Evaluation of the Blood Pressure Load in the Diagnosis of Hypertension in Pregnancy
Abstract—— The use of a set of new end points obtained from ambulatory blood pressure monitoring, in addition to the blood pressure values themselves, has been advocated to improve sensitivity and specificity in the diagnosis of hypertension and the evaluation of a patient’s response to treatment. Among these parameters is the use of blood pressure load, the percentage of values above a given constant reference limit or computed by reference to daytime and nighttime limits. We examined the effectiveness of this parameter as a potential screening test for the detection of hypertension in pregnancy. We analyzed 2014 blood pressure series systematically sampled by ambulatory monitoring for 48 consecutive hours every 4 weeks from the first obstetric visit (usually within the first trimester of pregnancy) until delivery of 205 normotensive pregnant women and 123 women who developed gestational hypertension or preeclampsia. The blood pressure load was obtained as the percentage of values >140/110/90 mm Hg (systolic/mean arterial/diastolic blood pressure) during active hours or 120/95/80 mm Hg during resting hours, as well as by comparison with limits obtained by progressively reducing the previous limits by 5 mm Hg, up to a final threshold of 125/95/75 mm Hg (day) and 105/80/65 mm Hg (night). Sensitivity for the blood pressure load computed by reference to the highest limits used here is <55% in all trimesters of pregnancy. The best results were obtained when 130/100/80 mm Hg (day) and 110/85/70 mm Hg (night) were used as references in the third trimester, and when the lowest tested limits of 125/95/75 and 105/80/65 mm Hg were used as references in the first and second trimesters (sensitivity always >73%). The optimum reference limits for calculating the blood pressure load, markedly < mm Hg, must be defined as a function of gestational age, in keeping with the predictable trends in blood pressure along pregnancy previously documented.
Recent studies have tried to overcome the poor results from isolated blood pressure (BP) measurements in detecting hypertensive complications in pregnancy1,2 by relying on ambulatory BP monitoring (ABPM).3–8 Using this approach, a predictable trend of BP variation along pregnancy was demonstrated for normotensive pregnant women systematically monitored every 4 weeks from the first trimester of pregnancy until delivery. The trend of decreasing BP up to the middle of gestation followed by increasing BP up to the day of delivery could not be found in pregnancies complicated with gestational hypertension or preeclampsia.9 Moreover, differences between healthy and complicated pregnancies in the circadian pattern of BP, previously documented for the second trimester of pregnancy,3 can be observed by ABPM as early as in the first trimester of pregnancy, before the actual clinical diagnosis of gestational hypertension or preeclampsia took place for the women investigated.8 The use of the 24-hour mean of BP does not provide, however, a proper approach for an individualized early diagnosis of hypertensive complications in pregnancy.3–5 Poor results from the diagnostic test based on mean BP values have led many investigators to extrapolate erroneously that ABPM is not a valid approach in pregnancy.5
The use of a set of new end points, in addition to the BP values themselves or average values derived from them, has been advocated to improve sensitivity and specificity in diagnosing hypertension and the evaluation of a given patient’s response to treatment.6,10,11 The circadian pattern with large amplitude that characterizes BP in healthy pregnancies at all gestational ages8 suggests that the constant threshold currently used for diagnosing hypertension in pregnancy should be replaced by a time-specified reference limit that reflects the mostly predictable BP variability. Once the time-varying threshold, given for instance by the upper limit of a tolerance interval,12 is available, the hyperbaric index (HBI), as a determinant of BP excess,6,11,13 can be calculated as the total area of any given patient’s BP above the threshold. This tolerance-hyperbaric test has already been shown prospectively to provide high sensitivity and specificity for the very early identification of subsequent hypertensive complications in pregnancy.6
Apart from the HBI, other parameters have been defined previously for the quantification of BP excess. In 1988, the Mayo Clinic suggested the use of a BP load,10 defined as the percentage of BP values exceeding a given constant threshold. More recently, within the context of ABPM, it was suggested the use of the amount of excess determined with respect to daytime and nighttime averages (usually mm Hg for systolic/diastolic BP (SBP/DBP) during activity and mm Hg during resting hours).14,15 The HBI represents a better determinant of excess than the BP load, which is more frequently used in studies based on ABPM. The advantages of the HBI compared with the BP load or with parameters derived from the use of fixed reference limits have been previously documented.6,8,11,16 The calculation of the BP load, however, is much simpler than obtaining the HBI.
