Stephen Colbert was right: "Guts" actually do affect our emotional response

i-c7b5006aab9cfe1e8ad3e0c768e57b01-vianna1.jpgOn the opening episode of the Colbert Report, faux conservative Stephen Colbert expressed his preference for "guts" over facts:

That's where truth comes from -- the gut. Facts come from the brain -- and some people think that makes facts better. But did you know you have more nerve endings in your stomach than in your brain? You can look it up....

Anyone can tell the news to you. I'm going to feel the news at you.

In Colbert's signature parodic style, he appeals to the "common-sense" notion that "guts" matter more than "brains." Even his dubious claim about nerve endings has some merit -- though the brain clearly has many more neurons, it can't sense tactile input the same way the organs in the human gut can.

i-eca0cf2af9fc3ac4445c7dff7d8aab70-research.gifThe idea that "guts" literally serve some cognitive function isn't as far-fetched as it may seem. Some research has found that different visceral states (e.g. indigestion, heartburn) map on to specific brain areas associated with emotion. The relationship between the gastrointestinal system and the brain is particularly complex, but little research has explored whether there is a direct link between our physical "guts" and our emotional responses.

A team led by Eduardo Vianna found a novel way to explore the relationship between guts and emotion. The key was patients with Crohn's disease, an condition that increases feedback from the nerves in the gastrointestinal system to the brain. Crohn's disease can have active and silent phases, so otherwise similar individuals can be easily compared.

Patients in both active and silent phases were shown movie scenes known to induce a variety of emotional states: Neutral (documentaries); Happiness (couples hugging or learning of a pregnancy); Disgust (dog hip-replacement surgery, woman eating dog feces); Sadness (death of a father; wife undergoing surgery); and Fear (horror films). Each patient was monitored with an electrogastrogram (like an electrocardiogram for the gut), and after each clip, the viewers rated their levels of each emotion on a scale of 1 to 7.

Even though their baseline levels of the emotions had been measured and found equal prior to the study, patients expressed different levels of the emotional arousal depending on the phase of the disease:


This graph shows that patients in active phase of Crohn's disease experienced emotions that were significantly more intense after watching the disgust-, fear-, and sadness-inducing clips than patients in the silent phase.

So when you say that someone who shows little emotion in the face of danger has "guts," you may literally be talking about the condition of their gastrointestinal tract, not just their mental discipline. The electrogastrogram measurements support this notion as well:


As patients in the active Crohn's phase experienced more intense emotions, their electrogastrogram levels increased significantly. But there was no correlation between these levels and emotional arousal in the silent phase patients.

So when Stephen Colbert suggests that it's better to use your "gut" instead of your brain to make judgments, he may simply be reflecting the natural way humans experience emotions. What's less certain is whether using your gut instead of relying on facts to guide actions is actually a better strategy for success in the modern world.

Vianna, E.P.M., Weinstock, J., Elliott, D., Summers, R., & Tranel, D. (2006). Increased feelings with increased body signals. SCAN, 1, 37-48.

More like this

Your blog is EXTREMELY old and poor science....the book is called, if memory does not fail me,....'The Origins of Consciousness (sp?) in the Breakdown of the Bicameral Mind' by Julian it for fun at least

You mean this book? It's not a work of science - it's an unprovable theory.

By Brian Mingus (not verified) on 20 Mar 2007 #permalink

well holy shit, that explains a lot.
yea... crohn's humor. seriously, thanks for the cite.


Enteric "second brain" -- Gershon

# ISBN-10: 0060930721
# ISBN-13: 978-0060930721

And I just today saw (NYT?) an article on treating autism with nutrition, based on the genetics recently worked out that changes not the CNS but the gut innervation.

Makes sense.

By Hank Roberts (not verified) on 22 Mar 2007 #permalink

Your first bar graph is completely meaningless without estimates of error and their corresponding p values. Your second scatter plot made me physically cringe. The lines that are used to "fit" the data could just as easily be swapped, and still have similar R values.


I dispute the notion that the first graph is "meaningless." We indicate which results are significant in the text, and combined with the graph that should give you a good sense of the data. Error bars, while useful for experts, are confusing for non-experts.

