No, C-sections Are Not “Best With a Little Labor”

Children born by C-sections have about 20% higher odds of obesity, asthma, allergies, and Type 1 Diabetes, according several large research reviews.

But are children born by scheduled C-sections especially at risk for health problems, as a recent New York Times piece claims?

“the data showed more health problems among babies born by planned C-section than among those delivered by emergency C-section or vaginal birth, even though the planned surgery is done under more controlled conditions. The finding suggests that the arduous experience of labor — that exhausting, sweaty, utterly unpredictable yet often strangely exhilarating process — may give children a healthy start, even when it’s interrupted by a surgical birth.”

A reader, confused by this New York Times piece, wrote to ask for my take. “Are planned C-sections really less safe?” she asked. “The actual study… didn’t seem to support what the NYTimes article claimed.”

And after reviewing the research myself, I have to agree.

The study in question, led by Dr. Mairead Black of the University of Aberdeen, and one of the largest and best-designed studies on long-term health following delivery by C-section, actually did not find more health problems among children born by planned C-sections than those born by emergency C-sections.

(The sole exception was an unexpected–and probably artifactual–increase in Type 1 Diabetes; more on this in a moment).

How Does This Study Fit in With What We Already Know?

Although C-sections have been consistently linked with poorer long-term health in children, scientists are still not sure why.

One possibility is babies miss out on the “sweaty and exhausting” experience of labor. The physical trauma of birth kickstarts the baby’s internal stress response, pumping cortisol through their veins, and giving their organs, including the lungs, the final push to full maturity.

Another possibility, favored by many scientists, is that C-sections alter the baby’s gut microbiome. C-section  babies miss out on the messy, bacteria-laden, splash into every bodily fluid passage through the birth canal–the route by which nature normally seeds a baby’s gut microbiome.

“If a baby is born naturally, it comes into contact with bacteria from the mother, which might help with immune system development,” lead researcher Dr. Mairead Black told The New York Times.

Compared to babies born by C-section, babies born vaginally have a more diverse and healthy gut microbiome–believed to be critical for their development of a healthy, balanced immune system (one good at attacking pathogens, but not overly jumpy and prone to self-attack).

Or perhaps the issue is not C-section birth per se, but the hodgepodge of pregnancy and birth complications that often result in C-sections, such as stalled labor, intrauterine growth restriction, and preterm birth.

To study one piece of this puzzle, the importance of labor-induced fetal stress, Black and colleagues at the University of Aberdeen in the UK compared babies born by planned versus emergency C-sections. Babies born by planned C-sections experience no labor, while babies born by emergency C-section often experience some, even though it is cut short.

Black and colleagues followed over 300,000 full-term singleton babies born to first-time mothers in Scotland between the years 1993 and 2007. Roughly 4% were born by planned C-sections, and 17% by emergency C-sections.

Compared to children born by emergency C-sections, babies born by emergency C-sections were at no higher risk of virtually every health outcome Black and colleagues assessed–asthma, inflammatory bowel disease, obesity at age 5, cancer, or all-cause mortality. In fact, these children born by planned C-section had a wee bit lower risk of dying during their first year of life.

The one exception: Children born by planned C-section appeared to have 50% higher risk of developing Type 1 Diabetes. (A 50% higher risk sounds scary, but because Type 1 Diabetes is rare, this amounts to only 2 additional diagnoses for every 1000 children.)

As the researchers acknowledge, the apparent increase in Type 1 Diabetes was probably not caused by birth by planned C-section, but by some artifact of their study’s design, a third factor not adequately accounted for in their research.

Why did they think this effect was not real? Because children born by planned C-section were not at higher risk of Type 1 Diabetes compared to children born vaginally, a pattern of results inconsistent with prior research, and one which makes little sense. If anything, the researchers expected the opposite, planned C-sections would lower the risk of Type 1 Diabetes. Earlier research has found severe fetal distress stress during labor–something obviously more common during emergency C-sections than during planned ones–raises the risk of Type 1 Diabetes.

My guess? The researchers were unable to completely account for maternal Type 1 Diabetes. Having a mother with Type 1 boosts a child’s chances of Type 1 Diabetes by about 10-fold. (Black and colleagues did try to control statistically for maternal Type 1 Diabetes, but were missing this information for some of the mothers.) To avoid complications during labor, many women with Type 1 Diabetes deliver by planned C-section.

C-sections Versus Vaginal Births

How did the children born by C-section fare compare to those born vaginally?

