Depression During Pregnancy

Heather A. Bennett; Adrienne Einarson; Anna Taddio; Gideon Koren; Thomas R. Einarson


Clin Drug Invest. 2004;24(3) 

In This Article

Outcomes Associated with Prenatal Depression

Depression is associated with high numbers of somatic symptoms, some of which can be debilitating. Included are headache, nausea, stomach pain, shortness of breath, gastrointestinal problems, palpitations, dizziness and sexual dysfunction.[15] In the absence of adequate treatment the depression can accelerate. Episodes may become more frequent and severe, resulting in substantial maternal and infant morbidity.[25,64,82] Functional impairment,[75] inadequate prenatal care,[81] pre-eclampsia,[65] substance abuse,[25,83,84] increased risk of postnatal depression[18] and ultimately poor pregnancy outcomes have all been associated with depression during the obstetric period.[16,26,76,85,86,87,88,89] Consequently, a risk exists for potentially devastating consequences that impact upon both mother and baby.

Somatic symptoms among depressed patients in the general population have been associated with increased healthcare utilisation,[90] functional impairment and absenteeism from work.[91,92,93,94,95] Similarly, patients with subthreshold depressive disorder experience considerable impairment.[96] Similar outcomes are likely to be associated with depression experienced by women during gestation; however, few data to support such an assumption were found in the literature.

Birndorf et al.[75] examined functional impairment among pregnant women during the first trimester of pregnancy. They found that those who had screened positive for either anxiety or depression had significantly higher levels of functional impairment, independent of health status, than did those who did not have a positive screen. However, that finding was derived from a small study of upper middle-class women in the early stage of pregnancy and may not be generalisable to other groups of pregnant women. Further studies examining this aspect of depression in the obstetric population are necessary to determine the extent of functional impairment experienced by this patient group.

The disability and suffering associated with depression are not limited to the patient; rather, entire families are affected. Depression-related functional impairment in the mother is likely to affect her children negatively.[7] The degree of impairment due to antenatal depression needs further quantification so that its impact upon the course of pregnancy, the mother and her family can be assessed.

Nutritional deprivation and poor maternal weight gain during pregnancy are risk factors for intrauterine growth retardation (IUGR) and low neonatal birthweight.[97] IUGR is a major cause of perinatal mortality and morbidity,[98] and an important cause of developmental impairment in later life.[99] When depression is associated with weight loss, fetal growth may be negatively affected.[85]

Rather than weight, body mass index (BMI) is used to gauge a woman's health status during pregnancy. It is defined as weight in kilograms divided by height in square metres. Low prepregnancy BMI and low weight gain during pregnancy have been correlated with increased depressive symptom scores (as measured using the CES-D) in women of low SES.[81] For middle-class women, prepregnancy BMI has been correlated with increased depressive symptom scores (assessed using a modified CES-D) postpartum, but not during the gestational period.[100] Whether the depressive disorder is the cause or the result of inadequate nutritional status is not clear.

Nevertheless, data suggest that inadequate nutrition/weight gain associated with depression may be problematic for obstetric patients of low SES. Additionally, the clinical relevance of this health state and subsequent effects on the neonate, if any, have not been quantified.

Pregnancy-induced hypertension (PIH) is a major-complication in pregnancy. It exists in two forms, pre-eclampsia and eclampsia. The former is characterised by an increase in blood pressure (BP) accompanied by proteinuria, oedema or both.[101] Eclampsia is the occurrence of one or more convulsions in association with the syndrome of pre-eclampsia. Between 1979 and 1992 in the US, PIH accounted for 15% of hospitalisations for prenatal complications, and pre-eclampsia/eclampsia resulted in an estimated 1.5 deaths/100 000 live births.[102]

The aetiology of pre-eclampsia is unknown; however, it has been suggested that altered excretion of vasoactive hormones as a result of depression may increase the risk for PIH.[103] In Finland, pregnant women with depressive symptoms had a higher probability of developing pre-eclampsia than did nondepressed women (odds ratio [OR] = 2.5; 95% confidence interval [CI] 1.1, 5.4). Increased maternal age (>30 years) in combination with depression elevated the risk further (OR = 3.4; 95% CI 1.4, 7.1). Depression comorbid with bacterial vaginosis (BV) increased the risk more than 5-fold (OR = 5.3; 95% CI 1.8, 15.0) compared with women who did not have either condition.[65] BV is neither routinely tested for nor treated during pregnancy.[104] As psychiatric status is not routinely assessed during pregnancy, further research to determine the incidence of this comorbidity and associated outcomes is needed.

