Introduction

Traditionally, hypertensive disorders of pregnancy have been divided into chronic arterial hypertension (women with hypertension that become pregnant) vs. gestational hypertension (pregnancy-induced hypertension), using the office blood pressure (BP) before 20 weeks of gestation to differentiate both conditions [1,2,3]. Thus, according to the traditional definitions, a pregnant woman who has an office BP < 140/90 mmHg before 20 weeks of gestation and subsequently develops hypertension, should be defined as gestational hypertensive. However, using this approach the possibility to have masked chronic hypertension (normal office BP and elevated BP on ambulatory BP monitoring before 20 weeks) is not considered. Recently, our group showed that masked hypertension is a frequent condition in high-risk pregnant women. Moreover, masked chronic hypertension carries an important increase in maternal and fetal risk [4].

The clinical significance of different hypertensive disorders of pregnancy is not the same. While the risk of maternal and fetal complications of chronic hypertension has been shown [5], the risk associated with gestational hypertension is less well defined. Wu et al. using the National Inpatient Sample database, analyzed the association between different hypertensive disorders of pregnancy and adverse in-hospital maternal and fetal outcomes in more than 44 million deliveries. Women with chronic hypertension, but not those with gestational hypertension, had a higher risk of both, maternal and fetal adverse outcomes [6]. However, the International Society for the Study of Hypertension in Pregnancy (ISSHP) in their last position paper state that “outcomes in pregnancies complicated by gestational hypertension are normally good, but about a quarter of women with gestational hypertension will progress to preeclampsia and have poorer outcomes”. They also state that “gestational hypertension is not a uniformly benign condition” [3]. The risk of complications has been attributed to the gestational age at which it develops [7]. Also, some data suggest that different outcomes could be related to different out-of-office BP levels. Davis et al. found that pregnant women with gestational hypertension who developed preeclampsia/eclampsia (PE) had higher awake and 24-h systolic BP than those who did not [8]. However, the possibility that those women had masked chronic hypertension was not further analyzed.

We hypothesized that a proportion of women with gestational hypertension defined using current recommendations might have masked chronic hypertension, rather than pregnancy-induced hypertension. Furthermore, this distinction could explain the heterogeneity of PE risk and could be important for prognostic considerations. Consequently, the objectives of this study were 1—to evaluate the prevalence of masked chronic hypertension in high-risk pregnant women classified as gestational hypertension using the traditional definition. 2—to compare the risks of developing preeclampsia in true gestational hypertension vs those women classified as having gestational hypertension but who had had masked chronic hypertension in an evaluation performed before 20 weeks of gestation

Material and methods

This is a cohort study of consecutive women with high-risk pregnancies derived between 1st January 2016 and 31st March 2020 to the Cardiometabolic Diseases Unit (San Martín Hospital, La Plata, Argentina) and evaluated before 20 weeks of gestation using a predesigned protocol. They had been referred to the High-risk Pregnancy Office of the Obstetrics Department at San Martín Hospital (La Plata, Argentina) by primary care physicians either because of their comorbidities, such as diabetes, hypertension, chronic kidney disease or others, or because of certain findings detected during the current pregnancy (gestational diabetes, gestational hypertension or multiple pregnancy).

All women were also evaluated in the Cardiometabolic Disease Unit with a pre-defined protocol for office and ambulatory BP measurement. This protocol, that includes the routine use of ABPM after 10 weeks of gestation in all high-risk pregnancies, has been incorporated as usual medical practice at our hospital since 2016. The protocol has been previously described [9]. In brief: a specially trained nurse, at the end of a 15-min interview, performed three BP measurements employing a validated oscillometric automatic BP device (OMRON HEM 705 CP) [10], in seated position with the arm at heart level and using appropriate arm sleeves. Office BP was defined as an average of these three determinations. Immediately after, an ABPM was initiated with a validated monitor (Spacelabs 90207) [11]. Measurements were scheduled every 15 min during the day and every 20 min at night. Only ABPMs with at least 70% successful measurements and at least one record per hour were considered valid [12]. Only women who had at least one ABPM before 20 weeks of gestation and who were followed up until delivery at San Martín Hospital were included in this analysis.

Chronic hypertension was defined as office BP ≥ 140/90 mmHg and 24-h ABPM ≥ 130/80 mmHg, based on the ABPM before 20 weeks of gestation, or treatment with antihypertensive drugs started before the current pregnancy [3]. Women without chronic hypertension was classified as normal BP (office BP < 140/90 mmHg and 24-h ABPM < 130/80 mmHg), white-coat hypertension (office BP ≥ 140/90 mmHg and 24-h ABPM < 130/80 mmHg) or masked hypertension (office BP < 140/90 mmHg and 24-h ABPM ≥ 130/80 mmHg).

According to current guidelines, women without chronic hypertension who developed hypertension (office BP ≥ 140/90 mmHg and/or antihypertensive treatment) after 20 weeks of gestation were defined as gestational hypertension. According to the ABPM before 20 weeks of gestation, these women were divided in two subgroups: subgroup 1- if their ABPM was normal (“true” gestational hypertension), and subgroup 2- if they had hypertension in their ABPM (masked chronic hypertension unmasked in the second half of gestation).

