Abstract
Maternal mental health is an important public health issue because of its effects not only on the mother's well-being and functional status, but also her relationship with her partner and the development of her children. There is accumulating evidence of the adverse sequelae of maternal anxiety on fetal development, obstetrical complications, pregnancy outcomes such as low birth weight, and subsequent child development. Evaluation of maternal anxiety and intervention to reduce these symptoms, may ensure optimal developmental outcomes, particularly in high-risk infants such as those born at very low birth weights. This article will outline recent advances in our understanding of the etiology, assessment and impact of maternal anxiety, and describe intervention strategies to promote maternal well-being.
Pregnancy and childbirth are among the major life cycle transitions in a woman's life. This transition is accompanied not only by dramatic changes in virtually every organ system in the body, but also by changes in social role definitions, self-concept, and relationships with partners, parents, friends and coworkers. While our societal expectations are that pregnancy and childbirth are times of joy and well-being, for a significant number of women this life stage is associated with considerable distress. Maternal mental health has become an area of increasing concern for researchers, clinicians and public policy-makers since evidence has been accumulating that the consequences of perinatal mental illness are not limited to the suffering of the affected women, but extend to marital quality, the mother–infant relationship and the developing child. Much of the research in this domain has focused on postpartum depression. More recently, maternal anxiety has become the subject of extensive investigation. Evidence is accumulating that maternal anxiety, both pre-natally and postpartum, has a significant impact on fetal development, obstetrical outcomes, the mother-infant relationship and infant development. In addition, prenatal anxiety is strongly predictive of postpartum depression [1,2]. Assessment of maternal anxiety and intervention to reduce these symptoms may ensure optimal developmental outcomes, particularly in high-risk infants such as those born at very low birth weights (VLBWs). This article will outline recent advances in our understanding of the impact of maternal anxiety and describe intervention strategies to promote maternal well-being.
Defining maternal anxiety
One of the challenges in determining the prevalence and consequences of maternal anxiety lies in its definition. The term ‘anxiety’ may encompass a wide variety of constructs ranging from clinical diagnosis to self-report measures of symptoms to more general measures of stress. It is important to note that there is significant comorbidity between anxiety and depressive disorders such as major depression and bipolar disorder, with many depressed patients also suffering from anxiety symptoms [3]. Indeed, it is often difficult to differentiate between anxiety and depression since they both have many symptoms in common, and the two types of disorders reflect dysregulation of the hypothalamic–pituitary-adrenal (HPA) axis [4]. The limited data on childbearing women who meet diagnostic criteria for anxiety disorders indicate prevalence rates ranging from approximately 1–4% (panic disorder and obsessive compulsive disorder) to 8.5% (generalized anxiety disorder) [5]. Elevated symptoms of anxiety have been reported in 15–16% of pregnant women and 8–9% of postpartum women [6]. Rates of post-traumatic stress disorder (PTSD) in pregnant and postpartum women range from 2 to 7.7% [7–10]; some studies report that up to 18% of childbearing women have PTSD symptoms [11,12]. While a diagnosis of an anxiety disorder is more prevalent during pregnancy than postpartum, one study found that women were more likely to obtain treatment in the postpartum period [13]. Moreover, remission was more likely among women with a new onset of a disorder, as compared with those with a history of psychiatric disorder.
In measuring anxious symptomatology, it is important to distinguish between ‘state’ and ‘trait’ anxiety [14]. State anxiety refers to a person's momentary or situational anxiety, which can vary according to time and setting. Trait anxiety is defined as a person's more stable, characteristic anxiety level that may reflect a heritable personality profile [15]. In the face of stressful life circumstances, most people may experience some state anxiety. On the other hand, individuals high in trait anxiety may be predisposed to appraise situations as stressful and to react to them more negatively.
A further distinction has been drawn between pregnancy-related versus more generalized anxiety [16]. A woman's concerns about the health of the fetus and the experience of labor and delivery may contribute to unique variance in explaining mental health as well as pregnancy outcomes [17].
