Evaluation of a definition of pre-eclampsia
Correspondence: Dr R. A. North, Department of Obstetrics and Gynaecology, National Women's Hospital, Claude Rd, Auckland 1003, New Zealand.
Objectives To determine: 1. whether an alternative definition of gestational hypertension and pre-eclampsia stratifies women according to their risk of maternal and fetal complications; 2. whether pregnancy outcome in women with gestational hypertension differs in the presence or absence of ‘+’ proteinuria; and 3. whether a blood pressure rise of ≥ 30/15 mmHg during pregnancy is associated with adverse outcome in women who remain normotensive.
Design Prospective, nested case–control study.
Setting Community based.
Population Healthy, nulliparous women (n= 1496).
Methods Women recruited into a study investigating serum markers predictive of pre-eclampsia were classified as having gestational hypertension (systolic blood pressure ≥ 140 mmHg with a rise of ≥ 30 mmHg and/or diastolic blood pressure ≥ 90 mmHg with a rise of ≥ 15 mmHg) or pre-eclampsia (gestational hypertension plus proteinuria ≥ 2+on dipstick or > 0.3 g/24 h). Maternal and fetal complications in gestational hypertension or pre-eclampsia were compared with a control group of 223 randomly selected normotensive women. The main outcome measures were severe maternal disease, preterm birth and small for gestational age infant.
Results A stepwise increase in adverse maternal and fetal outcomes occurred in gestational hypertension (n= 117, 743%) and pre-eclampsia (n= 71, 4.8%). Severe maternal disease developed in 26.5% (21.4% severe hypertension alone, 5.1% multisystem disease) of women with gestational hypertension and 63.4% (21.1% severe hypertension alone, 42.3% multisystem disease) of women with pre-eclampsia (OR 4.8; 95% CI 2.4–9.5). Preterm birth and small for gestational age infants were more frequent in gestational hypertension (OR 1.7; 95% CI 0.5–5.4, and OR 2.0; 95% CI 1.0–3.7, respectively) and pre-eclampsia (OR 14.6; 95% CI 5.8–37.8, and OR 2.6; 95% CI 1.2–5.3) than in the normotensive group. Among women with gestational hypertension severe maternal disease was more common in women with ‘+’ proteinuria (41.7%) than in those with no proteinuria (15.9%): OR 3–8; 95% CI 1.5–9.8. Pregnancies were uncomplicated in the 27% of normotensive women who had a rise of ≥ 30 mmHg systolic blood pressure and/or ≥ 15 mmHg rise in diastolic blood pressure.
Conclusions In the nulliparous population studied our definition of gestational hypertension and pre-eclampsia identified women at increasing risk of maternal and fetal complications. In gestational hypertension, the presence of proteinuria ‘+’ was associated with a 3.8-fold increase in severe maternal disease. Normotensive women who have a rise in blood pressure ≥ 30/15 mmHg had uncomplicated pregnancies.
No consensus exists on the definition of hypertensive disorders in pregnancy, which has resulted in several parallel classifications in use around the world17. The definitions of gestational or transient hypertension and pre-eclampsia need to be simple, easily applicable and classify women into groups at stratified risk of maternal and fetal complications8. All classifications, except the proposal of the Australasian Society for the Study of Hypertension in Pregnancy (ASSHP), have defined the disease on two or three signs of the disorder, namely hypertension, proteinuria and oedema. The nonspecific finding of oedema has been dropped from most definition1,3,4,6,7. ASSHP proposed any manifestation of maternal multisystem disease be used to diagnose pre-eclampsia, and this classification has recently been refined6,8
Each system of classification has used different definitions to diagnose hypertension and proteinuria. Hypertension has been defined by an absolute blood pressure threshold alone1,4,7, an absolute blood pressure threshold or a rise in blood pressure2,5,6, and a combination of absolute blood pressure with a rise in blood pressure3. Under the National High Blood Pressure Education Program (n HBPEP) and ASSHP classifications, hypertension may be diagnosed when there is only a rise in systolic blood pressure of at least 25–30 mmHg or a rise in diastolic blood pressure of at least 15 mmHg5,6. This has arisen because severe, multi-systemic disease can occur without the blood pressure reaching the arbitrary level of 140/90 mmHg9. The prevalence of maternal or fetal complications in women who have a rise in blood pressure, but whose blood pressure remains c 140/90 mmHg, is unknown. Furthermore, Redman and Jefferies3 reported that 28% of women had at least a rise of 20 mmHg in diastolic blood pressure during pregnancy. This suggests that if a definition using an isolated rise in blood pressure of 30115 mmHg was strictly applied, over one-quarter of all pregnant women would be diagnosed with transient or gestational hypertension.