Accordingly, we examined the effectiveness of the BP load as a potential screening test for the detection of hypertension in pregnancy in women who were systematically studied by 48-hour ABPM from the first obstetric visit to the hospital until delivery. Taking into account the predictable trend of decreasing BP in the second trimester compared with the first or third trimesters in normotensive pregnant women,9 the BP load was calculated for each BP series by comparison with the standard limits of mm Hg for SBP/DBP and by comparison with limits obtained by progressively reducing the previous constant threshold by 5 mm Hg at each step. Moreover, in keeping with the circadian variation of BP in both healthy and complicated pregnancies,8 sensibility and specificity for the BP load were calculated using different reference constant thresholds for day and night.
We studied 328 (176 primipara) untreated white pregnant women (205 who had uncomplicated pregnancies, 92 who developed gestational hypertension, and 31 who developed preeclampsia) who fulfilled all required criteria for this trial. Gestational hypertension was defined as conventional BP values >140 or 90 mm Hg for SBP or DBP, respectively, after the 20th week of gestation, without a clinical record of hypertension previous to pregnancy and with an HBI consistently above the threshold for diagnosis of hypertension in pregnancy6 after the 20th week of gestation for further corroboration. Preeclampsia was defined as gestational hypertension and proteinuria and >300 mg/24 h of urine collection, with or without edema, diagnosed after the 20th week of gestation in a previously normotensive woman. Diagnosis of gestational hypertension and preeclampsia was performed with information from conventional obstetric examinations, including monthly ABPM, and routine analyses of urine.
All women received obstetric care at the Obstetric Physiopathology Unit, Hospital Clínico Universitario, Santiago de Compostela, Spain. Reasons for receiving medical care at this unit included familial or personal history of either gestational hypertension or preeclampsia; chronic hypertension; cardiovascular, endocrine, or metabolic disease; bleeding; a personal history of spontaneous abortion; multiple pregnancy; obesity; and adolescent or middle-aged nulliparous pregnancy (<18 or >35 years). The relative risk of gestational hypertension and preeclampsia in this unit is ≈3.5 times that of the general obstetric population in our setting.8 All women participating in this trial were, thus, highly motivated and of relatively high risk for developing complications in pregnancy; the actual percentage of women with complicated pregnancies in this study does not necessarily reflect, therefore, the actual incidence of hypertension in pregnancy in our setting.
All issues related to ABPM, including handling and preparation of the monitors, individualized explanation about their use to each patient, and processing of the data provided by any given pregnant woman after monitoring, were always performed by the same member of the research group in 1 room of the unit. Conventional obstetric examinations of the pregnant women, usually done on the same day just before starting ABPM, were performed by other members of the research group in different rooms of the unit.
Inclusion criteria were absence of any condition requiring the use of antihypertensive medication, maternal age (18 to 40 years), and gestational age (<16 weeks at the time of inclusion). Exclusion criteria were, among others, multiple pregnancy, chronic hypertension, chronic liver disease, any disease requiring the use of antiinflammatory medication, diabetes or any other endocrine disease such as hyperthyroidism, and intolerance to ABPM. Apart from the 328 women providing all required information, 15 subjects who provided <4 profiles of ABPM (3 spontaneous abortions and 12 who withdrew from the trial) were eliminated from the study. The State Ethics Committee of Clinical Research approved the study. All women signed consent forms before entering the study.
In this trial, the SBP, mean arterial BP (MAP), and DBP of each woman were scheduled to be measured by ABPM every 20 minutes during the day (7:00 am to 11:00 pm) and every 30 minutes during the night for 48 consecutive hours with an SpaceLabs 90207 device at the time of recruitment (usually within the first trimester of pregnancy) and then every 4 weeks until delivery. BP series were eliminated from analysis when the subjects showed an irregular rest-activity schedule during the 2 days of sampling, an odd sampling with spans of >3 hours without BP measurement, or a night resting span <6 hours or >12 hours. The total number of BP series provided by the 328 women under investigation fulfilling all mentioned requirements set a priori was 2014. During sampling, all women lived their usual diurnal waking (∼9:00 am to approximately midnight) and nocturnal resting routine, following everyday life conditions with minimal restrictions. They were told to follow a similar schedule during the days of sampling and to avoid the use of medication for the duration of the trial.
The clinical evaluation of this oscillometric monitor for use in pregnancy according to the standards published by the Association for Advancement of Medical Instrumentation and the British Hypertension Society has been previously established.17 The BP cuff was worn on the nondominant arm. ABPM was performed in addition to the woman’s routine antenatal care, and no person was hospitalized during monitoring. Cuff size was determined by upper arm circumference at the time of each visit. ABPM always started between 10:00 am and 1:00 pm During monitoring, each subject maintained a diary, listing the times they went to bed at night, woke in the morning, and ate meals; exercise and unusual physical activity; and events and mood/emotional states that might affect BP.