The timing of the assessment of anxiety is also of importance. Anxiety during pregnancy may have differential effects, depending on the stage of development of the fetus [18]. Postpartum anxiety may influence mother–infant interaction, which in turn may affect the child's social, emotional and cognitive development.
Maternal anxiety: who is at risk?
Sociodemographic factors, such as lower levels of educational attainment and living alone, have been found to be associated with anxiety in childbearing women, as is past history of mental health problems [13,19]. Previous adverse pregnancy outcomes including a history of pregnancy loss, obstetrical complications and preterm birth are also related to symptoms of anxiety, both general and pregnancy-related [19–22].
In addition to self-reported stress, physiological indicators of stress such as Cortisol levels have been examined in relation to maternal anxiety. The findings are inconsistent, with some studies showing no association of Cortisol levels with pregnancy-related anxiety [15,23] or general anxiety [4], and another finding a negative correlation with morning Cortisol levels but a positive one with evening levels [24]. Pluess and colleagues reported a significant negative correlation of Cortisol levels and trait anxiety in pregnant women [15], and in a small pilot study Seng and colleagues found that women with symptoms of PTSD had lower Cortisol levels [25] The complex association between biological and psychological measures, as well as the differential effects that may be associated with various types of diagnoses or symptomatology, remains to be elucidated in further research.
Prenatal anxiety
Effects on the course of pregnancy
There is accumulating evidence of the adverse sequelae of prenatal anxiety on fetal development, obstetrical complications and pregnancy outcomes such as low birth weight. Women with higher levels of anxiety during pregnancy are likely to report more physical symptoms, make more frequent obstetrical visits and miss more days of work; they also require more pain relief during labor and delivery [18]. Psychological distress during pregnancy may also be associated with poorer self-care, including poor nutrition and substance use [26,27]. These types of behaviors may have a deleterious affect on pregnancy outcomes. Interestingly, a recent study by Martini and colleagues indicated that it was self-perceived stress, not a diagnosis of an anxiety disorder that was associated with adverse pregnancy outcomes such as preterm delivery and obstetrical complications [28]. A systematic review by Littleton and colleagues found little evidence for associations between general or pregnancy-specific maternal anxiety and a range of obstetrical complications; there were small effect sizes for length of labor and 5 min Apgar scores [29]. The quality of the research studies, many of which had small sample sizes, used diverse measures of symptomatology and assessments at different points in time during pregnancy, makes it difficult to draw definitive conclusions.
Maternal symptoms of anxiety including those that are specifically pregnancy related have been found to be associated with preterm birth [22,30,31]. This is particularly true for mothers reporting severe anxiety throughout the course of their pregnancies [32]. Elevated levels of corticotropin-releasing hormone in more anxious mothers have been implicated in the relationship between anxiety and gestational age at birth [33]; however, the correlations are modest at best. Diego and colleagues showed that maternal anxiety was associated with lower fetal weight via elevated Cortisol levels [34]. They hypothesize that anxiety is associated with dysregulation of the HPA axis, which in turn results in elevated Cortisol levels that may affect fetal growth, either by reducing blood flow to the fetus or by hyper-activation of the fetal HPA that may result in increased movements and calorie expenditure. Given the lack of consistency in results concerning the relationship between Cortisol and anxiety, this hypothesis remains to be confirmed.
Effects on the child
There is some evidence for the effect of prenatal anxiety on infant health, with higher anxiety being predictive of more respiratory illness; pregnancy-specific hassles (an indicator of maternal stress) were associated with digestive problems and general ill health [35]. There is a growing body of research on the impact of prenatal anxiety on child behavior and development, with temperament and emotional problems showing the clearest links to maternal anxiety [36]; associations have been found with more difficult infant temperament [37], infant negative behavioral reactivity [38], lower state organization in male neonates [39], sleep problems at 6–30 months of age [40], more behavioral and emotional problems at ages 4–5 [41,42], impulsive and externalizing disorders such as attention deficit disorder at ages 8–9 [43], and depressive symptoms in adolescent daughters [44]. Prenatal anxiety has been found to contribute unique variance to behavioral outcomes even when controlling for maternal depression [42]. One study has reported some benefits of prenatal anxiety in a well-functioning community sample of women [45], in that the children of more anxious mothers exhibited better motor and cognitive development at 2 years of age. On the other hand, negative attitudes towards the pregnancy were associated with poorer motor development, attention and emotional regulation. The authors acknowledge that the levels of anxiety in the study sample were quite low, but they suggest that mild levels of stress in utero may prepare the infant to be more adaptable to the postnatal environment. By contrast, maternal negativity during pregnancy may have had enduring effects on the mother–child relationship, and in this way had an adverse effect on child development. A lack of emotional involvement with the infant during pregnancy is likely to persist in the postpartum period [46].
