Evolutionary Psychology

Evolutionary Psychology is moving to SAGE. The new address is evp.sagepub.com. Submissions here.
Robert Kurzban

The Evolutionary Psychology Blog

By Robert Kurzban

Robert Kurzban is an Associate Professor at the University of Pennsylvania and author of Why Everyone (Else) Is A Hypocrite. Follow him on Twitter: @rkurzban

Glucose Is Not Willpower Fuel

Published 29 August, 2011

In my last post, I talked about the idea that there is a resource that is necessary for self-control. People discussing this idea have  used various metaphors – self-control is a muscle, willpower is a reservoir – but few suggestions have been made about what, exactly, the resource in question actually might be. Not long ago, one proposal was made to make the idea “more than a metaphor.”1 In this post, I want to talk a little bit about the candidate for this resource, glucose. Could willpower fail because the brain is low on sugar?

Let’s look at the numbers. A well-known statistic is that the brain, while only 2% of body weight, consumes 20% of the body’s energy. That sounds like the brain consumes a lot of calories, but if we assume a 2,400 calorie/day diet – only to make the division really easy – that’s 100 calories per hour on average, 20 of which, then, are being used by the brain. Every three minutes, then, the brain – which includes memory systems, the visual system, working memory, then emotion systems, and so on – consumes one (1) calorie. One. Yes, the brain is a greedy organ, but it’s important to keep its greediness in perspective.

The next question is the comparison between a brain exerting willpower and one that’s not. Suppose, for instance, that a brain in a person exerting their willpower – resisting eating brownies or what have you – used twice as many calories as a person not exerting willpower. That person would need an extra one third of a calorie per minute to make up the difference compared to someone not exerting willpower.

Over the years, a number of researchers have asked whether the brain uses more glucose during tasks that feel effortful compared to tasks that don’t. Here is a sample of the conclusions drawn by researchers reviewing the evidence. Does exerting “self control” burn more calories?

  • Clarke and  Sokoloff (1998) remarked that although “[a] common view  equates concentrated mental effort with mental work…there appears to be no increased energy utilization by the brain during such processes” (p. 664), and “…the areas that participate in the processes of such reasoning represent too small a fraction of the brain for changes in their functional and metabolic activities to be reflected in the energy metabolism of the brain…” (p. 675).
  • Gibson and Green  (2002), talking about a possible link between glucose and cognition, wrote that research in the area “…is based on the assumption that, since glucose is the major source of fuel for  the brain, alterations in plasma levels of  glucose will result in alterations in brain levels of glucose, and thus neuronal function. However, the strength of this notion lies in its common-sense plausibility, not in scientific evidence…” (p. 185).
  • Lennie (2003) concluded that “[t]he brain’s energy consumption does not change with normal variations in mental activity” and that “overall energy consumption is essentially constant” (p. 495).
  • Messier (2004) concluded that it is “unlikely that the blood glucose changes observed during and after a difficult cognitive task are due to increased brain glucose uptake” (p. 39).
  • Gibson (2007), concluded that “task-induced changes in human peripheral blood glucose are unlikely to reflect changes in relevant areas of brain glucose supply” (p. 75).

Now, there is some evidence that might be taken to pull the other way, but these summaries suggest that there is more or less a consensus among researchers who investigate brain metabolism about this issue.

Nonetheless, as I indicated above, some have suggested that exerting self control drains the brain of glucose, which in turn leads to a diminished ability to exert self control. Specifically, I want to address a paper on this topic that is supposed to provide key evidence. A careful reading of this manuscript, which has been heavily cited as providing evidence in favor of the model, actually undermines the claim. In one study in this paper,1 roughly 100 subjects fasted for three hours prior to the experiment. Glucose was measured before and after these subjects were asked to watch a film for five minutes; half were told not to look at words appearing on the screen. (This was the “self control” task.) Recall that in five minutes, the whole brain consumes about 1.5 calories. The paper reports that subjects who watched the film without looking at the words did show the glucose drop, while those who didn’t saw a (non-significant) rise in glucose. This seems to demonstrate that exerting “self control” does consume glucose.

Given that the experts on brain metabolism find that blood glucose doesn’t change depending on what task subjects do, this finding seems striking. There are, however, two reasons to worry. First, the subjects in the self-control condition (N=52) had initial blood glucose readings well above those of the other group’s as well as above the level found in other studies. In other words, the reason these subjects’ blood sugar went down might have to do with the fact that they were abnormally high.2 Second, when I asked for the raw data, I was told that the data were “corrupted,” and so – like Joseph Smith’s Golden Plates –we just have to take the authors’ word for it.

The same paper reported a series of additional studies in which subjects (who had not fasted for three hours) had their glucose reading taken before and after doing a “self-control” task. In these studies, glucose didn’t go down at all. In fact, it went up (though not statistically significantly). (See Table 1 in this paper.)

