Materials and Methods
Prospectively over 18 months, we enrolled 576 nulliparas who attended their first prenatal visits at our hospital or one of two affiliated health centers. The study was approved by the Human Subjects Committee of our hospital. Subjects were followed through their pregnancies until 6–8 weeks after delivery. At each monthly visit, prenatal flow sheets were completed with the following information: gestational age, weight, systolic and diastolic blood pressures (BPs), and proteinuria results. We excluded women with histories of chronic hypertension, recent ingestion of antihypertensive medication, systolic BP of at least 140 mmHg or diastolic BP of at least 90 mmHg before 20 weeks' gestation, multiple gestations in current pregnancies, and histories of diabetes or current gestational diabetes.
At each prenatal and single postpartum visit, women had their BPs measured by a nurse or nurse assistant who had training based on published guidelines.9 Measurements were taken after 2–3 minutes of resting, and rightarm BPs were measured using a mercury manometer with appropriate-size cuffs while subjects sat. Typically only a single measurement was taken. Readings were repeated if BP was at least 140 mmHg systolic or 90 mmHg diastolic. Each reading was recorded, and discrepant (at least 5 mmHg) readings were repeated until two readings were at most 5 mmHg apart. In that situation, the last two readings were averaged and entered for analyses.10 The first and fifth Korotkoff phases were used to define systolic and diastolic BPs, respectively. Pulse pressure for each visit was calculated by subtracting systolic from diastolic BP (in mmHg). Mean arterial pressure (MAP; in mmHg) was calculated as systolic BP + (2 × diastolic BP)/3.
Demographic information including age, height, and race was assessed at first prenatal visits. Weights at first prenatal visits were used to calculcate body mass index (BMI), and smoking status at first prenatal visits was used in analysis. Medication use was ascertained at first prenatal visits and updated at subsequent visits.
Main outcomes were gestational hypertension and preeclampsia. Prenatal flow sheets were examined, and each woman had BP measurements recorded during their first trimesters. The latter requirement was necessary to exclude women with chronic hypertension because BP reaches a nadir at the end of the second trimester.8 Gestational hypertension was defined as BP of at least 140/90 after 20 weeks of pregnancy, and preeclampsia was defined as BP of at least 140/90 after 20 weeks of pregnancy with proteinuria (at least 2+ by dipstick or at least 300 mg/24 hours).11 Indications for 24-hour urine collections were systolic BP of at least 140 mmHg or diastolic BP of at least 90 mmHg and routine dipstick (at each visit) of at least 1+ protein on two separate occasions or 2+ protein on one occasion without urinary tract infection. Measurements were repeated if dipstick measurements suggested increased proteinuria in subsequent visits or worsening hypertension. Among women with any hypertensive disorder, approximately 93% had 24-hour urine collections. Blood pressure measurements were not used if taken during active or induced labor and after initiation of antihypertensive medications (among cases).
Blood pressure measurements during prenatal visits were recorded and linked to the estimated gestational age. Gestational age was determined by last menstrual period confirmed by first- or second-trimester ultrasound unless ultrasound estimation varied by more than 7 (first trimester) or 10 (second trimester) days, in which case sonographic estimation alone was used to set gestational age. Measurements from each subject taken during weeks 7–15, 16–24, and 25–38 were pooled to determine the arithmetic mean for that subject during that period. Pulse pressure was examined as a continuous variable and divided into three equal categories, tertiles (up to 39 mmHg, 40–45 mmHg, and at least 46 mmHg), according to the distribution of the entire population.12 That method, used in cardiovascular studies to examine pulse pressures in nonpregnant women,2,13 avoids a direct assumption that there is a linear relation between exposure and outcome and thus uncovers a potential nonlinear association, should it exist.
Three-way comparisons were done with analysis of variance. Thereafter, pairwise comparisons were done using t tests and Wilcoxon ranksum tests. χ2 was used to assess categoric variables. The Mantel extension test was used to evaluate linear trends across categories of pulse pressure. In that test, the smaller the P, the more likely the null hypothesis (that the association is horizontal) is false. Multivariable analyses were done with logistic regression techniques, with pulse pressure modeled as indicator variables (tertiles) and as a continuous variable. Dependent variables in those analyses were gestational hypertension and preeclampsia, and potential confounders included age, BMI, race, and smoking status. Relative risks (RRs) and 95% confidence intervals (CIs) were calculated, and all statistical analyses were done with SAS (SAS Institute, Cary, NC). All P values were two-tailed.
