Clinical Psychopharmacology · Paper A
Mood stabilisers in pregnancy
Mood stabilisers used in bipolar disorder carry very different risks in pregnancy, so the choice of drug matters as much before conception as during it. A woman of childbearing age newly diagnosed with bipolar disorder has her future pregnancy risk shaped by the drug chosen at diagnosis, long before any pregnancy is planned. Mood stabiliser choice in women who could become pregnant is a decision made in advance, not adjusted only once pregnancy is confirmed.
Sodium valproate is the worst teratogen (drug causing birth defects) of the group. It carries about a 10% risk of neural tube defects, plus a dose-dependent reduction in childhood IQ and increased risk of autism and ADHDAttention-deficit hyperactivity disorder, a condition of inattention, impulsivity and sometimes hyperactivity.Read more → (attention deficit hyperactivity disorder). It must be avoided in women of childbearing potential unless a formal pregnancy-prevention programme, mandated by the MHRA, is in place. Carbamazepine also causes neural-tube and craniofacial defects. Lithium in the first trimester raises the risk of Ebstein's anomaly (a heart defect), though the absolute risk is low. Lamotrigine is comparatively low-risk and often preferred, with antipsychotics such as olanzapine or quetiapine as reasonable alternatives. A woman on long-term valproate who becomes pregnant unexpectedly before any medication review shows exactly why the pregnancy-prevention programme exists: e.g. an unplanned conception turning a stable prescription into a teratogenicCapable of causing a birth defect if taken during pregnancy.Read more → exposure overnight. Clinically, valproate is avoided from the outset in women who could conceive, rather than switched only after a pregnancy test is positive.
A fresh doctor might assume mood-stabiliser safety in pregnancy is decided once a positive pregnancy test appears. By then, the riskiest exposures, valproate and carbamazepine in particular, have often already happened during the first trimester, when major organs are forming. Hold the phrase DECIDE BEFORE CONCEPTION through every section below.
Mood stabilisers (lithium, valproate, carbamazepine, and lamotrigine, plus certain antipsychotics used off-label) treat bipolar disorder, but several carry risk (the potential to cause birth defects) that varies enormously between drugs. TERATOGENIC here specifically means the drug can disrupt normal fetal development if taken during pregnancy, most dangerously during the first trimester when the neural tube and major organs are forming.
A common mistake is reflexively stopping all mood stabilisers as soon as pregnancy is confirmed or suspected. Untreated bipolar relapse in pregnancy carries its own serious risks to mother and baby, so the correct approach is a specialist perinatal psychiatry review weighing the specific drug's risk against relapse risk, not a blanket stop: for example, abruptly stopping lithium can precipitate a severe relapse within days, which is its own danger to a pregnancy.
Consider a woman of childbearing age newly diagnosed with bipolar disorder. The drug started at diagnosis, potentially years before any pregnancy is planned, already shapes her future pregnancy risk if she is not specifically counselled. Valproate carries around a 10% risk of neural tube defects (failures of the spine or brain to close properly in early development) plus a dose-dependent reduction in childhood IQ and an increased risk of autism spectrum disorder and (attention deficit hyperactivity disorder): risks large enough that the MHRA (Medicines and Healthcare products Regulatory Agency) mandates a formal Pregnancy Prevention Programme before valproate can be prescribed to anyone of childbearing potential.
The MHRA Pregnancy Prevention Programme requires that a woman of childbearing potential is only prescribed valproate if she is using effective contraception, has had the risks explained and documented, and has an annual specialist risk review. This is not a suggestion or a soft warning: it is a regulatory requirement introduced because of how large and well-documented valproate's and neurodevelopmental risks are compared with the alternatives.
Carbamazepine also causes neural-tube and craniofacial defects, though generally considered somewhat less risky than valproate; folate supplementation is recommended. Lithium taken in the first trimester raises the risk of Ebstein's anomaly (a congenital heart defect involving the tricuspid valve), though the absolute risk remains low, and stopping lithium abruptly in pregnancy carries its own relapse risk: the decision needs specialist risk-benefit counselling rather than automatic discontinuation. Lamotrigine is comparatively low-risk and often the preferred mood stabiliser in women who may become pregnant, though its blood levels can fall as pregnancy progresses, needing dose monitoring.
Valproate's risk (neural tube defects, dose-dependent IQ reduction, autism, ) is substantially higher than lamotrigine's, so the two should never be treated as interchangeable "mood stabiliser" risk. Lithium's specific risk (Ebstein's anomaly, a cardiac malformation) is mechanistically and statistically different from valproate's neural-tube risk: different investigations follow each, which is exactly why naming the correct drug, not just "a mood stabiliser", matters in any vignette.
There is no single "diagnosis" here in the sense of confirming a disease: instead, risk is identified through the drug history itself and confirmed exposures are actively screened for in pregnancy. Fetal echocardiography (an ultrasound scan of the fetal heart) around 18-20 weeks is the specific investigation used to look for Ebstein's anomaly after lithium exposure, since standard anomaly scanning is not detailed enough to reliably pick up this particular cardiac defect. Detailed anomaly scanning is used after valproate or carbamazepine exposure, specifically to look for neural tube and craniofacial defects. Matching the right investigation to the right drug exposure (echocardiography for lithium, detailed anomaly scan for valproate or carbamazepine) is itself the diagnostic skill being tested.
Any woman of childbearing potential started on valproate should be enrolled in the MHRA Pregnancy Prevention Programme, with contraception and annual review documented. Pre-conception planning: ideally switching away from valproate and carbamazepine to a lower-risk option such as lamotrigine, starting high-dose folic acid, and arranging specialist perinatal psychiatry review: should happen well before a planned pregnancy, not after a positive test. Where lithium continues in pregnancy, levels need closer monitoring since renal changes across pregnancy, and fetal echocardiography around 18-20 weeks screens specifically for Ebstein's anomaly. Throughout, management is a continuous risk-benefit conversation between drug risk and relapse risk, run by specialist perinatal psychiatry and following NICE guidance on antenatal and postnatal mental health (NICE NG192), not a single one-off decision.
Vignettes often describe a woman of childbearing age newly starting valproate without any mention of contraception or counselling: the expected recognition is a breach of the MHRA Pregnancy Prevention Programme requirements. Another trap describes a pregnant woman on lithium whose family wants it stopped immediately: the expected answer weighs relapse risk and specialist review rather than automatic discontinuation, and a further trap tests whether candidates can match Ebstein's anomaly specifically to lithium rather than to valproate or carbamazepine.
Valproate carries the highest and neurodevelopmental risk of the mood stabilisers and needs the MHRA Pregnancy Prevention Programme in any woman of childbearing potential, lithium carries a smaller but real risk of Ebstein's anomaly needing fetal echocardiography rather than automatic stopping, and the safest path for all of them is planning before conception, with specialist perinatal psychiatry involved early.