PE was defined as the presence of any of the following: 1—preeclampsia (BP ≥ 140/90 mmHg associated with proteins in urine ≥ 300 mg/24 h), or 2—eclampsia (seizures in a patient with preeclampsia or gestational hypertension), or 3—HELLP syndrome (hemolysis, elevated liver enzyme levels, and low platelet count). Data of the delivery and perinatal outcomes -APGAR score (Appearance, Pulse, Grimace, Activity, and Respiration) and birth weight- were extracted from the default protocol used by the Obstetrics Department.

Continuous variables were expressed as mean and standard deviation (SD) and compared between groups using “t” test or ANOVA and post-hoc Scheffe test, as appropriate. Ordinal variables were expressed as median and interquartile range (IQR) and compared with Kruskal–Wallis H test. Categorical variables were expressed as percentage and were compared with ×2 or Fisher’s exact test, as appropriate.

The relative risks of women with normotension, chronic hypertension, and subgroups 1 and 2 of gestational hypertension, were estimated using logistic regression models and expressed as odds ratio with 95% confidence interval (OR, 95% CI). Normotension was the reference category. Two models were constructed: Model 1, unadjusted; Model 2, adjusted by relevant covariables (age, diabetes mellitus, gestational diabetes, and use of low doses of acetylsalicylic acid (ASA) and/or calcium supplements).

The data were analyzed using SPSS (IBM, Armonk, NY, USA); P values < 0.05 (two-tailed) were considered significant.

This was an observational study that carries no risk for the patients and conducted in accordance with the Declaration of Helsinki. The study protocol was approved by the medical bioethics committee of the Faculty of Medical Sciences, National University of La Plata (UNLP), Buenos Aires, Argentina (COBIMED 0/27).

Results

A total of 262 high-risk pregnant women (age 31 ± 7 years, with a median of 15 weeks, range 7–19 weeks) were evaluated before the 20th week of gestation. Of these, 30 patients were excluded because had not been followed up until the delivery, and 5 because they did not have a valid ABPM. The remaining 227 are included in the present analysis (age 32 ± 6 years, median gestation week 15, range 8–19).

At the evaluation in the first half of gestation, 67 (29.5%) women had chronic hypertension; 45 were under pharmacological treatment and 34 had office BP ≥ 140/90 mmHg, of which 12 had white-coat hypertension. The prevalence of masked hypertension in untreated normotensive pregnant women was 20.9% (31/148). The characteristics of the patients are shown in Table 1.

Table 1 Comparison of patients with and without chronic hypertension at the initial evaluation.

In the second half of pregnancy, 39 developed gestational hypertension defined by traditional criteria; 16 were classified as subgroup 1 (had had a normal ABMP performed before 20 weeks of gestation), and 23 (59%) were classified as subgroup 2 because of had had masked chronic hypertension in the ABPM performed before 20 weeks of gestation (Fig. 1). Table 2 compares the characteristics of subgroups 1 and 2 of gestational hypertension.

Fig. 1: Flow chart of the study.
figure 1

Flow chart of the study including the classification of pregnant women before 20 weeks of gestation using ambulatory blood pressure monitoring.

Table 2 Comparison of pregnant women without chronic hypertension at the initial evaluation according to their evolution after 20 weeks of gestation.

The prevalence of PE in the whole cohort was 23.3%. The prevalence of PE analyzed by categories were 15.7%, 12.5%, 43.5%, and 32.8% for women with normotension, subgroup 1 of gestational hypertension, subgroup 2 of gestational hypertension and chronic hypertension, respectively. Table 3 shows the adjusted and unadjusted risks of developing PE according to the condition in the second half of pregnancy. Compared with pregnant women with normotension, subgroup 1 of gestational hypertension (normal ABPM before 20 weeks of gestation) did not increase the risk of developing PE (OR 0.76, 95% CI 0.16–6.65). Conversely, subgroup 2 of gestational hypertension (those women who developed office hypertension after 20 weeks of gestation but had had masked chronic hypertension in the first half of gestation) increased the risk for PE more than 4 times (0R 4.47 CI 1.16–12.63). Risk estimation did not change substantially after the adjustment for multiple possible confounders.

Table 3 Absolute risk and adjusted and not adjusted odds ratio for developing preeclampsia/eclampsia according to the different categories of blood pressure in pregnancy.

Discussion

Our study shows that more than half of the women that had been classified based on office BP measurement as gestational hypertension had masked hypertension according to the results of an ABPM performed before 20 weeks of gestation. Thus, these women did not have a “true” gestational hypertension because they were not strictly normotensives in the first half of gestation. Indeed, these women have a masked chronic hypertension unmasked in the second half of gestation. Moreover, women with masked chronic hypertension, but not those with “true” gestational hypertension, had a very high risk to developed PE (~4 times more risk). Physiological BP decrease in the first half of the pregnancy could contribute to disguising chronic hypertension.

