Depressive symptoms in adulthood and intrauterine exposure to pre-eclampsia: the Helsinki Birth Cohort Study

Authors


K Räikkönen, University of Helsinki, PO Box 9, 00140 University of Helsinki, Finland. Email katri.raikkonen@helsinki.fi

Abstract

Please cite this paper as: Tuovinen S, Räikkönen K, Kajantie E, Pesonen A, Heinonen K, Osmond C, Barker D, Eriksson J. Depressive symptoms in adulthood and intrauterine exposure to pre-eclampsia: the Helsinki Birth Cohort Study. BJOG 2010;117:1236–1242.

Objective  We studied whether pre-eclampsia predicts depressive symptoms in offspring.

Design  Retrospective longitudinal cohort study.

Setting  Participants in the Helsinki Birth Cohort 1934–44 Study.

Population  We classed 788 women and men born at term after a normotensive, hypertensive or pre-eclamptic pregnancy, by using the mother’s blood pressure and urinary protein measurements, at maternity clinics and birth hospitals.

Methods  Linear and logistic regression analyses. We made adjustments for the mother’s age and body mass index (BMI) at delivery, the participant’s body size at birth/length of gestation, sex and childhood socio-economic status, age and educational attainment at testing.

Main outcome measures  Beck depression inventory (BDI) scores completed twice, at the ages of 60 and 63 years.

Result  Participants born after a primiparous pregnancy complicated by pre-eclampsia had over 30% (P < 0.04) higher depressive symptom scores in adulthood compared with those born after a primiparous normotensive pregnancy. We found no evidence of the association between pre-eclampsia and depression among participants born after multiparous pregnancies. Gestational hypertension and depressive symptoms were not significantly associated. The models adjusting for mother’s age and BMI at delivery, the participant’s body size at birth/length of gestation, sex, childhood socio-economic status, age and educational attainment at testing did not change the results.

Conclusion  Pre-eclampsia is associated with later depressive symptoms in individuals born at term and after a primiparous pregnancy. These findings are compatible with the adverse fetal ‘programming’ by a suboptimal prenatal environment.

Introduction

Mounting empirical evidence suggests that an increased risk for depression and for its subclinical symptoms may lie in a suboptimal prenatal environment. Indeed, low birthweight, shorter length of gestation, or both, which reflect suboptimal prenatal conditions, have been shown to predict clinical diagnosis of depression in adolescents,1,2 and depressive symptoms in populations of various ages.3–10 Furthermore, recent studies have emphasised that the risk for depression11 and mood disorders12 may be particularly increased in individuals born both preterm and small for gestational age. Although contradictory findings also exist,13–17 the findings tip in the positive direction, and thus lend credence to the developmental origins of health and disease (DOHaD) hypothesis.18 According to this hypothesis, a suboptimal prenatal environment may permanently alter organ structure and function, and render an individual susceptible to diseases in subsequent life. Yet, the mechanisms through which these prenatal influences operate remain largely unknown.

Pre-eclampsia is a pregnancy disorder characterised by elevated blood pressure and proteinuria that occurs after 20 weeks of pregnancy, complicating 3–5% of all pregnancies.19 In addition to posing a serious threat to maternal wellbeing, pre-eclampsia also poses a threat to the wellbeing of the fetus. It has been estimated that nearly 15% of all preterm births are caused by pre-eclampsia,20,21 and it is among the key reasons for restricted fetal growth.22,23 Pre-eclampsia could thus serve as an experimental platform to study the underlying prenatal ‘programming’ mechanisms. Here, we report the long-term consequences of pre-eclampsia on depressive symptoms in offspring, measured at the ages of 60 and 63 years among men and women who participated in the Helsinki Birth Cohort Study (HBCS).

In order to disentangle the interconnections between pre-eclampsia and the consequent shorter length of gestation/lower birthweight, we tested the associations among individuals born at term. Indeed, pre-eclampsia occurring later compared with earlier in pregnancy may be less severe and differ in etiology.21 Furthermore, as pre-eclampsia occurs more frequently and may be qualitatively different in primi- than in multiparous pregnancies,24 we tested the associations between pre-eclampsia and depressive symptoms separately in those born after primi- and after multiparous pregnancies.