In contrast, PIH has not be found to be correlated with depressive symptoms for low-income women in the US.[81] However, in that study, many of the women with depressive symptoms engaged in adverse health behaviours such as cigarette smoking. It is known that smoking provides a protective effect for PIH and pre-eclampsia.[105,106,107] Whether the risks for PIH are altered for depressed women who smoke is unclear. However, few studies have investigated the relationship between depression and pre-eclampsia, and meaningful conclusions cannot be made at present.

In the US, the relative risk (RR) of preterm birth for those who do not receive adequate prenatal care has been estimated to be 2.8-fold higher in both African- and Caucasian-American women than in women with prenatal care.[108] Infants of African- and Caucasian-Americans who did not receive appropriate prenatal care had RRs for post-neonatal death (i.e. death of the infant between 28 and 365 days of life) that were 1.8 and 1.6, respectively, when compared with offspring of women who did receive such care.[109] As well, Kogan and associates found that women who had received inadequate prenatal care were also underutilisers of paediatric care following the birth of their babies.[110]

The association between psychiatric diagnosis and prenatal care was examined in a US population-based study of more than one million women between 1994 and 1995.[111] Psychiatric status was determined by the International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) diagnostic codes as recorded on hospital discharge summaries. After controlling for age, education, race, marital status, parity (i.e. previous births) and source of hospital payment, the likelihood of inadequate care during the prenatal period was significantly elevated for women with documented mental disorders as compared with those without psychiatric diagnoses. They were twice as likely to delay seeking care, and they routinely attended <50% of recommended prenatal visits.[111] That study did not examine depression specifically and the percentage of women with an actual diagnosis of depression is unknown. However, depression is one of the most common of all psychiatric disorders in the community,[92] and it is assumed that a large percentage of study participants would have had a diagnosis of depression.

It is important that when depressed patients do seek care they be encouraged to continue to undertake routine medical assistance both during and after pregnancy. For women in the US who receive adequate care during pregnancy, the risk of poor pregnancy outcomes (e.g. low birthweight infants) was reduced 3.4- and 1.6-fold for African-Americans and Caucasian-Americans, respectively.[112] Monitoring of depressed patients for the risk of preterm birth may increase the opportunity for timely intervention, thereby affording improvement in their pregnancy outcomes. Additionally, research to determine the most appropriate interventions for women who experience prenatal depression is necessary.

Estimates in the US of the prevalence of smoking, alcohol use and illegal substance use during pregnancy are 20.4, 18.8 and 5.5%, respectively.[113] Smoking during pregnancy is associated with placental insufficiency, which can lead to adverse pregnancy and birth outcomes such as low birthweight, sudden infant death syndrome (SIDS), preterm delivery and infant mortality.[114,115]

In a study of 2432 pregnant women in Copen-hagen, IUGR with a neonatal birthweight lower than the 10th percentile was associated with maternal smoking habits (adjusted OR = 2.4 [95% CI 1.51, 3.80] for 0–9 cigarettes, 2.68 [95% CI 1.52, 4.68] for 10–15 cigarettes, and 2.88 [95% CI 1.36, 6.09] for >15 cigarettes daily).[99] Wisbourg et al.[114] reported that the risk of SIDS for children of smokers was three times that of children of non-smokers (OR = 3.5; 95% CI 1.4, 8.7) in a prospective study of 24 986 women in Denmark. Additionally, the risk of SIDS in that study increased with the number of cigarettes smoked per day (p < 0.05). In the US, the infant mortality rate for infants of smokers was 10.5 per 1000 live births in 2001, 62% higher than the rate of 6.5 per 1000 live births for infants of mothers who did not smoke.[115]

Consumption of alcohol by pregnant women in the US decreased between 1988 and 1992; however, that trend reversed during the early 1990s. Alcohol use by pregnant women increased from 9.5% in 1992 to 15.3% in 1995, and frequent alcohol use increased from 0.9% in 1991 to 3.5% in 1995.[116,117,118]

Similarly, the prevalence of prenatal alcohol consumption in Canada during 1996–97 was 17%.[119] A higher prevalence of 23% was observed in older women (>35 years) than in women who were <25 years old, in whom it was 12%. It must be remembered that, due to the stigma of substance abuse during pregnancy, these figures may under-represent the true magnitude of the problem.