Overall, children born by C-section, planned or emergency, were more likely to be hospitalized for asthma and had higher mortality rates during the first year of life as well as throughout childhood.

Contrary to earlier research, though, children born by C-section were no more likely to develop inflammatory bowel disease, Type 1 Diabetes, obesity, or cancer.

The Bottom Line

The NYT headline is misleading: Planned C-sections do not lead to worse health outcomes than emergency ones.

The one exception: children born via planned C-sections had a 50% higher risk of Type 1 Diabetes, but only compared with unplanned C-sections. No difference was seen when comparing children born by planned C-section with those born vaginally, a pattern of results which, as the researchers themselves acknowledge, does not make any sense. In fact, this pattern runs counter to prior research, which suggests severe fetal distress during labor ups the odds of Type 1 Diabetes, and a recent meta-analysis which found that C-sections of all types up the odds of Type 1 Diabetes by about 20%.

Why is birth by C-section associated with poorer health? We still do not know. Given the impossibility of randomized controlled trials for childbirth, we may never know.

But this study does have one take-away: missing out on labor-driven stress response is probably not the critical issue. If it were, we would see significantly worse health outcomes among children of planned C-sections than emergency C-sections.

As for the risks of C-sections overall, that’s too big of a topic for me to tackle here. But I will say this: C-section-driven health risks are minor. They are almost certainly swamped out by who we are–the genetic blessings and curses we bestow on our offspring–and what we do as parents.

(Not reassured? You can always swab your C-section-born baby’s skin and mouth with your vaginal secretions, as widely-respected gut microbiome researcher Rob Knight did after his wife’s emergency C-section. I certainly would.)


Giving Birth Takes Twice As Long It As It Did 50 Years Ago

Giving birth today takes twice as long as it did 50 years ago, and this can affect your chances of a c-section.

My first labor was long. Really long. It lasted from Friday night to Sunday evening. Pain I had anticipated, prepared for, given myself multiple pep talks for, but the duration… It undid me.

As I recently watched a good friend go through a similar labor, some old nagging questions resurfaced: How uncommon is it for women to labor for days? What is a “normal” length of labor, if such a thing exists?

Oddly enough, the medical answers to these questions have just changed dramatically. This is because of a recently completed landmark study of nearly 100,000 labors. The study, which used medical record data collected between 2002 and 2008 from hospitalsm across the U.S., showed unequivocally that we labor much more slowly than we used to. Much, much more slowly.

In fact, our labors have slowed down so much that in 2014 the American College of Obstetricians and Gynecologists (ACOG) revised its definitions of normal and overly slow laborUntil then, the definitions were based on data from the 1950s and 1960s. These data were used to define a “normal” labor duration, how long it takes most women in active labor to reach a full 10 cm of dilation and then to push the baby out. By the same token, these data were used to defined abnormal labor: labors that lasted longer than 19 out of 20 of these labors (the 95th percentile for duration) were considered overly slow or stalled.

Continue reading “Giving Birth Takes Twice As Long It As It Did 50 Years Ago”

Inducing Labor Past 39 Weeks Does Not Increase Your Chances of Having a C-Section

Pregnant women often fear that having their labor induced will lead to a c-section. In her bestselling book, Expecting Better, Emily Oster cites fear of a c-section as the primary reason she opted not to have her labor induced. Other reputable sources like the Mayo Clinic Online and March of Dimes claim that inductions increase the odds of a c-section.

But, according to the latest scientific research, these fears are unfounded. Past 39 weeks, labor inductions do not appear to raise the risk of having a c-section. Instead, being induced lowers your chances of a c-section by about 20%.

How did we get this so wrong? Until about 5 years ago, almost all the evidence on inductions and c-sections came from observational studies, which were all subtly biased, because they compared women who went into labor on their own with women who were induced.

What was wrong with this comparison?

Observational studies typically control for gestational age. This means that, in effect, they match women by the week they delivered, because the risks associated with giving birth climb as pregnancies continue past 39 weeks. Women who went into labor spontaneously at 39 weeks were compared to women who were induced at 39 weeks, and so on.

And, when matched by delivery week, women who go into labor spontaneously are less likely to have c-sections.

So what’s the problem? Pregnancies that go into labor spontaneously by a certain week are different from those that don’t.

Think about it this way: What is choice women actually face past 39 weeks? Women cannot will themselves into labor, or there would be a lot fewer pregnancies going past 40 weeks. Instead, pregnant women can only choose either to be induced or to wait it out.

So the proper study design compares these two options: being induced versus waiting it out. This is the approach taken by recent randomized controlled trials on inductions.

