Postpartum depression

Article written and reviewed by Dr Luca Moretti

Postpartum depression is a depressive episode that occurs after childbirth. In DSM-5-TR it is not a completely separate diagnosis from major depressive disorder: when a major depressive episode begins during pregnancy or within the first four weeks after delivery, the specifier with peripartum onset may be applied. In clinical practice and perinatal health guidelines, however, depression surveillance extends well beyond the first four weeks, frequently across the entire first year after birth.

This temporal distinction should be made explicit to avoid confusion. The four week limit belongs to the DSM-5-TR specifier, whereas the clinical term “postpartum depression” is often used for episodes beginning in the months after childbirth. ACOG considers perinatal mental health throughout pregnancy and the first postpartum year, a period during which new depressive episodes may emerge or episodes that began during pregnancy may persist.

Postpartum depression is common, but estimates depend on diagnostic criteria, screening instruments, population and time interval. A meta analysis of mothers considered healthy and without previous depression estimated an incidence around 12% and an overall prevalence around 17%. These values do not represent an individual's probability and may differ in high risk populations, across socioeconomic contexts and when subthreshold depressive symptoms are included.

The clinical picture should be distinguished from maternity blues or baby blues, a very common transient condition in the first days after childbirth, characterized by emotional lability, easy crying, irritability and increased sensitivity, generally self limited within about two weeks and without the pervasive impairment of a major depressive episode. Intense symptoms, symptoms persisting beyond this period, or symptoms associated with suicidality, severe anxiety or loss of functioning require clinical assessment.

It is equally essential to distinguish depression from postpartum psychosis, a rare but extremely severe condition typically characterized by rapid onset, marked insomnia, disorganization, mania or severe depression, delusions, hallucinations and possible risk to mother and infant. Postpartum psychosis is a psychiatric emergency and has a strong relationship with the bipolar spectrum; it should not be interpreted simply as a more intense form of postpartum depression.

The birth of a child changes sleep, circadian rhythms, caregiving burden, relationships and the endocrine environment. These changes do not make depression a “normal” reaction to motherhood. The disorder emerges when affective, cognitive and neurovegetative symptoms reach clinically significant intensity and duration and impair functioning, self care or the ability to manage daily life safely.

Recognition is also important because maternal depression can interfere with quality of life, the couple relationship, breastfeeding when desired, social functioning and early interactions with the infant. These associations should not fuel guilt or stigma: postpartum depression is a treatable medical condition and seeking care is a protective factor for mother and family.

Etiology, Pathogenesis and Pathophysiology

There is no single cause of postpartum depression. The current model is multifactorial and integrates genetic and psychiatric vulnerability, peripartum neuroendocrine changes, sleep, stress, social support, obstetric complications and individual characteristics. No hormone, event or psychosocial factor is necessary or sufficient in an individual case.

The most important group of risk factors includes a personal psychiatric history of depression, anxiety or other disorders, especially previous peripartum episodes. The presence of bipolar disorder is particularly important because the postpartum period is associated with a high risk of affective relapse and psychosis; apparently unipolar depression may represent the depressive phase of previously unrecognized bipolar disorder.

Depression and anxiety during pregnancy increase the risk of persistence or recurrence after childbirth. Adverse events, intimate partner violence, poor social support, financial difficulties, traumatic pregnancy or childbirth, neonatal complications and high stress are also associated with increased risk. These are probabilistic factors: their presence does not demonstrate causality and their absence does not exclude the disorder.

After childbirth there is a rapid decline in progesterone and estrogens and progesterone derived neurosteroids. Allopregnanolone is a positive modulator of GABA-A receptors, and abrupt changes in the neurosteroid system have contributed to the development of postpartum specific pathophysiological models. The clinical efficacy of neuroactive drugs such as zuranolone supports the relevance of this system, but does not demonstrate that depression is caused simply by a “low” allopregnanolone level.

Vulnerability may depend more on individual sensitivity to hormonal changes than on the absolute concentration of a single hormone. Experimental and clinical studies suggest that some women with a history of perinatal depression may respond differently to fluctuations in reproductive steroids, but there is currently no endocrine test that can accurately identify who will develop the disorder.

