Section 6: Special Populations: Perinatal Anxiety

Perinatal mental health screening has historically centered on depression. That focus was necessary but incomplete. Contemporary epidemiological data reveal that perinatal anxiety disorders are at least as prevalent as perinatal depression, frequently more functionally impairing, and carry independent risks to obstetric outcomes and fetal development.

6.1 Prevalence: Anxiety Is the Dominant Perinatal Psychiatric Condition

The standard perinatal screening paradigm focuses on depression. The data suggest it should focus equally on anxiety. A Bayesian meta-analysis by Fawcett and colleagues found that approximately 1 in 5 women (20.7%) meet diagnostic criteria for at least one anxiety disorder during the peripartum period, a rate higher than the commonly cited perinatal depression prevalence of approximately 12%. The data establish that specific phobias and GAD are the most prevalent individual diagnoses during pregnancy and the postpartum period, with OCD showing the highest relative increase compared to pre-pregnancy baseline rates. (Fawcett et al., 2019, Journal of Clinical Psychiatry)

Comorbid anxiety and depression during the perinatal period (CAD) affects approximately 9% of women globally (95% CI: 8-10%), based on a 2025 meta-analysis of 122 studies across 43 countries. This rate was stable across income levels and pandemic context, suggesting a consistent underlying biological vulnerability. (Ou et al., 2025, Psychological Medicine)

The burden extends to the paternal dyad. A meta-analysis of 23 studies found that up to 3.18% of parental dyads experience concurrent psychiatric distress. (Roddy Mitchell et al., 2020, Journal of Affective Disorders)

6.2 Clinical Presentations: The Diagnostic Traps

Three perinatal presentations generate frequent diagnostic confusion:

Infant-focused health anxiety vs. perinatal OCD. It is clinically established that 70-100% of new mothers experience fleeting, unbidden intrusive thoughts about harm to their infant. (Fairbrother & Woody, 2008, Journal of Anxiety Disorders) These are normative. The critical distinction is functional: in normative maternal vigilance, the thoughts are briefly distressing but the mother can redirect her attention. In perinatal OCD (affecting approximately 3-5% of perinatal women), the intrusive thoughts are ego-dystonic, repetitive, intensely distressing, and drive compulsive behaviors (checking, avoidance, mental rituals). The mother with perinatal OCD typically avoids being alone with her infant, not because she wants to harm the child, but precisely because she is terrified of the intrusive thought that she might. (Fairbrother & Woody, 2008, Journal of Anxiety Disorders)

Perinatal OCD vs. postpartum psychosis. This distinction is safety-critical. In perinatal OCD, the mother is horrified by her intrusive thoughts, recognizes them as irrational, and takes protective action (avoidance of the trigger). In postpartum psychosis, the mother may experience command hallucinations or delusional beliefs about the infant (e.g., "the baby is possessed") with impaired reality testing. The ego-dystonic nature of OCD intrusions is the key differentiator. Mothers with OCD are at no elevated risk of acting on their intrusive harm thoughts.

Tocophobia (severe fear of childbirth). This condition drives maternally requested cesarean sections (accounting for a clinically meaningful proportion of elective C-sections globally), avoidance of future pregnancies, and severe perinatal distress including insomnia, sexual dysfunction, and somatic complaints. (O'Connell et al., 2017, Acta Obstetricia et Gynecologica Scandinavica)

In Plain Language for Families:

Almost every new parent has the occasional scary thought about > something bad happening to their baby. That is normal; the brain is > running a constant safety check on the thing it is most invested in > protecting. It becomes a problem when those thoughts start running > on a loop, when the parent cannot stop checking or avoiding > situations because of the thoughts, or when the distress becomes so > severe that it interferes with bonding. If a new mother says she is > afraid to be alone with her baby because of unwanted, frightening > thoughts, she is almost certainly experiencing perinatal OCD, not > a danger signal. She needs reassurance that these thoughts do not > reflect her character, and she needs treatment.

6.3 The Stakes: Untreated Anxiety Harms Pregnancy Outcomes

The question of whether to treat anxiety in pregnancy is sometimes framed as a trade-off between medication risk and maternal comfort. The obstetric data reframe it as a trade-off between medication risk and fetal risk. A meta-analysis by Grigoriadis and colleagues quantified the obstetric risks of untreated antenatal anxiety. (Grigoriadis et al., 2018, Journal of Clinical Psychiatry)

Preterm birth. Antenatal anxiety is associated with significantly increased risk of overall preterm birth (OR = 1.54, 95% CI: 1.39-1.70). Spontaneous preterm birth, excluding medically induced deliveries, also shows a significant association (OR = 1.41, 95% CI: 1.13-1.75). These effect sizes are comparable in magnitude to established obstetric risk factors like maternal smoking and advanced maternal age. (Grigoriadis et al., 2018, Journal of Clinical Psychiatry)

Fetal growth. Increased odds of low birth weight (OR = 1.80, 95% CI: 1.48-2.18), with a mean absolute weight reduction of approximately 56 grams. Increased odds of small-for-gestational-age classification (OR = 1.48, 95% CI: 1.26-1.74). Statistically significant reduction in fetal head circumference (mean difference = 0.25 cm). (Grigoriadis et al., 2018, Journal of Clinical Psychiatry)

The biological mechanism is fetal programming through the HPA axis: chronic maternal cortisol crosses the placenta, prematurely activating fetal stress response systems.