Methods

Participants

The HBCS includes 13 345 men and women who were born as singletons between 1934 and 1944 in one of the two maternity hospitals in Helsinki (the University Central Hospital and the City Maternity Hospital), who attended child welfare clinics in the city, and who were still living in Finland in 1971, by which time a unique personal identification number had been assigned to each resident of the country. The study population has been described in detail elsewhere.25,26

In 2001–2004, hereafter referred to as period 1, a randomly selected sample of 2003 people of the initial population participated in a clinical examination,9 including the beck depression inventory (BDI).27 In 2004, hereafter referred to as period 2, a more detailed psychological survey, again including the BDI, was sent to the randomly selected population, with 1735 returning the questionnaire.9 Of these, 981 participants in period 1 (453 men and 528 women; 49.0%) and 815 participants in period 2 (349 men and 466 women; 47.0%) belonged to the subpopulation with data from medical records on maternal blood pressure and urinary protein tests during pregnancy for the diagnosis of pre-eclampsia: 817 participants for period 1 and 684 participants for period 2 were born at term (between 37 and 42 weeks of gestation). Of them, 29 and 23, respectively, were born to mothers suffering from chronic hypertension, and they were excluded from the analyses. Our analytic sample thus consisted of 788 and 661 women and men born at term after normotensive pregnancies, or pregnancies complicated by hypertension or pre-eclampsia (see the definition of pre-eclampsia below), respectively, for period 1 and period 2. The included and excluded participants had a similar maternal weight, height and parity; the participants that were included had a 90.7 g higher birthweight, their mothers were 0.7 years younger and had 6% more often a father who was a manual worker (P < 0.03).

The Ethics Committee of the National Public Health Institute and the Ethics Committee of Public Health and Epidemiology at Helsinki and Uusimaa Hospital District approved this project, and all participants gave written informed consent.

Measures

Pre-eclampsia

For the identification of pre-eclampsia, data of the mothers’ blood pressures and the results of urinary protein tests were recorded after 20 weeks of pregnancy at antenatal clinics or at the birth hospital. Based on this information, we defined four groups of mothers: (1) those who had severe pre-eclampsia, with proteinuria ‘+’ (∼1 mg/ml of albumin),28 and a systolic blood pressure ≥160 mmHg or a diastolic pressure ≥110 mmHg; (2) those who had non-severe pre-eclampsia, with proteinuria ‘+’ and a systolic pressure ≥140 mmHg or a diastolic pressure ≥90 mmHg in the absence of severe pre-eclampsia; (3) those who had gestational hypertension, with a systolic pressure of 140 mmHg or more, or a diastolic pressure of 90 mmHg or more but no proteinuria; (4) those who were normotensive, with neither systolic pressure attaining 140 mmHg or more, nor diastolic pressure attaining 90 mmHg or more. These definitions are described in detail elsewhere.29 The two groups of mothers with severe and non-severe pre-eclampsia were combined for the purposes of this study. These definitions are consistent with the National High Blood Pressure Education Program Working Group on High Blood Pressure in Pregnancy (NHBPEP) 2000 criteria,29,30 with two exceptions: first, we considered one high blood pressure measurement to be sufficient for diagnosis because our data did not allow as to require two separate measurements, as the NHBPEP criteria do; second, our data included only a qualitative measurement of proteinuria ‘+’.28,30

Other neonatal, maternal and adult characteristics

Data on the date of birth and birthweight (g) of offspring, the mother’s height (cm), weight (kg), age, parity, and the date of the last menstrual period were extracted from hospital birth records. The fathers’ occupational status (manual workers, lower middle class, upper middle class) was derived from the birth records. Adult height (cm) and weight (kg) were measured, and the level of educational attainment (elementary school, vocational school, senior high school and college/university degree) was determined in conjunction with the clinical examination in 2001–2004. BMI was calculated as weight divided by height2 (kg/m2).

Depressive symptoms

The BDI27 was used to evaluate the severity of depressive symptoms. The BDI consists of 21 items assessing symptoms of depression during the previous 2 weeks. Each item contains four statements reflecting the varying degrees of symptom severity. Respondents are instructed to circle the number that corresponds with the statement that best describes them, ranging from zero to three, indicating increasing severity. Ratings are summed to calculate a total BDI score that can range from 0 to 63. Although the BDI is designed to screen but not diagnose depression, there are commonly used and well-defined cut-off points: none or minimal depression is <10; mild to moderate depression is 10–18; moderate to severe depression is 19–29; and severe depression is 30–63 points.27

Statistical analysis

We used multiple regression analyses to test associations between pre-eclampsia and depressive symptoms (BDI scores) in period 1 and period 2, and their average, as the outcomes. The BDI scores were log transformed to attain normality, with regression coefficients reverse transformed and expressed as the percentage change in the dependent variable for each unit change in the independent variable. Logistic regression analyses were used with the BDI scores categorised (BDI score >10 for mild depressive symptoms as the cut-off). The associations were adjusted for factors known to be associated with an increased risk of pre-eclampsia and/or depression as follows: the crude model adjusted for mother’s age31 and BMI31 at delivery, father’s occupational status in childhood,31 participant’s sex,27 age at testing9 and level of education attained in adulthood.9 A further model adjusted for birthweight and for length of gestation,9,22,33 and length of gestation.9,20,21 We also tested if any potential associations between pre-eclampsia and depression varied by birthweight, sex or father’s occupational status in childhood. The analyses were conducted separately in primi- and in multiparous pregnancies.