In the US, Kelly and co-workers found that 8% of 186 pregnant women had both psychiatric and substance abuse disorders.[120] Similarly, in a Finnish study of 391 women, it was estimated that the prevalence of depression (EPDS score >12) combined with substance abuse in obstetric patients (n = 391) was 6.4%.[64] Researchers who examined 1014 women of low SES in the US, found depressive symptoms (CES-D scores ≥16) in early pregnancy were significantly associated with adverse health behaviour including cigarette, alcohol and cocaine use.[81] Also, a high proportion of US women who abuse drugs during pregnancy have been found to experience depressive episodes accompanied by suicidal ideation.[121]

Although not extensively researched,[64] available data suggest that depression is associated with licit and illicit substance abuse throughout pregnancy. It should be noted that researchers have examined women during routine prenatal visits; the association between depression and substance abuse for those who do not attend antenatal care is likely to be higher than that estimated above. Depression during gestation may elevate the potential for abuse, that is, these behaviours may reflect attempts by women to cope with their depression.

Regrettably, outcomes associated with substance abuse during pregnancy are often severe and sometimes, for the child, long-lasting. For example, it is well known that consumption of alcohol during the obstetric period may result in alcohol-related birth defects (ARBDs) including spontaneous abortion,[122] IUGR, fetal alcohol syndrome (FAS),[119] and fetal alcohol effects (FAEs).[123] FAS is characterised by craniofacial abnormalities, mental retardation, cognitive and behavioural problems.[124,125] Children with FAE have better cognitive abilities than do children with FAS but similar behavioural disabilities. Such a combination leads to problems in later life, including disruptive behaviour, delinquency, drug use and inappropriate sexual behaviour.[123]

Not all women who drink have offspring with ARBD. This suggests that the amount of alcohol consumed, timing of the consumption, metabolism of the mother, and genetic makeup of the fetus may all be factors affecting the outcomes of alcohol intake by pregnant women.[119,122,126,127] While low and moderate levels of alcohol consumption have been shown to be 'safe'[119,122,126,127] an inability to define 'excessive' alcohol intake accurately remains a major dilemma.

There has been no conclusive evidence that prenatal illicit drug abuse results in morphological teratogenesis; however, negative long-term outcomes for the baby have been reported. For example, in a cohort study of 1760 cases of SIDS that occurred between 1979 and 1989 in New York City, the rate of SIDS in drug-exposed babies was 5.83 per 1000 infants compared with 1.39 per 1000 infants for non-exposed babies.[128] Furthermore, when maternal drug abuse continues after the birth of the baby, children may be removed from the parent by the authorities, causing negative social implications.

As well as the direct effect on the fetus from in utero alcohol exposure, substance abuse of any nature may result in inadequate prenatal care and the inability of the mother to take adequate care of herself during the prenatal period.[64,129] Developmental and behavioural problems in the infant may result. Because depression may be a risk factor for substance abuse, it is important that the psychiatric status of pregnant women be considered. If depression does lead to substance abuse, it may be that treating the depression itself will reduce the abuse and the associated outcomes. Appropriate and timely intervention may prevent the dire consequences associated with such harmful health behaviour.

Suicidal ideation and attempts have been consistently reported in the depressed population.[130] Both those with MDD and those with depressive symptoms are at risk.[131] Rates of suicidal ideation and attempt in the general population have been compared across the US, Canada, Puerto Rico, France, West Germany, Lebanon, Taiwan, Korea and New Zealand.[132] Lifetime prevalence for suicidal ideation in females varied from 2.9% in Beirut, Lebanon, to 14.8% in Christchurch, New Zealand. Rates for suicide attempts were lower and more consistent across countries. They ranged from 1.0% in Beirut, Lebanon, to 7.0% in Puerto Rico and Savigny, France.

Fewer data are available on suicidality during pregnancy. Rates of suicidal ideation among depressed obstetric patients have ranged from 3% in Finland[64] to 17.6% in the US.[75] However, those estimates have been determined based on small samples (n = 391 and 57, respectively). Importantly, women who abruptly discontinue antidepressants upon becoming pregnant very often report suicidal ideation.[133]

Completed suicides during pregnancy are less common. UK researchers[134] reported that during the period 1997–99, mental illness accounted for 12% of maternal deaths, of which 10% were suicides. Overall, 42 deaths were identified as having resulted from psychiatric disorders, with suicide (28 identified cases) accounting for one death per million maternities. Six maternal suicides occurred during pregnancy, seven occurred within 42 days of delivery, and a further 15 deaths occurred more than 42 days from delivery. Of those suicides, ten were attributable to a probable diagnosis of severe depression.

