Methods
We searched Medline, Google Scholar, and Science Direct databases on original research and review articles with the search words “hypertensive pregnancy disorder” or “preeclampsia” or “gestational hypertension” and “mental disorder” or “psychiatric,” “schizophrenia” or “depression” or “bipolar” or “anxiety” or “autism” or “eating disorder” or “substance use disorder” or “ADHD” or “conduct disorder” or “personality disorder”. We also examined reference lists of the identified articles for additional references. We focused our search on articles published since 2015. The corresponding author went through the search results and excluded duplicates, narrative and systematic reviews, and other studies not providing any data on our study question. We also excluded studies focusing solely on the symptoms of mental and behavioral disorders.
Two authors (RR, MLP) conducted a quality of evidence assessment of the new original findings according to the Newcastle-Ottawa Scale (NOS) assessment criteria. The evaluated studies were cohort, cross-sectional, and case-control studies, each rated according to the criteria appropriate for the particular study type [
24,
25]. The NOS scales for case-control and cohort studies yield a maximum of nine stars and the scale for cross-sectional studies a maximum of ten stars. A higher number of stars indicate higher methodological quality. In one cohort study, one evaluator, MLP, was an author, and hence to avoid bias, RR conducted this NOS assessment together with AL. In cases of disagreement in assessment, we reached consensus by discussion.
Supplementary Table
1, Supplementary Table
2, and Supplementary Table
3 in the Online Data Supplement provide our specific assessment criteria for the study questions at hand, which we predefined before the start of the assessment and systematically applied in duplicate to all studies. To sum, we assessed the statistical methods based on whether the study used sibling comparisons and whether the study accounted for familial confounding by maternal and/or paternal mental disorders, took into account cardiometabolic conditions of maternal prepregnancy overweight/obesity and/or diabetes disorders, and considered the mediating or moderating effects of preterm and/or SGA birth. Additional assessment criteria included whether maternal hypertensive pregnancy disorder and/or offspring mental and behavioral disorder diagnoses were physician-diagnosed from structured interviews, medical records, or health registers vs. retrospective self- or maternal self-reports of diagnosis. We assessed the representativeness of the exposed and selection of the non-exposed groups, attrition bias, and the adequacy of the length of follow-up for the child to develop the outcome in question.
Discussion
The findings of the recent meta-analyses and cohort studies consistently point to the predisposing effects of maternal hypertensive pregnancy disorders and especially preeclampsia on offspring mental and behavioral disorders in childhood. The expanding evidence base includes findings among altogether millions of participants. Findings from cross-sectional and case-control studies, in turn, are very inconsistent, but notably, the same studies have had important limitations in methodological quality.
Hence, several cohort studies and meta-analyses yield a coherent picture of replicated associations between maternal hypertensive pregnancy disorders and increased risk of mental and behavioral disorders in children. The same increasing body of evidence suggests that these effects of maternal hypertensive pregnancy disorders are independent of maternal overweight/obesity and diabetes disorders and familial confounding by maternal or paternal mental disorders. However, only the Swedish population-wide studies on ASD and ADHD and the case-control study in Iran on ASD assessed familial confounding more soundly via comparisons of differentially exposed siblings [
9,
10,
42], and no sibling comparison data exists on other mental and behavioral disorders than ASD or ADHD. Furthermore, while preterm and SGA birth have emerged as possible moderators or partial mediators of the effects of hypertensive pregnancy disorders on offspring mental and behavioral disorders [
8‐
11], mediation or moderation by preterm birth was not addressed in any of the sibling comparisons [
9,
10,
42]. Although maternal hypertensive pregnancy disorders consistently predicted increased risks of ASD and ADHD, the effect sizes for these most commonly studied disorders were relatively small in the most representative studies. Maternal hypertensive pregnancy disorders thus constitute one of many risk factors for these neuropsychiatric disorders, with small but significant effect sizes. Interestingly, the authors of the Swedish cohort studies later showed that offspring risks of ASD and ADHD were even higher if both the grandmother and mother had had preeclampsia, suggesting multigenerational effects [
54], and a novel avenue for research.