Proteinuria has been used to separate the more severe disease of pre-eclampsia from gestational or transient hypertension. The significant number of false positive dipstick tests at ‘+’ or 0.3 g/L proteinuria10,11. has resulted in two classifications using the stricter criterion of at least ‘2+’ or 1 g/L proteinuria on dipstick4,6. Despite the high false positive rate with ‘+’ proteinuria, between 33% and 83% of these women have a 24 h urinary protein excretion 2 0.3 g/24 h10,11. Use of strict criteria to define proteinuria will improve specificity, which is important when women with pre-eclampsia are identified for research purposes. On the other hand, if the definition is for clinical practice, the stricter definition of proteinuria may exclude women with significant disease.
We had conducted a prospective, community-based study of healthy nulliparous women to investigate potential serum markers predictive of preterm birth and pre-eclampsia12. Rigorous criteria were used to define gestational hypertension and pre-eclampsia. Gestational hypertension was defined as systolic blood pressure of at least 140 mmHg with a ≥ 30 mmHg rise and/or diastolic blood pressure of at least 90 mmHg with a rise of ≥ 15 mmHg occurring on two or more occasions after 20 weeks of gestation. Pre-eclampsia was defined as gestational hypertension and proteinuria of at least ‘2+’ or 1 g/L on dipstick or 24 hour urinary protein excretion ≥ 0.3g.
The main aim of the present study was to investigate whether these definitions of gestational hypertension and pre-eclampsia identified women at increasing risk of maternal and fetal complications compared with normotensive women. To assess the clinical importance of proteinuria ‘+’ or 0.3 g/L on dipstick, we compared pregnancy outcome in women with gestational hypertension subdivided according to the presence or absence of ‘+’ proteinuria. Pregnancy outcome was also determined in the subgroup of women who had a rise in systolic blood pressure of at least 30 mmHg and/or a rise in diastolic blood pressure of at least 15 mmHg, but whose maximum blood pressure was c 140/90 mmHg. Finally, we determined the clinical significance of classifying our women according to the definition of pre-eclampsia proposed by Redman and Jefferies3: “first diastolic pressure below 90 mmHg, a subsequent increase of at least 25 mmHg, and a maximum reading of at least 90 mmHg”.
A prospective, community-based, nested case-control study was carried out to determine serum markers predictive of preterm birth and pre-eclampsia12. Sixteen hundred and fifty-one nulliparous women, less than 18 weeks of gestation, were recruited. Fifty were excluded for the following reasons: fetal congenital malformations (n= 17), multiple pregnancy (n= 14), aspirin therapy (n= 4) and essential hypertension (n= 15). A further 33 women had miscarriages, 32 women had no outcome data available as they had moved house, and 40 women withdrew from the study. The final study group comprised 1496 healthy nulliparous women. The regional health authority's ethics committee approved the study, and informed consent was obtained from all participants.
The women received routine antenatal care from a range of private and public obstetricians, midwives and general practitioners. Deliveries occurred at tertiary level teaching hospitals, general district hospitals, birthing units and at home. Medical records were reviewed after delivery. All women with sustained systolic blood pressure of at least 140 mmHg or diastolic blood pressure of at least 90 mmHg before the onset of labour were identified. In these women blood pressure at booking (before 18 weeks of gestation) was checked to determine the women who met the rise in blood pressure criterion according to our definition. Gestational hypertension was defined as systolic blood pressure of at least 140 mmHg with a ≥ 30 mmHg rise or a diastolic blood pressure of at least 90 mmHg with a rise of ≥ 15 mmHg on two or more occasions after 20 weeks of gestation, but before the onset of labour. Pre-eclampsia was defined as gestational hypertension and proteinuria of at least ‘2+’ or 1 g/L on dipstick or a 24-hour urinary protein excretion ≥ 0.3 g. After excluding women with gestational hypertension and pre-eclampsia from the study cohort, the control group was identified by random number selection (Paradox 4.5, Borland, Scott Valley, California, USA). Detailed pregnancy data were collated on women with gestational hypertension, pre-eclampsia and the randomly selected controls. As this was a community-based study, with women receiving routine antenatal care, not all laboratory tests were measured in all women.
