29 Maternal stress, anxiety, and depression during pregnancy: effects on the fetus and the child

The mental health of a woman can be important not only for herself but also for her child. This is especially so during pregnancy and the early postnatal period. While postnatal maternal depression can impair her mothering and thus affect the child via psychological mechanisms, changes in her mood during pregnancy can change her physiology, and this in turn can affect fetal development.

Symptoms of anxiety and depression are frequent during pregnancy. Indeed, they occur more frequently in late pregnancy than in the postpartum period (Heron et al. 2004). Lee et al. (2007), in a Hong Kong sample, found that more than half the women had symptoms of anxiety and one-third had symptoms of depression on at least one antenatal assessment. Pregnancy is also a period in which major life events cluster, as women, and their partners, begin to adjust to a major life transition. Several studies show that domestic violence in pregnancy is common (Chhabra 2007; Macy et al. 2007).

There is very strong evidence from animal studies that maternal stress in pregnancy has a long-term effect on the neurodevelopment of the offspring (Weinstock 2001). With animals it is possible to cross foster the prenatally stressed pups to control mothers after birth, and thus establish the timing of the exposure. Prenatal stress in animal models has been linked with a wide range of outcomes, including altered cerebral laterality and abnormal sexual behaviour (Weinstock 2001). However, the most widely reproduced effects are on cognition, including reduced memory and attention, and increased anxiety and emotional dysregulation. Work with non-human primates has identified brain structures altered by prenatal stress. For example, Coe et al. (2003) have shown that exposure to unpredictable noise, either early or late in pregnancy, resulted in reduced volume of the hippocampus in the offspring. This is a part of the brain that is important for memory. The responsiveness of the hypothalamic–pituitary–adrenal axis (HPA) which produces the stress hormone cortisol, was also increased in the offspring. Other experiments have shown that the effects of prenatal stress may be moderated and even reversed by positive postnatal rearing. This suggests that although there may be persisting effects of prenatal stress, it is not inevitable (Maccari et al. 1995). One notable result with the animal studies is that the effects of prenatal stress on male and female offspring are often different (Weinstock 2007). Learning deficits are more readily seen in prenatally-stressed males, while anxiety, depression, and increased response of the HPA axis to stress are more prevalent in females.

In this chapter we will discuss the evidence that maternal mood during pregnancy can affect the outcome for her child. We will particularly focus on neurodevelopmental outcomes that have implications for psychology and psychiatry, although there is evidence that antenatal maternal stress and anxiety can also have other adverse effects such as causing preterm labour (Homer et al. 1990; Wadhwa et al. 2001)

There is now good evidence that if a mother is stressed or anxious while pregnant, her child is substantially more likely to have emotional, behavioural, or cognitive problems, including an increased risk of symptoms of attention deficit/hyperactivity, anxiety, or language delay (Van den Bergh et al. 2005; Talge et al. 2007). These findings are independent of effects due to maternal postnatal depression and anxiety.

An immediate link between antenatal maternal mood and fetal behaviour is well established, from 27–28 weeks of pregnancy onwards (Van den Bergh et al. 2005). For example, if the mother carries out a stressful task such as mental arithmetic or the Stroop test, the heart rate of her fetus changes, especially in more anxious women (Monk et al. 2003). The mechanisms underlying this are not known.

An early observational study linking antenatal stress with longer term effects on the child was published by DH Stott in 1973. Information was collected from 200 women in Scotland in 1965–6, at the end of their pregnancy. The questions asked of the mother concerned her physical and mental health, the course of the pregnancy, and her social circumstances. The health, development, and behaviour of the child were followed for the next four years. Stott’s major conclusion was that stresses during pregnancy involving severe and continuing personal tensions, in particular marital discord, were closely associated with child morbidity in the form of ill health, neurological dysfunction, developmental delays and behaviour disturbance (Stott 1973). Another early study was that of Meijer (1985) which examined the outcome for two cohorts of boys, one group consisting of those born in the year of the Israeli Six-Day War and a second group born two years later. The children from the ‘war-exposed pregnancies’ had significant developmental delays and showed regressive behaviour. However, these studies, although pioneering, did not control for possible confounding factors such as smoking or postnatal mood.

In the last ten years, several independent prospective, and better controlled, studies have examined the effects of antenatal stress, anxiety, or depression on social/emotional and cognitive outcomes during childhood. Even though these studies used a wide range of different methods, both for measuring antenatal stress or anxiety, and for assessing the child, they all support a link between prenatal mood and the development of the fetal brain. These studies are mainly European with two in North America (Field et al. 2003; Laplante et al. 2004); none are from developing countries or countries at war, where one might predict that the effects would be even more marked.

A wide range of different outcomes can be affected by prenatal stress. Several studies have shown links between antenatal stress/anxiety and behavioural/emotional problems in the child. The most consistent adverse outcome is in symptoms of attention deficit hyperactivity disorder (ADHD) (O’Connor et al. 2002b; Van den Bergh and Marcoen 2004; Rodriguez and Bohlin 2005), but an increase in anxiety is also often observed (O’Connor et al. 2002b; Van den Bergh and Marcoen 2004). Other studies show an effect of prenatal stress or anxiety on the cognitive development of the child, as assessed by scores on the Bayley Mental developmental Index (MDI) (Huizink et al. 2003; Bergman et al. 2007) or language development (Laplante et al. 2004).