Baseline characteristics taken at 10 ± 1 weeks' gestation and delivery characteristics of all women are presented in Table 1. Mean age and BMI were higher in women who developed hypertensive disorders of pregnancy compared with normotensive women. Delivery characteristics also differed between groups.
Systolic and diastolic BPs and MAPs differed throughout gestation among groups (P < .01, analysis of variance). Table 2 summarizes mean BP characteristics according to each period. Although pairwise comparisons of those same characteristics found differences between cases and normotensive women, there were no significant differences between women who developed gestational hypertension and those who developed preeclampsia. Pulse pressure at 7–15 weeks, in contrast, did differ between the two groups of cases. During subsequent periods (16–24 and 25–38 weeks), pulse pressure remained elevated in women who developed any hypertensive disorder of pregnancy, but differences between the two groups of cases diminished and did not achieve statistical significance. At 6-week postpartum visits (6 ± 2 weeks), systolic and diastolic BPs and MAPs had increased among women with any hypertensive disorder of pregnancy, but pulse pressures were similar between groups.
Association between pulse pressure and risk of hypertensive disorders of pregnancy then was examined in more detail. Examined as tertiles and after adjusting for age, BMI, race, and smoking status, increasing pulse pressure was associated with progressively increasing RR of preeclampsia: pulse pressure up to 39 mmHg, 1.0 (reference); 40–45 mmHg, 1.3 (95% CI: 0.6, 3.0); at least 46 mmHg, 1.6 (95% CI: 0.8, 3.2); P for trend = .01. In contrast, assessment of pulse pressure in tertiles did not show positive association with gestational hypertension: pulse pressure up to 39 mmHg, 1.0 (reference); 40–45 mmHg, 1.1 (95% CI: 0.5, 2.7); at least 46 mmHg, 1.0 (95% CI: 0.4, 2.2); P for trend = .95. When examined as a continuous variable in multivariable analysis (Table 3), a 1-mmHg increase in pulse pressure was associated with a 6% increased risk of preeclampsia but not with increased risk of developing gestational hypertension (RR: 1%, 95% CI: −1, 6). Although uneven sampling per period might have altered the estimates for each period, controlling for numbers of visits in the multivariable model did not substantially change results (data not shown).
We found that pulse pressure early in pregnancy was elevated in women who developed preeclampsia but not in those who developed gestational hypertension or remained normotensive. This association was independent of age, BMI, and other potential confounding factors. Differences in pulse pressure were evident as early as 7–15 weeks' gestation. By 6–8 weeks post-partum, although systolic and diastolic BPs and MAPs continued to differ between women who developed any hypertensive disorder of pregnancy and normotensive women, pulse pressure was similar across groups.
Vessel compliance has been examined in women with preeclampsia. Using Laplace transform analysis of Doppler waveforms, recent cross sectional studies found evidence of increased vessel stiffness in women with hypertensive disorders of pregnancy compared with controls.7,14
Although women with preeclampsia or gestational hypertension had evidence of increased vessel wall stiffness compared with controls, at least one study7 suggested that finding was more marked in women with preeclampsia. At a given stroke volume and velocity of ventricular ejection, mechanisms that influence pulse pressure are related to the status of the conduit arteries, ie, the viscoelastic properties of the arterial wall and the timing of the reflected waves.3 Increased stiffness and earlier wave reflections increase pulse pressure. We had found that pulse pressure was elevated early in pregnancy when arterial compliance was expected to markedly increase.15 Our results are consistent with physiologic studies that found alterations of vascular reactivity in women who developed preeclampsia were evident well before clinical manifestations were present.5,6 Although pulse pressure is a crude assessment of arterial compliance, it predicted adverse cardiovascular and cerebrovascular events in large epidemiologic studies.1,2 In those studies, pulse pressure was associated with significant adverse outcomes in normotensive2 and hypertensive1 women and with adverse outcomes in a linear fashion,2 similar to our findings. We excluded women with histories of chronic hypertension because they were at increased risk of preeclampsia,11 and antihypertensive medications might have altered pulse pressure.16 Selecting agents that preferentially improve vessel compliance ultimately might affect prevention and management of preeclampsia, and pulse pressure might be used to identify women at increased risk and as a marker of effective therapy.
