Women who take antidepressants during the first trimester are almost twice as likely to give birth to children with autistic spectrum disorders, a nested case-control study has found.

The Stockholm youth cohort comprises all children aged up to 17 years resident within Stockholm County between 2001 and 2007. After exclusion of those with incomplete data, there were 4,429 cases of autistic spectrum disorder. For each case there were ten age- and sex-matched controls (n = 43,227).

Maternal or paternal depression before the birth of the child was identified from the Stockholm psychiatric outpatient register and the national inpatient register. For children born from 1995 onwards, maternal antidepressant use was identified from records of the first antenatal interview (median ten weeks’ gestation).

A history of maternal depression was associated with an increased risk of autistic spectrum disorder adjusted odds ratio (OR) 1.49 (95% CI:1.08-2.08). This association was restricted to those who reported antidepressant use in early pregnancy. There was no association with paternal depression.

Antidepressant use in early pregnancy nearly doubled the odds of autistic spectrum disorder adjusted OR 1.90 (95% CI:1.15-3.14). The increased risk was seen with both SSRIs and tricyclics.

Cases were identified from secondary care registers so the results may not be applicable to patients with milder depression treated in primary care.

Failure to identify primary care patients also meant that for children born from 1995 onwards, the number of mothers with recognised depression was very small, 249, and of these only a small proportion were prescribed antidepressants, 36 of 249 (14%).

The study therefore lacked power and adjustment for confounders may not have been very effective.¹

The authors acknowledge the problem of confounding by indication i.e. those women who were prescribed antidepressants may have had more severe depression. Therefore, it is not possible to say whether the observed association between antidepressant use and autistic spectrum disorder is due to the medication or the severity of the depression.

Serotonergic neurones play an important role in the regulation of neurodevelopment, and animal studies suggest that intrauterine exposure to high serotonin levels may lead to autistic behaviours.2 Therefore, a direct causal role of antidepressants is plausible.

However, even if this is the case, the authors estimate that less than 1% of autism cases can be attributed to antidepressant use.

Should we warn women that taking antidepressants in pregnancy may increase the risk of autism? The UK prevalence of autistic spectrum disorders is around 1%,³ so a doubling of risk would equate to an absolute risk increase of 1 in 100. However, if this increased risk is due to depression severity, antidepressants may in fact reduce the risk.

This study provides further evidence that we should be very wary of initiating antidepressant treatment during pregnancy, particularly during the first trimester.

For those women who become pregnant while taking antidepressants, it can be difficult to decide whether or not to stop treatment. Rapid discontinuation of antidepressants is associated with a high rate of relapse4 and this has to be set against the uncertain evidence of harm and the low level of absolute risk.

 Dr Phillip Bland

• Rai D, Lee BK, Dalman C et al. Parental depression, maternal antidepressant use during pregnancy, and risk of autistic spectrum disorders: population based case-control study. BMJ 2013;346:f2059