Materials and Methods
Since 1967 medical information on all births in Norway with gestational ages at least 16 weeks has been forwarded to the Medical Birth Registry of Norway. Thus, information on previous pregnancy outcomes is collected prospectively. The national identification number allows linkage of all women's subsequent deliveries. The population of pregnant women in Norway is relatively homogeneous. More than 99% of them receive standardized antenatal care.11 At each antenatal visit, blood pressure (BP) is measured and urine is examined for protein with a reagent strip. Medical data are collected in the pregnancy record brought by the women to the delivery unit, and selected items of data are transferred to the Registry notification form. By the ninth day postpartum, the form is sent to the Registry. Almost no changes have been made in the form since the start of the Registry.
The study was based on records of all births in Norway from 1967 through 1992 and comprised 779,642 women and more than 1,500,000 births. This file has been linked to population census files containing socioeconomic data from the Central Bureau of Statistics and the National Education Register. We excluded women who had only one delivery between 1967 and 1992 (n = 161,618), a first delivery before 1967 (n = 179,300), sibships with multiple births (n = 2457), and those without information on the first day of the last menstrual period (LMP) in at least one pregnancy (n = 65,893). That left 370,374 women with first and second pregnancies for study.
Data used in the present study were pregnancy outcome, obstetric complications, and sociodemographic characteristics. Gestational age was based on the first day of the LMP. Interpregnancy interval was calculated by subtracting the date of birth of the first pregnancy from the first day of the LMP in the second pregnancy. Data on education were collected from census files and the education register. Preeclampsia is usually recorded by the registry as a specific diagnosis. The notification form also contains information on separate symptoms of preeclampsia, such as hypertension, proteinuria, and edema. In Norway, the diagnosis is in accordance with the recommendations by the ACOG in 1972,12 which defined preeclampsia as increased BP after 20 weeks' gestation with proteinuria, edema, or both. According to clinical practice in Norway, pregnancies with edema but without proteinuria are not included in the definition.13 Hypertension is defined as BP of 140/90 mmHg or greater, or increased diastolic BP of at least 15 mmHg or systolic BP of at least 30 mmHg from average values before 20 weeks' gestation. Proteinuria is defined as excretion of 0.3 g or more per day, usually equivalent to at least 1+ on a urine reagent strip. Our case definition included all pregnancies with specified diagnoses of preeclampsia and pregnancies recorded with hypertension and proteinuria in combination and preeclampsia with chronic hypertension.
Preeclampsia in second pregnancies was identified and associations with birth weight and birth weight percentiles in first deliveries were assessed. Birth weight of the first infant was divided into gradations of 500 g (Table 1). To classify birth weight for gestational age we used birth weight ratio, ie, 100 * (the actual birth weight)/(the mean birth weight for that gestational age [completed weeks] and gender among the 370,347 first births). Birth weight ratios were divided into eight groups: cutoff points were equivalent to the second, third, tenth, 50th, 90th, 97th, and 98th percentiles (Table 2), ie, at birth weight ratios 70%, 74%, 83%, 100%, 117%, 126%, and 129%. The effect on preeclampsia was adjusted for education in years (fewer than 10, 10–12, more than 12, and unknown), maternal age in years (19 or less, 20–29, 30–34, and 35 or more), chronic hypertension, diabetes mellitus, change of partner, year of birth (1967–80 and 1981–92) in the second pregnancy, and interpregnancy interval in years (under 0.5, 0.5–3.9, and at least 4) (Table 2).
The relative risk (RR) of preeclampsia in the second pregnancy was used to estimate the contribution of birth weight percentiles in the first pregnancy. To assess the independent contribution of birth weight adjusted for gestational age, RRs were estimated by adjusted odds ratios (OR), using logistic regression analysis (Statistical Package for the Social Sciences; SPSS Inc, Chicago, IL). To avoid assumptions of linear relationships in the logistic regression analyses, the independent variables were represented with indicator variables. Confidence intervals (CI) for proportions of preeclampsia were calculated by the score method.14
To estimate the proportion of all cases of preeclampsia attributable to histories of birth weight below the tenth percentile and preeclampsia in the first pregnancy, adjusted population-attributable risk percentages were calculated as (proportion of exposed among all cases) * (RR − 1)/RR, where RR is the adjusted relative risk.15 Adjusted ORs obtained from logistic regression were used to calculate approximate adjusted RRs.16
In first pregnancies, preeclampsia occurred in 3.5% of women (12,489 of 370,374) whereas in second pregnancies, the incidence rate was 1.6% (6054 of 370,374). In first pregnancies, preeclampsia was superimposed on chronic hypertension in 1% of the cases (100 of 12,489) and 2% (112 of 6054) in second pregnancies. In women without preeclampsia in their first pregnancies who delivered infants under 1500 g, the risk of having preeclampsia in their second pregnancies was about 3.5%, compared with 1.1% if their infants were at least 3500 g (Table 1). Also in women at risk of recurrent preeclampsia, the risk depended strongly on birth weight of first deliveries. After delivery of infants of at least 3500 g, recurrence risk was about 10%, whereas recurrence risk after delivery of infants under 1000 g was 33.5% (Table 1). In women without preeclampsia in their first pregnancies who delivered infants whose weight was less than the second birth weight percentile, the risk of having preeclampsia in second pregnancies was 3.1% (95% CI 2.7%, 3.6%) (Table 2), compared with 1.2% (95% CI 1.1%, 1.3%) for birth weights in at least the tenth percentile. After delivery of an infant whose birth weight was below the second and in at least the tenth percentile, recurrence risk was 19.6% (95% CI 17.5%, 21.9%) and 12.2% (95% CI 11.5%, 12.8%), respectively.