The major outcome measures were severe maternal disease, preterm birth and small for gestational age infants. Severe maternal disease was defined as any of the following: systolic blood pressure ≥ 160 mmHg; diastolic blood pressure ≥ 110 mmHg; platelets less than 100 × 109/L; renal insufficiency (serum creatinine ≥ 0.10 mmol/L); liver involvement (hepatic pain or aspartate transaminase concentration ≥ 40 IU/L); imminent eclampsia (headache with visual disturbance and clonus) or eclampsia; and pulmonary oedema5. Gestational age was calculated from the date of the last menstrual period or, if there was a discrepancy in gestation greater than plus or minus five days between the last menstrual period and an ultrasound scan performed before 20 weeks, then gestation was calculated from the scan. Preterm birth was defined as delivery before 37 weeks of gestation. Infants were classified as small for gestational age if their birthweight was less than the 10th centile for gestational age, according to the birth-weight centile tables of Guaran et al.13.
Analysis of variance was used in comparisons of continuous variables, and the χ2 or Fisher's exact test was used to compare categorical variables. Odds ratios with 95th centile confidence intervals (OR, 95% CI) were calculated. Differences of P 0.05 were considered significant.
Of the 1496 women, pre-eclampsia developed in 71 (4.8%) and gestational hypertension occurred in 117 (743%). Among these hypertensive women, 164 (87.2%) had both systolic and diastolic hypertension, 18 (9.6%) had diastolic hypertension alone, and six (3.2%) had systolic hypertension alone. The 223 randomly selected controls included 206 women whose maximum blood pressure was < 140/90 mmHg and 17 women with a maximum blood pressure ≥ 140/90 mmHg but without a rise in blood pressure of 30/15 mmHg. Six of these 17 women had an elevation in both systolic and diastolic blood pressure, three had only diastolic hypertension, and eight women had a systolic blood pressure ≥ 140 mag. Among these 17 women, one had 1 g/L proteinuria on dipstick and none had severe maternal disease. All 17 women delivered at term, but three babies were small for gestational age.
Maternal outcome according to hypertensive diagnosis is shown in Table 1. The definitions used for gestational hypertension and pre-eclampsia identified women at increasing risk of adverse maternal outcome compared with normotensive women. The odds ratio for severe maternal disease in women with pre-eclampsia compared with gestational hypertension was 4.8 (95% CI 2.4–9.5). In the gestational hypertension group, 25 of the 31 women with severe maternal disease had severe hypertension alone. In a further six women (5.1%) with gestational hypertension, three developed a mild rise in aspartate transaminase (45–40 IU/L), two developed abnormal renal function, and one thrombocytopenia. None developed the HELLP syndrome. Forty-five of the 71 women (63.4%) with pre-eclampsia had severe maternal disease, 21.1% with severe hypertension alone and 42.3% with evidence of multisystem disease.
Table 1. Maternal outcome according to hypertensive classification using our definition. Values are given as mean (SD) or %. BP = blood pressure; AST = aspartate transaminase.
