11th Greek Australian Legal and Medical Conference
Crete, Greece 2007

Severe depression during pregnancy: A clinical and an ethical challenge

Dr Iannis M. Zervas, M.D.

Depression, the 5th leading cause of morbidity in 1999 and expected to reach second place by 2020, affects approximately one out of ten pregnant women with varying intensity. Women are reported to have lifetime rates of depression of any severity reaching 24 % and this includes types of depression specific to women, such as premenstrual dysphoric disorder, perinatal depression, depression related to infertility, depression related to menopause, and chronic pelvic pain. In women depression is more frequent than men, has a younger age of onset, has more frequently a seasonal distribution. Ironically, marriage reduces risk of depression only for men. In addition, 20-30% of women of reproductive age receive psychotropics; 2/3 of antidepressants and tranquillizers in the US are prescribed to women. A clinician needs to take into account as well: that more than 50% of pregnancies are unexpected and that all medications cross the placenta and affect the fetus; that 60% of women with a history of mood disorder discontinue their medication close to conception or pregnancy but half of them have to restart; that rates of depressed pregnant women are similar to non-pregnant women of the same age and that the risk of major depression during pregnancy is 3.6 – 6.1 % ( risk of depression of any severity during pregnancy being 10 - 15%).

These well documented figures make one realize that depression in women during their fertility years and during pregnancy pose several clinical and ethical challenges. In addition the mere numbers involved should make this an issue for public health policy. If Greece is used as an example, with approximately 100,000 pregnancies reported per year, this means that about 5,000 women will manifest significant depression during their pregnancy, each year! We verified this in a recent study conducted by our department: 5% of women examined at two months post partum with a structured psychiatric interview had major depression and 7.4 % had minor depression. Although the sample was recruited from the general obstetric population, 70% of the women who developed depression postnatally had a history of a previous depressive episode, unrelated to pregnancy, which is 10 times more frequently than women who did not develop depression who had such a at a low rate (7%) (data submitted for publication). Depression in pregnancy therefore is a predictable, potentially preventable condition, which definitely presents an acute demand for adequate and safe treatment.


Risk factors include, besides an unwanted pregnancy, a personal history of mood disorder, a family history of depression, marital discord, poor support network, and recent negative life events, exemplifying the fact that in depression both biological (mainly genetic) and environmental (mostly life events) factors are important, probably in an interplay such as the one outlined by Caspi et al ( Science, 301, 386–389, 2003) and by Kendler et al (American Journal of Psychiatry, 163:109-114, 2006). These researchers show that what is inherited appears to be a genetic vulnerability, perhaps accentuated by the particular hormonal environment, on which life events impinge. If one lacks the vulnerability or the events, the possibility for developing depression is much smaller.

In the 2007 guidelines (guideline 45) of National Institute for Clinical Excellence for the prediction, detection and treatment of mental disorders in women during pregnancy and the postnatal period (up to 1 year after delivery) advice is included on the care of women with an existing mental disorder who are planning a pregnancy, and on the organisation of mental health services. It is recommended that:

“At a woman’s first contact with services in both the antenatal and the postnatal periods, healthcare professionals (including midwives, obstetricians, health visitors and GPs) should ask questions about: 1. past or present severe mental illness including schizophrenia, bipolar disorder, psychosis in the postnatal period and severe depression, 2. previous treatment by a psychiatrist / specialist or mental health team including inpatient care, and 3. a family history of perinatal mental illness. Other specific predictors, such as poor relationships with her partner, should not be used for the routine prediction of the development of a mental disorder”.


One should keep in mind the classical symptoms of depression (depressed mood, loss of interest, lack of energy or easy fatigue, weight changes, sleep changes, agitation or retardation, feelings of worthlessness and guilt, indecisiveness and / or lack of concentration, suicidal ideation, multiple physical symptoms). However, one should also keep in mind that diagnosis may be difficult: sleep and appetite, libido and energy disturbances are common in normal pregnancy, and anemia, diabetes and thyroid disease in pregnancy may present with depressive symptoms. One should look out, instead, for more specific indicators such as lack of interest in the pregnancy, guilt feelings and intense lack of pleasurable feelings (anhedonia).

Education of those involved in the care of pregnant women and screening of high risk groups are important steps toward better diagnosis.


A psychiatric consultant is usually called in when psychiatric symptoms appear during pregnancy, when there are worries about prenatal exposure to psychotropic drugs or when there are worries about relapse of psychiatric illness due to possible discontinuation of psychiatric treatment. The consultant needs to do a risk-benefit analysis in terms of treatment options, to help discuss these options and inform about risks all involved: the pregnant women themselves of course, but also the couple, the family, as well as the treating specialists (obstetricians, psychiatrists, maternity and mental health workers etc) Decisions for the perinatal psychiatrist are very difficult during pregnancy and require, besides updated knowledge, particular communication skills, a familiarity (and flexibility) with available community resources, as well as a good sense of measure and reason. Often, however, even in the presence of all of the above, deficiencies in the local mental health system in addressing the particular needs of this patient population make the delivery of optimal care particularly distressing for patients and physicians alike.

