Whether arterial elasticity is reduced in pre-eclampsia has been investigated only rarely. This study aimed to characterize in vivo the carotid arterial intima–media thickness (IMT) and mechanical properties in women with pre-eclampsia by employing a radiofrequency ultrasound technique.
We included 22 late-onset pre-eclamptic pregnant women and 28 normotensive pregnant women who were matched for age (29 ± 6 vs 27 ± 3, P = 0.09) and gestational age (36.0 ± 3.2 vs 35.8 ± 2.4 weeks, P = 0.802). All women were nulliparous with singleton pregnancy. The pre-eclamptic women had a significantly higher arterial pressure than did the normotensive women (P < 0.0001). All women underwent right common carotid arterial measurements with an ultrasound machine equipped with automatic Quality IMT (QIMT) and Quality Arterial Stiffness (QAS) capability. At follow-up examination 18 months after parturition, measurements were repeated in 10 of the pre-eclamptic women and 11 of the normotensive women.
In pre-eclamptic compared with normotensive pregnancy, carotid arterial IMT (459 ± 95 vs 351 ± 85 µm, P = 0.0001), internal diameter (7.8 ± 0.5 vs 7.2 ± 0.4 mm, P < 0.0001), pulse wave velocity (7.1 ± 1.7 vs 6.0 ± 1.5 m/s, P = 0.007), augmentation index (7.9 ± 9.2 vs −5.0 ± 5.6%, P < 0.0001) and arterial wall tension (55.0 ± 6.5vs 38.6 ± 4.9 mmHg/cm, P < 0.0001) were significantly greater, and the distensibility coefficient (0.020 ± 0.009 vs 0.029 ± 0.011 1/kPa, P = 0.006) was significantly smaller, remaining so after adjusting for body mass index and carotid arterial pressure. Eighteen months after parturition, carotid arterial internal diameter, pressure and wall tension remained greater in the pre-eclamptic group.
Pre-eclampsia, a condition specific to pregnancy, is a leading cause of maternal and perinatal mortality and morbidity. It has been shown in non-pregnant patients that hypertension increases the tensile stress applied on the carotid artery; in essential hypertension this leads to increased carotid arterial intima–media thickness (IMT) and stiffness[1, 2]. However, whether this holds true for pregnant women with pre-eclampsia has not been elucidated. Comprehensive evaluation of arterial structure and function of the peripheral vessels may enhance our understanding of the pathophysiology and management of women with pre-eclampsia. In some studies, the pre-eclamptic group had an increased IMT of the common femoral arteries, and in other studies, the pulse wave velocity (PWV), a measure of stiffness in the large arteries, was significantly increased in the carotid–femoral segment in pre-eclamptic compared with normotensive pregnancy (NP)[3-7]. However, in these studies the parameters of PWV and augmentation index (AIx) obtained by applanation tonometry likely represent only global characteristics of the arterial tree rather than local vascular alterations. We therefore hypothesized that arterial remodeling would also occur in local, elastic arteries, such as the carotid artery, and that this would increase the risk for future cardiovascular disease.
High-resolution ultrasound acquisitions based on radiofrequency signal, which preserves all the characteristics of the acoustic wave in terms of amplitude and phase, give the opportunity to assess precisely vessel-wall properties. This newly developed radiofrequency-based ultrasound vessel-wall tracking technique has been validated and should be capable of giving a rapid, sensitive (resolution, 17 µm) and highly reproducible estimate of the thickness between the blood–intima and the media–adventitia acoustic interfaces, i.e. the IMT, as well as automatically measuring arterial stiffness between systole and diastole. The purpose of this cross-sectional study was to use this technique to test the hypothesis that pre-eclampsia is accompanied by increased carotid arterial IMT and arterial stiffness.
This prospective study included 22 women with late-onset pre-eclampsia who had received no antihypertensive treatment prior to admission, and 28 age-matched (29 ± 6 vs 27 ± 3 years, P = 0.09) and gestational age-matched (36.0 ± 3.2 vs 35.8 ± 2.4 weeks, P = 0.802) normotensive pregnant women, all recruited from the routine antenatal clinic of Tangdu Hospital, China, between January 2010 and April 2012. The diagnosis of pre-eclampsia was based on the guidelines of the International Society for the Study of Hypertension in Pregnancy. All women were nulliparous with singleton pregnancy. Four pre-eclamptic women had a growth-restricted fetus. No woman had other risk factors (e.g. smoking, diabetes, hypercholesterolemia) for arterial stiffness. Women with gestational hypertension or chronic hypertension were excluded. The study was approved by the Human Subjects Ethics Committee of the Fourth Military Medical University and all women gave informed consent to participate.
All examinations were carried out in a quiet room before noon. Women were instructed to rest for 10 min, then brachial blood pressure (BP) measurements were performed by a single investigator (N.Z.), using an oscillometric device (Yuyue, Yuyue Equipment & Supply Co., Ltd. Jiangsu, China). Measurements were taken at 3-min intervals for 20 min and the average was taken as the resting BP level. Mean arterial pressure (MAP) was calculated as: DBP + ((SBP – DBP)/3), where DBP is diastolic BP and SBP is systolic BP.
Women were then placed in a supine position prior to ultrasound examination. All measurements were performed by one of two investigators (D.X. and L.-J.Y.), who were blinded to the diagnosis of pre-eclampsia, using a MylabTwice (Esaote, Genoa, Italy) ultrasound machine equipped with a 12-MHz vascular probe (LA523) and with automatic Quality IMT (QIMT) and Quality Arterial Stiffness (QAS) capability. Traditional echocardiography was performed first, with cardiac output (CO) measured separately from the carotid arterial measurements. CO was calculated using the Teichholz equation method using M-mode echocardiography. Total peripheral resistance (TPR, in dynes/s/cm5) was calculated as TPR = MAP/CO × 80.
Six consecutive right common carotid arterial measurements were then performed and only when all six met the quality standard (Quality Control shown as a green number on-screen during scanning) was their average taken as the final result for this patient. These measurements and average calculations were done automatically and displayed on the left side of the screen, as shown in Figure 1.
QIMT and carotid arterial internal diameter measurements were performed in a longitudinal section, strictly perpendicular to the ultrasound beam, with both arterial walls clearly visualized. A high-quality standard B-mode image was acquired along a section of the right common carotid artery at least 1.5 cm in length, and the automatic QIMT calculation was activated, measuring the IMT in real time, using the radiofrequency reception signal (Figure 1a).
Automatic QAS measurements were performed on the same right common carotid arterial segments as those used for the measurement of IMT, evaluating the modification of arterial internal diameter between systolic and diastolic phases (Figure 1b) as follows. Carotid arterial internal diameter waveforms were assessed by means of ultrasound and converted to carotid arterial pressure waveforms using an empirically derived exponential relationship between pressure and arterial cross-section. The derived carotid arterial pressure waveform was calibrated to brachial end-diastolic pressure and MAP by iteratively changing the wall rigidity coefficient (Figure 2). This allowed calculation of the arterial stiffness and arterial wall tension.
The following carotid arterial stiffness indices were obtained (Figure 2): PWV (m/s), distensibility coefficient (DC (1/KPa)) and AIx (%), together with local systolic blood pressure (SBPLoc) and local diastolic blood pressure (DBPLoc).
PWV was calculated using the following equation: , where D is diastolic arterial diameter, ΔD is change of diameter in systole, Δp is local pulse pressure, ρ is blood density and DC is the fractional change in arterial cross-sectional area relative to the change in arterial pressure, calculated as: , where A is diastolic area, ΔA is change of area in systole.
AIx was calculated as: AIx = (AP/(SBPLoc – DBPLoc)) × 100, where AP is augmented pressure, calculated as: AP = SBPLoc – PT1, where PT1 is pressure at inflection point (T1).
Arterial wall tension (T (mmHg/cm)) was calculated as: T = P (mmHg/cm2) × r (cm), where P is the pressure imposed on the arterial wall and r is the radius of the carotid artery.
Follow-up examination was performed in 10 of the pre-eclamptic women and 11 of the normotensive women 18 months after parturition and the above parameters were remeasured using the same methods.
Reproducibility and variability
For evaluation of intraobserver variability of IMT measurements and arterial stiffness parameters of PWV, one of the two investigators repeated the measurements in 10 randomly selected normal pregnancies a week after they had performed the first set of measurements. For interobserver variability, we randomly selected a further 10 cases, five of which had been measured by one observer, and five by the other, and the measurements were repeated by the observer who had not performed the previous set of measurements.
All continuous variables are expressed as mean ± SD. Associations between arterial parameters were analyzed with Pearson's correlation coefficient (r). As body mass index (BMI) or blood pressure could have affected measurements, the results for IMT and arterial stiffness were adjusted for these two covariates, using a general linear model. AIx was also corrected for heart rate. Group data were compared with paired or unpaired Student's t-test. Bland–Altman analysis was performed to compare agreement between two measurements. P < 0.05 was considered a statistically significant difference.
A power calculation performed before the study, using IMT and PWV, indicated that 20 participants were required for each group. For the purpose of this calculation, values of 450 ± 96 μm and 7.5 ± 2 m/s were used for IMT and PWV, respectively, for the pre-eclamptic group and values of 350 ± 90 μm and 6 ± 1 m/s were used, respectively, for the normotensive group. The difference that we originally planned to detect was therefore 100 µm (450–350 μm) for IMT and 1.5 m/s for PWV (7.5–6 m/s). The significance level was defined as alpha = 0.05. The sample size determined gave a power of approximately 0.91 for IMT and 0.83 for PWV.
The data were tested for normality using the non-parametric Kolmogorov–Smirnov method. Statistical software package SPSS 12.0 (SPSS Inc., Chicago, IL, USA) was used for all data analyses.
Table 1 presents the demographic and hemodynamic characteristics of the pre-eclamptic and NP women. Eighteen (82%) of the pre-eclamptic women delivered by elective Cesarean section and four (18%) by emergency Cesarean section. In the NP group, three (11%) women were delivered by emergency Cesarean section. The gestational age at delivery was 37 ± 1 weeks in the pre-eclamptic group and 40 ± 1 weeks in the NP group (P = 0.03), and the median birth weight was 2600 (interquartile range (IQR), 2495–2762) g and 3300 (IQR, 3200–3387) g, respectively (P < 0.0001).
Table 1. Demographic and hemodynamic data in 22 late-onset pre-eclamptic and 28 normotensive (NP) pregnant women
Data are given as mean ± SD, mean ± SD (range) or n/n. BMI, body mass index; TPR, total peripheral resistance.
29 ± 6
27 ± 3
Gestational age at examination (weeks)
36.0 ± 3.2 (28–40)
35.8 ± 2.4 (31–40)
1.59 ± 0.03
1.61 ± 0.04
70.4 ± 12.0
66.1 ± 6.6
BMI at examination (kg/m2)
27.8 ± 4.0
25.5 ± 2.2
Brachial pressure (mmHg)
150 ± 14
117 ± 10
102 ± 9
76 ± 9
118 ± 9
90 ± 8
Heart rate (bpm)
89 ± 15
92 ± 15
1667 ± 452
1347 ± 212
TPR index (dynes/s/cm5/m2)
921 ± 298
752 ± 115
Proteinuria (+ ∼ ++/ +++)
Table 2 compares geometric (internal diameter and IMT) and mechanical (stiffness) carotid arterial parameters between pre-eclamptic and NP women. Carotid arterial IMT, internal diameter and the carotid arterial stiffness parameters of PWV, AIx and PT1, as well as carotid arterial wall tension, were all significantly greater, while DC was significantly smaller, in pre-eclampsia compared with NP, even after adjusting for BMI. Most parameters, apart from PWV and DC, remained significantly different after adjusting for SBP and DBP. Significant difference between the two groups still existed in AIx after correction for maternal heart rate (P < 0.0001) and in carotid arterial IMT after adjustment for the carotid arterial internal diameter (P = 0.001). Carotid arterial IMT and arterial stiffness parameters correlated significantly with SBP, MAP and DBP (Figures 3 and 4).
Table 2. Carotid arterial parameters in 22 late-onset pre-eclamptic and 28 normotensive (NP) pregnancies
Carotid arterial parameter
P after adjusting for:
Data are given as mean ± SD. AIx, augmentation index; BMI, body mass index; DC, distensibility coefficient; DBP, diastolic blood pressure; IMT, intima–media thickness; MAP, mean arterial pressure. PT1, pressure at inflection point; PWV, pulse wave velocity; SBP, systolic blood pressure.
459 ± 95
351 ± 85
Internal diameter (mm)
7.8 ± 0.5
7.2 ± 0.4
141 ± 13
108 ± 12
99 ± 7
75 ± 9
113 ± 8
86 ± 9
0.020 ± 0.009
0.029 ± 0.011
134 ± 11
107 ± 11
7.1 ± 1.7
6.0 ± 1.5
7.9 ± 9.2
−5.0 ± 5.6
55.0 ± 6.5
38.6 ± 4.9
As shown in Table 3, 18 months after parturition, the carotid arterial internal diameter was still significantly larger in women who had been pre-eclamptic compared with those who had had a normal pregnancy (P = 0.0301). Carotid arterial SBP and MAP also remained higher in the pre-eclamptic group compared with the NP group (P = 0.0114 and P = 0.0413, respectively). The carotid arterial tension was significantly decreased 18 months after compared with before parturition in the pre-eclamptic group (P = 0.0153) but was still higher (P = 0.0037) than was the tension in the NP group after parturition. However, there were no significant differences in any arterial stiffness parameters.
Table 3. Comparison of carotid arterial parameters in 10 women at the time of late-onset pre-eclamptic pregnancy and 18 months postpartum, and in 11 women 18 months postpartum whose pregnancy had been normotensive (NP)
Carotid arterial parameter
NP: 18 months postpartum
18 months postpartum
P < 0.05 vs before delivery.
P < 0.05 vs NP postpartum.
P < 0.01 vs NP postpartum. AIx, augmentation index; BMI, body mass index; DC, distensibility coefficient; DBP, diastolic blood pressure; IMT, intima–media thickness; MAP, mean arterial pressure. PT1, pressure at inflection point; PWV, pulse wave velocity; SBP, systolic blood pressure.
We observed good agreement between measurements taken by the same observer and by two independent observers for PWV and IMT values. The mean ( ± SD) difference was −0.008 ( ± 0.032) m/s for repeated measurements of PWV taken by the same observer and 0.004 ( ± 0.119) m/s for those taken by two independent observers. The mean ( ± SD) difference was 0.90 ( ± 8.10) µm for repeated measurements of IMT taken by the same observer and 1.20 ( ± 9.52) µm for those taken by two independent observers (Figure 5).
In this study, using ultrasound QIMT and QAS techniques, we assessed elastic arterial remodeling in women with pre-eclampsia. We found that, in these women, carotid arterial IMT and arterial stiffness were significantly increased and closely correlated with carotid arterial blood pressures. Our findings suggest that, in addition to the vasoconstriction of smaller peripheral arteries, abnormal arterial remodeling and mechanics occur in large, elastic arteries in pre-eclampsia.
Measurement of carotid arterial IMT in a clinical setting can identify individuals with advanced subclinical atherosclerosis and can quantify its severity non-invasively. Lorenz et al. performed a systematic review and meta-analysis of data and verified that carotid arterial IMT is a strong predictor of future vascular events. Our present study showed that carotid arterial IMT was significantly increased in women with pre-eclampsia even after adjusting for the carotid arterial internal diameter, raising the possibility that subclinical atherosclerosis might have occurred in these women. Though acute changes in IMT in response to acute blood pressure and vascular tone modifications have been observed in healthy subjects, it has been shown that such alterations in pre-eclampsia could linger beyond parturition and even leave a persistent defect in the systemic and pulmonary circulation of the offspring[17-20].