In the recently published CHAP study, Tita et al. [13] showed that the treatment of mild chronic hypertension (office BP 140–160/90–100 mmHg before 20 weeks of gestation) was associated with better pregnancy outcomes without an increase in the risk of low birth weight, highlighting the importance of identifying and early treat pregnant women with chronic hypertension. Regarding women with office BP < 140/90 mmHg, an observational study performed on low-risk pregnant from China shows that women with office BP between 130–140 and/or between 80–90 mmHg (measured before 20 weeks of gestation) had more than 2 times risk of PE, compared with those with lower values of office BP [14]. In this sense, in previously published studies we communicated a high prevalence of masked hypertension in high-risk pregnant women with office normotension [9]. Thus, it could be possible that some of the risks for PE observed in pregnant women with office hypertension lower than 140/90 mmHg could be attributed to masked chronic hypertension.

Gestational hypertension is also associated with cardiovascular disease in the long-term follow-up. In a populational study from Sweden including more than 400,000 women, the adjusted incidence rate ratio for later development of ischemic heart disease was 1.6 (95% CI 1.3–2.0) when the first pregnancy was complicated by gestational hypertension without proteinuria [15]. In a retrospective cohort study, women with gestational hypertension (without PE) showed a higher risk for all-cause and cardiovascular mortality than normotensive women matched by age, year of childbirth, and parity at the time of the index pregnancy [16]. Again, untreated masked chronic hypertension could be a plausible explanation for the relationship between gestational hypertension and long-term cardiovascular disease. Furthermore, in the study by Saudan et al, 70% of women with hypertension gestational had had hypertension at previous gestation, suggesting the possibility that these women had indeed chronic hypertension [6].

Our findings could partially explain the heterogeneity in the risk for PE associated with gestational hypertension. Indeed, our cohort of high-risk pregnant women with gestational hypertension was composed of two subgroups with very different risks of PE development: women without chronic hypertension who developed true gestational hypertension and had low risk of PE (OR 0.72, 95% CI 0.15–3.45). Conversely, women who had masked, untreated, hypertension, and had a very high risk for PE (OR 4.47, 95% CI 1.16–12.63). These subgroups could be easily identified by an ABPM performed in the first half of pregnancy.

Remarkably, the risk of PE of masked chronic hypertension was higher than that associated with chronic hypertension (OR 4.47 vs 2.81, Table 3). In the general population, a similar phenomenon has been described by Banegas et al. [17] for the risk of cardiovascular disease. It has been attributed to the fact that masked hypertension is an undiagnosed and untreated condition. The benefits of treating mild hypertension in pregnant women with chronic hypertension shown in the previously mentioned CHAP study, could support our findings. Indeed, Table 1 shows that the average baseline values BP of office and ABPM were normal in women with chronic hypertension, suggesting that, on average, they were adequately treated. Conversely, although women with pseudo gestational hypertension have average normal office BP, they remain hypertensives as evaluated by ABPM.

Although the results of our study are straightforward, certain limitations must be addressed. First, this study was performed on a cohort of high-risk pregnant women, and therefore, our findings are not necessarily applicable to pregnancies without this condition. Indeed, the high prevalence of PE observed might be explained by selection bias. Second, the diagnosis of hypertension by ABPM was achieved using the same threshold as for the general population. However, a recently published study of pregnant women in a southern Chinese population defined similar ABPM thresholds using a maternal and fetal outcome-derived approach [18]. Third, this is an observational study; consequently, some bias could be not discharged. Thus, the use of low doses of aspirin, calcium supplements, or antihypertensive drugs may influence the results. However, the OR values were not altered by adjustment for covariates. Fourth, no studies showed the benefits of treating masked hypertension in pregnant women. However, the CHAP study showed in the analysis for subgroups that women with chronic hypertension diagnosed and receiving medication previously, had a significantly lower risk than those newly diagnosed and those with chronic hypertension diagnosed but without receiving medication [13]. Finally, the number of events was modest and further studies are necessary to confirm our results.

In conclusion, gestational hypertension seems to be a heterogeneous condition. More than half of women diagnosed as gestational hypertensives using only office BP had masked chronic hypertension when evaluated using AMBP; these women had a very high risk of PE. Conversely, women with gestational hypertension who had a normal ABPM before 20 weeks of gestation had a low risk of developing of PE. Thus, an ABPM performed before 20 weeks of gestation in office normotensives appears necessary to appropriately identify these subgroups, at least in high-risk pregnancies.

Summary

What is known about topic

  • Traditionally, hypertensive disorders of pregnancy have been divided into chronic arterial hypertension (women with hypertension that become pregnant) and gestational hypertension (pregnancy-induced hypertension).

  • This classification is defined based on the office blood pressure before 20 weeks of gestation to differentiate both conditions.

  • The clinical significance of different hypertensive disorders of pregnancy is not the same, being the chronic hypertension has been shown, while the risk associated with gestational hypertension is less well defined.

What this study adds

  • ABPM in high-risk pregnancies performed before 20 weeks of gestation in office normotensives identify than half of women diagnosed as gestational hypertensives using only office BP had masked chronic hypertension.

  • Women with masked chronic hypertension, but not those with “true” gestational hypertension, had a very high risk to developed PE (~4 times more risk).