The cause of elevated pulse pressure or vessel wall stiffness in preeclampsia is unknown, but several features of the disorder suggest potential mechanisms. Increased pulse pressure is associated with arterial wall thickening and might promote development of atherosclerosis by hastening development of intimal injury.17 Preeclampsia also is notable for placental vessels that are thickened and atherosclerotic.6 Dyslipidemia is associated with impaired arterial distensibility,18 and hyperlipidemia is characteristic of preeclampsia.19 Recent evidence suggested that preeclampsia is a state of insulin resistance,20 and insulin resistance has been associated with poor arterial compliance.21 Therefore, elevated pulse pressure is consistent with the other pathophysiologic changes described in preeclampsia and might be an important component in the pathogenesis of this disorder.
Previous studies that examined use of BP parameters such as systolic and diastolic BPs and MAPs as predictors of hypertensive disorders of pregnancy suggested that they were altered early in pregnancy in women who developed any hypertensive disorder of pregnancy.10,22 There has been much debate about the predictive value of one measurement compared with another, and whether a characteristic can predict gestational hypertension, preeclampsia, or both.23–25 Among women who later developed preeclampsia or gestational hypertension in our study, systolic and diastolic BPs and MAPs were elevated as early as 7–15 weeks' gestation, and all remained elevated at 6–8 weeks postpartum. As also reported,26 we found that a MAP of approximately 90 mmHg in the middle trimester was associated with development of hypertensive disorders in the third trimester. However, we did not find significant differences in BP characteristics between women who developed preeclampsia and those who developed gestational hypertension. Although that might have been caused by limited sample size, it also supports lack of consensus about the ability of those parameters to distinguish women destined to develop preeclampsia from gestational hypertension. In contrast, pulse pressure early in pregnancy was significantly different between groups, supporting the notion that the pathogenesis of the two disorders differs.27
Limitations of this study deserve mention. We did not observe a substantial drop in BP characteristics in the second trimester as described.8 We restricted our cohort to nulliparas, who tend to have less a drop in systolic and diastolic BPs during the second trimester compared with multiparas.8,26 The periods were chosen based on achieving adequate samples of cases within each period. Second-trimester BP tends to reach a nadir at 20–22 weeks,8 so pooling measurements during that period with measurements taken slightly earlier and later minimized the nadir we observed. Like others,8,10,23 we used the technique of pooling BP measurements during pregnancy primarily because women do not attend prenatal clinics at the same weeks of gestation, and prenatal visits early in uncomplicated pregnancies are usually not more frequent than once per month. Adjusting for the number of visits per period did not materially change the results. The average age of primigravid women in this cohort was slightly older than in other cohorts.10,23 We adjusted for age in multivariable models, and our results were consistent with findings from a large study of a similar group of women.26 We also did not report sensitivity, specificity, and predictive values of pulse pressure measurements with respect to risk of preeclampsia. Our purpose was to test the hypotheses that a simple surrogate measure of arterial compliance could identify women at risk of preeclampsia. There were 34 cases of preeclampsia, and we believe a larger study is warranted to accurately assess pulse pressure as a screening test. Such analysis will require dividing pulse pressure at a specific cutpoint, which also will require a larger study.
1. Donamski MJ, Davis BR, Pfeffer MA, Kastantin M, Mitchell GF. Isolated systolic hypertension: Prognostic information provided by pulse pressure. Hypertension 1999;34:375–80.
2. Benetos A, Rudnichi A, Safar M, Guize L. Pulse pressure and cardiovascualar mortality in normotensive and hypertensive subjects. Hypertension 1998;32:560–4.
3. London GM, Guerin AP. Influence of arterial pulse pressure and reflected waves on blood pressure and cardiac function. Am Heart J 1999;138:S220–4.
4. O'Rourke MF, Kelly RP. Wave reflection in the systemic circulation and its implications in ventricular function. J Hypertens 1993;11:327–37.
5. Gant NF, Daley GL, Chand S, Whalley PJ. A study of angiotensin II pressor response throughout primigravid pregnancy. J Clin Invest 1973;52:2682–9.