In these trials, pregnant women are randomly assigned to be induced past a certain point in pregnancy, for example, at 41 weeks, or to wait it out. Women assigned to wait may go into labor on their own or be induced at a later date. Using this approach, studies find, almost without exception, that relative to waiting and delivering at a later point in pregnancy, inducing labor leads to a lower chance of a c-section.

A 2009 meta-analysis (combined analysis of multiple studies) was the first to convincingly reveal the problem.

Screen Shot 2014-10-03 at 12.42.57 PM

The analysis compared the results of observational studies and randomized controlled trials (RCTs). The observational studies found a roughly 20% increase in the risk of c-section following inductions; the better-designed RCTs found a roughly 20% reduction in the risk.

Continue reading “Inducing Labor Past 39 Weeks Does Not Increase Your Chances of Having a C-Section”

Trying to Bring On Labor? Forget Sex and Spicy Foods. Only Nipple Stimulation Works.

Walking, eating spicy foods, and having sex are unlikely to bring on labor. Only nipple stimulation has any scientific support, but it could be risky for your baby.

In February, I hit my due date. 40 weeks! I thought. Let’s get this show on the road. Then… nothing. 40 weeks 1 day came and went. 40 weeks 2 days. 40 weeks 3 days…

My belly felt enormous, like an overripe melon. And in a not-so-funny coincidence, this was exactly how my favorite pregnancy app described my baby’s current size.

We needed to get this baby on schedule.

So my husband and I went out to eat spicy food. Thai, Indian, Mexican. The hotter the better.

And in return for a scorched tongue and numbed taste buds, all I ended up with was indigestion.

Then I tried walking. Lots of walking. I dragged my heavy, swollen body on long hike after long hike, completing them at a pace best described as a fast waddle.

In the end, I went almost a week late—barely avoiding a medical induction—and only after having my membranes swept twice.

At my 39 weeks appointment for my second pregnancy, my OB predicted that I would again go a week late. “That’s just how you bake them,” she chirped at me, while giving a matter of fact state-of-the-cervix rundown: I was not dilated or effaced, and my cervix was still posterior. The only thing I was, it seemed, was deeply disappointed.

But this time around I had done research on natural methods of inducing labor. That weekend, I pulled my breastpump out of the closet, sterilized my nipple shields, plopped them on, and sat down to pump for a couple of hours while watching a movie. Coincidentally or not—I’ll never really know—I went into labor two days later and delivered right on my due date.

In retrospect, though, by inducing labor in this way, I may have unwittingly taken a bigger risk with my baby’s health than I am completely comfortable with.

In my efforts to bring on labor, I am hardly alone. According to a recent survey, 50% of pregnant women report trying at least one natural method to bring on labor. Walking is the most popular approach (43%), followed by intercourse (29%), eating spicy foods (10.5%), and nipple stimulation (7.5%). The grand irony here is that nipple stimulation, the least common approach, is the only one with solid scientific support.

Using nipple stimulation to bring on labor has a long history. Midwives in the 18th and 19th centuries used the practice to speed up stalled laborsNipple stimulation brings on labor by releasing oxytocin, the hormone that causes contractions. Hospitals use a synthetic version of the same hormone, pitocin, to induce or augment a stalled labor.

A 2005 meta-analysis of six randomized control trials (the gold standard for medical studies) found that nipple stimulation increases the likelihood of labor. A total of 719 pregnant women at term were randomly assigned to stimulate their nipples for several hours or not. 37.8% went into labor within the 3 days following nipple stimulation, compared to 6.4% of the controls.

This difference, large in relative terms, was statistically significant. Still, two-thirds of women failed to go into labor even after nipple stimulation—underscoring that the method is no guarantee of labor.

(Note to the do-it-yourselfers or those fearful of accidentally inducing labor: Nipple stimulation in these studies did not consist of a few random suckles, squeezes, or tugs. It was a serious, concerted effort. Women were instructed to gently stimulate one nipple for 10–15 minutes before switching to the other side, and to continue alternating between breasts for an hour, 3 times a day.)