While there are numerous studies on mental and behavioral disorders in childhood and adolescence, and meta-analytic evidence of associations between maternal preeclampsia on offspring schizophrenia in adulthood, very few studies have examined the effects on other adulthood mental disorders. An early cohort study showed that maternal gestational hypertension but not preeclampsia predicted an increased risk of severe mental disorders in adult offspring [
55]. Two case-control studies reviewed here had adulthood follow-ups, one on major depression, schizophrenia, and bipolar disorder [
43] and the other on eating disorders [
44]. These studies produced mixed findings in a restricted number of exposed individuals [
43,
44]. Hence, no clear conclusions can be made of effects on other adult mental disorders. Also regarding child and adolescent mental disorders, the studies have either focused on ADHD, ASD, any mental disorder, psychological development disorders, childhood behavioral and emotional disorders, anxiety, and depression as outcomes. In contrast, our literature search yielded no studies specifically on conduct disorders, personality disorders, or substance use disorders. Thus further research needs to examine how widespread the effects of maternal hypertensive pregnancy disorders are on different mental and behavioral disorders, particularly on externalizing disorders.
An additional question of the effects of hypertensive pregnancy disorders on offspring mental and behavioral disorders is whether dose-response associations exist, i.e., the effects become more evident when the hypertensive pregnancy disorder is more severe. According to the ICD-10, preeclampsia can be classified according to its severity to mild/moderate and severe subtypes [
23]. Some of the international guidelines for the treatment of hypertensive pregnancy disorders do not recommend the use of the severity classification in clinical practice as all preeclampsia cases can have dire consequences for the mother and her child [
56]. However, the severity classification, dose-response effects, may provide important insights on potential causality. Three studies since 2015 assessed preeclampsia severity effects on offspring mental and behavioral disorders [
11,
38,
46]. In PREDO, the more severe the maternal preeclampsia, the higher the offspring risk for childhood mental disorders [
11]. Also, severe but not mild/moderate preeclampsia had effects that were independent of maternal early pregnancy BMI and diabetes disorders [
11]. One study included in the meta-analyses on ASD defined severe preeclampsia as either a note of severe preeclampsia on a medical record, presence of HELLP syndrome, or preeclampsia combined with placental insufficiency [
46]. This exposure was associated with strong effects on ASD and developmental delay [
46]. In contrast, in the Israeli cohort study, preeclampsia severity was not associated with offspring ASD or eating disorders [
38].
While maternal hypertensive pregnancy disorders have now repeatedly shown effects on offspring mental and behavioral disorders that are independent of maternal diabetes and/or prepregnancy obesity, only one study assessed additive effects of these three types of cardiometabolic conditions [
11]. In that study, maternal hypertensive pregnancy disorders, diabetes disorders, and overweight/obesity in current pregnancy additively increased the risk of mental and behavioral disorders in children. While the cumulative incidence of childhood mental disorders was 7% among offspring of women with no maternal adverse cardiometabolic conditions in pregnancy, it was over 22% among offspring of women with all of these conditions [
11]. Further studies are needed to replicate these findings.
The evidence of preterm birth, SGA birth, and low birth weight partially mediating the effects of preeclampsia on offspring mental and behavioral disorders [
8,
11] suggests partially shared biological pathways underlying the effects of these conditions and maternal hypertensive pregnancy disorders on offspring mental health. Preeclampsia is a placental disorder characterized by placental insufficiency and SGA is often used in research as a proxy for placental insufficiency [
9,
10,
46]. Placental insufficiency and structural changes are associated with offspring psychopathology risk [
46,
57] and these placental modifications may be among the biological pathways leading from hypertensive pregnancy disorders, particularly preeclampsia, to offspring psychopathology risk. Furthermore, preterm birth predicts an increased risk of mental and behavioral disorders [
11,
35], possibly via structural and functional alterations in brain development [
58,
59]. Such neurodevelopmental alterations may contribute to the effects of maternal hypertensive pregnancy disorders on offspring mental and behavioral disorders [
60].
