Professional Documents
Culture Documents
Methods Conclusions
We analyzed public-use data from the National Hospital Discharge The increase in gestational hypertension may be exaggerated
Survey (NHDS), which has been conducted by the Centers for because of the revised clinical guidelines published in the 1990s;
Disease Control and Prevention, National Center for Health these same revisions would likely have reduced diagnoses of
Statistics since 1965. We calculated crude and age-adjusted preeclampsia. Therefore, our observation of a small but consistent
incidence rates and estimated the risk associated with available increase in preeclampsia is a conservative indication of a true
demographic variables using Cox regression modeling. population-level change.
Am J Hypertens 2008; 21:521-526 © 2008 American Journal of Hypertension, Ltd.
Results
Rates of preeclampsia and gestational hypertension increased
significantly (by 25 and 184%, respectively) over the study
Preeclampsia is a dangerous, multisystem complication of with onset of preeclampsia before 34 weeks of gestation. It is
human pregnancy and a leading cause of fetal and maternal important to note that preeclampsia is the underlying cause of
morbidity and mortality worldwide. This hypertensive disorder about one-quarter of all medically indicated preterm deliveries
of pregnancy is characterized by sustained de novo hyperten- in the United States.1 Because preeclampsia is a multisystem
sion and proteinuria after 20 weeks of gestation. The mater- disorder that can progress rapidly, it requires prompt interven-
nal syndrome is associated with pitting edema, particularly of tion that may include observation in a tertiary care setting and
the hands, face, and feet, abnormal clotting, and endothelial induction of delivery, which is the only known cure for this
abnormalities, as well as liver and renal dysfunction. With pro- condition.2–4
gression of preeclampsia to eclampsia or occurrence of HELLP Epidemiologic research suggests that preeclampsia has a
(Hemolysis, Elevated Liver enzymes, Low Platelets) syndrome, multifactorial etiology that includes an immunogenetic com-
the risk of maternal death increases. The fetal syndrome can be ponent.5 Established risk factors include nulliparity, family
manifested as preterm delivery, growth restriction, placental history of preeclampsia–eclampsia, preeclampsia in a previ-
abruption, fetal distress, and in some cases, death, especially ous pregnancy, obesity, increased insulin resistance, hyperlipi-
demia, increased trophoblastic mass (i.e., multiple gestation,
1Department of Community and Behavioral Health, College of Public Health,
molar pregnancy), and change of sexual partner between
University of Iowa, Iowa City, Iowa, USA; 2Department of Epidemiology,
College of Public Health, University of Iowa, Iowa City, Iowa, USA; 3Division of pregnancies.5 Protective factors include a prior completed
Reproductive Health, National Center for Chronic Disease Prevention and Health pregnancy and a terminated pregnancy for women conceiving
Promotion, Centers for Disease Control and Prevention, Atlanta, Georgia, USA;
4Office of the Director, National Center on Birth Defects and Developmental
again with the same partner.6
Disabilities, Centers for Disease Control and Prevention, Atlanta, Georgia, USA. In a previous publication, our group summarized nation-
Correspondence: Anne B. Wallis (anne-wallis@uiowa.edu) ally representative data on the incidence of preeclampsia and
Received 27 November 2007; first decision 23 December 2007; accepted 26 January eclampsia from 1979 through 1986.7 A subsequent publica-
2008; advance online publication 13 March 2008. doi:10.1038/ajh.2008.20 tion by Zhang et al. provided an update from the years 1988
© 2008 American Journal of Hypertension, Ltd. through 1997.8 There is now a pressing need to update these
reports so as to track changing rates and observe trends that and range checks for non-medical variables and consistency
may be related to population-level increases in pre-pregnancy checks for dates and ICD-9-CM codes. The National Center for
obesity,9,10 diabetes,11 maternal age,12 and multiple births12 Health Statistics reports high accuracy of discharge diagnostic
or changing social and economic conditions that may affect categories in comparison with hospital census reports.14
access to health care.
In this article, we present trends in incidence rates and Study population. The study population included all NHDS-
selected demographic risk factors during 1987–2004 so as to sampled discharges for delivery admissions among women
provide continuity with our earlier publication and enable who delivered one or more live or stillborn infants between
observation of long-term trends. Knowledge of these trends at 1987 and 2004. Hospital admissions for which a delivery
the population level should improve understanding of mater- occurred were identified by the ICD-9-CM code V27. In 2004,
nal risk, encourage preconception and interconception coun- the NHDS represented ~4.1 million hospitalizations for deliv-
seling, and increase appropriate monitoring and treatment, ery with an average length of stay of 2.6 days.15
particularly for high-risk women.