Using nipple stimulation for inducing labor would probably be a more widely known and recommended practice were there not concerns about its safety. One of the studies in the 2005 meta-analysis, conducted with high-risk women delivering at a hospital in India, was stopped early because of 4 fetal deaths: 3 in the nipple stimulation group, and 1 in the oxytocin induction group. (None of the other 5 studies reported any deaths.) Because of this, the meta-analysis concludes:

“Until safety issues have been fully evaluated it [nipple stimulation] should not be considered for use in a high-risk population”

It’s hard to know what to make of these safety concerns. The women in the Indian study were considered high risk because they had intrauterine growth retardation, high blood pressure, or were post-term. But the researchers failed to report which of these risk factors was associated with fetal death, or to provide any additional details about these deaths, other than to say one of fetuses lost did not appear to have any congenital abnormalities.

That nipple stimulation contributed to these deaths is certainly plausible. Nipple stimulation can result in uterine hyperstimulation—contractions that are too frequent or prolonged, and which can lead to fetal distress.* This is why women are told to alternate between breasts instead of stimulating both nipples simultaneously and to pause during contractions.

German study found uterine hyperstimulation occurred in 10% of women during nipple stimulation, and in 1% this was accompanied by reversible abnormalities in fetal heart rate patterns. Based on this, the authors caution against using nipple stimulation without medical supervision:

Due to the conflicting reports in the literature and because of the potential hazards involved, the use of nipple stimulation for the induction of uterine contractions can be advocated only in a controlled clinical setting. Its application without medical supervision, as propagated in the lay press, is definitely contraindicated.

In sum: nipple stimulation does increase the likelihood of going into labor. But because it can cause contractions that are too long or too intense, it is not recommended for women with high-risk pregnancies.

Below is a quick rundown on the evidence (or really, lack thereof) for other natural induction methods:

Walking. Walking is the most recommended and most commonly attempted method of inducing labor. It is true that in late pregnancy, walking seems to increase the frequency and strength of Braxton Hicks contractions. There is no evidence, though, that walking speeds the onset of labor.

Eating spicy foods. There is no evidence that spicy foods cause labor.

Acupuncture. review article of 3 randomized control trials, consisting of 212 women, found that acupuncture reduced the number of women who needed to a medical induction (33% versus 54%), but failed to find a difference in when the women went into labor. Because of the inconsistent results and methods (the type of acupuncture varied between studies), the reviewers felt no conclusions could be drawn without additional research. Since then, other randomized control trials have found no benefit of acupuncture for bringing on labor (see herehere, and here).

Having sex. As one researcher put it, intercourse would be a “safe, effective, and perhaps even fun” [italics mine] way to induce labor. Fun aside, it’s biologically plausible that sex would bring on labor. Semen contains prostaglandins, which are known to ripen the cervix. Orgasms promote the release of oxytocin and can therefore stimulate uterine contractions during late pregnancy. Foreplay involving nipple stimulation could also release oxytocin. A small observational study found that women who reported having sex after 39 weeks were much less likely to go post-term or need to be induced.

Sounds pretty good, right? Alas, the most recent, best-designed, and largest study to date, a randomized clinical trial conducted in Southeast Asia, failed to replicate these findings. Two hundred ten women were recruited from a prenatal clinic, where they had been scheduled for inductions. They were randomly assigned to be counselled by a physician to have sex or to receive standard counselling. The women kept daily diaries tracking when they had sex and whether they experienced an orgasm. Although women who were told to have sex to bring on labor were more likely to have sex before delivery (60% versus 40%), they did not differ in cervical ripeness or in their likelihood to go into labor on their own.

It’s possible that sex does bring on labor, but the Southeast Asian study was too small and therefore underpowered. A sample of 210 women is not large. Compounding this problem, a fair number of the controls still had sex, albeit at a lower rate than cases, watering down the comparison between the two groups.

Knowing this, in a separate analysis of the same data, the researchers compared all the women who reported having sex with the women who did not, regardless of whether they had been cases or controls. They still found no effect of intercourse. If anything, women who had sex during the study period went into labor later than women who refrained. An earlier but smaller clinical trial similarly found no benefit of sex.

In sum: sex late in pregnancy—after 39 weeks—does not appear to bring on labor. But it perhaps just might be fun.

Did you try any natural methods to induce labor? And do you think they worked?


* The evidence on nipple stimulation and uterine hyperstimulation comes mainly from a series of studies conducted in the 1980s. These studies examined whether nipple stimulation was a viable alternative to pitocin for use in contraction stress tests. Constraction stress tests monitor fetal heart rate during contractions to ensure the fetus can safely withstand labor. The studies variously found no instances of hyperstimulation, hyperstimulation in 10% of women, and hyperstimulation in 12% of women. A final study, which also reported on fetal distress, found hyperstimulation with fetal heart rate changes in 2.9% of women using nipple stimulation, compared to 1% of women using pitocin.