This week must have been a terrifying one for anyone who is pregnant or thinking about having kids. Not just because pregnancy sometimes seems to be one long list of things that can go wrong, no; this week we’ve added yet another fear to the agenda.
It’s a pretty scary thing to hear, especially when you consider that many people have C-sections because they have to — either their own life is at risk, or the baby is in trouble. It’s also just another thing to add to the already-lengthy list of issues that mums are expected to feel guilty about — don’t eat soft cheese, don’t drink, and heaven forfend you choose to have pain relief during one of the most physically traumatic experiences of your life lest your baby be forever scarred.
Fortunately for expecting mums everywhere, the headlines were wrong.
C-sections probably aren’t causing autism after all.
The new study that has everyone up in arms was what’s known as a meta-analysis, which is a kind of large meta study that looks for every study that has ever been done on a topic in order to put all of their results together in a big statistical model. The basic idea is that instead of just doing more big studies, you can pool the results from lots of smaller ones to answer the same question.
In this case, the scientists took over 60 individual studies and combined their results to look at the mental health conditions in kids that are associated with a C-section birth. Altogether, they compared over 20 million records, with tens of thousands of cases of autism and mental health problems such as ADHD and eating disorders.
The researchers found that C-sections were associated with autism and ADHD — in particular, kids who were born through a C-section were 33% more likely to be autistic than kids who were born vaginally. They also found that most mental health issues were not associated with C-sections, including eating disorders, OCD, tic disorders, intellectual disability, depression, and psychosis.
Cue the terrifying headlines.
The first thing to note about this study is that it was observational, or rather that it pooled the results of observational research. What this means is that it’s impossible to know whether the relationship between autism and C-sections seen in this research is causal or not.
As I’ve said before, there are innumerable things that can possibly confound associations like this, and it’s entirely possible that something that we know causes both autism and C-sections — for example, the mother’s age when she delivers — could be behind the association seen in this research. We can control for what we know and measure, but we can’t control for what we don’t know or don’t measure.
This research was particularly open to this sort of bias, because it didn’t do any controlling of its own. The researchers simply chucked the results from a lot of other trials — some of which controlled and some of which didn’t — into one big statistical model. It’s not unlikely that things that were not measured by the studies could’ve been behind the relationship, rather than C-sections themselves causing autism.
Another thing that the headlines forgot to mention was that, while meta-analyses themselves are considered to be the most reliable form of evidence, they are only as robust as the included studies. In this case, almost half of all the studies included were classified as “low” quality, which makes the results much less meaningful.
Perhaps the biggest issue was that while the relative risk increase reported everywhere was scary — 33% increase! — the absolute risk was tiny. If you add up all the numbers and calculate the absolute risk increase of autism from having a C-section, it comes out at about 0.17%.
To put it another way, for every 1,000 kids born vaginally, about 8 will be autistic. For every 1,000 C-section babies, that number goes up to 10.
Not nearly as terrifying as the headlines might’ve lead you to believe.
There were other issues with this study — there was significant statistical heterogeneity, which is basically a problem caused by lumping together lots of different studies — but overall the message remains the same. While there’s probably an association between autism and C-sections, it’s a very tiny increase in risk, and probably caused by things other than the C-sections themselves.
Funnily enough, that’s almost exactly what the authors of the study itself said, although to be fair they said it in a much more confusing way.
Something else that no media anywhere mentioned is that, while there was an association with autism/ADHD and C-sections, there were no associations for a whole host of other issues. For those worrying about eating disorders, depression, OCD, and more, this study should be reassuring, rather than scary.
If you’re worried about C-sections, you should probably talk to your doctor, but realistically the risks here are very small. They might be meaningful to a population health person looking at birthing, but to the average individual they really don’t mean much at all.
Don’t worry about C-sections any more than you already have to (they’re still major surgery).
The headlines were wrong.
C-sections probably aren’t causing autism after all.
You can now listen to Gid on the Sensationalist Science podcast for your weekly dose of scientific shenanigans and media muddling: