1/07/2013

What can functional connectivity studies of the brain tell us about gender?

You've all heard this one before: men can't help but think about sex every few minutes, while women think about gossip and shopping instead.  The idea that when mind-wandering, men and women think about different things is nicely satirized in this image:
But, is it true?

We all have a personal opinion about this issue, because of its broader implications for men's and women's roles.  Personally, I think this is a vast exaggeration like most stereotypes, applying to some men and women but not all, and probably not the majority.  I've known both women who think about sex all the time and men who rarely do.  But the question of how many men and how many women fit these stereotypes is, like most statistical questions, best answered by experiments, not personal experience.

As Neurokuz describes, a neuroimaging study recently took on this question in an ingenious way, and found that men and women may not daydream that differently after all.  Case closed, then?  Let's take a closer look at the study and find out.

The basic idea was to test what men and women think about when idly daydreaming--letting their minds wander without thinking about anything in particular.  One could simply ask men and women to close their eyes and think of nothing in particular, then interrupt them at random intervals and ask them to report what they were thinking.  However, one would have to be very careful to keep them from knowing the purpose of the study, or they would distort their reports in line with either their own beliefs about gender or their beliefs about what the experimenter would want to hear.  (Participants trying to oblige the experimenter plagues any psychology study that uses self-report).  Instead, these researchers used fMRI to look at a brain network called the "default mode network," a network of many regions spread across the brain thought to underlie thinking about nothing in particular while at rest.  The researchers predicted that if men really defaulted to thinking about sex while women pondered gossip and shopping, the pattern of activity in their default mode network should look different.

Notice that they didn't expect differences in the amount of default mode network activity--both men and women should have similar amounts because both were equally engaged in daydreaming.  Rather, they looked for differences in "functional connectivity," or in what brain regions are communicating with each other.  Functional connectivity isn't a direct anatomical measure of the white matter connections between regions.  Rather, it measures co-activation: the greater the correlation between activity in different regions, the higher the functional connectivity.  The idea was that thinking about sex and thinking about shopping should involve some different areas, so the pattern of functional activity would be different between men and women if they were thinking about different things.  Or as Neurokuz put it,
"if you accept the classic theory of males versus female brain function, this is the network that represents thoughts of sex and lame excuses for men, and thoughts of shopping and musical sitcoms for women."
When the 26 women and 23 men were scanned, their functional connectivity was virtually identical.  The researchers concluded only that men and women have identical default mode activity, and so sex need not be controlled for in future studies of the default mode network.  This seems to follow straightforwardly from the results, as far as I can tell.  They drew no conclusions about the implications for gender stereotypes.  Neurokuz, however, proclaimed that this study demolished the stereotype that men think about sex and lame excuses while women think about shopping.  I like this conclusion, but I'd recommend caution, because it's based on some assumptions that may or may not be true.

The assumption is that functional connectivity straightforwardly reflects the content a person is thinking about.  Or in other words, you have some areas used to think about sex, other areas used to think about shopping, they don't overlap at all, and so areas thinking about sex working together should look really different than areas thinking about shopping.  It's intuitively plausible, I guess, but we know so little about what brain networks are involved with thinking about specific topics that we really can't assume.

Suppose the content-specific parts of the thought are outside the default mode network altogether, and the part that's in the default-mode network is a set of regions for thinking about rewarding things.  If men find sex rewarding and women find shopping rewarding, then the same default-mode regions will light up.  And if the analysis limits itself to the default mode network and doesn't look at regions outside of it....the activation will look the same, even if the content is totally different.  Not saying this is true; just that it's possible and we just don't know enough yet to draw a strong conclusion.

And that's not even getting into all the controversies associated with the interpretation of the default mode network itself: is it a real, dedicated network?  (Dedicated meaning there is at least one function which it always performs).  Does default mode network activity directly reflect "what we're thinking about while letting our minds wander?" as Neurokuz has described it, or is there something else going on?  Is there even a consistent default mode network?  Maybe the topics we think about are so various and scattered in so many different ways around so many brain regions that there isn't enough consistency to talk about a default mode network. (Studies like this one that analyze a certain set regions preselected as the "default mode network" assume a consistent network).

And then there's the problem of functional connectivity.  It's bad enough inferring what someone's thinking from a regular fMRI study that just looks at activation in particular regions.  You're one step farther removed when you're looking at the correlations between activations in different regions.

Bottom line, if you want to draw strong conclusions about gender from a study, don't pick a neuroscience study like this one.  This study is interesting and tells us something new about the brain, but doesn't let us conclude much of anything about gender.  Instead, do a more traditional social psychology study where you measure people's behavior or get them to report on their thinking.  Who knows, maybe we'll still find that men and women think about the same things.