How do we account for the impact of prenatal anxiety on child outcomes? One prevailing theory implicates prenatal programming of elevated stress reactivity as a result of exposure to high levels of maternal Cortisol [47]. In support of such a mechanism there is evidence for long-term impact of prenatal anxiety on Cortisol levels in children up to 10 years of age [48]. The resulting dysregulation of the HPA axis is associated with negative emotional and behavioral outcomes in infants and children. Further evidence that the functioning of the maternal HPA axis may affect fetal and infant development comes from research showing that prenatal exposure to selective serotonin reuptake inhibitors was associated with Cortisol reactivity in 3-month-old infants [49]. As noted above, the evidence is mixed with regard to the relationship between maternal anxiety and Cortisol levels. Moreover, the focus on prenatal programming neglects to account for shared genetic factors that might affect both maternal and child behavior [37]. Maternal anxiety disorders diagnosed during pregnancy are associated with child diagnoses of anxiety disorders [28]. As Pluess and colleagues have argued, maternal predisposition to heightened stress reactivity may be a personality trait with a significant degree of heritability [15]. Moreover, there is evidence of differential genetic susceptibility to the effects of maternal prenatal anxiety: research by Oberlander and colleagues has found that the effect of maternal anxiety during the third trimester of pregnancy on child emotional development was moderated by the child's serotonin transporter promoter (SLC6A4) genotype [49]. In children with reduced serotonin expression (short alleles), prenatal maternal anxiety predicted greater child anxiety at 3 years, while children with two long alleles exhibited aggressive behaviors. It is evident that there is a complex interplay of genetic, hormonal and environmental factors that must be considered in order to understand the impact of maternal anxiety on child developmental outcomes.
Postpartum anxiety & the mother-infant relationship
After delivery, maternal anxiety may affect maternal behavior and the mother–infant relationship. Anxious mothers can be more intrusive and controlling, more overprotective and yet less able to parent their infants sensitively (i.e., to interpret and respond appropriately to the infant s communicative cues such as vocalizations, eye contact and body movements [50–52]). Anxiety affects the ability to attend to and process emotional information [53]; as a result, anxious mothers may be less sensitive [54] and more intrusive in interaction with their infants [55]. In rodent models, dams bred for high anxiety show more protective behavior and are highly motivated to retrieve their pups even in the face of aversive obstacles [56]. Whether this model may translate into maternal overprotection in human mothers is an empirical question.
Mothers of prematurely-born and low-birth weight infants often exhibit high levels of anxiety. Anxiety in mothers of VLBW (<1500 g) infants can affect mother–infant interaction in the neonatal intensive care unit (NICU) and at 9 and 24 months corrected age [57–59], in that more anxious mothers exhibited less sensitive and responsive behavior and provided less structure during free play. At 24 months the children showed poorer cognitive development and more internalizing behavior problems [60].
Hormonal factors have been proposed to account for the effects of maternal anxiety on mother–infant interaction. For example, the neuropeptide oxytocin (OT) may have an important role in the regulation of both mammalian social behaviors and emotional reactivity [61]. OT is made in and acts on the brain, especially in regions such as the hypothalamus that are involved in emotions and social relations [62,63]. OT level correlates with positive social behavior; it is released during positive social interactions and may facilitate the ability to be trusting [64], generous [65,66], socially perceptive [67] or to feel safe and relaxed. Elevated OT may also be an indicator of interpersonal distress, particularly in the relationship with a primary partner [68,69]. Such distress might be characterized by anxiety in relationships and perceptions that relationships are lacking in warmth. The role of OT is well established in labor, birth and lactation. Animal models show the involvement of OT in maternal affiliative behavior such as licking/grooming and suggest that the effects of prenatal stress on maternal behavior may be mediated by OT receptor binding [70]. Recent human research has shown that individual differences in plasma OT levels are highly stable across pregnancy and postpartum, and are significantly correlated with maternal interactive behaviors such as gaze, vocalizations, positive affect, affectionate touch and checking behavior, and maternal mental representations of attachment [71,72].
