Can we stick a fork in the “glucose-as-willpower-fuel” model?Published 1 October, 2012
A new paper is now available by Martin Hagger and Nikos Chatzisarantis (hereafter H&C) entitled, “The Sweet Taste of Success: The Presence of Glucose in the Oral Cavity Moderates the Depletion of Self-Control Resources,” published online in Personality and Social Psychology Bulletin. The paper addresses the idea that the reason that people can’t exert “self-control” – resisting tempting cookies, persisting on solving unsolvable anagrams, and so forth – is that their brain has insufficient glucose left. One line of evidence (Gailliot et al., 2007) marshaled in favor of this account is that people’s performance on such tasks has been shown to improve after subjects drank a sugary beverage. The interpretation of this evidence, however, runs afoul of an important confound. People who consume a sugary beverage have not only had their supply of glucose increased; they have also experienced the reward associated with the sweet taste in their mouth from the sugary drink. Is it the glucose, or the rewarding sensation of sweetness?
To explore this possibility, Molden et al. they ran some experiments to try to distinguish between these two possibilities in a paper published earlier this year in Psychological Science, which I discussed back in February. Here’s how they motivated the work:
Carbohydrate mouth-rinses activate dopaminergic pathways in the striatum–a region of the brain associated with responses to reward (Kringelbach, 2004)–whereas artificially-sweetened non-carbohydrate mouth-rinses do not (Chambers et al., 2009). Thus, the sensing of carbohydrates in the mouth appears to signal the possibility of reward (i.e., the future availability of additional energy), which could motivate rather than fuel physical effort.
Molden et al demonstrated, first, that when glucose was measured more accurately than in prior work, performing a self-control task did not, in fact, reduce the levels of circulating glucose. This is bad for the part of the theory that says that self-control tasks differentially reduces peripheral glucose. Second, in two experiments, drawing on findings in the exercise literature that showed that rinsing with a sugar solution improved cycling performance, they showed that subjects who swished but didn’t drink the beverage showed the same improvement. These data imply that it’s the reward rather than the glucose that’s causing the gains in performance.
The new paper by H&C adds additional data using the same technique. They report five studies, all of which follow along roughly the same lines. In Studies 2 and 3, for instance, subjects first performed a task that is purported to sap “self-control resources,” and then performed a task that requires self-control, either solving anagrams (Study 2) or drinking an unpleasant beverage (Study 3). Subjects rinsed with (but did not swallow) either a glucose beverage or a placebo between the two sets of tasks. Subjects who rinsed with the glucose drink spent more time on the anagrams and drank more of the unpleasant beverage than those who swished with the placebo drink, reproducing the prior Molden et al results. They conclude in their general discussion (p. 10):
Our findings provide an alternative explanation that not only provides a suggested mechanism independent of glucose metabolism but also accounts for the increasing body of research that has shown significant improvements in self control as a result of oral glucose supplementation (e.g., DeWall et al., 2008; Dvorak & Simons, 2009; Gailliot et al., 2007; Masicampo & Baumeister, 2008). As ingestion of a glucose solution in previous supplementation studies means that it will be present in the oral cavity, albeit for a short period of time, it is also likely to lead to the sensing of the glucose in the oral cavity and facilitation of self-control performance through a central mechanism.
Adding these five experiments to the prior work by Molden et al. – not to mention other reasons to be skeptical of the idea in the first place (see below) – might contribute to what I think will be the decline and ultimate disappearance of this idea in the literature. Two additional points. First, the sample sizes in the five reported studies are small. In Studies 2 and 3, for instance, the sample sizes are 32 and 34 respectively, so the individual cell sizes in the two conditions must have been modest. In addition to the usual worries about small sample sizes, this made me think of Ulrich Schimmack’s recent paper in Psychological Methods, which also went after the Glucose-as-Resource work, but on statistical grounds, which concluded that “from a statistical point of view, Bem’s (2011) evidence for ESP is more credible than Gailliot et al.’s (2007) evidence for a role of blood-glucose in self-regulation” (p. 9). Anyway, I find it puzzling that these cell sizes aren’t larger.