Similar rates have been reported in the US.[135] Of the 315 female suicides in New York City between 1990 and 1993, six cases were found to have been pregnant (gestation up to 34 weeks). The age- and race-adjusted risk of suicide among pregnant women, compared with the mean annual suicide rate for all female residents, was 0.33 (95% CI 0.12, 0.72).

These figures suggest that pregnancy may be protective against suicide. However, although suicide during pregnancy appears to be a rare event, this low risk should be regarded seriously. In addition to maternal death, it is necessary to consider the loss of potential life of the unborn child as a result of concurrent fetal loss. The number of pregnant women who attempt suicide as a result of depression is unknown, neither has the impact on the fetus of attempted suicides been reported. Further study is warranted to determine the level of risk that suicidal ideation and suicide attempts pose to those women and their babies.

It has been suggested that negative pregnancy outcomes could be the reflection of a psychosomatic disorder.[103] However, evidence to support such a theory has been conflicting and consideration of the existing data is warranted. In 1983, Steer et al.[89] reported that depression (BDI scores ≥21) was associated with a 3-fold increased risk of low birthweight (<2500g), preterm delivery (<37 weeks gestation), and a small-for-gestational-age infant for adults but not adolescents.

Kelly et al.[98] conducted a population-based cohort analysis of 521 490 live births in the US during 1995.[98] They found a significantly higher risk of low (OR = 2.0) and very low (<1500g) birthweight (OR = 2.9) and preterm infants (OR = 1.6) for women who had a psychiatric diagnosis (ICD-9-CM diagnostic codes recorded on maternal hospital discharge summaries).[98] A significant positive association existed between poor pregnancy outcomes and a prenatal diagnosis of psychiatric disorder.

In a community-based study of 666 pregnant women during 2000 in the US,[66] elevated depressive symptom scores (CES-D scores ≥16) were associated with a decrease in fetal growth for women of low SES. In contrast, no relationship was found between elevated CES-D score and decreased fetal growth or gestational duration in other high-risk groups, for example smokers, women with a history of adverse outcome, and those with social vulnerabilities.

The relationship between maternal depressive symptoms and spontaneous preterm birth in 1399 low-income African-American women was recently examined in the US by Orr and colleagues.[136] Spontaneous preterm delivery occurred among 8.4% of the total sample compared with 12.7% of those with a CES-D >33. For those with a CES-D ≥16 and <33, 8.0% had spontaneous preterm deliveries. The odds ratio (OR), adjusted for behavioural, clinical and demographic variables, for spontaneous preterm birth for those with an elevated CES-D score compared with those without such a score was 1.96 (95% CI 1.04, 3.72).

Two studies were identified that dealt specifically with depression and preterm labour (gestational age between 20 and 37 completed weeks of gestation, at least one contraction every 10 min for at least an hour, and at least one cervical change such as dilatation or effacement). In the first study of 1515 women attending obstetric units in England during 1993, depression (determined by the General Health Questionnaire, a screening instrument designed to detect the severity of psychological disturbance in general population surveys) was unrelated to the onset of non-spontaneous preterm labour.[137] Conversely, in a 2002 study of 634 women in France, those with depression (EPDS score ≥15) had a risk of spontaneous preterm labour that was twice that of women who were not depressed (OR = 2.1, 95% CI 0.99, 4.0).[16]

Research into the association between depressive disorder and operative deliveries (Caesarean section and instrumental vaginal delivery) has been sparse. Indeed, only three such studies could be identified. In a study of 959 Chinese women in Hong Kong, Chung et al.[76] found a 2-fold increase in the risk for operative deliveries (p = 0.02) and epidural analgesia (p = 0.01) for patients with depressive symptoms (BDI scores ≥15). In addition, they reported a similar risk increase for admission to neonatal care units for babies of patients with depressive symptoms (p = 0.03). On the other hand, an association between depression (CES-D ≥16) and mode of delivery was not found in a US study of 1697 women.[67] Similarly, in a study of low SES minority women in the US, no association was found between depressive symptoms (CES-D scores ≥16) and operative deliveries.[81]

In summary, the majority of evidence suggests that depression during the antenatal period is associated with preterm delivery, low infant birthweight, and small-for-gestational age infants. Those outcomes are major factors related to infant death and morbidity in the US.[97] Intervention during pregnancy to address depressive disorder may reduce infant mortality and morbidity. Divergent results regarding the relationship between depression and spontaneous preterm labour, and between depression and operative deliveries, have been reported, and further research to assess the risk is needed.