In sum, the paper frequently touted as the best evidence in favor of the glucose-as-resource-for-self-control model shows 1) a drop in glucose for people with abnormally high levels of glucose and 2) no drop in glucose among a larger sample of subjects.

In short, the idea that more sugar is consumed when exerting “self control” is wrong. Actually, Hockey (2011) would say that the emphasis on glucose as the fuel for self control is not just wrong, but destructive of progress, writing:

Remarkably, given that fatigue has been studied formally for well over 100 years, there is still no scientifically mature theory of its origins and functions.

One part of the reason for this lack of progress, he suggests, is that there is an “irresistible tendency to think of it in terms of a loss of energy resources,” but that there is “no evidence” for the claim that “fatigue is the result of glucose depletion.” Hockey concludes that “there is little doubt that the energy-depletion perspective has been a source of distraction in the search for a theory of fatigue” (p. 167).

Van den Berg (1986) pointed this out more than a quarter of a century ago. The right answer to the question about why exerting willpower is hard is not going to be packets of sugar:

Descriptions of behavioral acts or activities using a terminology in which concepts like energy play a major role are very likely based on a false analogy. The assumption that the brain is a machine like the muscles, albeit more complex, is ill-founded. Once we have developed a theory of the dynamics of behavior in which no traces remain of nineteenth century conceptual frameworks of energetics many of our current problems will vanish. We will not then be misled by the term mental energy, nor will we advise someone to take sugar when he or she complains of being tired after writing a paper! (p. 134)

1 Gailliot et al., 2007

2 For those who care about such things, the average was reported as 107.1 mg/dL. As one comparison from an entirely different group of undergraduates, Dvorak and Simons (2009) also had subjects fast and obtained a mean of 97.7 mg/dL (SD=11, N=180). So, the mean reported in the paper in question is around 10 standard errors above this mean.


Clarke, D. D., and Sokoloff, L. (1998) Circulation and energy metabolism of the brain. In  G. Siegel, B. Agranoff, R. Albers, S. Fisher, and M. Uhler (Eds.), Basic neurochemistry: Molecular,  cellular, and medical aspects (6th Ed.) (pp. 637-669). Philadelphia, PA: Lippincott Raven

Dvorak, R.D., and Simons, J.S. (2009). Moderation of resource depletion in the self-control strength model: Differing effects of the two modes of self-control. Personality and Social Psychology Bulletin, 35, 572-583

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.

Gibson, E. L. (2007). Carbohydrates and mental function: Feeding or impeding the brain? Nutrition Bulletin, 32, 71-83.

Gibson, E.L., and Green, M. W. (2002). Nutritional influences on cognitive function:  Mechanisms of susceptibility. Nutrition Research Reviews, 15, 169–206.

Hockey, G. R. J. (2011). A motivational control theory of cognitive fatigue. In P. L. Ackerman (Ed.), Cognitive fatigue: Multidisciplinary perspectives on current research and future applications (pp. 167-187). Washington, DC: American Psychological Association.

Lennie, P. (2003). The cost of cortical computation. Current Biology, 13, 493-497.

Messier, C. (2004). Glucose improvement of memory: A review.  European Journal of Pharmacology, 490, 33-57

Van den Berg, C. J. (1986). On the relation between energy transformation in the brain and mental activities. In R. Hockey, A. Gaillard, & M. Coles (Eds), Energetics and Human Information Processing (pp. 131-135). Dordrecht, The Netherlands: Martinus Nijhoff.

  • DiscoveredJoys

    Let me acknowledge straight away that I bring only an interested onlooker’s viewpoint to the debate. That said, I was interested in the idea that the resource analogy for willpower misleading. I’m pretty sure that many of the other analogies about the brain and its process are misleading too.

    The first thing that struck me was that the resource analogy starts with ‘willpower’ as a particular activity in the brain. I wonder if willpower does not actually exist as a ‘process’ – perhaps the ability to suppress distracting processes provides the experience interpreted by our consciousness as willpower?

    This could explain why we ‘run out of willpower’. The suppressed distractions become more distracting over time (the increasing feelings of hunger, discomfort, fatigue, desire for a change of activity, desire for novelty etc.) needing more and more suppression. Where you move from one willpower task to another many of of the same suppressed distractions continue, undermining the new task focus and providing the illusion of using up a resource .

    One of the tests for this hypothesis would be a reduced willpower exhaustion when two sequential tasks require suppression of markedly different distractions.

    It would be interesting to speculate whether or not this suppression of distraction mechanism is involved in the ‘flow’ experience – something that does not lend itself to the finite resource analogy.

    • Ian

      The trouble is that people are so used to seeing medical arguments showing how function is a byproduct a physiology. ie, I don’t have superstrength bones or muscles because they can’t get any stronger.