The release of OT during pregnancy may have an effect on maternal mood. For example, recent research suggests that there is a relationship between OT and mood in humans. For example, OT rises during lactation and many women report that during breastfeeding they experience increased calmness, a more positive mood state and reduced emotional responsiveness to stressful life events. This in turn makes breastfeeding easier [73]. Mezzacappa and Katkin compared self-report measures of stress and mood in mothers who either breastfed or bottle-fed [74]. Over 1 month, women who breastfed reported significantly less stress than mothers who bottle-fed. In addition, mothers who both breastfed and bottle-fed were found to display a greater decrease in negative mood after breastfeeding compared with bottle-feeding. Researchers have attributed these positive feelings to the release of OT coinciding with milk letdown. The release of OT during breastfeeding is believed to have many physiological effects on the body indicative of reduced stress, such as reduced systolic blood pressure, increased cardiac vagal tone and short-term reduction in systolic and diastolic blood pressure [73,75,76]. Thus, OT can directly affect maternal care-giving behavior, and also act indirectly by reducing anxiety and thereby promoting maternal care [56]. This is of particular interest in relation to mothers of premature infants, who are less likely to breastfeed and therefore do not have the benefits of the anxiolytic properties of OT release [77]. Women who experience emotional abuse and neglect in childhood have been found to have lower levels of OT, which in turn were associated with elevated symptoms of anxiety [78]; we may speculate that this may result in intergenerational transmission of maladaptive parenting behavior. The investigation of OT in relation to maternal anxiety and parenting behavior holds promise, both in terms of explanatory models as well as possible treatments involving the administration or induction of OT release.
Intervening with anxious mothers
Given the deleterious effects of anxiety in childbearing women on the health and well-being of both mothers and their infants, the need for effective treatments is evident. In fact, it appears that anxious women may be less likely than those suffering from depression to seek treatment for their mental health concerns [79,80]. Such women may be embarrassed to disclose their concerns or may not know whom to consult [80]. There have been few studies of treatments specifically targeting either prenatal or postpartum anxiety. Anxiety symptoms are sometimes included as secondary outcomes in treatment studies of perinatal depression. For example, a study of telephone-based peer support to prevent postpartum depression assessed anxiety as well, but there was no significant effect of the intervention on this outcome [81]. Psychopharmacological treatment is often unacceptable to pregnant and lactating women owing to their concerns about medication effects on the fetus and infant [82]. Moreover, such treatment might not be appropriate for women who are experiencing symptoms of anxiety but do not have a clinical diagnosis of an anxiety disorder [83]. Nonetheless, such women might benefit from other forms of support and intervention; for example, there is evidence that exercise helps to reduce anxiety in pregnant women [84]. There is limited evidence of sustained benefits to maternal mental health of different forms of individual and group psychotherapy, including cognitive behavioral therapy and interpersonal psychotherapy [83]. From a public health perspective, it would be important to disseminate accurate information about the relative risks of psychopharmacological treatment for anxiety and depression as compared with the risks of untreated anxiety for the developing fetus and infant. Efforts to destigmatize mental health problems in general, and in childbearing women in particular, might promote greater willingness to obtain appropriate services, as well as adherence to effective medical treatments.
Several intervention programs have been designed to reduce psychological distress, including but not limited to anxiety, in mothers of preterm infants. Melnyk and colleagues sought to reduce parental distress by teaching parents of infants born at weights under 2500 g how to cope with stressful aspects of the NICU experience [85]. Parents learned about specific stressors in the NICU environment and about strategies to cope with them, as well as ways to be involved in their infant's care. Intervention group mothers reported fewer symptoms of anxiety when the infant was 2 months corrected age. Telephone support from an experienced mother of a VLBW infant has also been shown, in a small sample, to reduce levels of stress, depression and state anxiety, although not trait anxiety [86].