Results

Table 1 presents characteristics of the sample according to pre-eclampsia status. Pre-eclampsia was associated with a higher BMI of the mother and lower birthweight of the babies. Pre-eclamptic pregnancies were more often primiparous. In this sample pre-eclampsia did not associate with other neonatal characteristics or father’s occupational status.

Table 1.   Characteristics of the sample
 Offspring born to pregnant mothers with
Normotension n = 513Hypertension n = 239PPre-eclampsia n = 36P
Mean (SD)/n (%)Mean (SD)/n (%)Mean (SD)/n (%)
Gender
Men232 (45)120 (50)0.2019 (53)0.38
Women281 (55)119 (50)17 (47)
At birth
Length of gestation (days)278.6 (8.5)278.6 (8.6)0.98276.3 (7.7)0.12
Birthweight (g)3454.8 (452.3)3371.4 (472.8)0.023159.2 (526.6)<0.01
Mother’s age28.0 (5.1)29.2 (5.8)<0.0127.6 (5.4)0.60
Mother’s BMI26.3 (2.8)26.9 (3.2)0.0227.7 (2.5)<0.01
Year of birth1940.1 (3.2)1940.5 (2.8)0.111939.3 (2.8)0.13
Parity
Primiparous237 (46)125 (52)0.2528 (78)0.02
Multiparous276 (54)114 (48)8 (22)
Father’s occupation
Manual worker343 (68)180 (76)0.0623 (64)0.89
Lower middle class115 (23)39 (17)9 (25)
Upper middle class48 (9)17 (7)4 (11)

Shorter length of gestation and low birthweight (≤2.5 kg) were associated with an increased risk of depressive symptom scores (P < 0.01). Women and those with lower educational attainment in adulthood reported higher depressive symptom scores. Depressive symptoms were not associated with parity, maternal age or BMI at delivery, father’s occupational status in childhood or participant’s age at testing (P > 0.27).

Table 2 shows that after adjustments for mother’s age and BMI at delivery, father’s occupational status in childhood, participant’s sex, age at testing and level of education attained, the depressive symptom scores were more than 30% higher in period 1 and period 2 among participants born after primiparous pregnancies complicated by pre-eclampsia compared with the participants born after primiparous normotensive pregnancies. Adding birthweight adjusted for gestation to the model had little effect on the results (Table 2). When length of gestation was added to the model the results were slightly attenuated (period 1, P = 0.06; period 2, P = 0.04; average of period 1 and period 2, P = 0.08). When we applied the mild cut-off criterion for the BDI scores (>10 versus ≤10), as compared with those born after primiparous normotensive pregnancies, the participants born after primiparous pregnancies complicated by pre-eclampsia had an odds ratio of 3.7 for having depressive symptom scores above the cut-off (Table 3). Further adjustment taking birthweight for gestation into account had little effect on the results (Table 3). When length of gestation was added to the model the results were slightly attenuated (P = 0.05).

Table 2.   Comparison of the depression symptom scores in offspring born at term to mothers with normotensive versus pre-eclamptic pregnancies
 nModel AModel B
Percentage difference (95% CI)PPercentage difference (95% CI)P
  1. Both models are adjusted for gender, father’s occupational status, mother’s age at delivery, mother’s BMI, age at clinical questionnaires and level of education attained; additionally, model B is adjusted for birthweight for gestational age.