|Age (years)||28.1 (4.8)||28.4 (4.5)||27.5 (4.4)|
|Booking BP (mmHg)|| || || |
| Systolic||113 (12)||114 (11)||117 (12)*|
| Diastolic||67 (9)||70 (8)**||69 (9)|
| ||(n=222)|| || |
|Maximum BP (mmHg)|| || || |
| Systolic||120 (11)||147 (10)‡||158 (14)‡|
| Diastolic||75 (8)||98 (7)‡||103 (10)‡|
|Proteinuria|| || || |
| ≥ 2+ or > 0.3 gl24 h||10||0||98.6†§|
| ||(n=211)|| || |
|Gestation at initial elevation in BP (weeks)||—||35.9 (4.0)||34.3 (4.8)|
|Thrombocytopenia|| || || |
| < 100 x109/L||0||0.9||4.2*|
| ||(n= 206)|| || |
| || ||(n= 78)||(n= 58)|
|Severe maternal disease¶||0||26.5†||63.4†|
There was a stepwise increase in the frequency of fetal complications in gestational hypertension and pre-eclampsia (Table 2). The odds ratio for preterm birth with gestational hypertension was 1.7 (95% CI 0.5–5.4) and with pre-eclampsia 14.6 (95% CI 5.8–37.8) compared with the normotensive group. All preterm deliveries, except one, in the gestational hypertension and pre-eclampsia groups were iatrogenic. A third of the women with pre-eclampsia were delivered preterm, 11 women (15.5%) before 33 weeks and 14 (19.7%) between 33 and 36 weeks. The risk of an infant with small for gestational age compared with a normotensive pregnancy was increased in gestational hypertension (OR 2.0; 95% CI 1.0–3.7) and pre-eclampsia (OR 2.6; 95% CI 1.2–5.3). The perinatal deaths were two babies delivered at 22 and 25 weeks due to early onset pre-eclampsia and one stillbirth secondary to placental abruption at 38 weeks in a woman with severe gestational hypertension.
Table 2. Fetal outcome according to hypertensive classification using our definition. Values are given as mean (SD) or %, unless otherwise indicated. SGA = small for gestational age.
|Gestation at birth (weeks)||40.0 (1.6)||39.2 (1.7)*||36.9 (4.2)*|
|Preterm birth < 37 weeks||3.6||6.0||35.2†|
|Birthweight (g)||3421 (469)||3246 (561)**||2789 (941)†|
|SGA (< 10th centile)||11.7||20.5*||25.4**|
|Perinatal deaths (n)||0||1||2|
Maternal and fetal complications in women with gestational hypertension, with and without ‘+’ or 0.3 g/L proteinuria on dipstick, are shown in Table 3. Severe maternal disease, predominantly severe hypertension, was more common in women with ‘+’ proteinuria compared with women with no proteinuria (OR 3.8; 95% CI 1.5–9.3). Multisystem disease occurred in five women (10.4%) with gestational hypertension with proteinuria ‘+’, compared with one (1.5%) in those without proteinuria (P= 0.04). Women with gestational hypertension and no proteinuria were still at risk, with 11 (15.9%) developing severe maternal complications and a quarter of their babies being small for gestational age.
Table 3. Pregnancy outcome in women with gestational hypertension and proteinuria ‘+’ or no proteinuria on dipstick. Values are given as mean (SD) or %, unless otherwise indicated. BP = blood pressure; SGA = small for gestational age.
|Gestation BP elevated (weeks)||36.1 (4.1)||35.6 (3.8)|
|Maximum BP (mmHg)|| || |
| Systolic||144 (10)||150 (10)**|
| Diastolic||97 (6)||100 (7)*|
|Severe hypertension||14 5||33.3*|
|Severe maternal disease||15.9||41.7**|
|Gestation at birth (weeks)||39.5 (1.9)||38.9 (1.4)|
|Premature birth (< 37 weeks)||5.8||6.3|
|Birthweight (g)||3233 (600)||3267 (505)|
|SGA (< 10th centile)||24.6||14.6|
|Perinatal deaths (n)||0||1|
Among the 205 women with a maximum blood pressure < 140/90 dg, 56 women (27.3%) (95% CI 21.2–33.4) had a rise in either systolic blood pressure of at least 30 mmHg or in diastolic blood pressure of at least 15 mmHg. No booking blood pressure was available in one normotensive woman. The outcome of pregnancy in these women was similar to the outcome in women with a maximum blood pressure less than 140/90 mmHg, but whose blood pressure rose < 30/15 mmHg during pregnancy (Table 4). Seventeen women (30.4%) in the group with a significant rise in blood pressure were delivered by caesarean section for failure to progress or fetal compromise in labour (n= 1l), breech (n= 5) and placenta praevia (n= 1).