Each individual’s clock-hour BP values were re-referenced from clock time to hours before and after awakening from nocturnal sleep. This transformation avoided the introduction of bias due to differences among subjects in their sleep/activity routine.18 BP values were then edited according to commonly used criteria for the removal of outliers and measurement errors.19
The BP load was obtained for each profile of ABPM as the percentage of values >140/110/90 mm Hg (in SBP/MAP/DBP) during diurnally active hours, or >120/95/80 mm Hg during nocturnally resting hours, as well as by comparison with limits obtained by progressively reducing the previous limits by 5 mm Hg at each step, up to a final threshold of 125/95/75 (daily activity) and 105/80/65 (nightly rest). Because the conventional assessment of hypertension relies of casual BP measurements >140 or 90 mm Hg for SBP or DBP,20,21 results will be expressed as a function of the maximum BP load, defined as the maximum of the 3 values of BP load determined for SBP, MAP, and DBP, respectively, for any given ABPM profile. Additionally, we also calculated the mean BP load, the average of these 3 values of BP load obtained for each cardiovascular variable. Sensitivity and specificity of the maximum and mean BP load calculated with respect to each of the 4 different constant thresholds used in this study, were obtained for each trimester of pregnancy by comparing distributions of values obtained for healthy and complicated pregnancies, without assuming an a priori threshold for diagnosis of gestational hypertension based on the BP load. Sensitivity is defined as the probability that the clinical test is positive given that the person has the disease (ie, the proportion of persons with the disease identified as such by the clinical test). Specificity is defined as the probability that the clinical test is negative given that the person does not have the disease (ie, the proportion of healthy subjects identified as such by the clinical test).
Additionally, we also calculated the positive and negative predictive values (proportion of subjects with a positive test that really have the disease, and proportion of subjects with a negative test who are really healthy, respectively), as well as the relative risk, defined as the proportion of subjects with a positive test who really have the disease divided by the proportion of subjects with a negative test who really have the disease. A good clinical test should have sensitivity, specificity, and positive and negative predictive values close to 100%, whereas the relative risk is markedly larger than 1. If the relative risk is, for instance, 10, then persons with a positive test are 10 times as likely to have the disease than persons with a negative test.22
Figure 1 represents the frequency histograms with the distributions of the maximum BP load obtained for each of the 48-hour BP profiles as the percentage of values exceeding the lowest of the reference limits tested here, ie, 125/95/75 mm Hg for SBP/MAP/DBP during diurnally active hours and 105/80/65 mm Hg during nocturnally resting hours. Distributions of the maximum BP load obtained using higher reference limits are markedly biased to the left (lower values of BP load). As an example, when the highest tested limits of 140/110/90 mm Hg (day) and 120/95/80 mm Hg (night) were used for calculating the maximum BP load for women sampled in the first or second trimesters of their gestation, all values obtained from normotensive women and 72% of the values obtained from women who developed gestational hypertension or preeclampsia later in pregnancy were <10%. The comparison of histograms in Figure 1 between healthy (top) and complicated pregnancies (bottom) does not show a clear separation between both populations at any given trimester. Even in the third trimester of pregnancy, complicated pregnancies can be characterized by a maximum BP load <20% by using as a reference limit a constant value as low as mm Hg for SBP/DBP. The histograms represented in Figure 1 indicate that, despite the large percentage of overlap between the distributions obtained for healthy and complicated pregnancies, the later group shows, from the population point of view, significantly higher values of BP load than those of normotensive pregnant women at all stages of gestation.
Because no reference value for the BP load has yet been provided to be used as an upper limit for diagnosis of gestational hypertension or preeclampsia on the basis of the results of ABPM, sensitivity and specificity was evaluated directly by comparison of the distributions obtained for groups of women with uncomplicated or complicated pregnancies. The values thus calculated for the maximum and mean BP load in each trimester of pregnancy and for the composite of all trimesters, for each of the 4 tested reference limits, are given in the Table. For all the parameters included in the Table, sensitivity corresponds to the largest possible value found for an assumed maximum specificity of 100%. Correspondingly, the values of specificity provided in the Table are calculated assuming a possible sensitivity of 100%. A similar approach was also used to calculate positive and negative predictive values. Results from the Table indicate similar sensitivity and specificity at all stages of gestation for the maximum and the mean BP load. The best results, specially for sensitivity, were obtained when 130/100/80 mm Hg (day) and 110/85/70 mm Hg (night) were used as reference in the third trimester, and when the lowest tested limits of 125/95/75 and 105/80/65 mm Hg were used as references in the first and second trimesters. With regard to these preferred reference limits, sensitivity of the mean BP load ranges from 73% in the first trimester to 78% in the third trimester. Although specificity and positive predictive value are generally lower at all stages of gestation, the negative predictive value is as high as 83% in the third trimester of gestation. The relative risk increases with gestational age. The results for this parameter were again similar for the maximum and mean loads, with a highest value of 4 obtained for the mean BP load in the third trimester of pregnancy.