6.4 SSRI Safety in Pregnancy: Absolute Risk, Not Relative Fear

When non-pharmacological interventions are insufficient, SSRIs are indicated. ACOG (2023) establishes SSRIs as the recommended first-line pharmacotherapy for moderate-to-severe perinatal anxiety. (ACOG Clinical Practice Guideline No. 5, 2023)

Sertraline is the preferred agent. A Bayesian meta-analysis of National Birth Defects Prevention Study data found no confirmed associations between first-trimester sertraline exposure and any specific birth defect. (Reefhuis et al., 2015, BMJ)

Paroxetine carries an FDA warning. The same analysis confirmed associations between paroxetine and specific malformations: right ventricular outflow tract obstruction defects (OR approximately 2.4), anencephaly (a neural tube defect, OR approximately 3.2), and atrial septal defects (OR approximately 1.8). Paroxetine should be avoided in patients who are or may become pregnant. (Reefhuis et al., 2015, BMJ)

The absolute risk imperative. Even for paroxetine, the absolute risk of cardiac malformation remains below 1%. The paroxetine OR of approximately 2.4 applies to right ventricular outflow tract obstruction defects specifically, not all cardiac defects. The baseline rate for these specific malformations is far lower than the overall cardiac defect rate of approximately 1 per 100 births. The resulting absolute risk increase, even at the upper bound of the OR, remains well under 1%. Clinicians should counsel with absolute numbers, not relative risk multipliers.

PPHN (Persistent Pulmonary Hypertension of the Newborn).

Late-pregnancy SSRI exposure is associated with a 2.5 to 4.9-fold relative increase. The baseline rate is approximately 15 per 10,000 births (0.15%); even at the upper bound, the absolute risk rises to approximately 0.3-0.7%. The psychiatric consensus is that this risk does not justify discontinuing effective treatment for moderate-to-severe anxiety. (Huybrechts et al., 2015, JAMA)

Neonatal Adaptation Syndrome (NAS). A 2024 retrospective cohort of 280,090 term infants found NAS in 11.2% of SSRI-exposed neonates versus 4.4% of unexposed neonates (adjusted OR = 2.14), with symptoms (jitteriness, weak cry, respiratory distress) that are typically mild and self-limiting within 48-72 hours. Note: older guideline estimates (AAP/AACAP) cite NAS rates of 20-30% using broader symptom definitions. The lower figure from the 2024 cohort reflects tighter diagnostic criteria in a large, controlled study. Either way, the syndrome is transient and not associated with long-term neurodevelopmental harm. (Malm et al., 2024)

Warning:

Never advise a psychiatrically stable patient to abruptly > discontinue her SSRI upon discovering a pregnancy. This practice > carries a high risk of psychiatric relapse, which itself is > associated with worse obstetric outcomes than continued SSRI > exposure. If a medication change is warranted (e.g., switching from > paroxetine to sertraline), do so with a structured taper plan, not > by abrupt cessation. The risks of untreated severe anxiety in > pregnancy (preterm birth, low birth weight, impaired fetal > development) are well-quantified and consistently exceed the > medication-related risks when SSRIs are prescribed appropriately. > (ACOG, 2023)

6.5 Safety During Lactation

Sertraline has the most favorable lactation profile, with minimal infant serum levels documented across multiple studies. Escitalopram and paroxetine also show low relative infant doses. Fluoxetine has the highest reported infant exposure among SSRIs due to its long half-life and active metabolite.

The general principle: the benefits of breastfeeding combined with maternal mental health treatment typically outweigh the minimal medication exposure for most SSRIs. If a mother is stable on a specific SSRI that successfully controls her anxiety, switching agents solely for breastfeeding is rarely justified; the risk of destabilizing her psychiatric status outweighs the marginal difference in infant exposure between agents.

For detailed SSRI safety profiles in pregnancy and lactation, see Module 3, Section 6.6.

6.6 CBT Adaptations for the Perinatal Period

CBT is effective for perinatal anxiety, with a meta-analytic NNT of approximately 4-5 and response rates around 40-50% compared to treatment-as-usual. (Maguire et al., 2018, Journal of Anxiety Disorders)

However, standard protocols require adaptation:

Sleep restriction must be modified. Absolute sleep deprivation is unavoidable with a newborn, so standard CBT-I sleep restriction protocols (which limit time in bed to compress sleep) cannot be applied rigidly. Involve the partner or support person in nighttime feedings to protect one consolidated sleep block.

Exposure exercises must account for the maternal-infant context. Exposure for perinatal OCD may involve gradually reducing avoidance of being alone with the infant, which requires careful therapeutic scaffolding.

Delivery format matters. Perinatal patients face significant logistical barriers to attending weekly in-person sessions. Telehealth, internet-delivered, and guided self-help CBT formats show equivalent effect sizes and reduce the access barrier.

6.7 Screening Guidance

Do not rely solely on the Edinburgh Postnatal Depression Scale (EPDS). While the EPDS captures some anxiety (the anxiety subscale has modest sensitivity), it is insufficient for diagnosing specific anxiety disorders. Administer the GAD-7 alongside the EPDS at perinatal visits. If OCD is suspected, deploy the FOCI or OCI-R.

PsychHQ

Special Populations: Perinatal Anxiety