Normotension (referent) versus pre-eclampsia
Primiparous
BDI in period 1237 vs 2836.9 (2.0 to 71.8)0.0436.5 (1.1 to 71.9)0.04
BDI in period 2203 vs 2252.2 (7.8 to 96.6)0.0249.1 (3.8 to 94.4)0.03
Average BDI201 vs 2139.7 (1.6 to 77.7)0.0438.4 (–0.4 to 77.1)0.05
Multiparous
BDI in period 1276 vs 8−15.8 (−71.5 to 40.0)0.58−13.4 (−69.3 to 42.5)0.69
BDI in period 2237 vs 5−58.9 (−142.2 to 24.4)0.17−54.8 (−138.7 to 29.0)0.20
Average BDI236 vs 5−41.7 (−109.2 to 25.8)0.23−38.8 (−106.9 to 29.2)0.26
Table 3.   Comparison of the risk for having depressive symptom scores above at least mild severity in offspring born at term to mothers with normotensive versus pre-eclamptic pregnancies
Model AModel B
 Odds ratio (95% CI)POdds ratio (95% CI)P
  1. BDI scores ≥10 indicate at least mild severity, and are compared with scores of <10 that are indicative of no to low severity.

  2. Both models are adjusted for gender, father’s occupational status, mother’s age at delivery, mother’s BMI, age at clinical questionnaires and level of education attained; additionally, model B is adjusted for birthweight for gestational age.

Normotension (referent) vs pre-eclampsia
Primiparous
Average BDI vs ≥103.7 (1.2 to 11.3)0.023.8 (1.2 to 11.9)0.02
Multiparous
Average BDI vs ≥100.9 (0.1 to 8.6)0.920.8 (0.1 to 8.3)0.86

Pre-eclampsia was not associated with depressive symptom scores in multiparous pregnancies (Tables 2 and 3). Gestational hypertension did not have any significant effects on depressive symptoms (P > 0.34; data not shown).

Analyses testing whether birthweight, sex or father’s occupational status modulated any associations between pre-eclampsia and depressive symptoms revealed no significant results (P > 0.19).

Discussion

Our findings demonstrate that pre-eclampsia occurring in term primiparous pregnancies is associated with higher depressive symptoms in offspring several decades later. The depressive symptom scores were more than 30% higher, and the odds ratio for the risk of having depressive symptom scores above the lower cut-off point, indicating mild severity, was over 3.7 among those born at term after primiparous pregnancies complicated by pre-eclampsia, compared with those born at term after primiparous normotensive pregnancies. Pre-eclampsia did not predict depressive symptoms among participants born at term after multiparous pregnancies. Gestational hypertension and depressive symptoms were not significantly associated. The associations were not confounded by factors associated with the risk for pre-eclampsia and/or depressive symptoms, namely mother’s age and BMI at delivery, socio-economic status in childhood, level of education achieved in adulthood and age at testing. Adjustments for birthweight and length of gestation, consequences of pre-eclampsia, had little effect on the results.

In one recent study, pre-eclampsia was found to reduce internalising behaviours, including withdrawal and anxious/depressed behaviours and somatic complaints at ages 5 and 8 years, whereas gestational hypertension was found to be associated with higher internalising behaviours at age 14.32 Our findings are thus not in line with these intriguing recent findings in children. However, we are unaware of any previous studies testing associations between pre-eclampsia and depressive symptoms in the late adult life of offspring. Our findings thus provide novel information on the prenatal origins that may pose a risk for depressive symptoms in subsequent life. Our findings also show that although pre-eclampsia is known to be associated with lower birthweight and shorter length of gestation, a finding also evident within our sample, birthweight and length of gestation attenuated the associations only by a little. The conclusion from these findings is that pre-eclampsia adds to the list of known risk factors for depressive symptoms, and that birthweight and length of gestation may only partially be the mediating pathway.

We restricted our analysis to subjects born at term. Restriction of the sample to those born at term was deemed necessary because of practical and empirical reasons. First, very few subjects were born premature after pregnancies complicated by pre-eclampsia (n = 4 out of 52; three after primiparous and one after multiparous pregnancies). Second, pre-eclampsia at earlier and at later stages of pregnancy differ in symptoms, severity and etiology.21,33 Third, as premature birth and restricted intrauterine growth are among the frequent consequences of pre-eclampsia, disentangling their unique contribution to the offspring development is more complex in premature than term births. Furthermore, testing associations separately in those born after primiparous and multiparous pregnancies was deemed necessary, as there is evidence that pre-eclampsia may be qualitatively different in primiparous and multiparous pregnancies.24,34 However, it is important to note that we found that pre-eclampsia was more common among primi- (11%, n = 29) than multiparous pregnancies (3%, n = 8). Thus, we clearly had more statistical power for detecting differences among the primi- than among the multipara. Although the overall prevalence rate of pre-eclampsia (41/981, 4.2%) in the current study fell within the range of prevalence rates reported in previous studies,19,35 the associations need to be re-tested in bigger samples in order to obtain further insight into the associations between pre-eclampsia and the mental health of offspring.

