Andrew Solomon has a fascinating piece on the subject:
While the drugs are risky, depression during pregnancy is at least as problematical. Animal studies show that stressed mammalian mothers are likely to have offspring with poor neurodevelopment. Pregnant women experiencing depression or anxiety are under greater stress and may have altered neurobiology themselves, which could affect fetal development via changes in the uterine environment. Indeed, untreated depression during pregnancy is associated with increased miscarriage rates, preterm birth, and low birth weight—some of the very risks associated with maternal use of S.S.R.I.s. Depressed mothers are at increased risk for preeclampsia. Recent research has shown that the fetus of a depressed, expectant mother has alterations in the microstructure of the right amygdala. There is even some evidence that mothers who are extremely stressed during their first trimesters may be more likely to have children who later develop schizophrenia. …
At the same time, it is important not to blame mothers for their children’s neurological challenges. The shadow of the “refrigerator mother,” who was said to cause autism, falls long across this research. It may be counterproductive to tell women under stresses they cannot avoid that they are damaging their children by being unhappy—or by being treated for their distress. Blaming some women for injuring their children by taking antidepressants and others for injuring their children because they are depressed creates a no-win situation that is itself depressing.
There is no universal right answer here, and, under those circumstances, quoting the studies may seem counterproductive. But women need the leeway to make their own choices—to look, as one does in many areas of health care, at two unsatisfactory options and select between them, and to do so with as much information as possible.
Previous Dish on the subject here.