Wontpower: New Data Further Undermine the Glucose Model of Self-ControlPublished 10 February, 2012
When I was in graduate school, there was a running joke about writing your dissertation. Suppose that you have this great idea, you run some careful, well-controlled experiments, and you find… nothing. It seemed only fair that if that happened – if you came up with a big donut – you shouldn’t have to write a whole section conducting a lengthy post-mortem on the study (“…one possibility is that the research was insufficiently powered…”), so you should just be able to end with:
Anyway, there has been tremendous discussion recently of the resource model of self-control, which I’ve discussed (here, here, and here) in the past, including a book (Willpower) by Roy Baumeister and John Tierney – not to mention a recent (Feb 1) piece (subscription) in New Scientist – as well as a tremendously large scientific literature drawing on these ideas. I won’t go through the model in detail; the gist is that there is a mysterious resource that you need to exert self-control, and that this mysterious resource isn’t mysterious at all, but it’s actually glucose. You need glucose in your blood to feed your brain to be able to exert self-control and, when you do exert self-control, you burn glucose, which in turn makes it harder to exert self-control in the future. This explains why you can’t solve word puzzles after you don’t eat cookies, or something like that.
I have an interest in this model, having published on it both in the pages of the online journal that hosts this blog as well as in my book, voicing, in both places, a somewhat, let us say, skeptical view of the idea. The basic problems are these: 1) it’s metabolically implausible, 2) the data in the key paper purporting to support the idea don’t, actually, support the idea, and 3) it isn’t even the right sort of explanation for how the mind works, akin to thinking that the reason that a web page is loading slowly is because your laptop battery is only at 80%.
Having said all that, there is a related phenomenon that seems to be genuine: people behave differently when they’ve ingested calories recently as opposed to when they haven’t. Less hungry organisms might be more patient, less punitive, and better able to concentrate on a task before them, for instance. This should seem intuitive to all of us who get really grouchy when we haven’t eaten in a while (as in the hungry people in commercials from the clever series of Snickers ads) and makes a lot of sense from the standpoint of thinking about how to design an organism that has many possible priorities. Hungry organisms should be expect to behave differently from full ones, generally shifting their attention and energy toward getting food, to the exclusion of other priorities.
Anyway, a new paper by Molden et al. to appear in Psychological Science, “The Motivational versus Metabolic Effects of Carbohydrates on Self-Control,” reports four experiments that put these ideas to the test.
In the first study, the authors wanted to address a shortcoming of prior work by using very accurate tools to measure glucose. Subjects fasted for 4 hours, had their blood glucose measured, performed a self-control task frequently used in this literature, and then had their blood glucose measured again. For the glucose model of self-control to be correct, the readings must be lower after the task than before. Not only did glucose not go down, but it went up (from 81.27 mg/dL to 82.39 mg/dL), though not statistically significantly.
To connect this back to the key paper in this literature (cited 287 times, as of this writing), in that paper (Gailliot et al., 2007), across four studies reported (for subjects who did not fast), glucose also went up about 1 mg/dL. So that’s consistent with the prior results. It contrasts, however, with the report from that same paper of a drop of 5.88 mg/dL, results from subjects who had fasted. This implies that there was indeed a problem with the measurement in that study, or there was some other problem with the data.
In a second experiment, the authors drew on some studies from the exercise literature that I pointed out in the paper in Evolutionary Psychology. In this work, it was found that merely rinsing with sugar solutions increases performance on physical activity (e.g., bicycling), suggesting that increased performance on cognitive tasks might be due to the sensation of reward when one drinks a beverage with carbohydrates in it, rather than willpower fuel.
Molden et al used a similar procedure, having people complete a self-control task and then swish – but not swallow – solution with sugar or a non-sugar sweetener. Briefly, they found that swishing the glucose solution – but not the sweetener – yielded effects frequently seen in this literature, suggesting that, just as in the exercise case, it’s the reward, not the glucose itself, that’s affecting behavior. A third study replicated the second; the last study showed that swishing with the sugary solution doesn’t give rise to more glucose in the blood.
In sum, these data, as well as other sets of results, clearly show that exerting self-control does not, in fact, reduce glucose, a finding which is not surprising, given that it is consistent with what is known about brain metabolism. The results that apparently contradict this result, those reported by Gailliot et al (Study 1), might very well, it seems, be due to error in the measurement device, or some other factor.
The findings that exerting self-control doesn’t reduce blood glucose critically undermines the larger model, which turns on the idea that whatever the resource is, performing self-control task uses it up. If self-control tasks do not reduce blood glucose, then the reason for subsequent drops in self-control tasks can’t possibly be due to a drop in glucose.
As for other results showing that drinking a sugary beverage changes behavior – making one more patient or less punitive – these can be understood as a phenomenon similar to what one observes in the exercise literature. Quoting myself quoting others:
[Chambers et al., 2009] conclude that “improvement in exercise performance that is observed when carbohydrate is present in the mouth may be due to the activation of brain regions believed to be involved in reward…” (p. 1779). Carter, Jeukendrup, and Jones (2004) showed a similar result, concluding that “the mechanism responsible for the improvement in high-intensity exercise performance with exogenous carbohydrate appears to involve an increase in central drive or motivation rather than having any metabolic cause” (p. 2107).
So, anyway, to return to the running joke I mentioned at the top of this entry… about that whole, “glucose is willpower fuel” thing… never mind…
Carter J. M., Jeukendrup A. E., and Jones D. A. (2004). The effect of carbohydrate mouth rinse on 1-h cycle time trial performance. Medicine and Science in Sports and Exercise, 36, 2107–2111.
Chambers, E. S., Bridge, M. W., and Jones, D. A. (2009). Carbohydrate sensing in the human mouth: Effects on exercise performance and brain activity. Journal of Physiology, 587, 1779-1794.
Gailliot, M. T., Baumeister, R. F., DeWall, C. N., Maner, J. K., Plant, E. A., Tice, D. M., Brewer, L. E., and Schmeichel, B. J. (2007). Self-control relies on glucose as a limited energy source: Willpower is more than a metaphor. Journal of Personality and Social Psychology, 92, 325-336.
Molden, D. C., Hui, C. M., Scholer, A. A., Meier, B. P., Noreen, E. E., D’Agostino, P. R., & Martin, V. (in press). The Motivational versus Metabolic Effects of Carbohydrates on Self-Control. Psychological Science.