Today’s media features another story on the effects of fatherhood on men that is tied back to the earlier report about fatherhood and low testosterone levels. But both stories also make me think about the role of science and science reporting in what we know (or think we know) about parenting. That’s something of a perennial theme here. I want to take the time to talk about the science (and the reporting) in a little detail.
This new story is about the relationship between fatherhood and life expectancy, and if you poke around the web, you will frequently find it reported this way: “Fatherhood Helps You Live Longer.” I’m sure that attracts readers.
Now that seems a particularly bad headline because if you actually read the reports of the study, it isn’t what it says at all. Here’s another press account, one with a somewhat better headline ,and a discussion of the methodology, which is striking. (The title of the study won’t make the headlines.)
It seems researchers really wanted to study the relationship between infertility and lifespan, but while they had a huge database, it didn’t directly tell them about infertility–it told them whether they actually had children. Thus the childless group would include both infertile men and men who simply didn’t want children. (It also includes men who had children but didn’t know it.)
To maximize the chances they were really comparing fertile/infertile (as opposed those thinking they had children with those thinking they did not have children) The researchers made an assumption. Here is how its described in Time:
To increase the likelihood that all the men included in the study actually wanted children, the researchers excluded those who had never been married; by including only men who were married or had been married at least once, the researchers could be reasonably sure they had the desire and opportunity to have kids.
Now I do wonder a bit about the assumption. There are in fact married couples who are childless by choice. But that said, it looks to me like 92% of the married men did have at least one child (that they knew about or believed to be theirs?) This takes me to the next thing I wonder about though–8 married men out of a hundred didn’t have (or didn’t think they had?) children. If one out of 100 married men is childless by choice (which seems possible to me) that one is in those 8. That means 12% of the men now assumed to be infertile aren’t infertile. Doesn’t that mess things up just a bit?
I’m not an expert and I cannot criticize the methodology, really, so I won’t linger here. Instead, I’ll consider the findings of the study. Here I’ll just quote the conclusions noted in the abstract of the study:
Married men who have no children have a higher risk of dying from cardiovascular disease contracted after the age of 50 than men with two or more children.
Points to the researchers here for their careful parsing. You might notice that men with one child seem to have fallen out of the picture. I think that’s because the analysis for men with one child looked a lot like the analysis for men with no children. (The abstract says: “this elevated risk appeared to extend also to men with only one child.” Doesn’t that knocks the infertility point rather out of the box? Nevermind.)
It also strikes me that the conclusion may be an instance of the negative pregnant. If you say “have a higher risk of dying from cardio vascular disease” then the negative implication is that there isn’t a higher risk of dying generally–just of dying from that cause. Consistent with this, I cannot tell if the men presumed to be infertile actually had a higher risk of dying generally.
Finally, an important point to keep in mind: The study doesn’t purport to make any findings about causation–it is at most about correlation. Thus, there’s no suggestion that infertility causes heart disease or that having children wards off heart disease. It’s quite possible that heart disease and fertility are both caused by some other factor that isn’t explored.
Now I’m not going through all this to trash the study. I think it is interesting and I’d like to know more. But like a great deal of science, it is limited in what it can tell us. Reporters (and after reporters, headline writers) may not be so careful with the limits and so by the time the story reaches us on the web it can look rather different. That’s a good reason why some caution is warranted.