Are you pregnant, or just tired?

A reader pointed me to a website that's collecting some interesting statistics on women who are trying to get pregnant. Countdown to Pregnancy is a site that tells women, "It's never too early to speculate!" The site encourages women to track symptoms that they're perceiving as signs of possible pregnancy, starting the first day after ovulation. The women then report whether or not they were actually pregnant that month, so the site has been able to put together some charts that show how the perception of pregnancy quickly diverges from the reality of pregnancy. I've got a feeling (hah!) that there's some really interesting statistics buried in here, waiting for anybody with enough time to dig them out.

For instance, contrary to Countdown to Pregnancy's slogan, there really doesn't seem to be much of a point in speculating about whether symptoms are a sign of pregnancy until a couple weeks after ovulation. On the first day after ovulation, 6.8% of women who later turned out to be pregnant reported feeling fatigued. But, on that same day, about 3.5% (Countdown to Pregnancy does not display its data in the best possible way) of women who weren't pregnant also felt fatigued. There's a 64% difference between those two, but the absolute difference is very small. It's still pretty likely, at this point, that if you feel fatigued, it's not a sign of pregnancy.

Contrast that to 15 days after ovulation, when about 2.5% of non-pregnant women felt fatigued, while 9.5% of pregnant women felt that way. Now there's a 116% difference between the outcome of the same symptom. The absolute difference is still in the single digits, but by 15 days, you're starting to really see a serious divergence between how pregnant women are feeling and non-pregnant women feel. There's enough of a difference that the symptoms start to take on meaning you can actually speculate about.

It's interesting. And I'd be really curious to see how the voluntary survey results from this site line up with the same sort of survey done in a more controlled, scientific way.