The second graph gives a good sense of the variance present in the data and reflects a significant positive correlation between arousal and EGG levels in active-phase patients.

Diarrhea is a common symptom of nervousness-frequently experienced before performing or speaking to an audience. Vomiting is experienced by some people as a result of receiving extremely disturbing information or a negative emotional interaction with another person. So, yes, empirically it's safe to say that there's a connection.

wow -- look at that graph! if you removed a single data point (the yellow dot in the upper-left quadrant), the two lines would be remarkably close together. looks like a difference that's probably in the noise...

(not that it's not interesting in many ways)

I think there's a lot of misreading going on with these graphs. Take a look at the second graph. On the left hand side, all the yellow points are above all the red points. In the middle, all the points are close together. On the right, all the red points are above all the yellow points. That's why there's a significant difference between the silent-phase results and the active-phase results. This is borne out by statistical calculations reported in the article; the graph is just an illustration of those results.

Before taking this study seriously, one would have to ask: did the researchers control for medications? Many active Crohn's patients will be taking steroids which cause more intense emotional states. And what kind of sample size are we talking about? A minimum of 200 active and 200 inactive participants are needed to draw any statistically significant correlations.

Also, as a long-time Crohn's sufferer, I think it's fair to say that if you are suffering an active flare, then you are experiencing a higher level of stress already and that it may easily increase your emotional reaction to a stimulus. It is very difficult to function normally and get work and family commitments fulfilled when one has to constantly and in great pain use the restroom.

Yes, it's easy to say the gut influences emotions - but I don't think comparing Crohn's disease sufferers is the way to do it. This is degrading and can easily be used to support the many wrong people who insist that such diseases are all in the mind.

By Michelle Gray (not verified) on 23 Mar 2007 #permalink


Sample size is related to significance, but statements like "A minimum of 200 active and 200 inactive participants are needed to draw any statistically significant correlations" are inaccurate. The sample size in this case was much lower: 8 active and 12 inactive. Yet the results were statistically significant. This is because the task was repeated many times, with many different stimuli, and the results were consistent enough to be significant.

Re: controlling for medication. Medication was considered, but an even more important control was conducted: 20 people who do not have Crohn's disease were also tested, and their results matched those in silent phase.

I'm surprised you believe these results could be used to make the claim that Crohn's is "all in the mind." The results show significantly more nerve activity in the abdomen for active phase patients than for silent phase patients; surely this demonstrates that Crohn's is a real medical condition.

Also, there are plenty of real conditions such as depression which are entirely "in the mind," in the sense that no symptoms are observed other than in the brain and behavior of an individual.

I'm sorry - I have to make another comment after looking again at the second graph. Obviously the sample size was teeny tiny - honestly my first grader had a larger sample size for his science fair project and we still made sure to include lots of comments about it not being significant due to sample size.

AND, all good researchers (and yes, I used to be one) know to throw out their outliers, otherwise it will skew their data. There are at least two outliers that need to be removed and then the two lines will look a lot similar.

By Michelle Gray (not verified) on 24 Mar 2007 #permalink

Michelle --

I think you're confusing the number of participants with the statistical power of an experiment. If you only have one data point per participant, then it's true that a sample size of 20 would be too small. However, in this study, each participant viewed ten different movie clips. So that means there were 200 different data points.

Each dot on the graph represents an average of all the participants in the condition, not just one participant. So an "outlier" in this case isn't the same as in other studies where you may assume one participant didn't follow directions.

In medical studies, it's often assumed that a certain number of participants is required to get valid results -- because in that case, each person indeed represents a single data point (i.e. "the treatment worked or it didn't"), so that may be where you came up with the 200 number.

However, this isn't a medical study; it's a study where a medical condition helped researchers draw conclusions about a behavioral and cognitive phenomenon.

I'd argue that it's more important to show a variety of different video clips, rather than, for example, to show the same clip to each of 200 different participants -- because the phenomenon might be due to some particular property of the single clip, instead of a true difference among participants.