Serotonin may help you recognize the sad and bad, rather than glad

Jan 18 2011 Published by under Behavioral Neuro, Neuroscience

Sci has RETURNED from a bangup AWESOME time at #scio11. Sadly, my trusty netbook was not particularly trusty, and so I wasn't able to crazy tweet up the conference like other attendees did. But I learned a lot and had a spectacular time meeting and having deep conversations with everyone. It's like the first week of freshman year of college, only with more booze. Bigger writeup coming soon, as I finalize the google doc for the session that I helped with on "how to explain science in blog posts".

In the mean time, let's talk about serotonin, and recognition of emotional faces. Alves-Neto, et al. "Effect of escitalopram on the processing of emotional faces" Brazilian Journal of Medical and Biological Research, 2010.

Let's start with serotonin. Serotonin is a neurotransmitter (a chemical messenger) which is found ALL over your body. It has major effects in the gut, but most people like it for its effects in the BRAIN. Most particularly for its effects on mood and our processing of emotions. Most antidepressants on the market today (like Prozac and Celexa and others) work by targeting the serotonin system, blocking a molecule that recycles serotonin, and thus keeping serotonin around and increasing its levels in the brain.

Now, I mostly deal with the mechanics of how we think antidepressants work (via things like increased serotonin or increases in the birth of new neurons in the brain), but there are also psychological theories as to how antidepressants work, in how they affect your thought processes. These affected thought processes will always have a physical basis of changes in the brain, but as yet the two mechanisms haven't really connected.

Right now, the hypothesis is that antidepressants exert their physical effects on the brain, and these effects change how you perceive things around you, and how you REMEMBER things that have happened to you. Some scientists think that depression (which also comes down to physical changes in the brain, we just don't know what they are yet) is characterized by a negative outlook on life. Well, duh, you might think. But it's more than just that. It's a negative outlook on the way you perceive things, causing negative bias. When someone looks angry, for example, a person without depression may think they just look angry at something, while someone with depression might assume the angry person is angry at THEM, whether or not they actually are. Not only that, depression can put a negative cast on how you think of things in the past. You may remember someone frowning, and while someone without depression may think they were disappointed about something, someone with depression will think the person is disappointed with THEM, specifically. People with depression pay less attention to happy faces, and even interpret neutral faces as sad, which suggests that they have a significant negative spin on their emotional processing. This effect is usually referred to as rumination, where someone focuses their attention on negative perceptions and memories.

And so if you have a negative cast on your perceptions and emotional processing, it makes sense that the changes produced by antidepressants would work to change that, by maybe getting rid of the negative boas and allowing you to perceive things more as they are. This hasn't been proven, and the hypothesis has a long way to go. And looking at this paper will kind of tell you why.

The authors wanted to look at the effects of escitalopram (a serotonin selective antidepressant) on how people process emotional faces. To do this, they took a bunch of dudes (no women because they didn't want to deal with effects of the menstrual cycle), and gave them a dose of citalopram. They then put them in front of a video monitor and had them look at faces. The faces start out neutral, and morph slowly into expressions of happiness or sadness (they also used disgust and a few other measures, but those didn't pan out). The participants were instructed to determine which emotion it was and hit the relevant key for it on the computer as SOON as possible. The idea here is that if you become better at processing certain emotions while on escitalopram, your responses will become faster for those emotions.

And they did indeed get something.


You can see above that they tested reactions to both male and female faces. The responses to female faces never changed, which I think is MASSIVELY interesting and which they mention as being weird, but never discuss (it seems like every other paper I cover has sex differences that aren't discussed). But when guys given citalopram looked at MALE faces, they were able to determine that the faces were sad FASTER than when they were given a placebo. You can see how the top right graph has the dark line that shoots up quickly, it means that the citalopram condition had the participants identifying sad faces when they were only 30% morphed, while placebo condition took until 40% to notice and they never identified them with as much certainty.

The bottom right graph shows the exact opposite for HAPPY faces. You can see that the citalopram participants showed less response to happy faces, they never identified them as strongly.

Now, you are probably sitting there thinking: that isn't RIGHT! If antidepressants WORK, they are supposed to make you less negative and decrease your recognition of sad faces and increase recognition of happy faces!