Results try, while women delivering in the west are at the lowest risk.
The crude average annual incidence rates (cases per 1,000 During the second 9-year time period of the study, the risk of
deliveries) of preeclampsia and gestational hypertension over preeclampsia is lower in each of the geographic areas relative
the 18-year study period were 27.4 and 21.3, respectively (data to the northeast.
for individual years are shown in Table 1); for eclampsia, the Women <20 years of age are at significantly increased risk of
rate was 0.92. gestational hypertension. Women ≥35 years of age were also at
The rate of preeclampsia and gestational hyperten- increased risk, although this rate ratio was not statistically sig-
sion increased significantly over the 18-year study period nificant. In all sub-categories of maternal age and geographic
(Figure 1). The age-adjusted rate (per 1,000 deliveries) of area, the rates of gestational hypertension increased uniformly
preeclampsia rose by 24.6% from 23.6 in 1987–1988 to 29.4 over time (Table 2).
in 2003–2004. Over the same time period, the rate of gesta- 35.0
tional hypertension nearly tripled from 10.7 to 30.6. In con-
trast, the age-adjusted rate of eclampsia decreased by 22.0%
Table 1 | Weighted number of cases and crude incidence rates, gestational hypertension and preeclampsia, 1987–2004
Diagnosis 1987 1988 1989 1990 1991 1992
Gestational hypertension (ICD-9-CM 642.3)
No. of cases 41,249 40,671 57,484 60,005 64,958 64,374
Rate (95% CI) 10.5 (8.2, 12.9) 10.8 (8.0, 13.5) 14.6 (11.3, 17.9) 14.9 (12.0, 17.9) 16.3 (13.1, 19.6) 16.5 (13.3, 19.6)
Preeclampsia (ICD-9-CM 642.4–5)
No. of cases 98,017 83,046 109,846 93,168 99,446 87,263
Rate (95% CI) 25.1 (21.5, 28.6) 21.9 (18.2, 25.7) 27.9 (23.1, 32.6) 23.2 (19.3, 27.0) 25.0 (20.7, 29.3) 22.3 (18.6, 25.9)
Diagnosis 1993 1994 1995 1996 1997 1998
Gestational hypertension (ICD-9-CM 642.3)
No. of cases 79,380 81,864 68,184 86,902 85,985 83,698
Rate (95% CI) 19.8 (16.0, 23.5) 21.0 (17.3, 24.7) 18.1 (14.8, 21.4) 22.7 (18.8, 26.6) 22.6 (18.9, 26.3) 20.9 (17.3, 24.5)
Preeclampsia (ICD-9-CM 642.4–5)
No. of cases 91,931 100,936 108,689 111,551 117,770 125,262
Rate (95% CI) 22.9 (18.7, 27.1) 25.9 (21.9, 29.9) 28.9 (24.4, 33.3) 29.0 (24.6, 33.5) 30.9 (26.4, 35.4) 31.2 (26.0, 36.5)
Diagnosis 1999 2000 2001 2002 2003 2004
Gestational hypertension (ICD-9-CM 642.3)
No. of cases 100,586 119,338 105,522 110,123 125,630 122,667
Rate (95% CI) 26.4 (21.9, 30.9) 31.9 (27.2, 36.6) 27.5 (23.2, 31.7) 27.9 (23.3, 32.4) 31.2 (26.0, 36.5) 29.7 (25.4, 33.9)
Preeclampsia (ICD-9-CM 642.4–5)
No. of cases 119,221 104,437 117,248 115,924 106,282 132,800
Rate (95% CI) 31.2 (26.6, 35.8) 27.9 (23.8, 32.1) 30.5 (26.1, 35.0) 29.3 (24.8, 33.9) 26.4 (22.3, 30.6) 32.1 (27.6, 36.6)
Table 2 | Weighted number of cases, rates, and rate ratios for gestational hypertension, preeclampsia, and eclampsia, 1987–2004
Time Period
1987–1995 1996–2004
Demographic variables No. of cases Rate Rate ratio (95% CI) No. of cases Rate Rate ratio (95% CI)
Preeclampsia (ICD-9-CM 642.4–5)
Age group (years)
<20 167,883 37.1 1.65 (1.47, 1.86) 175,270 41.8 1.73 (1.57, 1.90)
20–24 232,811 25.5 1.15 (1.03, 1.28) 274,537 31.5 1.31 (1.20, 1.42)
25–29 221,740 20.9 0.95 (0.85, 1.06) 267,442 28.2 1.18 (1.08, 1.28)
30–34 82,862 24.7 1.00 (referent) 138,560 29.8 1.00 (referent)
Women <20 years of age had a 2.6-fold higher risk of eclamp- Discussion
sia than older women during the first time period of the study, Our analysis indicates that important population-level changes
and a slightly lower risk during the second time period. In our in the rates of preeclampsia, eclampsia, and gestational hyper-
regional analysis, the risk of eclampsia was higher in the south tension occurred during the 18-year period ending in 2004.
