Article Text

Download PDFPDF

Prescribing for pregnancy: managing prescribing for women with mental health diagnoses
  1. Louise Page
  1. Obstetrics and Gynaecology, Chelsea and Westminster Hospital NHS Foundation Trust, London, UK
  1. Correspondence to Louise Page, Obstetrics & Gynaecology, Chelsea and Westminster Hospital NHS Foundation Trust, London, UK; Louise.Page{at}

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Key learning points

  • Mental illness during pregnancy or in the first year after birth is common

  • Preconception care is important to optimise and stabilise medication and the mental health condition prior to conception

  • Aim to prescribe the lowest effective dose, with monotherapy where possible

  • Advise women not to stop medication without discussion with a practitioner experienced in caring for women with mentalillness in pregnancy

  • A multi-professional approach, with the woman and her family at the centre is key

  • There are many safe options for pharmacological treatment for women with mental illness in the perinatal period

Introduction and general principles

Up to 20% of women will experience a mental health problem during pregnancy or in the first year after birth.1 The impact can be widespread affecting not only the woman but also her partner and their children. The economic impact of mental illness for a 1-year cohort of births in the UK is over £8 billion.2 Women may enter pregnancy with an existing mental health diagnosis, for others illness can develop for the first time during pregnancy or in the postpartum period. Many women will benefit from psychological interventions and certainly for women with mild to moderate illness this should be the first-line therapeutic intervention.3 Some women will be taking psychotropic medication prior to pregnancy and ideally, they would have received preconception care in order to optimise and stabilise their medication and their condition prior to conception. The preconception consultation should also take the opportunity to share general advice for women such as optimising their body mass index (BMI), stopping smoking and starting folic acid supplementation. For more details on general prepregnancy care, please refer to the first article in this series (Girling J. Prescribing for pregnancy: general prepregnancy care. DTB 2019;57:168–72).4

When a pharmacological intervention is required in pregnancy, the aim should always be to prescribe the lowest effective dose, with monotherapy where possible and it should not be taken for longer than needed. Due to the increased volume of distribution and accelerated metabolism in pregnant women, the dose of some medications, such as lamotrigine may need to be increased during pregnancy to gain the same therapeutic effect.

There are significant risks of relapse if women suddenly stop medication. Women should never be advised to stop medication without discussion with a practitioner experienced in caring for women with mental illness in pregnancy. The decision about continuing, stopping or changing medication should be made by a multi-professional team that includes the woman. Consistent advice and approach to care is key. Communicating recommended medication changes to the woman’s general practitioner is vital as they may continue to be the primary prescriber for the woman during pregnancy and the postnatal period. Changing medication during pregnancy exposes the fetus to an additional agent and so should only be recommended where there is clear clinical benefit expected from altering a woman’s regime.

Nausea and vomiting are not uncommon in pregnancy, especially in the first trimester and therefore a woman may require an antiemetic medication in addition to her psychotropic medication. First-line antiemetic choices include cyclizine, prochlorperazine, promethazine or chlorpromazine.5 In the UK, a product containing doxylamine and pyridoxine is licensed for the treatment of nausea and vomiting in pregnancy in women who do not respond to conservative management.6

Women’s compliance with medication may change during pregnancy for many reasons, most commonly due to her concern about the effect of the psychotropic medication on the developing fetus. Women may also make infant feeding decisions based on concerns of medication passing to the infant through breast milk. It is important to discuss the risks (or indeed the safety) of medication in the context of the potential risks to the baby if a mother experiences worsening mental health during pregnancy or in the postnatal period.

How might the mental illness affect the pregnancy?

Women with mental illness are more likely to smoke tobacco, drink alcohol or misuse substances, all of which are known to have a deleterious effect on pregnancy. In addition, they may not make use of all available antenatal care opportunities.7

There are reported associations between maternal mental illness and pregnancy outcomes such as fetal growth restriction and preterm birth. The degree to which these risks can be ameliorated by treatment is less clear.7 Increasingly the association between maternal mental illness and paternal mental illness is becoming recognised.

Women with mental illness are more likely to have a raised BMI and in addition to this risk factor for gestational diabetes, those who are being treated with certain antipsychotic medications notably olanzapine, quetiapine and clozapine are at a further increased risk and so should be offered screening with a glucose tolerance test at 28 weeks and women with additional risk factors for gestational diabetes may be offered an additional screening test at 34 weeks.8

How might being pregnant affect maternal mental illness and what might happen after birth?

Women with serious mental illness should be under the care of a specialist perinatal mental health team, working collaboratively with the midwifery and obstetric teams and the woman’s GP.

A care plan should be developed by the healthcare professionals in combination with the woman and ideally her partner or family. The care plan should clearly outline the treatment plans both pharmacological and non-pharmacological for the pregnancy and post-natal period.