Actually, no! Other studies have shown that giving antidepressants like this will actually increase recognition of fear, so additional recognition of happiness is not unexpected. But it still probably leaves you wondering how this works. It leaves me wondering how it works, too. The results are not very clearly discussed as to WHY they got the results they got, so it's possible the scientists themselves don't know. But I have some hypotheses.

First, I am not going to lie to you, I had some issues with this paper. I think it illustrates some cool things about serotonin and emotional processing, but there's a lot of little issues about this paper that maketh a young scientist wince. First off, the paper is TERRIBLY cited. There's a bunch of stuff in the introduction that isn't cited at all and definitely should be, not to mention more stuff in the discussion. The journal this paper is published in isn't a very good one (impact factor 1.075), but still, that's something that should have been caught. Always cite your work!

Also, what with WITH the acute administration? Several groups have done this, giving acute antidepressants and looking at emotional face processing, and I always wonder why they just give the one dose. Antidepressants don't WORK for depression after the first dose. Many of them take several weeks to produce antidepressant effects. Why not administer chronic antidepressant treatment for three weeks or so and see how that differs when people look at faces? I think the responses might be quite different from the acute administration. After all, a one time dose of escitalopram is going to increase serotonin levels in your brain. But three WEEKS of escitalopram is going to change lots of other things, like receptor densities, neuron birth, and more. While it's not ethical to give antidepressants to people who are not depressed for a test like this, citalopram is prescribed for other issues such as anxiety, and some people with anxiety do not exhibit depressive symptoms. You could take some of them following their antidepressant treatment and see what happened. You could also look at people with depression and see how they are changed.

Finally, what does the processing of emotional faces MEAN? The scientists went on in the discussion about possible interactions in the raphe and how maybe their participants were too psychiatrically normal, but many of their arguments don't make a lot of sense (they talk about how maybe escitalopram is activating 5-HT1A receptors in the raphe and decreasing serotonin neuron firing, and this is decreasing serotonin post-synaptically...and while that certainly does happen, its easily overwhelmed by the increase in extracellular serotonin from the escitalopram blockade of the serotonin transporter.../geekrant). I'd be much more interested to see a discussion of the acute vs chronic effects of citalopram in emotional processing, and some hypotheses as to why INCREASES in serotonin can increase reactions for negative faces.

But in the future, Sci wants to see some studies with CHRONIC antidepressant administration! And while you're at it, get some ladies in there. Considering that women suffer depression at rates higher than that of men, it might make for interesting and highly useful science.

Alves-Neto, W., Guapo, V., Graeff, F., Deakin, J., & Del-Ben, C. (2010). Effect of escitalopram on the processing of emotional faces Brazilian Journal of Medical and Biological Research (ahead) DOI: 10.1590/S0100-879X2010005000007

6 responses so far

  • becca says:

    I don't think they were actually afraid of blood cooties, but that women are too good at facial recognition; they would totes skew the curve. 😛

  • Samantha Vimes says:

    I think the most interesting thing about this is *definnitely* that drugs or no drugs, men read women's faces *way* better than the do men's. I wonder if this is a cultural thing-- men are expected to look more at women, or men are taught to see women as more mysterious and therefore a code to be cracked-- or is it a more universal thing? I'd be tempted to do the study without drugs with non-WEIRD participants to get a better idea of what's going on.

  • Kees says:

    I would like to see how the recognition of sad faces works out for homosexual men;
    -If their recognition of male faces with the antidepressant doesn't increase and if their recognition of female faces does increase, I would guess that the difference in recognition is related to how men look at the faces.
    -If their response to the drug is very similar to heterosexual men, I'd guess the differences in recognition are due to the differences in how men and women express their emotions.

    Would it make any sense to investigate that? Neuroscience is totally not my field of expertise, so please comment if you think my proposal is total BS.

    • scicurious says:

      My hypothesis is that women may be easier to read regardless, and that this may be due to social factors. But other studies have shown that women read women better than men, so it may be a matter of reading your own sex better, which doesn't really make intuitive sense to me. Absolutely I think that investigation would make sense in conjunction with other studies.

      • James Kornell says:

        Don't know about easier to read, per se, but if as a baby your predominant caretaker was a woman (as she probably was), then understanding her moods was a matter of life or death, regardless of your own gender. (This would predict no diff between heterosexual and homosexual men. Testable!)

Leave a Reply