Our group has developed and tested an intervention specifically designed to reduce anxiety and promote sensitivity in mothers of VLBW infants [87]. The Cues Program is a brief, six-session intervention, implemented during the infant's NICU hospitalization, which employs empirically-based techniques from the domains of cognitive behavioral therapy and parent sensitivity training and is a unique combination of two components: training in anxiety reduction strategies and sensitivity. In one-to-one sessions with a trained intervener, mothers are taught to read their own cues and recognize signs of anxiety; to utilize a number of strategies to reduce their anxiety, including muscle relaxation, imagery and cognitive reframing; to read their infant's communication cues; and to respond sensitively to infant cues and distress. The program targets mothers of VLBW infants, who are at greater biological risk than heavier premature infants born at weights below 2500 g, and compares the intervention to an attention-control condition, in order to determine whether it is specific skills or nonspecific attention that help to reduce maternal anxiety. In the attention-control or care condition, mothers are given an equal number of sessions with an intervener. However, the content of the sessions is restricted to general information about infant care, including such topics as immunization and sleep position. A randomized, controlled trial of this intervention demonstrated that mothers in both groups exhibited high levels of anxiety at baseline and significantly reduced levels of anxiety at the immediate postintervention follow-up when the infant was 6–8 weeks corrected age, but there was no difference between the two groups [88]. These results suggest that nonspecific attention and information can be just as beneficial as specific anxiety-reduction skills training for mothers during the stressful NICU hospitalization. The key to success of an early intervention to reduce anxiety in mothers whose infants are in the NICU may be the availability of a supportive intervener who reaches out to mothers and who provides information and reassurance. Future research needs to include both attention control and treatment as usual comparison groups in order to determine the crucial components of the intervention.
Conclusion
Anxiety in childbearing women is a significant concern, owing to its implications for the health and well-being of mothers and their children. It is essential to incorporate genetic, hormonal and psychosocial factors in studying both the determinants and the consequences of maternal anxiety. In order to better inform the development of effective approaches to treatment, research must carefully distinguish between personality dispositions and situational determinants, as well as between subclinical manifestations of anxious symptoms and clinical diagnoses of such syndromes as generalized anxiety disorder, panic disorder and PTSD. Not all mothers respond to adversity with psychological distress [30]. For example, while most women would find the birth of a preterm infant to be a stressful life circumstance, there is evidence to show that symptoms of anxiety are not necessarily related to the severity of neonatal morbidity [57,89]. Identifying the most vulnerable women, as well as the factors that make them vulnerable, will permit the development of more targeted intervention strategies. There is also a need to reduce the stigma associated with seeking treatment for mental health problems during pregnancy and postpartum in order to promote optimal outcomes for mothers and infants.
Future perspective
The field of perinatal mental health has grown exponentially in the past 25 years. The initial research focused on postpartum depression, its psychosocial and hormonal determinants, as well as its course and treatment. A natural extension of this work has looked to earlier determinants of postpartum distress, by studying women during pregnancy, and to other syndromes including anxiety. While anxiety and depression are often comorbid, a better understanding of the causes and consequences of these disorders in childbearing women requires that they be defined and measured with greater accuracy and consistency. Much research in perinatal mental health focuses on psychological distress, which can include depression, anxiety and other indicators of stress. As a result, it may be difficult to compare results across study populations. The use of physiological measures such as cortisol is one means to adopt more objective measures of anxiety. Given the mixed results of research using cortisol, there is a need to explore other biological measures (e.g., OT) and to standardize and validate self-report measures as well since situational and psychological factors must continue to be considered. This is highlighted by the fact that pregnancy-related anxiety seems to be a good predictor of maternal outcomes.