Second, and this is more of a sociology of science note, even though H&C’s studies are conceptually the same as Molden et al.’s, the latter paper is, surprisingly, not reviewed or even cited. This is a puzzling omission, given how close the two sets of studies are, and especially given how important credit for priority is to the first author, Hagger. Close followers of the blog might recall that when I discussed the Molden et al. work, Hagger posted the following comment, claiming priority and advising me that “It would be worthwhile citing this in order to ensure that your literature review is as comprehensive as possible.” Here is the comment in full:
I wonder if you are familiar, Dr. Kurzban, with our review published in Health Psychology Review [Hagger, M. S., Wood, C., Stiff, C., & Chatzisarantis, N. L. D. (2009). The strength model of self-regulation failure and health-related behavior. Health Psychology Review, 3(2), 208-238. doi: 10.1080/17437190903414387] that supercedes the one you published in Evolutionary Psychology that suggests this possibility. In the review, we state explicitly: “While Galliot et al.’s (2007a, 2009) series of studies provides consistent evidence for glucose supply as a mechanism for self-control resource use and depletion, there may also be a central mechanism controlling self-control resource depletion. As a consequence, supplementation with a carbohydrate mouth rinse may moderate the ego-depletion effect.” We elaborate on this further citing the same exercise physiology research that you did. It would be worthwhile citing this in order to ensure that your literature review is as comprehensive as possible and that you are not ignorant of the fact that other researchers may have also had this idea and, in fact, published it before you. Of course, people arriving at ideas in parallel to (or slightly in front of) others should never be a hinderance to publication, it happens all the time, and it is important academics are open minded about parallel submissions of similar work submitted to journals and give them fair opportunity to be published.
My point here isn’t about the issue of parallel submissions, and it seems to me that many journals, for better or worse, prefer to publish only the first instance of a new idea – which is why scientists get so irked when they are scooped – but it seems to me that it’s unfortunate that H&C didn’t ensure that their literature review was as comprehensive as possible. The record should, in my view, reflect the close connection between these two sets of studies.
So, to return to the question in the title, should this new evidence, taken together with prior data, put a punctuation mark on the glucose-as-willpower-fuel model? In my view, it certainly should, and its demise is overdue. The new data should be understood in the context of other difficulties the model faces. First, the math makes no sense. The number of calories burned by a self control task compared to a similar but slightly different task could have been guessed to have been trivially small from what is already known about brain metabolism. Second, the glucose-as-willpower model is the wrong sort of explanation, similar, as I’ve argued before, to locating the cause of a slow computer application to a low battery. The right explanation for effects in this literature is, ultimately, going to have to be a computational explanation.
H&C adroitly recall Baumeister et al.’s (1998) remark that it is “implausible that ego depletion would have no physiological aspect or correlates at all” (p. 1263). If the glucose-as-willpower model is wrong, then advocates of the more general willpower-as-a-resource model should be worried that there is no good candidate for the “physiological aspect” of the resource.
Is it time to stick a fork in the willpower-as-resource model as well? I think so. Still, there is an interesting phenomenon to be explained; why do the tasks in this literature evoke the unpleasant phenomenology of “effort.” For those attending the Society for Personality and Social Psychology conference in January in New Orleans, I’ll present an alternative interpretation of mental fatigue in the evolutionary psychology preconference.
Gailliot, M. T., Baumeister, R. F., DeWall, C. N., Maner, J. K., Plant, E. A., Tice, D. M., 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(2), 325-336.
Hagger, M. & Chatzisarantis, N. (in press) The Sweet Taste of Success: The Presence of Glucose in the Oral Cavity Moderates the Depletion of Self-Control Resources. Personality and Social Psychology Bulletin.
Schimmack, U. (2012). The Ironic Effect of Significant Results on the Credibility of Multiple-Study Articles. Psychological Methods. Advance online publication. doi: 10.1037/a0029487