The characteristics of the test for diagnosing hypertension based on the maximum BP load are described in Figure 2. The graphs on top of Figure 2 provide the receiver operating characteristic (ROC) curves for the maximum load calculated using 4 different tested reference limits for BP in each trimester of pregnancy. The poorest results, corroborating the information provided in the Table, were obtained at all stages of pregnancy for the highest limits of 140/110/90 mm Hg for SBP/MAP/DBP. Results from the ROC curves further indicate the advantages of the lowest tested limits for the diagnosis of gestational hypertension or preeclampsia, with the maximum BP load in the second trimester of pregnancy. Although the differences between limits 3 and 4 indicated in Figure 2 are small for diagnosis in the first and third trimesters, the ROC curves also corroborate results from the Table, indicating that the lowest limits should be preferred to be used in the calculation of the BP load for women sampled in the first trimester of pregnancy; and the third limits (130/100/80 mm Hg), for women sampled in the last trimester.
The graphs on the bottom of Figure 2 represent sensitivity and specificity determined as a function of the threshold value for the maximum BP load to be used for diagnosis, calculated with respect to the lowest reference limits tested here. Results from these graphs indicate that as we decrease the threshold value for the maximum BP load, sensitivity will increase. Specificity, on the contrary, decreases very rapidly because, as indicated in Figure 1, there is a considerable amount of overlap between the distributions of the maximum BP load obtained for normotensive and hypertensive pregnant women. The values shown in the Table represent an average result, indicating how much sensitivity can be improved and how much specificity will be lost by increasing sensitivity. Figure 2 also shows that the slopes for increasing sensitivity and decreasing specificity while lowering the threshold values of BP load are very pronounced, although somehow different. This indicates that a relatively small change in the optimal threshold value of BP load used for diagnosis would result in an important loss in either sensitivity or specificity. In any event, for the lowest tested reference limits, there is a reasonable range of maximum BP load (18% to 28% in the first trimester; 13% to 27% in the second trimester; 25% to 43% in the third trimester) that provides both sensitivity and specificity >80%.
The values of sensitivity and specificity provided in the Table need to be considered carefully. Irrespectively of the reference limits used for calculating the BP load, if one uses a “high” threshold value for diagnosis of gestational hypertension on the basis of maximum BP load (values between 30% and 50% are currently used for diagnosing hypertension15,16), sensitivity will be very low or even 0, as shown in Figure 2. Specificity, on the contrary, will be very high, reflecting the fact that the criterion will identify practically all subjects as normotensive. This is particularly relevant if the BP load is calculated using as reference the generally accepted limits of mm Hg for SBP/DBP,20,21 because for these limits all normotensive women systematically sampled by 48-hour ABPM in this study have a maximum BP load <10%, and 91% of the BP profiles sampled from women who developed gestational hypertension or preeclampsia have a maximum load <30%. The simultaneous change in sensitivity and specificity with respect to the threshold value of maximum load potentially used for diagnosis in graphically displayed in Figure 2. The ROC curves on top of Figure 2 provide further evidence that the use of mm Hg as a reference limit for calculating the BP load results in an unstable and poor diagnostic test.