Our results are compatible with the DOHaD theory.18 Being a placental disorder that may induce fetal malnutrion and oxidative distress, pre-eclampsia may compromise the fetal developmental milieu. Pre-eclampsia is also associated with the reduced function of the placental 11–β–hydroxysteroid-dehydrogenase–2 (11βHSD–2),36,37 an enzyme that catalyses the conversion of maternal circulating cortisol to inactive cortisone. Thus, in pre-eclampsia, the fetus may become overexposed to maternal circulating glucocorticoids. All of these influences have the potential to ‘program’ the body’s organ structure and function, and physiological feedback systems, the hypothalamic–pituitary–adrenocorticoid system in particular, which is highly relevant to the biology underlying depression.38 The extent to which these mechanisms explain the associations remains an unanswered question. A further possibility is that a common genetic predisposition underlies the associations: maternal prenatal depression may increase the risk of pre-eclampsia,39 and pre-eclampsia33 and depressive symptoms40 are at least moderately heritable.

Limitations

Only half of the cohort had antenatal records after 20 weeks of pregnancy, and were therefore eligible for the present study. Their mothers tended to be slightly shorter and younger, more of them were primiparous, and more were married to manual workers. These differences would, however, introduce a bias only if the association between hypertensive disorders in pregnancy and depressive symptoms in offspring were different in people who have antenatal records as compared with those who do not. This seems unlikely but cannot be excluded. Moreover, our data did not allow us to require two elevated blood pressure measurements to establish diagnosis.30 However, the overall prevalence rate of 4.2% is consistent with the commonly reported rates of 3–5%. We have no data on the complications of pre-eclampsia, including eclampsia, which at that time complicated 0.6% of all pregnancies at Helsinki University Central Hospital,41 and could be associated with more severe consequences for offspring. Subjects who died before 1971 were excluded from our cohort. Loss of follow up across decades is also inevitable. However, over 86% of the randomly selected, still traceable cohort members were available for the psychological survey. A higher proportion of women than men agreed to participate in the follow up. Participation in the follow up was not, however, related to any of the neonatal or maternal characteristics. Furthermore, it is possible that obstetric complications predispose the mother to postnatal depression, which in turn may pose a risk for depression in the offspring by hereditary mechanisms or by compromising the early attachment relationship. We do not have data on maternal depression, and therefore cannot rule out this possibility. Neither do we have data available on maternal stress and smoking during pregnancy. Some evidence suggests that stress during pregnancy may favour the development of pre-eclampsia,42 whereas smoking during pregnancy may protect against pre-eclampsia.43 Furthermore, stress44,45 and smoking during pregnancy are associated with less optimal developmental outcomes in the offspring. Hence, we cannot rule out that stress and smoking during pregnancy may, to some extent, explain our findings. Finally, our study focussed on depressive symptoms. Therefore, our findings may not generalise to more severe forms of depression. Neither does our study determine if the participants have suffered from depression earlier in life, and been treated.

Conclusion

We conclude that pre-eclampsia is associated with depressive symptoms several decades later in offspring born at term after a primiparous pregnancy. These findings are in line with the DOHaD theory, suggesting that susceptibility to depression has its origins in the fetal developmental milieu.

Disclosure of interests

The authors declare no conflict of interest.

Contribution to authorship

Conception and design, or acquisition of data, or analysis and interpretation of data: ST, KR, EK, AKP, KH, CO, DJPB and JGE. Drafting the article or revising it critically for important intellectual content: ST, KR, EK, AKP, KH, CO, DJPB and JGE. Final approval of the version to be published: ST, KR, EK, AKP, KH, CO, DJPB and JGE.

Details of ethics approval

The Ethics Committee of the National Public Health Institute and the Ethics Committee of Public Health and Epidemiology at Helsinki and Uusimaa Hospital District approved this project, and all participants gave written informed consent. The institutional review board project approval #79/08/95 was obtained on 28 June 1995 and 31 January 2001.

Funding

This study was supported by grants from the Academy of Finland, European Science Foundation (EuroSTRESS), University of Helsinki, the British Heart Foundation, the Finnish Foundation of Cardiovascular Research, the Finnish Diabetes Research Foundation, the Finnish Medical Society (Duodecim), Finska Läkaresällskapet, the Päivikki and Sakari Sohlberg Foundation, the Juho Vainio Foundation, theYrjö Jahnsson Foundation, the Signe and Ane Gyllenberg Foundation, the Jalmari and Rauha Ahokas Foundation, the Emil Aaltonen Foundation, the Finnish Ministry of Education and the Finnish Foundation for Paediatric Research.

Ancillary