The hypothalamic pituitary adrenal axis undergoes profound changes during pregnancy and the puerperium. Cortisol, placental CRH and the stress response change physiologically and may interact with affective vulnerability. Even in this area, however, findings are heterogeneous and cortisol or CRH are not diagnostic biomarkers for postpartum depression.

Inflammatory, immune, thyroid and metabolic systems have been studied. Pregnancy and the puerperium involve major immunological adaptations, and associations between inflammatory markers and depressive symptoms have been observed in some patient subgroups. The low specificity of these markers and the influence of childbirth, infections, obesity and other conditions prevent their use for clinical diagnosis.

Sleep is a particularly relevant factor. Frequent awakenings to care for the infant and sleep fragmentation are common and are not equivalent to depression, but severe sleep deprivation can worsen emotional regulation and anxiety. A marked reduction in the need for sleep without feeling tired, especially when accompanied by increased energy, racing thoughts or disinhibition, should instead raise suspicion of mania or postpartum psychosis.

The relationship between breastfeeding and depression is bidirectional and complex. Breastfeeding difficulties may increase stress and feelings of inadequacy, while depression may make initiation and continuation of breastfeeding more difficult. It is incorrect to present breastfeeding as certain prevention of depression or to attribute the illness to a decision not to breastfeed.

Psychological processes such as unrealistic expectations about motherhood, perfectionism, guilt, loss of a sense of competence and difficulty with the role transition may contribute to maintaining symptoms. Interpersonal psychotherapy is particularly consistent with these mechanisms because it addresses transitions, relationships and social support, while CBT targets dysfunctional thoughts, avoidance and loss of rewarding activities.

The final pathophysiology therefore results from the interaction of preexisting vulnerability, neuroendocrine adaptation, sleep, stress and environment. This heterogeneity explains why two women with similar obstetric experiences may have completely different psychological outcomes and why depression may arise even after an uncomplicated pregnancy and delivery.

There are no validated blood, genetic, hormonal or neuroimaging biomarkers for diagnosing postpartum depression. Diagnosis remains clinical, while neurosteroid mechanisms now also have therapeutic relevance through the development of GABA-A receptor modulators.

Clinical Manifestations

Assessment begins by reconstructing the timeline: symptoms during pregnancy, timing of childbirth, the first days of the puerperium and subsequent course. It is important to establish whether the episode truly began after delivery or represents continuation of an unrecognized prenatal depression.

The cardinal symptoms are depressed mood and anhedonia. The patient may report sadness, crying, emotional emptiness, loss of interest, inability to experience pleasure or a sense of detachment from her own life. Depression does not have to manifest as rejection of the infant, and the presence of affection or moments of pleasure does not automatically exclude a depressive episode.

Assessment of neurovegetative symptoms is more complex than at other times of life. Insomnia, fatigue and appetite changes may be physiological consequences of caregiving. They are more suggestive when they are disproportionate to the infant's demands, persist even when there is an opportunity to rest, or occur together with other depressive symptoms.

Guilt and low self worth may focus on maternal competence: believing oneself to be a “bad mother”, to be harming the infant, or not to deserve the family. These experiences should be distinguished from normal uncertainties of the puerperium by their intensity, rigidity and impact. In psychotic forms, ideas may acquire delusional quality and require urgent management.

Anxiety is very common and may predominate. Persistent concerns about the infant's health, feeding or sleep may accompany depression, generalized anxiety disorder or perinatal obsessive compulsive disorder. Correct assessment requires defining the nature of the thoughts and the behaviors used to neutralize them.

Intrusive thoughts about harming the infant require rigorous distinction. In obsessive compulsive disorder they are typically unwanted, ego dystonic, a source of intense fear and accompanied by avoidance or compulsions; the patient does not want to act on the thought. In psychosis, by contrast, delusional beliefs or hallucinations may impair judgment and materially increase risk. Thought content alone does not allow dangerousness to be estimated.