The postpartum period is a time when all women are at risk of deterioration in their mental health. The risk of postpartum psychosis is highest in the first 2 weeks after birth. Certain women are at significantly increased risk of postpartum illness—the risk of postpartum relapse for women with bipolar disorder can be up to 75% if they have not taken prophylactic pharmacotherapy during pregnancy.9 The risk of a recurrence of postpartum psychosis rises to a 50% risk in women who have previously experienced the illness. Higher risk women may start an antipsychotic medication shortly after delivery with the aim to ameliorate the risk.10

Is her current medication suitable for pregnancy? Are the medications she uses for a flare/exacerbation safe?

There is a background rate of congenital malformation of 2%–4%. Studies in this area need caution with interpretation as many are retrospective, there may be reporting bias or increased screening leading to increased detection. Confounding factors need also to be considered, women with mental illness are more likely to be smokers and they may also exposed to other factors that can affect fetal development such as obesity, social adversity, alcohol and substance misuse.

Selective serotonin reuptake inhibitors (SSRIs)

There are conflicting and confounded data regarding any increased risk of congenital malformations in babies exposed to SSRIs. Although the subject of on-going research, advising the discontinuation of SSRIs is not supported by the current evidence base.11 GPs, who are often the first healthcare professional a woman will interact with in early pregnancy, should support women to continue treatment with an SSRI prior to her referral for maternity care.

There appears to be an increased risk (0.3% vs 0.2% in the background population) of persistent pulmonary hypertension of the new-born (PPHN) in babies exposed in utero beyond 20 weeks’ gestation. The absolute risk of PPHN in babies exposed to SSRIs is low.7 Sertraline carries the lowest risk.12 The new-born check for a baby exposed to SSRIs in utero should include pulse oximetry as part of the cardiovascular assessment.13

Babies exposed to SSRIs in late pregnancy are at risk of a transient neonatal withdrawal syndrome and particular attention should be paid to central nervous system, motor, respiratory and gastrointestinal symptoms in the new-born period.14 Symptoms occur between 8 and 48 hours after birth and generally have resolved by 72 hours. Up to 30% of babies exposed to SSRIs may show features of neonatal transient withdrawal syndrome, the symptoms are usually mild and often require observation only.13 Reassuringly follow-up studies up to 8 months of age show no difference between exposed and control infants. It is important to note that neonatal withdrawal syndrome is a diagnosis of exclusion. The symptoms overlap with other more common neonatal complications such as sepsis, hypoglycaemia and polycythaemia and care must be taken not to delay an alternative diagnosis.13

Second-generation antipsychotics

The data on the use of aripiprazole and olanzapine in pregnancy are limited. They do not appear to be associated with an increased risk of teratogenicity following exposure in the first trimester.7 A single study in women taking aripiprazole suggests an increased risk of low birth weight or preterm delivery but the data are insufficient to prove or quantify the risk. Maternal hyperglycaemia has been seen in women taking olanzapine, and therefore as previously mentioned screening for gestational diabetes is recommended.14


The evidence base for quetiapine is larger and does not suggest an increased risk of congenital malformations following first trimester exposure. There also does not appear to be a relationship with later pregnancy complications. As with olanzapine screening for gestational diabetes is recommended.14


There appears to be an increased risk of congenital malformation in babies exposed to risperidone (5% vs 3%) in unexposed population.7

All babies exposed to antipsychotic medication should be observed for withdrawal symptoms or signs of poor neonatal adaptation syndrome (PNAS). The frequency of withdrawal symptoms is unclear.7 15 Some babies require observation; others will require admission to a neonatal unit for supportive care.16 For women with mental illness who are well while taking second-generation antipsychotic medications, the benefits of continuing their medication during pregnancy may outweigh any potential risks of stopping or changing their medication.14


There is a potential small increase in the risk of a cardiac defect (Ebstein’s anomaly) associated with lithium exposure in the first trimester. The evidence came from a retrospective study and the findings have not been replicated in other studies.14 It has been suggested that an increase in the prevalence of Ebstein’s anomaly may be associated with maternal mental illness itself rather than lithium exposure.17 Later in the pregnancy lithium use may be associated with polyhydramnios due to fetal polyuria, third trimester ultrasound scanning should be considered.18

Lithium levels require more frequent monitoring in pregnancy, at least monthly initially and then weekly from 36 weeks’ gestation. It is common to need to increase the lithium dose during pregnancy in order to stay within the narrow therapeutic window. As women may be at particular risk of relapse after birth, clinicians may aim for a target lithium level at the upper level of the therapeutic window in the postnatal period.