Furthermore, our study showed a much larger difference in IMT between the pre-eclamptic and NP pregnancies (0.1 mm) when compared with the reported 0.0147-mm annual rate of progression of mean common carotid arterial IMT in a population of healthy men. Cosmi et al. showed that aortic IMT in fetuses and children with intrauterine growth restriction was increased[22, 23]. These studies and our results suggest that carotid arterial IMT should be assessed for whether it might be a predictor of future vascular events in certain complicated pregnancies and affected children.
Besides IMT, arterial stiffness and wave reflection are important in the development of cardiovascular diseases. The QAS technique provides several standard parameters for assessing arterial stiffness, including local carotid arterial pressure, DC and AIx. The assessment of local carotid arterial pressure profiles by the Esaote vessel-wall tracking technique uses arterial distension waveforms, which closely approximate pressure waveforms in the carotid artery[12, 24-26]. DC is the fractional change in cross-sectional area relative to the change in arterial pressure. We found that DC was lower in pre-eclamptic pregnancy compared with NP, indicating that carotid arterial stiffness was increased in these women. AIx is a parameter based on analysis of the pressure waveform, expressed as the ratio of augmented pressure (attributed to wave reflection) to pulse pressure. Kaihura et al. found no significant difference in AIx between pre-eclamptic and normal pregnant women by analyzing the carotid–radial and carotid–femoral parts of the arterial tree using applanation tonometry; in contrast, a recent systematic review and meta-analysis showed that AIx was significantly increased in pre-eclamptic women evaluated using the same method, in accordance with our findings by a different method: the QAS technique.
PWV is the most useful and robust index of arterial stiffness, and applanation tonometry the most commonly used method for its measurement, by either carotid–radial or carotid–femoral arterial pathways. Different from these traditional PWV measurements, we obtained ‘local’ pulse wave tracking and PWV measurements, using a radiofrequency-based ultrasound vessel-wall tracking technique to assess local elastic arterial PWV. To our knowledge, this is the first report in the literature of this technique.
Though we cannot exclude the possibility that the observed arterial remodeling was a result of hypertension alone and not pre-eclampsia, the study by Tihtonen et al. found that changes in arterial stiffness were smaller in chronic hypertensive than in pre-eclamptic pregnancies. Other studies have shown that carotid–femoral PWV and AIx on applanation tonometry are higher in women with pre-eclampsia compared with women with gestational hypertension[32, 33]. However, in order to determine whether these are independent findings representing some special pathology of pre-eclampsia or findings which appear inevitably, as blood pressure rises, further studies are required, comparing women with pre-eclampsia and those with gestational hypertension, after matching for blood pressure level, and perhaps repeating examinations postpartum. The studies by Robb et al. and Hausvater et al. demonstrated that, despite blood pressure returning to within the normal range, AIx and carotid–femoral PWV remained elevated 7 weeks or even several months postpartum in women who had been pre-eclamptic compared with those with uncomplicated pregnancies. Interestingly, in our current study, there were no significant differences in carotid arterial stiffness parameters between pre-eclamptic and NP groups 18 months after parturition. However, the carotid arterial systolic pressure and tension remained high 18 months postpartum in the women who had had pre-eclampsia, suggesting the possibility that the increased carotid arterial wall tension might also be associated with the increased cardiovascular diseases in these pre-eclamptic women in later life.
The main limitation of our study is the small number of patients. This was because only women with late-onset pre-eclampsia and with no antihypertensive treatment prior to admission were assessed, and women with gestational hypertension and chronic hypertension were excluded. A second limitation is that we did not measure preconception values; therefore, we could not define a true baseline for the parameters measured during pregnancy. A third limitation is that only limited follow-up data were available to determine whether the changes in carotid arterial IMT and arterial stiffness in pre-eclampsia are temporary. Further study with larger populations needs to be carried out to verify our results.
In conclusion, women with pre-eclampsia apparently have significantly increased carotid arterial IMT, arterial stiffness and arterial wall tension. QIMT and QAS show potential as techniques to assess local elastic arterial remodeling non-invasively and comprehensively in pre-eclamptic women.
This work was supported by a Shannxi Province Scientific Research Grant, China (2011K15-02-05) and a grant from the National Natural Science Foundation of China (NSFC 81170149).