The relationship with the infant may be characterized by difficulty engaging, excessive anxiety, a sense of distance or loss of confidence in one's abilities. These phenomena are not mandatory criteria for depression and should not be used to judge the mother's moral quality; they are functional aspects to explore because they can guide support and treatment.

Pain, childbirth complications, anemia, thyroid dysfunction, breastfeeding problems and other physical conditions should also be assessed. Genuine somatic symptoms can coexist with depression, and the puerperium is a period in which fatigue, tachycardia, weight loss, headache or other potentially organic symptoms should not automatically be attributed to the psychiatric disorder.

Suicide assessment should distinguish passive thoughts of death, active ideation, intent, plan, access to means and previous attempts. Infant safety should also be assessed when psychosis, severe disorganization, intent to harm or inability to provide essential care is present.

Postpartum psychosis can begin rapidly with extreme insomnia, lability, euphoria or depression, confusion, disorganized behavior, delusions and hallucinations. It may have prominent affective features and is strongly associated with the bipolar spectrum. Suspicion requires immediate psychiatric assessment and generally inpatient treatment.

Maternity blues, by contrast, involves lability and crying but generally preserves contact with reality and tends to resolve spontaneously within two weeks. Progressive worsening, marked loss of functioning or persistence beyond the expected period should prompt reconsideration of the diagnosis.

During history taking it is essential to investigate a history of mania, hypomania or psychosis, including episodes occurring after previous pregnancies. A patient may present as depressed without spontaneously reporting previous periods of euphoria or increased activity; a family history of bipolar disorder or postpartum psychosis further increases clinical concern.

Mental status examination assesses appearance, psychomotor activity, speech, affect, thought, perceptions, orientation, insight and judgment. Disorientation or marked alteration of consciousness also requires consideration of delirium and acute medical causes, especially in the presence of infection, hemorrhage, metabolic abnormalities or obstetric complications.

Investigations and Diagnosis

There are no separate diagnostic criteria that transform postpartum depression into a syndrome completely distinct from major depressive disorder. Diagnosis requires the presence of a depressive episode according to nosographic criteria and definition of the temporal relationship with pregnancy and childbirth. DSM-5-TR uses the specifier “with peripartum onset” when the episode begins during pregnancy or within four weeks after delivery.

DSM-5-TR criteria applicable to major depressive disorder with peripartum onset

• During the same period of at least two weeks at least five symptoms must be present, representing a change from previous functioning; at least one must be depressed mood or markedly reduced interest or pleasure.
• Depressed mood for most of the day and on most days.
• Markedly diminished interest or pleasure in all or almost all activities for most of the day and on most days.
• Significant unintentional weight loss or gain, or decreased or increased appetite.
• Insomnia or hypersomnia nearly every day.
• Psychomotor agitation or retardation nearly every day, observable by others and not limited to a subjective feeling.
• Fatigue or loss of energy nearly every day.
• Feelings of worthlessness or excessive or inappropriate guilt nearly every day.
• Reduced ability to think or concentrate, or indecisiveness, nearly every day.
• Recurrent thoughts of death, recurrent suicidal ideation, suicide attempt or suicide planning.
• The symptom cluster must cause clinically significant distress or impairment in social, personal, or other important areas of functioning.
• The episode must not be attributable to the physiological effects of a substance or another medical condition.
• To diagnose major depressive disorder, the presentation must not be better explained by schizophrenia spectrum disorders or other psychotic disorders, and there must be no history of a manic or hypomanic episode.
• The specifier with peripartum onset applies in DSM-5-TR when the mood episode begins during pregnancy or within four weeks after delivery.

ACOG recommends the use of standardized and validated instruments for screening of depressive and anxiety disorders during pregnancy and the postpartum period, linking screening to systems capable of providing diagnosis, treatment and follow up. Screening without the ability to provide care is insufficient.

The Edinburgh Postnatal Depression Scale is one of the most widely used instruments in the perinatal period; PHQ-9 may also be used. Neither establishes a diagnosis by itself. The score is used to identify who needs further assessment and to monitor symptoms; items relating to self harm must be evaluated clinically regardless of the total score.