Genetic and epigenetic studies will be important avenues for future research. Studies that draw the distinction between state and trait anxiety already point to the need to consider genetic predispositions to heightened reactivity to stressful life events, which may also be transmitted from mother to child. The investigation of genetic polymorphisms may also provide important information regarding susceptibility to adverse events, as well as amenability to different treatment modalities. Recent epigenetic research in animals showing that stress during pregnancy can affect neuroendocrine systems that are related to maternal care-giving behavior [70], demonstrates the importance of integrating biological and environmental perspectives.
Executive summary
Anxiety manifests itself in psychological symptoms as well as in a variety of clinical diagnoses, such as generalized anxiety disorder, panic disorder and post-traumatic stress disorder.
Approximately 15% of pregnant and postpartum women report symptoms of anxiety, while clinical diagnoses are present in up to 9% of pregnant and postpartum women.
It is important to distinguish between situational determinants and personality factors in assessing anxiety.
The timing of the assessment, during pregnancy or postpartum, is important in understanding the determinants and sequelae of anxiety.
Low socioeconomic status and past history of mental illness are associated with anxiety in childbearing women.
Obstetrical complications and pregnancy loss are related to higher levels of maternal anxiety.
Physiological indicators of anxiety such as cortisol have not been consistently related to maternal anxiety.
There is some evidence that anxiety is associated with adverse obstetrical outcomes including preterm birth.
Dysregulation of the hypothalamic–pituitary–adrenal axis is thought to be one mechanism whereby maternal anxiety affects fetal growth and the early onset of labor.
Infants born to mothers with elevated levels of anxiety during pregnancy have more physical health problems, more difficult temperaments, sleep disturbances, and behavior and emotional problems in childhood and adolescence.
Prenatal programming of stress reactivity has been proposed as one possible mechanism to account for the impact of prenatal anxiety on child outcomes.
Genetic factors may determine which infants are more likely to be affected by prenatal anxiety.
Some exposure to prenatal–maternal anxiety may promote adaptive behavior in infants and children.
Anxious mothers may be less sensitive and more intrusive and overprotective in interacting with their infants and young children.
The neuropeptide oxytocin, which has anxiolytic properties, is implicated in labor and lactation.
Oxytocin may play a role in lowering anxiety in breastfeeding women and promoting maternal care-giving behavior.
Oxytocin also promotes maternal care-giving behavior and may in the future be the basis for novel treatments of maternal anxiety.
Childbearing women are unlikely to seek professional help for anxiety symptoms.
Psychopharmacological treatments may be unacceptable to pregnant and lactating women.
Psychosocial interventions have been designed to reduce maternal anxiety but have produced mixed results.
Research must carefully distinguish between personality dispositions and situational determinants of maternal anxiety.
Generalizability of research results will be enhanced if studies adopt standard measures of anxiety symptoms, including both psychological and physiological indicators.
Genetic and epigenetic studies will further our understanding of the causes and consequences of maternal anxiety.
The current state of knowledge has demonstrated associations between maternal anxiety and negative outcomes; however, it is important to avoid ‘blaming the victim’ and there is a need to place these research findings in context.
We hope that future research will avoid a ‘blame the victim’ mentality in its examination of the implications of maternal anxiety. While it is important to investigate the biological, emotional and behavioral sequelae of anxiety, the evidence suggests associations and not ultimate causality. Many factors influence obstetrical outcomes, fetal development and the mother–infant relationship, some of which may mitigate the effects of maternal anxiety. As some researchers have argued [45], exposure to moderate levels of prenatal anxiety may confer some benefits on the developing infant. A comprehensive approach to studying maternal anxiety and its sequelae must refrain from raising undue concerns in young mothers. Nonetheless, mothers experiencing symptoms of anxiety should be encouraged to consult with their healthcare providers, who may be best able to evaluate them and offer reassurance as well as appropriate treatment, in order to reduce their suffering and promote the optimal development of their children.
Footnotes
Acknowledgements
The authors would like to thank Ninat Friedland for her assistance in preparing this manuscript.
The Canadian Institutes of Health Research, the Fonds de Recherche en Santé du Québec and the Conseil Québecois de la Recherche Sociale provided funding for the research conducted by the authors that is cited in this paper. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.
No writing assistance was utilized in the production of this manuscript.
