Take fertility. In 2011, British psychologists pulled together data from 14 studies of in vitro fertilization. In each study, researchers asked women to assess their emotional distress, anxiety or depression. Then they followed them through a single cycle of fertility treatment to see whether they got pregnant or not. The smaller individual studies arrived at disparate results, but the meta-analysis rolling up all the findings, which included over 3,500 women and appeared in the BMJ, was fairly definitive: Women's emotional state before IVF bore no relationship to whether the treatment worked. In other words, women with more extreme levels of anxiety or depression were just as likely to get pregnant after a single cycle as women with milder levels. "It was a great relief," said psychologist Jacky Boivin, who has counseled women struggling with infertility for years and who led the meta-analysis team. (Less clear is whether extreme levels of stress, as in war or famine, suppress natural fertility. They may, but the effect may also turn out to be temporary, according to Boivin: "At some point, even in harsh environments, reproduction tends to come back.")
What about during pregnancy? The old idea that stress causes miscarriage isn't supported by the data and seems, thankfully, to have lost some of its traction. The current angst, though — that emotional symptoms can lead to preterm birth — threatens to torment women at least as much. It's true that you can find smaller studies that fuel the fear. But consider this large, population-based work, in which researchers interviewed over 78,000 Danish women. Those who reported higher levels of life stress and more emotional symptoms like anxiety when they were 30 weeks pregnant did tend to give birth earlier. But the difference was pretty minimal: The women with the highest life-stress scores gave birth, on average, about two days before women with lower scores. Those who reported the most intense emotional symptoms had pregnancies that were just two-and-a-half days shorter. This isn't an effect that matters in children's lives.