Before starting an antidepressant, the possibility of bipolar disorder should be assessed. ACOG emphasizes the importance of screening or assessment for bipolarity before antidepressant pharmacotherapy in the perinatal period, because treating bipolar depression as unipolar depression may be inappropriate and the postpartum period is a high risk phase for mania and psychosis.

Assessment should include suicidality, ability to care for oneself, availability of support and infant safety. When suicidal intent, psychosis, mania, severe disorganization or risk of harm is present, management becomes urgent and may require hospitalization, ideally in specialized mother and baby units where available and clinically appropriate.

Laboratory investigations are guided by the clinical picture. A complete blood count may be useful in the presence of hemorrhage or suspected anemia; TSH and thyroid hormones when signs of thyroid dysfunction are present; electrolytes, liver or kidney function and other tests when indicated. There is no diagnostic blood panel for postpartum depression.

The differential diagnosis includes maternity blues, anxiety disorders, perinatal OCD, post traumatic stress disorder after traumatic childbirth, bipolar disorder, postpartum psychosis, substance related disorders and medical conditions. The temporal relationship among symptoms, childbirth, sleep and medications should be reconstructed precisely.

Postpartum thyroiditis and other thyroid dysfunctions can produce overlapping symptoms such as fatigue, anxiety, weight changes and cognitive difficulties; thyroid testing should be guided by context and not performed to prove depression. Anemia, infections, pain and other obstetric complications should be treated when present.

Intrusive thoughts should be assessed for ego dystonicity, compulsions, insight, intent and presence of psychosis. A mother terrified by an unwanted thought and trying to avoid it has a different profile from a patient who, because of a delusion, is convinced that the infant should be harmed. The distinction has immediate consequences for the level of risk.

The final formulation should specify diagnosis, severity, timing of onset, anxious, psychotic or mixed features, bipolarity, suicidality, comorbidity, functioning, breastfeeding and treatment preferences. These elements allow treatment to be selected without reducing the patient to a screening score.

Treatment and Prognosis

Treatment is individualized according to severity, risk, previous responses, bipolarity, breastfeeding, medical conditions, preferences and available support. Psychotherapy, pharmacotherapy and social interventions are not rigid alternatives: they can be combined according to the presentation and degree of impairment.

For mild or moderate depression, validated psychotherapies such as CBT and interpersonal psychotherapy are central options. IPT is particularly suited to role transition, changes in the couple relationship and social support; CBT targets self devaluation, thoughts of inadequacy, avoidance and loss of rewarding activities.

Treatment should also address sleep, pain, practical support, division of the caregiving burden, violence or social difficulties when present. Ensuring periods of protected sleep may have clinical value, especially in patients with bipolar vulnerability, without presenting sleep as sufficient treatment for depression.

In moderate to severe forms, or when psychotherapy is insufficient or not preferred, antidepressants constitute a validated treatment. SSRIs are frequently used; according to ACOG, when there is no previous history of pharmacological response, sertraline or escitalopram are reasonable first line options. If a patient has previously responded well to an antidepressant, that information carries substantial weight in the choice.

Breastfeeding is not an automatic contraindication to antidepressant therapy. The decision considers infant exposure, age and prematurity, maternal dose, previous efficacy and the risk of untreated depression. Sertraline generally has low concentrations in breast milk and is frequently preferred; other antidepressants may also be compatible depending on the case.

It is generally not appropriate to automatically replace an effective medication taken during pregnancy solely to begin breastfeeding. For example, fluoxetine may result in greater infant exposure than some other SSRIs, but when it was effective during pregnancy or other drugs have failed, changing treatment must be weighed against the risk of maternal destabilization.

A specific new development is zuranolone, a positive allosteric modulator of GABA-A receptors and a neuroactive analogue used as a 14 day oral course. Randomized studies have shown rapid symptom improvement in women with postpartum depression. In the European Union, Zurzuvae received authorization in September 2025 for treatment of postpartum depression in adult women after childbirth.

Zuranolone can cause somnolence, sedation, dizziness and impairment of abilities needed for hazardous activities; prescribing and counseling should follow updated regulatory information. Updated LactMed data indicate low amounts in breast milk and do not consider the need for zuranolone alone a reason to stop breastfeeding, but recommend monitoring the infant for excessive sedation, especially in newborn or preterm infants.