Three studies have shown an association between antenatal anxiety or stress and more mixed handedness in the child (Obel et al. 2003; Glover et al. 2004; Gutteling et al. 2007). Atypical laterality has been found in children with autism, learning disabilities, and other psychiatric conditions, including problems with attention as well as in adult schizophrenia. There is anecdotal evidence for a link between antenatal maternal stress and both autism and dyslexia, in addition to the evidence discussed already for ADHD (O’Connor et al. 2002b). It is an interesting possibility that many of these symptoms or disorders, which are associated with mixed handedness, share some neurodevelopmental components in common, which may be exacerbated by antenatal maternal stress or anxiety.

Different studies have examined the child at different ages, from newborn to adolescence. The newborn studies show effects that must be independent of postnatal experience; those with adolescents show the persistence of impairment (Van den Bergh et al. 2005; Talge et al. 2007). Two studies found impairment in the newborn using the Brazelton scale (Brouwers 2001; Field et al. 2002) and one study used the Prechtl neurological assessment (Lou et al. 1994). Field et al. (2002) reported that the newborns of mothers with high anxiety had greater relative right frontal brain activation (as measured by electroencephalogram) and lower vagal tone. The babies also spent more time in deep sleep and less time in quiet and active alert states, and showed more state changes and less optimal motor maturity, autonomic stability, and withdrawal. Lou et al. (1994) proposed there may be a ‘fetal stress syndrome’ analogous to the ‘fetal alcohol syndrome’, on the basis of their study which showed that antenatal life events resulted in a smaller head circumference, lower birthweight, and lower neurological scores on the Prechtl scale. O’Connor et al. showed a continuity of effects from four to seven years old (O’Connor et al. 2003), while Van den Bergh and colleagues (2007) have found links between antenatal anxiety and child depression in adolescence.

More needs to be understood about the exact period of gestation which is most important for all the effects described here. Different studies have found different periods of vulnerability. What is clear is that the effects are not confined to the first trimester. Although the basic body structures are formed early, the brain continues to develop, with neurons making new connections, throughout gestation, and indeed after birth. In the study of O’Connor et al. (2002b) anxiety was measured only at 18 and 32 weeks’ gestation, and the associations were stronger with the latter time point. It remains possible that the effects were actually maximal at midgestation, for example, about 24 weeks. It is also likely that the gestational age of sensitivity is different for different outcomes. Brain systems underlying different aspects of cognition or behaviour mature at different stages.

The size of the effects found in many of these studies is considerable, although it is important to emphasize that most children are not affected. In the general population study of O’Connor et al. we found that women in the top 15% for symptoms of anxiety at 32 weeks’ gestation had double the risk of having children with behavioural problems at four and seven years of age, even after allowing for multiple covariates (O’Connor et al. 2002b). It raised the risk for a child of this group of women having symptoms of ADHD, anxiety, depression, or conduct disorder, from 5% to 10%. These results imply that the attributable load in behavioural problems due to antenatal anxiety is of the order of 15% (Talge et al. 2007). Other studies have also found that antenatal maternal stress or anxiety accounted for 10–20% of the variance depending on the particular outcome studied (Laplante et al. 2004; Van den Bergh and Marcoen 2004; Bergman et al. 2007).

These are substantial effects, but there remains considerable variation across children. Bergman et al. (2007), for example, have found that although antenatal maternal stress increased the risk for both cognitive delay and raised anxiety, these did not necessarily occur in the same children. It is likely that the particular outcome affected depends on the timing of the stressor but also on the specific genetic vulnerabilities of the child. There is now some evidence for gene/environment interactions with respect to the postnatal development of psychopatholgy (Caspi et al. 2003); it is probable that the same occurs prenatally.

Some have suggested that the observed associations between maternal mood in pregnancy and child outcome are actually due to genetic factors or the postnatal environment; a mother prone to be stressed or anxious may pass these genes on to her child which in turn makes them more likely to be anxious or have other behavioural or cognitive problems. She may also show impaired postnatal parenting. Although both genetics and parenting certainly can contribute to child outcome the evidence suggests that there is also a direct contribution from antenatal maternal mood. The associations have been found with antenatal stress or anxiety independent of postnatal maternal mood or exposure to life events (e.g. O’Connor et al. 2002a; Bergman et al. 2007). If the association between mother and child were genetic one would predict as strong an association with postnatal factors. The animal evidence also shows that maternal stress during pregnancy can have long-term effects on fetal development, independent of the postnatal environment or genetics.

