Babies of women with bipolar disorder are more likely to be born preterm, irrespective of whether the mother had received mood stabilising drugs, a Swedish cohort study has found. Infants whose mothers had untreated bipolar disorder had an increased risk of microcephaly and neonatal hypoglycaemia.

Overall 332,137 women who delivered between 2006 and 2009 were identified from the national birth register. A total of 874 had bipolar disorder (defined as those with at least two ICD-10 bipolar diagnoses recorded in the national patient register). Of these, 320 were treated with mood stabilisers (lithium, antipsychotics or anticonvulsants) during pregnancy and 554 were untreated.

Compared with women without bipolar disorder, both untreated and treated women had a significantly increased risk of either induction or planned caesarean delivery; adjusted odds ratios (OR) 1.57 (95% CI:1.30-1.90) and 2.12 (95% CI:1.69-2.67) respectively. The risk of preterm birth was also significantly increased, OR 1.48 (95% CI:1.08-2.03) and 1.50 (95% CI:1.01-2.24) respectively.

The infants of untreated women, but not those of treated women, were significantly more likely to be microcephalic, OR 1.68 (95% CI:1.07-2.62) and were significantly more likely to develop hypoglycaemia, OR 1.51 (95% CI:1.04-2.43). Of the treated women 3.4% gave birth to an infant with a congenital malformation compared with 1.9% of the untreated women and 2.0% of the women without bipolar disorder.

The prevalence of bipolar disorder in this cohort was very low (0.26%, or 1 in 380). This may reflect the stringent definition but may also indicate a problem of underdiagnosis. There is evidence to suggest that a significant minority of depressed patients have unrecognised bipolar disorder,1 which may have serious consequences for female patients who subsequently become pregnant.

The authors conclude that overall their analyses do not indicate a significant difference between treated and untreated women. The increased risk of microcephaly in infants of untreated mothers was thought to be due to general growth restriction rather than an isolated phenomenon. Intrauterine growth retardation is a risk factor for neonatal hypoglycaemia. The lack of significant risk in the infants of treated mothers may have been because growth restriction was masked by enhanced fetal growth due to antipsychotics and valproate.

Women with bipolar disorder should be referred for specialist preconception advice if they are considering pregnancy. It can be very difficult to decide whether or not to continue mood stabilising medication.2 The teratogenic risk is highest with valproate and this should not be used in pregnancy. The other mood stabilisers are also associated with congenital malformations, and antipsychotics are associated with an increased risk of gestational diabetes. However, if mood stabilisers are stopped this increases the risk of relapse, both during pregnancy and in the postnatal period,3 and it is important to remember that bipolar disorder is the principal risk factor for puerperal psychosis.2

This study was unable to link the increased risk of pregnancy complications to mood stabilising treatment, which may provide some reassurance where a decision is made to continue treatment during pregnancy.

GPs have an important role in recognising bipolar disorder, advising female patients of the importance of specialist preconception advice, and ensuring close joint management by psychiatric and obstetric services during pregnancy.

Dr Phillip Bland

• Boden R, Lundgren M, Brandt L et al. Risks of adverse pregnancy and birth outcomes in women treated or not treated with mood stabilisers for bipolar disorder: population based cohort study. BMJ 2012;345:e7085