Those absolute risks can also be expressed as RRs. After adjusting for education, maternal age, chronic hypertension, diabetes mellitus, year of birth, interpregnancy interval, change of partner, and education, the increased risk of preeclampsia in women who delivered infants small for gestational age persisted. Women without preeclampsia in their first pregnancies who delivered infants whose weight was below the second percentile were 3.1 times more likely to have pre-eclampsia in second pregnancies (adjusted OR 3.1; 95% CI 2.5, 3.7) than those who delivered infants whose weight was between the 90th and 96.9th percentiles (Table 2). Under the third percentile, adjusted OR was 2.6 (95% CI 2.1, 3.0). The adjusted ORs of recurrent preeclampsia with delivery of infants under the second and third percentiles were 22.6 (95% CI 18.8, 27.3) and 21.2 (95% CI 17.8, 25.2), respectively.
Birth weights below the tenth percentile in the first delivery accounted for 606 (10.0%) of the total 6054 cases of preeclampsia in the second pregnancy (adjusted population attributable risk 10.0%) and 501 (29.7%) of 1687 recurrent cases. Of all cases, 1547 were attributable to histories of preeclampsia in the first pregnancy (adjusted population attributable risk 25.6%).
Based on the total Norwegian birth population, the study was most likely not affected by selection bias. Information on previous pregnancy outcomes was collected prospectively, thereby avoiding recall bias. The rate of preeclampsia in second pregnancies was consistent with earlier studies,8 but the rate of 3.5% in first pregnancies was slightly low compared with hospital-based studies17 and identical to the rate reported in Sweden,18 and in accordance with data from the National Hospital Discharge Survey,19,20 reporting an overall incidence of preeclampsia of about 2.5%.
Some of the effects of LBW and small for gestational age fetuses on later preeclampsia might be explained by shared risk factors for FGR and preeclampsia that were not considered in the present study, such as thrombophilia, which might compromise fetomaternal circulation.21 However, most established risk factors,17,22–24 such as short stature and overt chronic renal disease, are weak or uncommon and would not influence the association essentially.
Because of lack of data, we were not able to include excessive weight gain and prepregnancy weight in the study, which are strong and common risk factors for preeclampsia25 and tend to increase fetal weight.26 Adjusting for maternal weight or weight gain likely would increase the effect of LBW. Our database did not include smoking status. It is well known that smoking is a strong and common risk factor for LBW. However, smoking seems to be negatively associated with the diagnostic criteria of preeclampsia (proteinuria and hypertension).24,25 To the extent that such a negative association occurs, adjusting for smoking would increase rather than decrease the effect of LBW on later preeclampsia.
Women who delivered infants whose weight was below the third birth weight percentile were about three times more likely to have both initial and recurrent preeclampsia than those who delivered infants at least in the tenth percentile. Those findings are consistent with the hypothesis that a shared etiologic factor or recurrent pathophysiologic mechanism causes FGR in first pregnancies and preeclampsia in the next. Consistent with this, occlusive changes in the spiral arteries, likely caused by inadequate trophoblast invasion early in pregnancy, have been found in preeclampsia and normotensive FGR.2,3 Some cases of FGR might differ from preeclampsia only in the maternal response to a shared placental pathology. It is postulated that materials produced by the poorly perfused placenta enter the maternal circulation. Those materials result in endothelial activation, clinically manifested as the systemic maternal syndrome of preeclampsia.4 Our findings also agree with those of earlier studies that preeclampsia,7–10 FGR,5 and smallness for gestational age6 tend to recur. Our results also were consistent with those of a British study8 that reported an effect of birth weight: women who had preeclampsia and LBW infants (under 2500 g) in their first pregnancies had double the incidence of preeclampsia in their second pregnancies.
The present study confirmed earlier observations7–10 that a history of preeclampsia is the strongest predictor of preeclampsia among parous women, with 10–15 times higher recurrence risk. Twenty-six percent of all cases were attributable to history of preeclampsia. Tables 1 and 2 might be interpreted as indicating the positive predictive value of LBW or birth weight ratios (percentiles) and preeclampsia. The predictive value of birth weight under 1000 g and preeclampsia is as high as 34%, and the predictive value of birth weight below the second percentile and preeclampsia was 20%.
A history of LBW for gestational age was found in a substantial proportion of all cases of preeclampsia and seems to be important in its etiology. After adjustment, previous delivery of an infant whose birth weight was below the tenth percentile accounted for 10% of all cases and 30% of recurrent cases of preeclampsia.
