight centile was significantly associated with apolipoprotein A-1 in the normal pregnancy group. This finding is consistent with a previous report30 and of extreme interest and potential significance. Higher levels of apolipoprotein A-1 may protect from vascular disease in the placenta leading to good fetal growth. Absence of vascular disease may be a major determinant of size at birth even in normal pregnancy. It has been observed that the blood vessels of the placenta age a ‘lifetime’ in the nine months of pregnancy. Similarities in the degenerative arterial disease and apolipoprotein profile at the end of pregnancy and in adult life suggest a common pathogenic mechanism.
In conclusion, our study has shown that levels of fetal plasma apolipoprotein A-1, apolipoprotein B and the ratio of apolipoprotein B/A-1 are altered in the fetuses who are victims of umbilical placental insufficiency in the same direction as in adults in association with a high risk of atherogenesis. Whether the alteration of apolipoprotein A-1 and apolipoprotein B is a causal or permissive factor for umbilical placental vascular disease in placental insufficiency remains to be established. The effects of gestational age at delivery on the differences in apolipoprotein A-1 and apolipoprotein B levels between normal pregnancy and pregnancy with placental insufficiency needs to be further studied.
The study was funded by Department of Obstetrics and Gynaecology, Westmead Hospital. We wish to thank Ms N. Duarte and Mrs J. Lynch, Department of Cardiovascular Medicine, Prince Henry/Prince of Wales Hospital, Sydney, for their assistance in measuring plasma apolipoproteins and Dr X. Wang, Department of Obstetrics and Gynaecology, Westmead Hospital, Sydney, for collecting plasma samples.