Perinatal Mental Health: Pregnancy, Postpartum, and the Mental Health Crisis Nobody Talks About

Perinatal refers to the period from conception through 12 months postpartum — not just the weeks after delivery. Mood and anxiety symptoms often begin during pregnancy itself, and waiting to screen until the 6-week postpartum check means missing months of treatable illness.

Perinatal Depression

More than sadness. The hallmark is often anhedonia — an inability to feel joy or pleasure, including in interactions with the baby. Patients describe feeling flat, disconnected, like they are going through the motions. Irritability, not sadness, is frequently the dominant presentation. The expectation that new parents should feel overwhelmingly happy amplifies guilt and shame and keeps people from disclosing.

Perinatal Anxiety

Anxiety is actually more common than depression in the perinatal period and is more frequently missed by providers. Hypervigilance — constant scanning for threats to the baby, inability to relax, inability to sleep even when the baby is sleeping — is characteristic. The Edinburgh Postnatal Depression Scale (EPDS) screens primarily for depression; adding the GAD-7 is necessary to catch perinatal anxiety. Anxiety is the reason to screen with both instruments, not just the EPDS.

Perinatal OCD

Perinatal OCD is critically misunderstood and one of the most feared diagnoses to disclose. The defining feature is ego-dystonic intrusive thoughts — thoughts completely contrary to the person's values and deeply distressing. The most common: intrusive thoughts about accidentally or intentionally harming the baby. These thoughts are horrifying to the person having them precisely because they are so contrary to their feelings.

This is the clinical distinction that changes everything: ego-dystonic harm thoughts in perinatal OCD are not a risk factor for harming the baby. They are intrusive anxiety-driven content. The person is distressed by the thought, not drawn toward it. Patients with perinatal OCD avoid disclosing because they fear their baby will be taken away. This fear delays treatment and compounds suffering. Providers who understand this distinction can respond in a way that keeps patients in care.

Perinatal PTSD

Traumatic birth experiences — emergency C-sections, severe pain without adequate support, loss of control, feeling unheard — can produce genuine PTSD. Childhood trauma or prior sexual trauma can also resurface with pregnancy's physical intimacy demands and the vulnerability of labor. Flashbacks, hyperarousal, avoidance, and emotional numbing are all possible presentations. Birth trauma PTSD is underrecognized because medical focus shifts to the newborn after delivery.

Postpartum Psychosis

Rare (1–2 per 1,000 births) but a psychiatric emergency. Postpartum psychosis presents within 1–2 weeks of delivery — often within days — with rapidly shifting mood, confusion, hallucinations, delusions, disorganized behavior, and in severe cases, command hallucinations. This is not baby blues. It is not severe depression. It is a distinct condition requiring immediate psychiatric evaluation and typically inpatient care.

Women with bipolar disorder are at 25–50% risk of postpartum psychosis. Lithium prophylaxis started immediately postpartum significantly reduces this risk and should be discussed during pregnancy for at-risk patients.

The underdiagnosis of perinatal mood and anxiety disorders is not accidental. It is the product of several overlapping barriers:

• The “glow” expectation — cultural scripts about pregnancy and new parenthood as universally joyful make it psychologically costly to disclose that you are struggling. Patients minimize or conceal symptoms because they feel they “should” be happy.
• Provider assumption that pregnancy is protective — a historic belief (now disproven) that pregnancy hormones protect against psychiatric illness. They do not. Pregnancy is a high-risk period for mood disorder onset or recurrence, especially in women with a prior psychiatric history.
• Symptom dismissal as “normal hormones” — fatigue, tearfulness, and anxiety are attributed to normal perinatal changes rather than screened as potential disorder symptoms.
• Perinatal OCD especially missed — patients with intrusive harm thoughts rarely disclose them voluntarily. Providers rarely ask. The result is that perinatal OCD goes undetected and untreated for months or years.
• Screening gaps — the Edinburgh Postnatal Depression Scale is often administered only at the 6-week postpartum visit. Many providers screen once; best practice is screening at every prenatal and postnatal visit through 12 months.

• Prior psychiatric history — the single strongest predictor. A prior depressive episode carries a 50–80% recurrence risk in the perinatal period, especially if SSRIs are discontinued during pregnancy.
• SSRI discontinuation in pregnancy — stopping SSRIs during pregnancy to “be safe” carries substantial relapse risk. This is a decision that requires a careful risk-benefit analysis with a prescriber, not a default.
• Birth complications or NICU admission — NICU admission is a direct trauma trigger. Maternal anxiety, PTSD, guilt, and delayed bonding all spike in NICU families. This population is systematically under-screened, despite having some of the highest PMAD rates of any perinatal subgroup.
• Lack of social support — social isolation is a potent risk multiplier.
• Trauma history — childhood trauma and prior sexual trauma are strong risk factors. Pregnancy and delivery can trigger trauma responses in ways patients and providers do not anticipate.
• Thyroid dysfunction — postpartum thyroiditis affects up to 10% of women and can mimic or exacerbate mood disorders. TSH should be screened if depressive or anxiety symptoms emerge postpartum.