Results from the Table and the top graphs in Figure 2 further indicate that the optimum reference limits for calculating the BP load, markedly < mm Hg, must be defined as a function of gestational age, in keeping with the predictable trends in BP along pregnancy previously documented.9 For the best limits among those tested in this trial ( mm Hg for SBP/DBP in the first and second trimesters; mm Hg in the third trimester of pregnancy, with the corresponding values for the night resting span shown in the Table), sensitivity and specificity of the maximum BP load are higher than the values obtained for the 24-hour mean of BP for the same women participating in this study, especially for the daily mean of DBP. For this parameter, specificity is as low as 5% for women sampled in the second trimester of pregnancy, a result that corroborates previous findings.4,18 The BP load is also markedly superior to conventional BP measurements. These casual values have sensitivity below 7% and specificity below 3% at all stages of gestation.8,18
This study indicates that ABPM has clear advantages over conventional casual sampling for the diagnosis of hypertensive complications in pregnancy. Both sensitivity and relative risk are markedly higher for the BP load or the 24-hour mean than for casual BP measurements.4 Relying on the maximum BP load provides a test that seems to be more stable than the test based on the 24-hour mean of BP. The maximum BP load is characterized by a lower slope of decreasing sensitivity for increasing values of load used as threshold for diagnosis (Figure 2) compared with the results previously documented for the daily mean.4 The increasing sensitivity of the maximum BP load obtained when the parameter is calculated with regard to references limits markedly < mm Hg corroborates earlier findings concluding that a 24-hour mean BP of mm Hg for SBP/DBP in the first or second trimesters of pregnancy or mm Hg in the third trimester is markedly associated to an increasing risk of developing gestational hypertension or preeclampsia.8 The need to use different reference limits in the calculation of the BP load to increase sensitivity and specificity as a function of gestational age is associated with the previously documented predictable trends of BP throughout pregnancy9; changes in BP along pregnancy have been confirmed by many independent studies, but they are not yet incorporated in the definition of gestational hypertension20,21 and have been used only occasionally for the proper diagnosis of hypertension in pregnancy.6,13,18
In summary, the BP load provides a test for diagnosis of hypertension in pregnancy that could be useful if calculated by reference to limits for BP defined as a function of gestational age and rest-activity cycle that, according to the results in the Table, should be markedly below the currently accepted thresholds of normotension in pregnancy. The test based on the BP load for the diagnosis of gestational hypertension or preeclampsia, however, is limited by the use of constant reference thresholds and by the potential influence of outliers and measurement errors implicit in the definition of load as a percentage of values above the threshold. The tolerance-hyperbaric test, although computationally more demanding, avoids all those limitations.6,11,16,18 For the same women participating in this trial, sensitivity of the HBI was 91% in the first trimester and increased up to 99% in the third trimester. The positive and negative predictive values were >93% in all trimesters. Limitations of ABPM stem from the fact that instrumentation for automatic monitoring, although advanced, is not perfect, still quite expensive, and not well tolerated during pregnancy, specially when the sampling rate is high. Despite these limitations, sequential measurements of BP early in pregnancy to be used for the calculation of the BP load or, preferably, the HBI6 provide good predictors of subsequent gestational hypertension or preeclampsia, rendering ABPM an useful technique in pregnancy.
This research was supported in part by grants from Dirección General de Enseñanza Superior e Investigación Científica, DGES (PM98-0106); Xunta de Galicia (PGICT99-PXI-32202B and PGICT00-PXI-32205PN); and Vicerrectorado de Investigación, University of Vigo.
- Received March 26, 2001.
- Revision received April 19, 2001.
- Accepted May 10, 2001.
Hermida RC, Ayala DE. Diagnosing gestational hypertension and preeclampsia with the 24-hour mean of blood pressure. Hypertension. . 1997; 30: 1531–1537.
Hermida RC, Ayala DE, Mojón A, Fernández JR, Silva I, Ucieda R, Iglesias M. Blood pressure excess for the early identification of gestational hypertension and preeclampsia. Hypertension. . 1998; 31: 83–89.
Hermida RC, Ayala DE, Mojón A, Fernández JR, Alonso I, Silva I, Ucieda R, Iglesias M. Blood pressure patterns in normal pregnancy, gestational hypertension, and preeclampsia. Hypertension. . 2000; 36: 149–158.
Ayala DE, Hermida RC, Mojón A, Fernández JR, Silva I, Ucieda R, Iglesias M. Blood pressure variability during gestation in healthy and complicated pregnancies. Hypertension. . 1997; 30: 611–618.
Hermida RC, Fernández JR, Mojón A, Ayala DE. Reproducibility of the hyperbaric index as a measure of blood pressure excess. Hypertension. . 2000; 35: 118–125.
Hermida RC, Fernández JR, Mojón A. High sensitivity test for the early diagnosis of gestational hypertension and preeclampsia: III. computation of time-specified tolerance intervals as reference for blood pressure excess in the diagnosis of gestational hypertension. J Perinatal Med. . 1997; 25: 237–253.
White WB, Dey HM, Schulman P. Assessment of the daily blood pressure load as a determinant of cardiac function in patients with mild to moderate hypertension. Am Heart J. . 1989; 188: 782–795.
National High Blood Pressure Education Program Working Group report on high blood pressure in pregnancy. Am J Obstet Gynecol. . 1990; 163: 1689–1712.
Rosner B. Fundamentals of Biostatistics. 2nd ed. Boston, Mass: Duxbury Press; 1986: 1: 51–56.