Table 4. Pregnancy outcome in women with a maximum blood pressure below 140/90 mmHg, with and without a rise in systolic BP ≥30 mmHg or diastolic BP ≥ 15 mmHg during pregnancy. Values are given as mean (SD) or %. dBP = diastolic blood pressure; sBP = systolic blood pressure; SGA = small for gestational age.
|Age (years)||7.8 (4.6)||8.7 (5.3)|
|Booking BP (mmHg)|| || |
|Systolic||14 (11)||06 (12)**|
|Diastolic||9 (8)||9 (7)**|
|Maximum BP (mmHg)|| || |
|Systolic||17 (10)||21 (8)*|
|Diastolic||2 (7)||8 (5)**|
| ||(n= 141)||(n= 52)|
|Severe maternal disease||0||0|
|Gestation at birth (weeks)||40.0 (1.7)||40.1 (1.4)|
|Birthweight (g)||3432 (491)||3433 (421)|
|SGA (< 10th centile)||10.7||12.5|
When our data were reclassified according to the definition of Redman and Jefferies, 58 of the 71 women with pre-eclampsia (81.6%) and 77 of the 117 women with gestational hypertension (65.8%) were considered to have pre-eclampsia. These women were at high risk of maternal and fetal complications (Table 5). Fifty-three women (40 with gestational hypertension and 13 with pre-eclampsia by our criteria) were re-classified as normotensive by Redman and Jefferies. These reclassified normotensive women had a higher rate of complications (Table 5). Nine (17%) developed severe maternal complications compared with none of our original normotensive women (P= 0.0001). However, they represent only 3.9% of all women (n= 1361) in the study who would be considered normotensive by Redman and Jefferies.
Table 5. The characteristics of women and infants reclassified according to the Redman and Jefferies3 classification. Values are given as mean (SD) or %, unless otherwise indicated. BP = blood pressure; SGA = small for gestational age.
|Booking BP (mmHg)|| || || |
| Systolic||113 (12)||118 (11)†||114 (11)|
| Diastolic||67 (9)||75 (5)‡||67 (8)|
| ||(n= 222)|| || |
|Maximum BP (mmHg)|| || || |
| Systolic||120 (11)||147 (10)‡||153 (14)|
| Diastolic||75 (8)||94 (6)‡||102 (8)|
|Proteinuria|| || || |
| ≥ 2+ or > 0.3 g/24 h||1.0||23.1||43.0|
| ||(n= 211)|| || |
|Severe maternal disease||0||17.0‡||49.6|
|Gestation at birth (weeks)||40.0 (1.6)||39.3 (1.8)**||38.0 (3.4)|
|Premature < 37 weeks||3.6||9.4||20.0|
|Birthweight (g)||3421 (469)||3360 (602)||2961 (787)|
|SGA (< 10th centile)||11.7||9.4||27.4|
|Perinatal deaths (n)||0||0||3|
The definitions of gestational hypertension and pre-eclampsia used in this large prospective study of nulliparous women identified groups at stratified risk of adverse maternal and infant outcomes. Women with pre-eclampsia were at high risk of severe maternal disease (63.4%), preterm birth (35%) and small for gestational age infants (25%). Gestational hypertension was not benign, with a quarter of these women developing severe maternal disease, usually severe hypertension alone. Both this community-based study and a previous study in a tertiary referral population found gestational hypertension is associated with a small, but important, risk of multisystem disease14. In both studies proteinuria was defined as at least ‘2+’ or 1 g/L on dipstick or > 0.3 g/24 h. Consequently, women with ‘+’ or 0.3 g/L proteinuria were included in the gestational hypertension group.
Women with gestational hypertension and ‘+’ proteinuria on dipstick had a three-fold greater risk of severe maternal disease (predominantly severe hypertension) compared with women with negative or trace proteinuria. This suggests ‘+’ or 0.3 g/L proteinuria on dipstick may be a more appropriate cutoff when a definition of pre-eclampsia is applied to clinical practice. Given the high false positive rate at ‘+’ proteinuria with dipstick analysis, the presence of proteinuria should be confirmed by measuring 24 hour urinary protein excretion or a protein: creatinine ratio15. It may be preferable for future definitions of pre-eclampsia to use more accurate methods to diagnose proteinuria15. It is important that criteria used to define gestational hypertension identify women with a higher rate of adverse pregnancy outcomes than the general population. Otherwise, why make the diagnosis? Our definition of gestational hypertension identified a group at increased risk of adverse outcomes, even in women without proteinuria. Severe hypertension occurred in 14.5% of these women, multisystem disease in one and a quarter of their infants were small for gestational age.
