Mode of delivery and childhood illness: A randomized trial in 2004.

One of the biggest difficulties in proving a causal link between cesarean birth and chronic health problems in childhood is the type of research studies that can practically and ethically be done during pregnancy.

The gold standard in medical research is the randomized controlled trial (RCT). In an RCT researchers randomly place subjects into either “treatment” or “control” groups, then expose the treatment group to something—a new vaccine for Infection X, for example—and then compare outcomes in the two groups later on. (In this example, how many kids in the treatment [vaccinated] group came down with Infection X versus how many in the unvaccinated control group?) If there’s a significant difference in outcomes between the two groups, you’ve got a strong argument that the treatment made the difference.

How a randomized controlled trial works.**

As you can imagine, randomly assigning pregnant women to cesarean (treatment) or vaginal birth (control) groups is nigh onto impossible—ergo, you can’t do an RCT. This means that virtually all research studies on the issue of cesareans and chronic childhood have been observational in nature—looking backward in time at databases, for example, or trying to fish significant trends out of hospital registries, birth cohorts and the like. The best an observational study can tell you is that A and B are associated with one another, but that’s it—you can’t prove that A actually causes B. An observational study can’t prove that cesareans are a cause of asthma; it can only say that cesareans are associated with an increased risk of childhood asthma.*

So mode-of-delivery RCTs are out of the question…or are they? Actually, in 2004, a Canadian research team did one.

Well done, Northern Neighbors!

The multi-center, multi-nation Term Breech Trial wasn’t about whether cesareans might increase the risk of childhood asthma, diabetes and such. It was about trying to figure out whether elective cesarean section or vaginal birth was the safest way to deliver a breech baby at term. Since the existing research was somewhat murky at the time, it was considered ethical (with informed consent) to randomize women to have either a planned cesarean or attempt a vaginal birth.

The particulars of the breech birth debate are best left for another post, but tucked away in the study’s results section was this little nugget:

“…more parents in the planned cesarean birth group than the planned vaginal birth group reported that their children had had medical problems in the past several months…relative risk, 1.41; 95% CI, 1.05-1.89; P=0.2.”

Plain English version (mine): The toddlers who had been in the planned cesarean group were about 40% more likely to have been sick in the previous few months than those in the planned vaginal birth group. The types of medical problems—typical 2 year-old stuff like colds, ear infections and stomach flu—were no different between the groups. The only difference was in the numbers of children who’d gotten sick.

As is the case with all medical research, you can find things in the study to complain about: relatively small numbers, for example, the use of parental questionnaires and the fact that some mothers in planned vaginal birth group ended up having cesareans (and vice-versa), etc.

But here’s my bottom line:

In a randomized trial of pretty well-matched subjects, those babies whose mothers were in the planned cesarean group tended to get sick more often than those in the planned vaginal birth group.

This doesn’t address the issue of chronic illnesses like asthma, type 1 diabetes and the like, but it does support the theory that cesarean birth can mess with a baby’s developing immune system.

* * *

*Here’s an exaggerated example of the trouble with mistaking association for causation: Virtually all adults who die suddenly of heart attacks drank water in the 24 hours before they died. So, drinking water is associated (time-wise) with heart attacks. But you would be wayyyy wrong to say that, based on that association, a glass of water can cause a heart attack.

**Credit: SUNY Downstate Medical Center

The Brazilian preference?

What happens when a woman emigrates to a foreign country and then gives birth? Is her birth experience more likely to conform to the country she comes from, or the one she moves to?

For example, is an immigrant woman more likely to have a cesarean birth than a native-born woman, or less? How big a role do the cultural norms and expectations she brings with her from her home country play in determining mode of delivery?

Studies of immigrant birth experiences have been mixed to date. Immigrants do tend to have higher cesarean rates than natives, but interpretations of such findings are often complicated by things like language barriers and the difficulties new arrivals may have in accessing timely maternity care. Teasing out the effects of culture can be tricky.

A Portuguese article just published in the journal PLOS ONE helps to clarify this issue. The study compares cesarean rates between native-born and immigrant Brazilian women in northern Portugal. Two major potential confounders are quickly dealt with: the two groups of women both spoke Portuguese, which eliminates language barrier as a source of cesarean-inducing miscommunication, and all the women were drawn from five public hospitals, so that the care they received was more or less uniform.

The authors found that Brazilian immigrant women had a 50% higher cesarean rate than did native-born Portuguese women (48.4% vs. 32.1%), a difference that persisted even after controlling for such things as demographic, medical and obstetric risk factors. In fact, the cesarean rate for Brazilian immigrants was nearly identical to the overall cesarean rate in Brazil itself.

What explains the native-immigrant difference? The authors speculate it has much to do with attitudes about childbirth that the women brought with them from Brazil:

“This extremely high prevalence [of cesarean birth] seems to be a cultural consequence of attitudes towards labor and the perception of obstetric care among Brazilian women. The majority of Brazilian women perceive cesarean as the most adequate mode of delivery and as a symbol of high social status.”

In other words, culture strongly influences mode of delivery, even far from home. A woman raised to see cesarean birth as a desirable norm is much more likely to end up having one.

Home VBACs on the rise

The rise and fall of VBACs

A new study by Eugene DeClercq and Frances MacDorman–my favorite epidemiologists–looks at trends in home vaginal birth after cesarean (home VBAC) from 1990-2008. The results are interesting, if not surprising, given the trends in American maternity care as a whole.

Basically, Declercq et al found that home VBACs in the U.S., while still rare events, have been increasing–most recently from 664 in 2003 to 1000 in 2008. VBACs accounted for less than 1% of home births in 1996, but are now 4% of total home births. Meanwhile hospital VBACs have decreased from 3% of total hospital births in 1996 to 1% in 2008.

VBACs underwent a steep decline in the late 1990s after the American College of Obstetricians and Gynecologists came out strongly against VBACs being performed in hospitals that did not have 24/7 OB anesthesia services, in case of the need for an emergency c-section. For a small number of women–likely those with no local VBAC alternative–this simply drove VBACs out of the hospital and “underground” – ie, into the home setting.

The safety of home VBACs can (and will) be debated endlessly. The take-home point for me, though, is that this is one more reason to integrate home birth into the larger American maternity care system. Women can’t be forced to give birth in hospitals, and neither demonizing home birth  nor overstating the advantages of hospital birth is doing anyone any good. (For examples of some especially vocal opposition to home birth, see the replies to Dr. Alice Dreger’s article in The Atlantic I cited a few days ago.)