The earlier intravenous neurosteroid brexanolone demonstrated efficacy in phase 3 trials and was the first drug specifically approved in the United States for postpartum depression. However, it is no longer marketed in the United States and the FDA withdrew its approval in April 2025 at the manufacturer's request; it therefore now has mainly historical and scientific relevance and should not be presented as a currently available treatment.

In presentations with severe depression, psychotic features, catatonia, high suicide risk or need for a very rapid response, electroconvulsive therapy can be particularly effective. In the postpartum period it can also be used in patients who are breastfeeding, with specialist management of anesthesia and medications.

Postpartum psychosis requires a different pathway from unipolar depression. Hospitalization is generally necessary and treatment may include antipsychotics, lithium and, when indicated, ECT. Contemporary evidence and consensus emphasize the strong connection with bipolar disorder and risk of recurrence, requiring specialist planning for any future pregnancies as well.

When lithium is used postpartum, the decision about breastfeeding requires highly individualized assessment and appropriate monitoring because transfer into milk and neonatal renal immaturity may be clinically relevant. It is incorrect to extend the reassurance applicable to some SSRIs to all psychotropic medications.

Follow up should be close during the early phases and reassess symptoms, suicidality, sleep, functioning and safety. Effective antidepressant therapy is generally continued after remission to reduce relapse risk; duration depends on episode history, severity and recurrence risk, and discontinuation should be gradual.

Prognosis is generally favorable with appropriate treatment, but persistence and recurrences are possible. A history of perinatal depression identifies increased risk in subsequent pregnancies and justifies preventive planning, monitoring beginning during pregnancy and early intervention for prodromal symptoms.

Remission should include recovery of functioning, sleep to the extent possible within the family context, self care capacity and participation in relationships. The therapeutic goal is not to impose an ideal model of motherhood, but to restore health, safety and autonomy.

Complications

The most serious complication is suicide. Thoughts of death, ideation, intent and plan should be assessed directly and repeatedly. A previous attempt, severe depression, agitation, insomnia, substance use, psychosis and poor support increase clinical concern.

Postpartum psychosis is not properly an inevitable complication of depression, but a differential diagnosis and emergency that can include psychotic depressive episodes. The distinction is crucial because the risk of disorganized behavior, suicide or harm to the infant often requires immediate protection and hospitalization.

Depression can cause severe functional impairment, making eating, hygiene, healthcare appointments, household management and infant care difficult. In severe cases, concrete support must be organized rather than limiting treatment to medication or psychotherapy prescriptions.

Difficulties in the mother infant relationship may emerge through emotional detachment, intense anxiety or reduced responsiveness. These manifestations are clinical associations and should not be used to blame the patient. Early interventions can support both maternal remission and family interactions.

Depression is associated with a higher likelihood of early discontinuation of breastfeeding, but the relationship is bidirectional and influenced by many variables. The choice whether to breastfeed should be supported without stigma; the priority is the health of mother and infant.

Couple conflict, isolation and financial difficulties may worsen during the episode and become perpetuating factors. When violence or coercion is present, patient safety requires a specific pathway and cannot be treated as a simple “relationship difficulty”.

Untreated depression may become persistent and increase the risk of relapses and recurrences. Incomplete remission before a subsequent pregnancy may increase vulnerability and makes preconception and perinatal planning particularly important in women with a significant history.

An incorrect unipolar diagnosis in a patient with bipolar disorder can delay appropriate treatment and increase the risk of manic or psychotic episodes. Longitudinal reassessment is therefore a safety measure, especially with rapid onset, family history of bipolar disorder or antidepressant induced activation.

Obsessive thoughts of harm can cause avoidance of the infant and severe distress even when there is no intent to act. Failure to distinguish between OCD and psychosis can result in both underestimation and overestimation of risk and should be avoided through careful specialist interviewing.

The overall consequences may involve the mother, infant and family system, but they are not inevitable. Screening linked to care, correct diagnosis, timely treatment and social support reduce the duration of suffering and promote full recovery.