Maternal hypertensive pregnancy disorders may also increase the risk of offspring mental disorders via maternal and offspring changes in the inflammatory system and hypothalamus-pituitary-adrenal axis functioning. Such changes have been shown as a consequence of maternal hypertensive pregnancy disorders and in offspring with mental disorders [
60‐
63]. On a molecular level, there may be pleiotropic genetic effects between maternal hypertensive pregnancy disorders and offspring mental and behavioral disorders and epigenetic changes may mediate these associations. The genetic risk factors for mental disorders and hypertension partially overlap [
64] and epigenetic DNA methylation and gene expression changes are seen in offspring of women with hypertensive pregnancy disorders [
65] and patients with mental disorders [
62].
The findings reviewed here suggest a possible independent role for maternal hypertensive pregnancy disorders in the etiology of offspring mental and behavioral disorders. Considering the marked effects maternal hypertensive pregnancy disorders also have on maternal and offspring cardiovascular and neonatal morbidity and mortality, the public health impact of these common conditions is marked and widespread. Together, these findings indicate that the pharmaceutical and lifestyle interventions that have either proven effective or show promise on the treatment of maternal hypertensive pregnancy disorders [
56] also may have buffering effects on the somatic and mental health of the mother and her offspring.
The limitations of the available evidence include the case-control and cross-sectional studies not fulfilling most criteria to ensure unbiased reporting related to the definitions of exposures and outcomes, comparability of selection of cases and controls, and controlling for key covariates. In contrast to the cohort studies, which classified hypertensive pregnancy disorders according to standardized international diagnostic guidelines, the case-control and cross-sectional studies most often used retrospective maternal self-report questionnaires and did not specify the diagnostic criteria they used for hypertensive disorders (Table
2 and Supplementary Table
4 in the Online Data Supplement). The retrospective self-reports are prone to bias, which may limit the validity of the diagnostic categories and the generalizability of the findings of these studies. However, in the current review, the method of exposure assessment was accounted for in the NOS Quality of Evidence assessment (Table
2 and Supplementary Table
1, Supplementary Table
2, and Supplementary Table
3 in the Online Data Supplement). Furthermore, the large-scale studies on maternal hypertensive pregnancy disorders and offspring mental and behavioral disorders have been conducted in relatively affluent societies, and it remains uncertain how generalizable the findings are to different populations with varying healthcare coverage and guidelines [
66]. The guidelines for the treatment of hypertensive pregnancy disorders vary across countries [
66], and how this affects the prognosis of offspring born from pregnancies complicated by hypertensive pregnancy disorders remains unknown. For example, the US treatment guidelines suggest induced delivery after 34 gestational weeks in pregnancies complicated by preeclampsia, while there is no such recommendation in Europe [
1]. Studies on the similarities and differences of the associations of hypertensive pregnancy disorders with offspring mental and behavioral disorders in different countries are needed.
Furthermore, ethical reasons prohibit randomized controlled trials on the effects of maternal hypertensive pregnancy disorders on offspring mental and behavioral disorders. It is important to note that causality cannot be directly inferred from the epidemiological studies reviewed here. The prospective cohort studies nevertheless yield preliminary answers about the direction of associations. Future studies may approximate a causal design by examining in randomized clinical trials whether interventions that have proven effective for maternal hypertensive pregnancy disorders also prevent mental and behavioral disorders in the offspring. It also remains unknown whether the effects of maternal hypertensive pregnancy disorders on offspring mental and behavioral disorders are modified by familial risk for psychopathology. To our knowledge, no studies have examined interaction effects of polygenic risk scores or parental mental disorders with maternal hypertensive pregnancy disorders on offspring mental disorders. Studies using sibling comparisons while simultaneously taking into account all key confounders, mediators, and moderators will shed important new light on possible familial confounding. Furthermore, while two meta-analyses [
15•,
19•] and one original research study [
11] examined the specific effects of maternal gestational hypertension, chronic hypertension, and preeclampsia on offspring mental and behavioral disorders, most of the new research studies have focused either on preeclampsia as a sole exposure or on the combined effects of gestational hypertension and preeclampsia on offspring mental and behavioral disorders. More research is needed on the roles played by other maternal hypertensive pregnancy disorders. Finally, further studies should examine more thoroughly the effects of maternal hypertensive pregnancy disorders on offspring mental and behavioral disorders in adulthood and externalizing disorders at any age.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.