Reaching a decision is usually a complex process and at times becomes an exercise in clinical judgment. Existing data are based on retrospective data and case series and when the issue at hand is the possible effects of the drugs used on a particular pregnancy, one should keep in mind that there are multiple other determinants: Exposure to other drugs, substances, and toxins; nutritional status; mother’s age; genetic vulnerability; time frame of exposure; stressors; and many other pregnancy factors.

If the depression is left untreated, this may result in harm of the mother (suicide at worst, but more commonly poor self-care, poor nutrition and sleep, substance abuse – most commonly alcohol and tobacco – and exposure to additional medication, premature labor, postpartum depression and, after delivery, poor infant care due to disorganization, impulsivity, suicidality). Untreated depression during pregnancy has also been associated with an equal number of long-term post-delivery problems for the baby and the developing child that may impact on its physical and mental well being, such as lower birth weight, lower APGAR score, smaller head circumference, reduced neonatal functioning and possible dysfunction of the infant HPA axis.

Milder forms of depression can be treated with psychotherapy alone (usually by courses of brief therapy of a cognitive or interpersonal orientation) but more severe forms often require biological interventions ranging from medications to hospitalization and sometimes even electroconvulsive therapy. Available medical information on the safety of these interventions on the developing fetus have gone through several stages over the years: from being considered unsafe in older times – leading physicians to advise termination of pregnancy in women who had received such medication – to freely prescribing certain antidepressants from the nineties on. Recent studies show that these antidepressants may not be as safe as it was thought only a few years ago, as they may be responsible not so much for congenital malformations but rather for perinatal complications and for the adaptability of the developing newborn and perhaps older child.

In the background of all clinicians’ worst nightmares is, of course, the thalidomide story, released in 1957 as a “pregnancy safe” tranquilizer in Germany and another 46 countries, and retracted in 1961 after 40% of the exposed newborns had died and 12,000 survived with severe congenital abnormalities.

Although we have not witnessed such major catastrophes since, and medical science has become much more cautious, the initial optimism with the newer antidepressants has been seriously tempered in the last few years. Already since 1996 Christina Chambers was reporting a 15.5% incidence of more than 3 minor anomalies in newborns exposed to fluoxetine during the 1st trimester, showing that the drug indeed has an effect on fetal development and reporting that “Exposure to fluoxetine during the first trimester increases the probability of reduced adaptability syndrome (respiratory distress, irritability, tremor, cyanosis while being fed). It also increases the possibility of NICU admission”.

We now know that the use of SSRI drugs after the 20th week of pregnancy increases by 6-fold the probability of persistent pulmonary hypertension in the newborn (1-2:1000 becomes 6-12:1000) and that newborns with persistent pulmonary hypertension are full term without congenital anomalies and manifest severe respiratory failure after birth requiring intubation and ventilatory support. Mortality is 10-20%. As reported by the FDA Advisory Committee, Pediatric Subcommittee (June 9 2004), newborns exposed to SSRI/SNRI at the end of the 3d trimester had complications requiring prolonged hospitalization, ventilatory support and tube feeding. Complications may become evident immediately after birth. These complications include respiratory failure, cyanosis, apnea, convulsions, temperature instability, difficulties feeding, vomiting, hypoglycaemia, hypertonia, hyperreflexia, tremor, irritability, and non-stop crying. Many studies (Chambers C et al. NEJM 1996;335:1010-1015; Moses-Kolko EL et al. JAMA 2005;293:2372-2383; Laine K et al. Arch Gen Psychiatry 2003;60:720-726; Costei AM et al. Arch Pediatr Adolesc Med 2002;156:1129; Sanz EJ et al. Lancet 2005;365:482-487) have now shown that 25-30% of neonates exposed during the last part of pregnancy will develop neonatal withdrawal syndrome or serotoninergic syndrome. These appear between the 1st to 10th day after birth and last 1-14 days.

The recent NICE (April 2007) guidelines have finally incorporated these into their recommendations. The Royal Australian and New Zealand College of Psychiatrists, in its December 2005 guidelines has been cautious, however to recommend vigilance: “It is recommended that psychiatrists and general practitioners administering SSRI/SNRI during pregnancy check frequently the latest data on the neonatal discontinuation syndrome”.

The health system

Services should probably follow the Institute of Medicine recommendations (2001), in that they must be safe, effective, patient-centered, timely, efficient, and equitable. Care should also be based on a ‘continuous healing relationship’ ”. Specialized outpatient services, day-hospitals and inpatient services should be developed, as only few countries have made available services directed to the needs of this particular patient population.

In closing, it should be re-emphasized that the purpose of pregnancy is to produce a mother as well as child (and it is better still if it produces a father as well). Depression in pregnancy is frequent and can compromise the physical and mental health of the mother and of the child in the long-term. It is a preventable and treatable illness with a complex etiology; therefore treatment should be as complete as possible, address multiple biological, psychological and social variables and follow principles of medical ethics. Finally, medical education and medical services should be modified accordingly to provide a working knowledge and care base for these common, frequent and severe conditions with far reaching implications.