The most common thing I hear from patients with perinatal depression is: “I don't want to take medication because of the baby.” This is completely understandable and based on incomplete information. The clinical reality is this: untreated maternal depression is not safer for the baby than treated depression.

Untreated perinatal depression and anxiety are associated with poor prenatal care adherence, inadequate nutrition, preterm birth, low birth weight, impaired maternal-infant bonding, and disrupted infant neurodevelopment. These are not hypothetical risks. They are documented in the literature. The risk-benefit conversation is not “medication risk vs. no risk.” It is “medication risk vs. untreated illness risk” — and untreated illness frequently carries more risk.

SSRIs: First-Line with the Strongest Safety Record

Sertraline has the most reproductive safety data of any antidepressant — decades of evidence, the largest pregnancy registries, and the lowest transfer to breast milk. It is the appropriate first-choice SSRI for most perinatal patients. Escitalopram is also well-studied and a reasonable first-line option. Both carry strong safety evidence that should be part of every informed consent conversation.

Paroxetine has historically been flagged (Category D under the old system) due to an early signal for cardiac defects. Current data do not support avoiding it categorically, but it is reasonable to favor sertraline or escitalopram when starting a new SSRI in pregnancy. If a patient is stable on paroxetine, the case for switching needs careful risk-benefit analysis — discontinuation-related relapse risk is real.

Mood Stabilizers: Revised Risk Estimates

Lithium and the historic Ebstein anomaly concern: early case reports suggested a 400-fold increased risk of Ebstein anomaly. Revised data show a much smaller absolute risk (approximately 0.6%, compared to a background rate of ~0.1–0.2%). For women with bipolar disorder, lithium's efficacy for postpartum psychosis prevention may outweigh this risk in many cases. This is a nuanced conversation that should happen well before conception.

Lamotrigine levels drop 40–50% during pregnancy due to increased renal clearance and CYP450 induction. Dose increases are typically needed during pregnancy to maintain therapeutic levels, and levels should be monitored closely. After delivery, levels rebound rapidly — dose reductions are needed postpartum to avoid toxicity.

Valproate is associated with a 5–9% rate of neural tube defects, and neurodevelopmental risks including lower IQ. It is generally avoided in pregnancy when alternatives exist.

Benzodiazepines in Pregnancy

The concern: early data suggested teratogenicity. Revised evidence does not support significant teratogenic risk at low doses. The primary perinatal concern with benzodiazepines is transient neonatal symptoms (hypotonia, respiratory depression, withdrawal) if used near delivery. Low-dose short-term use for acute anxiety or sleep, when clinically necessary, is a different risk calculation than chronic high-dose use.

The best-studied medications in breastfeeding:

• Sertraline — the most well-studied antidepressant in breastfeeding. Minimal transfer to breast milk; infant plasma levels typically undetectable. First-line choice.
• Paroxetine — low transfer to breast milk despite its Category D pregnancy designation. Reasonable option for breastfeeding when a patient is already stable on it.
• Nortriptyline — well-studied in breastfeeding; minimal infant exposure. An option for patients who respond poorly to SSRIs.

The NIH's LactMed database is the best free resource for medication-specific breastfeeding data — both patients and prescribers can look up any medication and review the evidence on transfer and infant effects.

The principle that matters most: monitor the infant for sedation, feeding difficulties, or unusual behavior when any psychiatric medication is used during breastfeeding. Most infants show no adverse effects; monitoring provides the safety net.

The EPDS is a 10-item validated screening tool for perinatal depression. It is widely used, well-validated, and imperfect.

What it screens for: depressive symptoms — anhedonia, low mood, tearfulness, self-blame, suicidal ideation (question 10, which should always be reviewed directly).

What it misses: anxiety. The EPDS does not capture generalized anxiety, panic, or hypervigilance well. Anxiety is more common than depression in the perinatal period. Using the EPDS alone means systematically missing the most common perinatal mental health presentation. Adding the GAD-7 to every perinatal screening encounter is the minimum standard for comprehensive screening.

SSRIs — Same Efficacy as in Non-Perinatal Populations

SSRIs are as effective in the perinatal period as in other clinical contexts. There is no evidence that pregnancy or postpartum status reduces antidepressant efficacy. Onset of benefit follows the same 4–8 week timeline.

Interpersonal Therapy (IPT)

IPT is the psychotherapy with the strongest evidence base for perinatal depression specifically. It addresses role transitions, interpersonal conflicts, and grief — all of which are structurally present in the perinatal period. Effect sizes in randomized trials are comparable to antidepressants for mild-to-moderate perinatal depression.

Postpartum Support International (PSI)

PSI operates a free warmline (1-800-944-4773) and maintains peer support groups, provider directories, and population-specific resources (NICU families, partners, same-sex couples, fathers, adoptive parents). PSI is one of the best-organized patient support organizations in mental health and should be given to every perinatal patient as a resource.

Partner Education

Partners who understand PMADs are a clinical asset. Practical sleep support — a partner handling one nighttime feeding to provide a protected sleep window — is as clinically significant as medication in some mild-to-moderate presentations. In bipolar disorder, sleep deprivation is a direct episode trigger. Sleep protection is a medical intervention, not a comfort preference.