Specifically, the researchers found that women taking fluoxetine (Prozac) and paroxetine (Paxil) were slightly more likely to give birth to babies with specific types of heart defects.
But these results do not mean the drugs are never safe in pregnancy, cautioned study author Dr. Heli Malm.
The study does not show that the drugs themselves caused the birth defects, for one, and any risk - if it exists - would be small: affecting just 105 out of 10,000 babies born to mothers taking fluoxetine, and 31 out of 10,000 babies from mothers taking paroxetine.
Still, fluoxetine, paroxetine, and similar types of antidepressants - known as selective serotonin reuptake inhibitors -- "should only be used during pregnancy when clearly indicated," Malm told Reuters Health. "However, when clearly indicated, they seem to be relatively safe."
Consequently, there may be some instances in which it's safer for women with depression to stay on the drugs during their pregnancy, said Malm, based at the Teratology Information Service in the Hospital District of Helsinki and Uusimaa, Finland. "The absolute risk for the specific cardiac malformations is small and ensuring the mother's well-being is important."
Between 3 and 6 percent of women take SSRIs during pregnancy, raising concerns over whether these drugs pose risks to the baby.
The data have so far been somewhat conflicting, although some studies have already linked SSRIs to higher-than-average, though small, risks of certain birth defects. One study of nearly half a million Danish children found an increased risk of heart defects among those whose mothers had used SSRIs such as fluoxetine (Prozac), but also sertraline (Zoloft) and citalopram (Celexa).
Last year, a study among 12,700 U.S. infants found that mothers who used bupropion (Wellbutrin) during early pregnancy had more than double the risk of heart defects known as left outflow tract defects, compared with infants whose mothers had not used the drug.
During the latest study, Malm and her colleagues reviewed national data collected about 635,583 births that occurred between 1996 and 2006.
They found that mothers who took fluoxetine during early pregnancy were more likely to give birth to babies with isolated ventricular septal defects, in which they have a hole between the left and right sides of the heart. Babies whose mothers took paroxetine in early pregnancy were more likely to be born with right ventricular outflow tract defects, which affect the flow of blood from the heart's right chambers to the rest of the body.
Specifically, among 10,000 babies born to women who didn't take SSRIs, 7 developed a right ventricular outflow tract defect. But among the same number of babies born to women taking paroxetine, 31 had the heart defect.
For fluoxetine, 105 out of 10,000 exposed babies developed isolated ventricular septal defects, which were diagnosed in only 49 of 10,000 unexposed babies.
There also appeared to be a slight increased risk of neural tube defects among babies exposed to SSRIs - 22 out of 10,000 exposed babies were born with the defect, versus only 9 out of 10,000 babies born to moms who didn't take SSRIs.
Babies born to mothers taking the antidepressants were also significantly more likely to have disorders related to alcohol, the authors report in the journal Obstetrics & Gynecology. "The only explanation to this finding is that alcohol consumption is more abundant among women using SSRIs," said Malm. "This may result from psychiatric problems."
Even if certain antidepressants pose risks during pregnancy, so might untreated depression, noted Dr. Mette Norgaard at Aarhus University Hospital in Denmark, who reviewed the findings for Reuters Health.
"To the best of my knowledge no studies have examined how untreated depression will affect the fetus and the maternal health during pregnancy and most likely it is not safe to have an untreated depression," the researcher said in an email. "I would recommend women who are taking SSRIs and are already pregnant or are planning to become pregnant to talk to their doctor on how they should continue treatment."
One caution, added Norgaard, is that women taking SSRIs may be "in closer contact with the healthcare system," making it easier for doctors to diagnose mild malformations in their babies. "It is thus possible that more infants whose mothers used SSRIs are diagnosed with milder forms of congenital malformations that will remain undetected or become detected later in life in unexposed infants."