6. Roberts JM. Endothelial dysfunction in preeclampsia. Semin Reprod Endocrinol 1998;16:5–15.
7. Brackley KJ, Ramsay MM, Pipkin FB, Rubin PC. The maternal cerebral circulation in preeclampsia: Investigations using Laplace transform analysis of Doppler waveforms. Br J Obstet Gynaecol 2000;107:492–500.
8. Christianson RA. Studies on blood pressure during pregnancy. Am J Obstet Gynecol 1976;125:509–13.
9. Sibai B. Hypertension in pregnancy. In: Gabbe SG, Niebyl JR, Simpson JL, eds. Obstetrics: Normal and problem pregnancies. 3rd ed. New York: Churchill Livingston, 1996:941.
10. Moutquin JM, Rainville C, Giroux L, et al. A prospective study of blood pressure in pregnancy: Prediction of preeclampsia. Am J Obstet Gynecol 1985;151:191–6.
11. National High Blood Pressure Education Program Working Group. Report of the National High Blood Pressure Education Program Working Group on high blood pressure in pregnancy. Am J Obstet Gynecol 2000;183:S1–22.
12. Rothman KJ, Greenland S. Modern epidemiology. Boston: Little Brown and Company, 1998.
13. Verdecchia P, Schillaci G, Borgioni C, Ciucci A, Pede S, Porcellati C. Ambulatory pulse pressure: A potent predictor of total cardiovascular risk in hypertension. Hypertension 1998;32:983–8.
14. Stone PR, Wiesender CC, Murakami N. Maternal brachial arterial waveforms in gestational hypertension: Analysis using a Laplace transform technique. J Maternal Fetal Med 1998;3:113–20.
15. Poppas A, Shroff SG, Korcarz CE, Hibbard JU, Berger DS, Lindheimer MD, et al. Serial assessment of the cardiovascular system in normal pregnancy. Circulation 1997;95:2407–15.
16. Boutouyrie P, Bussy C, Hayoz D, Hengstler J, Dartois N, Laloux B, et al. Local pulse pressure and regression of arterial wall hypertrophy during long-term antihypertensive treatment. Circulation 2000;101:2601–6.
17. Lyon RT, Runyon-Hass A, Davis HR, Glagov S, Zarins CK. Protection from atherosclerotic lesion formation by reduction of artery wall motion. J Vasc Surg 1987;5:59–67.
18. Leeson CP, Whincup PH, Cook DG, Mullen MJ, Donald AE, Seymour CA, et al. Cholesterol and arterial distensibility in the first decade of life: A population-based study. Circulation 2000;101:1533–8.
19. Thadhani R, Stampfer MJ, Hunter DJ, Manson JE, Solomon CG, Curhan GC. Body mass index and hypercholesterolemia and risk of hypertensive disorders of pregnancy. Obstet Gynecol 1999;94:543–50.
20. Kaaja R. Insulin resistance syndrome in preeclampsia. Semin Reprod Endocrinol 1998;16:41–6.
21. Westerbacka J, Uosukainen A, Makimattila S, Schlenzka A, Yki-Jarvinen H. Insulin-induced decrease in large artery stiffness is impaired in uncomplicated type 1 diabetes mellitus. Hypertension 2000;35:1043–8.
22. Sibai BM, Ewell M, Levine RJ, Klebenoff MA, Esterlitz J, Catalano PM, et al. Risk factors associated with preeclampsia in healthy nulliparous women. The calcium for preeclampsia prevention (CPEP) study group. Am J Obstet Gynecol 1997;177:1003–10.
23. Villar MA, Sibai BM. Clinical significance of elevated mean arterial blood pressure in second trimester and threshold increase in systolic or diastolic blood pressure during third trimester. Am J Obstet Gynecol 1989;160:419–23.
24. Conde-Agudelo A. What does an elevated mean arterial pressure in the second half of pregnancy predict–Gestational hypertension or preeclampsia? Am J Obstet Gynecol 1993;169:509–14.
25. Chelsey LC, Sibai BM. Clinical signficance of elevated mean arterial pressure in the second trimester. Am J Obstet Gynecol 1988;159:275–9.
26. Page EW, Christianson R. The impact of mean arterial blood pressure in the middle trimester upon the outcome of pregnancy. Am J Obstet Gynecol 1976;125:740–6.
27. Roberts JM, Redman CW. Preeclampsia: More than pregnancy-induced hypertension. Lancet 1993;341:1447–51.