The best method to define hypertension is debated. A diastolic blood pressure of 90 mmHg is two standard deviations from the mean16. Two definitions, still widely in use, allow an isolated rise in systolic or diastolic blood pressure to classify a woman as hypertensive5,6. Our finding that a quarter of women, whose blood pressure remains below 140/90 mmHg, have a ≥ 30 mmHg rise in systolic or a 2 15 mmHg rise in diastolic blood pressure during pregnancy is in keeping with previous studies3,17,18. In one study this rise occurred in 67% of normotensive, primiparous women18. In our study normotensive women with a subsequent rise in blood pressure of ≥ 30/15 mmHg had a lower mean booking blood pressure. Their pregnancies were not associated with an increase in adverse pregnancy outcomes. The rate of caesarean section in these women was unusually high due to the high rate of breech presentations.
Pregnant women with a blood pressure < 140/90 mmHg can develop eclampsia or other manifestations of multisystem disease associated with pre-eclampsia. In the UK Eclampsia Survey, eclampsia occurred in 0.049% of pregnancies9. Approximately 20% of women with eclampsia have a blood pressure < 140/90 mmHg, giving an estimated rate of eclampsia in the presence of a normal blood pressure of 0.0 1%19. Assuming a rise in blood pressure of ≥ 30/15 mmHg occurred in the normotensive eclamptic women and, based on our data, a rise in blood pressure ≥ 30/15 mmHg occurs in 27% of normotensive women, then inclusion of an isolated rise in blood pressure in the definition of pre-eclampsia would result in 2700 women being classified as abnormal, to identify one normotensive woman with eclampsia. Assigning a disease label is not benign, as it is likely to increase the anxiety of both the woman and her obstetrician and may result in inappropriate intervention. It would, therefore, seem sensible not to include an isolated rise in blood pressure in future definitions.
The optimal definition of pre-eclampsia will, in part, depend on whether it is to be used in clinical practice or research. If the definition is for use in patient management, it should have excellent sensitivity to ensure all women with the disease are detected, perhaps at the cost of slightly reduced specificity. On the other hand, a definition devised for research studies, needs to be very specific for the disease. These two situations, clinical practice and research, may ultimately require slightly different definitions which identify overlapping, but not identical groups of women.
Only two previous classification systems have provided data on pregnancy outcome to justify their definitions3,8. Redman and Jefferies applied their definition of pre-eclampsia to a second population to show it identified women whose babies had lower birth weights and higher perinatal mortality. Our study and a previous study of nulliparous women confirm that the Redman and Jefferies definition selects women at high risk of maternal and fetal complications20. If Redman's and Jefferies' classification was applied in clinical practice, it would be of concern that 17% of women with isolated systolic hypertension or a diastolic blood pressure 190 mmHg and a rise of 15–24 mmHg (now classified as normotensive) would develop severe maternal disease.
In line with most previous definitions, our definition used hypertension and proteinuria to identify women with pre-eclampsia. Use of more accurate methods to detect and quantify proteinuria should be considered15. In view of the multisystem nature of the disease, incorporation of a multisystem approach into a definition of pre-eclampsia may improve the disease classification6,8.
These data highlight the importance of basing definitions of pre-eclampsia on scientific evidence, including correlation with pregnancy complications. The purpose of this study is not to propose yet another definition of pre-eclampsia, but to provide data on the clinical implications of choosing particular criteria to define the disease. The information in this study may assist in developing improved definitions in the future. Any proposed classification should be tested in different populations for generalizability (e.g. including multiparous women), before acceptance into clinical practice or research. This evidence-based approach may circumvent some of the difficulties associated with previous definitions of pre-eclampsia based on consensus.
Our definition of gestational hypertension and pre-eclampsia identified women at stratified risk of pregnancy complications. Women with gestational hypertension and proteinuria ‘+’ on dipstick had a three-fold increase in maternal complications compared with women with gestational hypertension and no proteinuria. In women whose maximum blood pressure is < 140/90 mmHg, a rise in blood pressure of 30/15 mmHg during pregnancy is not a marker of adverse pregnancy outcome. This information may be helpful in devising improved definitions of pre-eclampsia in the future.