Sodium valproate

Due to the association with sodium valproate exposure and both increased congenital malformations (up to 10%) and developmental disorders (up to 40%) sodium valproate should not be prescribed for mental health reasons for women of childbearing age. Sodium valproate should only be prescribed for girls and women with epilepsy when no alternative medication is available and when a pregnancy prevention plan is in place.3 Should a woman conceive while taking sodium valproate she should be advised to continue the medication and be referred for an urgent discussion with a practitioner experienced in caring for women taking sodium valproate in pregnancy. She should be supported in her choice to continue or to interrupt the pregnancy.19


First trimester exposure to benzodiazepines is not associated with increased major congenital malformations although may be associated with an increased risk of spontaneous miscarriage.7 20 Prolonged use of benzodiazepines near term increases the risk of neonatal withdrawal syndrome, the effect is dose related. Where possible it is advised to use shorter acting preparations such as lorazepam. Neonatal withdrawal can start two or 3 days after birth. Infants should be observed for symptoms but not separated from their mothers in anticipation of symptoms.13

Are any of the drugs problematic in labour?


Serum lithium levels and fluid balance require careful monitoring during the intrapartum and early postnatal window, as dehydration or hypovolaemia can lead to toxicity. The omission of multiple doses of lithium during a long intrapartum period requires careful consideration as this may increase the risk of postpartum relapse. Lithium should be stopped 24–48 hours before a planned induction of labour or caesarean section, or at the onset of labour with a lithium level taken 12 hours after the last dose.7 If this level is not above the therapeutic range then restart lithium on day one postnatal and recheck the level in 1 week. The baby should have a lithium level checked shortly after birth.14

Can she breast feed on these medicines?

The majority of psychotropic medications are safe for use during lactation. It is useful to consider the relative infant dose (RID) when discussing medication safety with mothers. The RID is calculated by dividing the infant’s dose via milk (in mg/kg/day) by the mother’s dose (in mg/kg/day). An RID of less than 10% is considered a safe level during breastfeeding.21 The RID for medications is available using the LactMed database.22

Breastfeeding may have an effect on sleep and daily routine that may be detrimental for a mother with physical or mental health problems. A lack of sleep is a common precipitating factor for a relapse of severe psychiatric illness. Mixed feeding or the use of expressed breast milk or artificial milk may be a way to optimise the mother’s well-being, which in turn may on balance offer the most benefit to the infant. Following an informed decision-making conversation, the mother’s choices for infant feeding should be respected.22

Breastfeeding while on lithium is generally not advised—the RID is 11%–42%. The main concern would be with neonatal toxicity due to the narrow therapeutic window. If a mother wishes to breastfeed while taking lithium, close monitoring of the maternal serum levels is required and some clinicians would also recommend monitoring of the infant’s lithium levels too.22

Sertraline has the lowest RID (<0.5%) for the SSRIs. However, all would be suitable for a mother to take during lactation, especially if her mood is already stablished on an alternative SSRI.

Second-generation antipsychotic medications are generally considered suitable for use during lactation. Olanzapine and quetiapine have a low RID, with a slightly higher RID for aripiprazole. Aripiprazole can also affect prolactin levels which may in turn have an effect on milk production.22 Breastfeeding is usually not recommended for women taking clozapine as there is an association with excessive sedation, seizures and agranulocytosis.13 22

Specialist input is required when a woman chooses to breastfeed while taking benzodiazepine as there are data to suggest an incidence of neonatal adverse events of 1.6% and an association with increased maternal sedation.7 If treatment with benzodiazepine is required for a lactating mother the use of a short acting agent such as lorazepam should be considered.13

Is the baby likely to develop mental illness?

There is an established association between perinatal mental illness and an increased risk of psychological and developmental difficulties in the child. There is growing evidence that the effect is not just related to maternal mental health but also related to paternal mental health. However, with recognition and timely intervention to include improving parent–child interactions these difficulties can be minimised.23 24


There are many safe options for pharmacological treatment for women with mental illness in the perinatal period. The evidence base surrounding medication safety is growing as is the availability of non-pharmacological interventions and specialist perinatal mental health teams. Collaborative working across specialities with the woman and her family at the centre of decision making is key to optimising outcomes for the mother and child.

This article is part of a series that discusses the need to actively manage prescribing before and during pregnancy. Women with long-term conditions need targeted preconception guidance about the impact of their health on pregnancy outcome, the potential impact of pregnancy on their medical or mental health disorder and use of their medication in pregnancy and breastfeeding.25 Abruptly stopping long-term medication because of pregnancy risks relapse at the very time that optimal control and wellbeing is needed for mother and fetus.4

Information for patients

  • Pregnancy and breastfeeding whilst taking medication for mental illness is safe. Occasionally changes may be needed to optimise your medications during this time. Please speak to your doctor before making any changes to your medication. If you are taking medication and are considering becoming pregnant, please discuss this with your healthcare team before stopping contraception.

Supplemental material



  • Competing interests None declared. Disclosure of conflicts of interest form(s) are published online as supplementary files.

  • Provenance and peer review Commissioned; externally peer reviewed.