A new study of the impact of pregnancy on elite runners includes great info, but doesn't add much

It's not the researchers' fault that their work was anything but "groundbreaking"

The other day, I described how pregnancy and childbirth appear to affect subsequent distance-running ability based on a sampling of world-class distance runners stretching back to the 1980s (for some reason, the Internet goes dark on women’s distance running if you try to reach much further back than that). I joined this to the proposition that basic evolutionary biology along with the body’s proven ability to fully recover and rebuild from diseases like cancer and serious injuries support the idea that we* should expect this to be true in general.

I raised the subject owing to the recent publication of “Impact of Pregnancy in 42 Elite to World-class Runners on Training and Performance Outcomes” in the journal Medicine and Science in Sports and Exercise. You can and should read the abstract to see what information the researchers were after and how they went about collecting it.

David Roche, the slapstick equivalent of Trail Runner’s science explainer, called the study “groundbreaking.” For this to be true of any study, it has to reveal something both new and important. For example, definitive refutation or confirmation of a hypothesis that’s been floating around, or evidence that a newly crafted antibiotic is clinically effective against a new strain of streptococcal bacteria. It is not sufficient for the topic itself to occupy a narrow niche, however interesting.

Also, a study of 42 subjects is unlikely to be groundbreaking in any way unless the researchers split them into experimental and control groups matched for relevant characteristics (e.g., age), and even then, you’d have to see a really big difference in outcomes between the experimental and control groups to obtain a statistically significant result. Depending on the trait under scrutiny, this might not be that difficult.

Clearly, the researchers who conducted the impact-of-pregnancy study couldn’t operate like that. In fact, they were hamstrung by a number of intrinsic limitations, notably the dearth of women fast enough to qualify who had access to the range of information the researchers needed. I am not portraying their project as either a waste of work or a misrepresentation of reality. It’s just not…very revealing.

The most obvious problem when I read the abstract as well as Roche’s analysis of the study is that there was no way to control (account) for the varying ages and career-arcs of the participants. For example, a 24-year-old with a marathon best of 2:30 and a 33-year-old with a fastest time of 2:24 might have precisely equal talent levels. But if both decide to start families at that stage, and both intend to reach the next Olympics, it’s plain that the younger runner has a far better chance of improving than the older one does, even assuming that they enjoy identical robust “recoveries” from the effects of their pregnancies.

Also, age and stage-of-career are obviously correlated, but far from perfectly. Someone who joins the Bowerman Track Club at age 22 is likely to have blazing, PED-propelled personal bests by the age of 25, and it’s possible to stay in the game for ten years at that point by hovering close to those bests. But someone who, say, sells real estate in the southern D.C. suburbs for a while after a lukewarm college career and only gets serious at age 35 is far better positioned to climb higher on the World Athletic points tables despite getting long in the teat.

I have a copy of the 23-page-long full study. The researchers did their best to include only women who had recorded, or remembered, details about their training throughout their pregnancies and statistically iron out the differences they found in this and related areas. A third of the 63 women who initiated the survey were dropped because they answered only a few of the study questionnaire’s 139 questions.

What I would expect, based on my own propositional and suppositional blather above and the other day, is this: If you’re a woman who still realistically has personal bests awaiting you, then if you get pregnant tomorrow, experience no pregnancy or labor complications, and can resume training soon after having your baby, you’ll be just as likely to run those personal bests as you would have otherwise.

What the researchers found is that 46 percent of the participants, or 19 of them, achieved better World Athletics point scores post-pregnancy. Only 25 of the 41 intended to remain at the same level, with three women responding that they intended to remain elite without necessarily retaining their full previous abilities and the other 14 basically saying they expected to be done. The 25 women with declared intentions to improve or at least not slow down on average improved their World Athletics scores by a tenth of a percent, while the 14 who were sort of retired fell by an average of 4.6 percent in this area (not quite the same as kicking back with a twelve-pack and a blunt.

I would say that the researchers more or less confirmed my own theories here, but presented the data somewhat pessimistically. I don’t think the “46 percent” figure means much unless you’re a woman who is at a point in your elite career where you probably have a 50-50 chance of getting faster whether you have children or not. I don’t think a healthy woman with a commitment to serious running has to worry about beating a slightly unfair coin toss if she’s hoping to get faster in a household made suddenly larger by one (and sometimes more) persons. Anyone with realistic goals a couple of years distant has no demonstrated reason to expect problems.

I also consulted David Couper, a biostatistics professor at the University of North Carolina-Chapel Hill, about the study’s findings and design. (Couper ran 2:26 and change for the marathon in pre-Cambrian times.) Here’s his input:

Interesting, though not particularly surprising results.

To me the biggest concern is this: “… inclusion criteria: a)  their first pregnancy carried to term at least five years before survey completion.” Because they were recruited long after their pregnancy, rather than prospectively (i.e., before or early in their pregnancy) the sample is likely to be biased in favor of those who managed to return to a high level of performance.

Another concern with the “at least five years before” part is that there may be recall issues with some of the responses to the 139-item questionnaire. Training volumes are probably reasonably accurate, if taken from logbooks, but memory of intensity of workouts and whether they had really planned to return to the same level of competition may not be.

As you mentioned, there is no control for age or other factors, including things such as how much support they have (in terms of things such as childcare, etc., etc.)

I would not have detected the bias that was likely introduced by the (obligatorily) long time gap between the women’s pregnancies and their participation in the study. But it makes sense. If you posted notices all over town saying, “We’re looking for men between 20 and 70 to get an idea of average bench press and penis size,” you wouldn’t see an equally distributed range of ages and physiques signing up for that study. In other words, you wouldn’t get a representative population sample of males aged 20 to 70. Couper is saying that the women who chose to be enrolled in the impact-of-pregnancy study probably had better outcomes than the typical elite woman who joins the ranks of motherhood.

But overall, allowing for the shortcomings of the study we both noticed, Couper wasn’t any more surprised than I was to see that women who intend to meld motherhood with top-level running turn out to be pretty damned good at it. I don’t see how anyone could consider any workout or race difficult after delivering a baby, but that’s another story.

Footnote: Perhaps the most significant thing I did in my incomplete medical education was scrub in in a cesarean section at Evans Army Community Hospital at Fort Carson, near Colorado Springs. The last clerkship I did was OB/GYN, and although I had no interest in surgery, I was ready for this one. As a med student in this situation, all you have to do is recite the seven layers of the abdominal wall when the presiding surgeon demands this information, not pass out, and hold a womb containing a baby in your hands for a few terrifying, beautiful seconds.

Also, because this was an Army hospital, there were unusual wrinkles in the OR set-up. For example, one of the doctors was a captain, but one of the assisting nurses was a lieutenant colonel who looked about 28, but was two officer ranks above the doc.

But the most striking thing after many hours that passes quickly in that steamy room was watching two distinct and vital operations unwind as soon as the baby was alive and separate from her mother. I had all but forgotten that an entire neonatal team was waiting for this moment, and oh my, can those people move. They whisked the women’s daughter to a far portion of the room to run health checks while the docs and nurses started closing up the “wound.”

It did not occur to me in that moment that the mother, unconscious and having been spared the pain of someone sledding through her vagina, had missed something critical: The chance to hold her baby in the first moments of its life. Being in my mid-twenties and focused on not fucking up in the OR suite played a role, but I mostly remain clueless in this and related areas, and it’s partly why I hesitate to yammer about something removed on multiple levels from my own experience.