Interestingly, researchers who used the BDI or the EPDS to determine the depression status of participants have consistently reported an association between depressive symptoms and poor pregnancy outcomes. Results reported by authors who have used the CES-D have been less definitive. Some have reported no association, while others have reported an association only for women of low SES. However, as the CES-D has not been validated for use in a pregnant population, caution must be used when interpreting results based on that instrument.

Of particular note is the similarity of methods used by researchers for participant accrual. Nearly every study that examined the association between depression and pregnancy outcomes identified and enrolled patients through obstetric and prenatal institutions. That is, nearly all participants were receiving prenatal care. For those depressed women who may not seek care, the association between depressive disorder and adverse pregnancy outcomes may be higher than reported in the literature. Thus, the risks of negative outcomes discussed above may be conservative. Overall risks can only be determined if all women who experience depression during the prenatal period are included.

Depression during the prenatal period has been identified as a risk factor for postpartum depression (PPD).[36,138] Approximately one-third of women who experience depression throughout their pregnancy remain depressed following the birth of their child.[12,13]

US researchers[139] who observed 465 postpartum women reported that four variables, including maternal age, prenatal depression, morbid thoughts and sleep difficulties at 1 month postpartum were predictive of depression at 4 months postpartum. Others have reported that BDI scores in the eighth month of pregnancy were significantly correlated with postpartum BDI scores (p < 0.001).[54] Thus, a number of small studies have reported consistent results. However, perhaps the most compelling evidence has been that provided by Beck's Postpartum Depression Predictors Inventory.[140] Based on the findings of a meta-analysis of 26 original research studies (n = 2189), the magnitude of the relationship between 13 predictors and PPD was described.[141] Prenatal depression was reported to be one of the strongest predictors of PPD (Pearson correlation coefficient r = 0.44–0.45).

The disabling effects and serious consequences of postnatal depression have been well documented.[142,143,144] Identification of those who have depression during the prenatal period, and the provision of efficacious treatment during pregnancy, may reduce the number of women at risk for the development of PPD.

Little is known about prenatal transmission of depression to the fetus. It has been assumed that negative effects on the infant have derived from early mother-baby interactions. However, infants of depressed mothers have exhibited 'depression-like' behaviour (i.e. fewer expressions of interest, more pre-cry expressions, excessive crying, lower orientation scores and motor behaviour cluster, inferior excitability, and more abnormal reflexes) prior to substantial mother-baby interactions.[86,87] Furthermore, elevated levels of cortisol and norepinephrine and reduced dopamine levels have been detected in the babies of the depressed mothers. Lundy et al.[87] reported that maternal depressive symptoms and altered catecholamine levels have been found to predict the 'depression-like' behaviour of neonates. Gitau and others[145] reported that between 40 and 50% of fetal cortisol may be derived from the mother. Hormone levels in the fetus, therefore, may be affected when maternal cortisol and catecholamines cross the placenta.

These data provide physical evidence of cause and effect, i.e. the increase in maternal hormones due to the depressive state of the mother may result in altered hormonal levels and subsequent neuro-behavioural changes in the fetus.

Zuckerman and colleagues reported similar outcomes when they examined the neonatal neuro-behavioural functioning of 1123 babies.[88] Maternal CES-D scores during pregnancy were associated with excessive crying in the newborn at 3 months postpartum (p < 0.01). Importantly, higher CES-D scores resulted in increased distress in the baby. Mothers with elevated CES-D scores (i.e. at the 90th percentile) during pregnancy were 2.6 times more likely to have inconsolable newborns than were women with scores at the 10th percentile (95% CI 1.54, 4.23).

Thus, emotion-based changes in the pregnant woman's physiology, which result in changes in the in utero environment, may also negatively impact upon the child's future temperament and ability to regulate emotions.[146,147] Maternal depression throughout pregnancy that results in negative neo-natal neurobehavioural functioning may precipitate early childhood problems.[88]

Conversely, Glover and O'Connor[148] have reported that behavioural and emotional problems in offspring were linked to antenatal stress or anxiety at 32 weeks gestation, but not to antenatal depression. Nevertheless, as MDD has been linked to the brain's response to stress,[50] further research is needed to determine the relationship between stress, anxiety, depression and neonatal behavioural outcomes.

The numbers of neonates who have impaired neurobehavioural functioning and resultant problems during the gestational period are as yet not known. Further research is needed to quantify such outcomes, and to assess their impact on the quality of life of those patients, their families and society.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.