The evidence shows that the effects described earlier are not specific to one type of stress or anxiety. It also seems that many neurodevelopmental effects can be observed with relatively low levels of anxiety or stress (O’Connor et al. 2002b). Most of the studies have used maternal self-rating questionnaires, some using anxiety questionnaires, and others measures of stress (O’Connor et al. 2002b; Van Den Bergh and Marcoen 2004). Some studies assessed daily hassles (Huizink et al. 2003), whereas others focused on life events (Lou et al. 1994; Bergman et al. 2007), perceived stress (Rodriguez and Bohlin 2005), or pregnancy-specific worries. Some have used exposure to an external trauma, such as a severe Canadian ice storm (Laplante et al. 2004), the Chernobyl disaster (Huizink et al. 2007), or September 11 (Yehuda et al. 2005).

In contrast to most of the findings, one study has found, in cohort of financially and emotionally stable women, that there was a small but significant positive association between antenatal stress and both the mental and physical development of the child (DiPietro et al. 2006). The authors suggest that a small to medium amount of antenatal stress may actually be helpful for the development of the child, although this remains to be confirmed.

Little is known about the types of anxiety or stress which may be most harmful for fetal development. Generalized anxiety, panic, specific phobia, post-traumatic stress, acute stress, and obsessive–compulsive disorders may involve quite different, or even opposite, physiological processes (Tsigos and Chrousos 2002). It is notable that whereas maternal anxiety has been found to be associated with raised cortisol in the child (O’Connor et al. 2005), maternal exposure to the trauma of September 11 was found to be associated with low cortisol levels in the infant (Yehuda et al. 2005).

It is interesting that the life events found in one study to be most linked with both low scores on the Bayley MDI and increased fear reactivity, were ‘separation/divorce’ and ‘cruelty by the partner’ (Bergman et al. 2007). This finding is similar to the conclusion by Stott (1973) that continuing personal tensions (in particular, marital discord) were a particular risk factor for later ‘neurological dysfunction, developmental delays and behaviour disturbance’ in the child.

The high co-occurrence of symptoms of anxiety and depression raise questions about the specific predictions from maternal anxiety. There is some evidence that the effect on the child derives more from prenatal anxiety than depression. O’Connor et al. (O’Connor et al. 2002b) found that although antenatal depression was associated with child behavioural problems in a similar way to prenatal anxiety, the effect was smaller; furthermore, when prenatal anxiety was included as a covariant, the association with depression was not significant. In contrast, the prediction from prenatal anxiety to child behavioural problems was substantial and not reduced when prenatal depression was covaried. The authors also found that the link between antenatal anxiety and child behavioural problems was separate and additive to the effects of postnatal depression (O’Connor et al. 2002a).

Thus the current evidence suggests that the risk most closely linked with adverse child outcomes is maternal anxiety/stress. There is also evidence that the effects on the child are not restricted to extreme anxiety or stress in the mother, but can occur along a continuum of stress or anxiety (O’Connor et al. 2002a).

The mechanisms underlying these effects have only just started to be studied in humans. In animal models, increased fetal exposure to glucocorticoids such as cortisol has been found to be one mechanism for such fetal programming (Weinstock 2001; Schneider et al. 2002) although many other systems, such as those involving dopamine and serotonin, have also been shown to be involved.

In humans, maternal anxiety during pregnancy has been associated with reduced blood flow to the baby through the uterine arteries (Teixeira et al. 1999), but we do not know whether this is of clinical significance. We have started to test the hypotheses that maternal anxiety is associated with raised maternal cortisol, that this in turn is related to increased fetal exposure to cortisol, and that this can affect fetal neurodevelopment. The maternal HPA axis becomes desensitized to stress as gestation increases (Kammerer et al. 2002) and the association between maternal anxiety and cortisol level is weak, especially in the second half of gestation (Sarkar et al. 2006), so the first part of this hypothesis remains to be proven. We know that there is a strong correlation between maternal and fetal levels of cortisol (Sarkar et al. 2007), suggesting that there is passage of cortisol across the placenta, at least from 18 weeks’ gestation. This correlation is increased with higher maternal anxiety (Glover et al. 2009), suggesting that placental function can be affected by the emotional state of the mother, and regulate the amount of cortisol that reaches the fetus. However, much remains to be understood with respect to the mediating mechanisms.

The clinical implications of this research are that the emotional state of the woman during pregnancy should receive much more attention, both for the sake of the woman herself, and for the development of her future child. Most antenatal anxiety and depression are currently undiagnosed and untreated.

One small randomized, controlled trial of 16 sessions of interpersonal psychotherapy for antenatal depression has been conducted (Spinelli and Endicott 2003). It was found to be successful in reducing the depression, and the authors recommend it as a first line of antidepressant treatment during pregnancy. A meta-analysis of treatments for depression in both pregnancy and postpartum (Bledsoe and Grote 2006), conclude that a range of treatments can be effective, with cognitive behavioural therapy producing the largest effect sizes. A recent study has shown the efficacy of cognitive behavioural therapy in reducing antenatal anxiety (Austin et al. 2008).

Effective interventions to reduce maternal stress and/or anxiety during pregnancy should help to decrease the incidence of cognitive and behavioural problems in children. The evidence shows that it is not only extreme levels of stress and anxiety that need attention, but possibly the most affected 15% of the population. Also, it is not only clinically diagnosed mental illness that is important here, but stress due to other factors too, especially the relationship with the partner.