1. Long PA, Abell DA, Beischer NA. Fetal growth retardation and pre-eclampsia. Br J Obstet Gynaecol 1980;87:13–8.
2. Robertson WB, Brosens I, Dixon HG. The pathological vascular response of the vessels of the placental bed to hypertensive pregnancy. J Pathol Bacteriol 1967;93:581–92.
3. Khong TY, De Wolf F, Robertson WB, Brosens I. Inadequate maternal vascular response to placentation in pregnancies complicated by pre-eclampsia and by small-for-gestational age infants. Br J Obstet Gynaecol 1986;93:1049–59.
4. Roberts JM. Endothelial dysfunction in preeclampsia. Semin Reprod Endocrinol 1998;16:5–15.
5. Patterson RM, Gibbs CE, Wood RC. Birth weight percentile and neonatal outcome: Recurrence of intrauterine growth retardation. Obstet Gynecol 1986;68:464–8.
6. Bakketeig LS, Hoffman HJ, Harley EE. The tendency to repeat gestational age and birth weight in successive births. Am J Obstet Gynecol 1979;135:1086–103.
7. More MP, Redman CW. Case-control study of severe pre-eclampsia of early onset. BMJ 1983;287:580–3.
8. Campbell DM, MacGillivray I, Carr-Hill R. Pre-eclampsia in second pregnancy. Br J Obstet Gynaecol 1985;92:131–40.
9. Sibai BM, Mercer B, Sarinoglu C. Severe preeclampsia in the second trimester: Recurrence risk and long-term prognosis. Am J Obstet Gynecol 1991;165:1408–12.
10. Lie RT, Rasmussen S, Brunborg H, Gjessing HK, Lie-Nilsen E, Irgens LM. Fetal and maternal contributions to the risk of pre-eclampsia: Population based study. BMJ 1998;316:1343–7.
11. Backe B. Studies on antenatal care
[doctoral thesis]. The University of Trondheim, Trondheim, Norway: Tapir, 1994.
12. National High Blood Pressure Education Program Working Group. High blood pressure in pregnancy. Am J Obstet Gynecol 1990;163:1691–712.
13. Øian P, Henriksen T, Sviggum O. Norwegian Society of Obstetrics and Gynecology. Clinical guidelines in obstetrics. Hypertension in pregnancy. In: Dalaker K, Berle EJ, eds. Oslo, Norway: The Norwegian Medical Association, 1999;101–2.
14. Vollset SE. Confidence intervals for a binomial proportion. Stat Med 1993;12:809–24.
15. Rockhill B, Newman B, Weinberg C. Use and misuse of population attributable fractions. Am J Publ Health 1998;88:15–9.
16. Zhang J, Yu KF. What's the relative risk? A method of correcting the odds ratio in cohort studies of common outcomes. JAMA 1998;280:1690–1.
17. Cunningham GF, MacDonald PC, Gant NF, Leveno KJ, Gilstrap LC III, Hankins GDV, et al. Hypertensive disorders in pregnancy. In: Gant NF, ed. Williams obstetrics. 20th ed. Norwalk, Connecticut: Appleton and Lange, 1997:693–744.
18. Cnattingius S, Mills JL, Yuen J, Eriksson O, Salonen H. The paradoxical effect of smoking in preeclamptic pregnancies: Smoking reduces the incidence but increases the rates of perinatal mortality, abruptio placentae, and intrauterine growth restriction. Am J Obstet Gynecol 1997;177:156–61.
19. Saftlas AF, Olson DR, Franks AL, Atrash HK, Pokras R. Epidemiology of preeclampsia and eclampsia the United States, 1979–1986. Am J Obstet Gynecol 1990;163:460–5.
20. Samadi AR, Mayberry RM, Zaidi AA, Pleasant JC, McGhee N Jr, Rice RJ. Maternal hypertension and associated pregnancy complications among African-American and other women in the United States. Obstet Gynecol 1996;87:557–63.
21. Kupferminc MJ, Eldor A, Steinman N, Many A, Bar-Am A, Jaffa A, et al. Increased frequency of genetic thrombophilia in women with complications of pregnancy. N Engl J Med 1999;340:9–13.
22. Baird D. Epidemiological aspects of hypertensive pregnancy. Clin Obstet Gynecol 1977;4:531–48.
23. Wen SW, Goldenberg RL, Cutter GR, Hoffmann HJ, Cliver SP. Intrauterine growth retardation and preterm delivery: Prenatal risk factors in an indigent population. Am J Obstet Gynecol 1990;162:213–8.
24. Eskenazi B, Fenster L, Sidney S. A multivariate analysis of risk factors for preeclampsia. JAMA 1991;266:237–41.
25. Rasmussen S, Øian P. Smoking, hemoglobin concentration and pregnancy-induced hypertension. Gynecol Obstet Invest 1998;46:225–31.
26. Rasmussen S, Øian P. First- and second-trimester hemoglobin levels: Relation to birth weight and gestational age. Acta Obstet Gynaecol Scand 1993;72:246–51.