Driven by concerns regarding fetal exposure to psychotropic medications, many women with psychiatric illness attempt to discontinue their pharmacologic treatment during pregnancy; however, recent studies indicate that this approach may not be appropriate for all women. Dr. Adele Viguera and her colleagues at the Center for Women’s Mental Health have reported that among pregnant women with bipolar disorder, relapse rates were very high (58%) in women who discontinued maintenance treatment with lithium during pregnancy (Viguera et al 2000). Given this risk of recurrent illness, many women may consider continuing lithium treatment during pregnancy. While the teratogenic effects of first trimester exposure to lithium have been well studied, data on the long-term outcome of children exposed to lithium during pregnancy are sparse. At the 61st Annual Meeting of the Society of Biological Psychiatry in Toronto, Dr. Viguera presented preliminary data on the neurobehavioral outcomes of children exposed to lithium in utero.
In this study, 61 mothers with bipolar disorder and their children were recruited. The final analysis included children exposed to lithium at any point during pregnancy (n=32) and children not exposed to lithium or any other mood stabilizer during pregnancy (n=24). Five subjects were excluded from the analysis due to missing data. Information on maternal demographic, obstetrical, and psychiatric history was collected, as well as information about the medical/developmental history of the child. Early childhood development was assessed by a blinded rater using the Bayley Scales of Infant Development-II (BSID-II) and the Wechsler Preschool and Primary Scale of Intelligence-Revised (WPPSI-R). Children were assessed at 2.5 to 3.5 years for the BSID-II and from 3.5 to 7.3 years of age for the WPPSI-R.
There were no significant differences in the mean or range of global IQ scores of children exposed to lithium in utero versus non-exposed children. However, it is notable that the average developmental score for the lithium exposed group was 7-8 points lower than the control group. Despite this point-difference between exposed and non-exposed children, the average score for the exposed group remained well within the average range. Exposed children were more likely to be born early ( 37.8 vs. 39.3 weeks), have slightly lower Apgar scores at birth (8.6 vs. 9.1), weigh slightly less (3240 vs. 3550 grams), and were more likely to be admitted to the NICU.
Though preliminary, these findings suggest no gross developmental deficits associated with lithium exposure. While lithium appeared to have a statistically significant impact on certain outcomes (i.e., gestational age, birth weight, Apgar scores, Bayley scores); the effect was relatively small and appeared to have little clinical relevance. In fact, the two groups behaved similarly with respect to meeting developmental milestones. Furthermore, all scores for the lithium-exposed children were within the normal range.
It should also be noted that differences in outcomes between the exposed and non-exposed groups may stem from unidentified differences between the two groups of women recruited for the study. More specifically, women were not randomly assigned to the two groups but made their decisions regarding their treatment based on their clinical histories and personal preferences. Thus, it could be hypothesized that the women who chose to maintain treatment with lithium were more severely ill than those who chose to discontinue lithium, and it is the illness itself – not the medication-that has a negative impact on outcomes. Similarly, other studies in women with unipolar illness have demonstrated that more severe depressive symptoms during pregnancy have been associated with shorter gestation, lower birth weight, and lower Apgar scores, absent exposure to any medication (Bonari et al, 2004). Clearly there is need for more systematic study in larger samples of women in order to better understand the impact of lithium on neurobehavioral outcomes; nonetheless, the observed absence of clinically significant cognitive deficits is reassuring.
Ruta Nonacs, MD PhD
*This post was originally published as an article in our July 2006 newsletter.