Monday, May 28, 2012

Correlation is not Causation, a.k.a. "Look at the Confounders Batman!"

In the past week or so, a study has been making the rounds. This one claims that babies born via c-section have twice the risk of being obese as those who are born vaginally. The anti-csection brigade is using it as yet another reason to clamp down on the epidemic of unneccessary c-sections, and at the same time making moms who would willfully choose c-section, absent any medical indication, feel as though their choice is posing some risk to their child.

I've had a brief look at the study in question - and here is my conclusion:

Correlation is not causation, and "Look at the confounders, Batman!".

There are a lot of reasons why the rate of c-sections has increased over the past few decades. Moms are older by the time they have children - there are things that seem to be pre-requisites to starting a family now, many women want to be married for a while before starting a family, many women want to own a house before starting a family, many women want to have a career before starting a family, and as a result of wanting a career many women must complete post-secondary education before starting a family. At the end of a day a woman is often in her thirties before she even tries getting pregnant. As a result of being in your thirties before you even start on the "mommy track" you might be more likely to need help getting pregnant in the first place. This might mean fertility drugs. This might mean IVF. Even if it doesn't mean those things, your risk of having multiples increases with age. Many moms might only be planning on having small families. Moms also seem to be more likely to be starting their pregnancies with higher BMI's than in the past and they also seem to be having higher rates of gestational diabetes. Women also seem to becoming more aware that vaginal birth is not risk-free and may also have some unpleasant risks. And all of this is in a context of having the risk associated with having a c-section plummet - surgical methods have improved immensely over the past few decades. So in short, I'm not shocked that the use of c-sections in birth has increased - quite simply because in an increasing number of cases the benefits of surgical birth outweigh the risks and costs associated with surgical birth.

Now the question that needs to be asked, and wasn't asked by this study - is whether or not all of those things that wind up causing an increase in the c-section rate might also increase the risk of childhood obesity. In which case, it wasn't the c-section that caused the babies to be beefier - and doing things to address the rate of c-sections alone (without addressing the underlying causes of the increase) won't do anything to address the rate of childhood obesity. You might just wind up with just as many beefy kids, but more birth injured moms and babies.

I also found it quite interesting, that the risk of obesity seemed to be HIGHER among those who were having urgent or emergent c-sections than among those having planned c-sections. This would seem to indicate to me, that the causes of the c-section in the first place are probably much stronger determinants of childhood obesity than the method of delivery.

In short, this study does little but add to the hysteria around the debate surrounding childbirth and further confuses the very complex problem of childhood obesity. I can see it now, the mother in line at McDonald's with her chubby little cherub playing their PS3, saying "The kid is a c-section baby - it's got nothing to do with everything else we do."

1 comment:

  1. The same phenomenon is at work with the study that "says" that maternal obesity results in autistic children. Unmentioned is that (1) obesity strongly correlates with depression and (2) depression correlates with autism spectrum disorders -- it is thought that they are related neurotransmitter disorders, and some antidepressants, such as SSRIs, have helped reduce the severity of autistic symptoms. Individuals with autism quite often have a strong family history of mood disorders.

    I'd speculate that the depression/autism spectrum disorders are inherited, and the obesity is a *result* of the depression, as the depressed individual (who may not realize that she has depression) tries to self-medicate with food and/or alcohol. My speculation at least provides a plausible mechanism to explain why obese mothers might be more likely than others to give birth to kids with ASD. (As opposed to the idea that mothers being fat causes their kids to have autism (because we all know fat is gross and icky-poo so of course it causes bad things) and if those lard-ass mamas would just GET THIN then their kids wouldn't have autism.)