than in the northeast (reference) during the first time period. The rates of both preeclampsia and gestational hypertension
The risk for women whose deliveries took place in the mid- increased significantly. In contrast, the eclampsia rate showed
west was comparable to the risk for women in the northeast, a nonsignificant decrease of 22% over the same time period.
whereas the risk was lowest for those in the west. During the This study is important because few others have reported
second time period, the risk in all three regions had decreased incidence rates based on population-level data, and none
to levels more comparable to or lower than those observed in reflect the results of analysis carried out through 2004, the lat-
the northeast (Table 2). est year for which NHDS data are available. The NHDS data
set is remarkable for its depth in terms of national represen- adopted new clinical guidelines and terminology, respectively,
tation of US short-stay hospitals and its breadth in providing for classifying the hypertensive disorders of pregnancy.2,38
annual data on hospital discharges since 1965. The incidence The American College of Obstetricians and Gynecologists’
rates we report for preeclampsia are similar to those reported 1996 Technical Bulletin provided stricter and more specific
from other population-based studies in the United States, definitions of preeclampsia than those previously recom-
Canada, and western Europe, range from 2 to 5%.7,8,11,18–21 mended, requiring that cases meet criteria for hypertension
Rates as low as 1.1% and as high as 7.5% have been reported in the presence of proteinuria after the 20th week of gesta-
from regional,22 clinical,23 and multicenter trials;24,25 however, tion. In 2002, the American College of Obstetricians and
the generalizability of these data is limited by the selective rep- Gynecologists adopted the recommendations of the National
resentation of hospitals and study populations.25,26 High Blood Pressure Education Program Working Group
It is interesting to note that our data indicate higher rates of on High Blood Pressure in Pregnancy to use the term “ges-
eclampsia than those reported from population-based stud- tational hypertension” rather than “transient hypertension of
ies in Canada, the United Kingdom, and western Europe, pregnancy” for a patient with hypertension with no evidence
were present in the first nine-year time period had dimin- 18. Basso O, Rasmussen S, Weinberg CR, Wilcox AJ, Irgens LM, Skjaerven R.
Trends in fetal and infant survival following preeclampsia. JAMA 2006;
ished, thereby suggesting that prevention efforts have become 296:1357–1362.
increasingly successful across the country. The reduction of 19. Dahlstrom BL, Engh ME, Bukholm G, Oian P. Changes in the prevalence of
risk in the south is most notable. pre-eclampsia in Akershus County and the rest of Norway during the past
35 years. Acta Obstet Gyn Scand 2006; 85:916–921.
The observed increases in the incidence of preeclampsia 20. Ros HS, Cnattingius S, Lipworth L. Comparison of risk factors for preeclampsia
and gestational hypertension represent important changes and gestational hypertension in a population-based cohort study. Am J Epidemiol
in the burden of maternal morbidity at the population level, 1998; 147:1062–1070.
21. Xiong X, Demianczuk NN, Saunders LD, Wang FL, Fraser WD. Impact of
raising both clinical and public health concerns. These study preeclampsia and gestational hypertension on birth weight by gestational age.
findings should serve as an important call for clinicians to Am J Epidemiol 2002; 155:203–209.
heighten their awareness of the increased population-level 22. Lawlor J, Osman M, Shelton JA, Yeh J. Population-based analysis of hypertensive
disorders of pregnancy. Hypertens Pregnancy 2007; 26:67–76.
risk for hypertensive disease originating in pregnancy. An 23. Eras JL, Saftlas A, Triche E, Hsu CD, Risch HA, Bracken MB. Abortion and its effect
increase in the risk for conditions as potentially dangerous on risk of preeclampsia and transient hypertension. Epidemiology 2000; 11:36–43.
as preeclampsia and eclampsia underlines the importance of