      However when you deal with patterns of behaviour, you have to remember that behaviour can have virtually any arbitrary shape. There are hard working people, lazy people, nervous people, calm people, etc.

      Certainly, our behaviour *could* be simply a reflection of underlying physiology. Maybe we really do get tired and unmotivated simply because we’re running low on glucose.

      However I believe (as does most of the EP profession) that it’s far more likely that we adjust things like our rate of effort based not on things internal to us (such as blood glucose, or self esteem), but based on things external to us. It is far more likely that there are strong selection pressures for people to fine tune their behaviour to their surroundings as effectively as possible, using as much outside information as possible.

      It’s far more useful, not to mention intellectually satisfying, to use these *external* explanatory models, and furthermore, they are actually much more testable and falsifiable in the long run. Therefore rather than trying to explain motivation in the context of “he quit the task because it hurt his self esteem, or because it didn’t satisfy his ego, or because his blood glucose was low”, it is far more useful to think in terms of “he quit the task because a module in his head which seeks to track the usefulness of the task calculated that that was no longer the case.” Initially, yes, the second option requires us to look for a subtler and more complicated pattern, but it is also far more likely to result in success.

      From there it’s simply a question of trying to predict function based on purpose, that is, what kind of things would a paleolithically evolved rate of effort calculator want to look for?

      • DiscoveredJoys

        I’ll happily agree with the thrust of your comments, but one quibble I’ll raise (which may only be use of language rather than concept) is the phrase:

        ” …using as much outside information as possible.”

        If you model the brain as a data processing device like a computer, this makes sense. But I rather expect the brain is an interpreting and associative device (no handy metaphor or analogy) evolved to refuse and ignore as much unimportant data as possible. This leaves the potentially salient and potentially meaningful information available for interpreting into appropriate behaviour.

        I also expect that our conscious thoughts are merely a feedback process which sits far down the chain of processing – but that is straying beyond the scope of the current article.

      • Ian

        The more common view I’ve come across is not that the brain is like a computer, but that it is a computer.

        Honestly, I might as well start a Rob Kurzban fan club, because I really do think that his book is one of the best explanations I’ve come across of how the brain works. That is, he uses a modular model to explain the various features of our cognitive processing. So yes, your brain does exist to throw out data, but that process by which that takes place is one in which specific algorithms that do their work unconsciously. One good metaphor I’ve come across is comparing it to digestive enzymes, each of which breaks down food a little bit more into more useful chunks. Certainly, computer scientist friends have assured me that within current comp.sci. theory this currently the only way to program computers to do similar tasks.

        Now, having said that, most of this takes place unconsciously, so unless you’re doing clever experiments you probably won’t notice that it’s there. In contrast, the parts of our thinking that we *are* consciously aware of are indeed largely associative and interpretive. But what they are associating and interpreting is the outputs of our lower level unconscious modules. If you go back to the digestion enzyme model, if our unconscious is our gut, then our conscious is the processes by which we build something useful from those constituents. Whether that’s an accurate metaphor or not I’m not sure, but it certainly is pretty.

        As for conscious thoughts, yes, they probably exist in part as a feedback process (see Jonah Lehrer’s How We Decide for a rather superficial but well-written explanation of this). However another large function of conscious thought is communication, which leads to people having all sorts of erroneous but politically useful beliefs about how great they are. Many bullies don’t realise they are bullies, at least not consciously; they just think losers are annoying. This is a great strategy because it gives you plausible deniability for your social aggression. But that’s a whole different topic.

        • DiscoveredJoys

          I reckon we’re on the same page here. I’d recommend Timothy D Wilson’s “Strangers to Ourselves” book for a very accessible explanation of just how much of our thinking is handled by the ‘Adaptive Unconscious’. In many ways a companion book…

          • Ian

            I’ll look that up, thanks for the tip

  • Alex

    These resource analogies are very intuitive as you point out. For example, many people thought that the sun was like a big furnace that burned resources to create light and heat. It was known at the time that it would be impossible for the sun, given it’s size, to burn on a resource for that long. However, people tried to invent new resources that could lead to a longer burn .

    Then scientists crafted this nifty idea called nuclear fission that actually explains how the sun and other starts keep burning. It’s cool stuff when we give up on our intuitive notions and actually do the hard work of figuring out how stuff works.

    • Alex

      *fusion… not fission.

    • Ian


      • http://vertebratesocialbehavior.blogspot.com Clara B. Jones

        Unless I am mistaken, though you are accurate to say that fusion is a neat idea, attempts to demonstrate empirically that it is, in fact, consistent with the laws of physics, have failed. This is one example of why it is necessary to empirically test the conclusions of math models.

        • Ian

          Sorry, unless I am mistaken Clara, are you arguing that the nuclear fusion model of stellar evolution is incorrect?

          I took Alex’s comment to say that scientists first thought the sun was powered by a chemical reaction, then realized that the sun would run out of fuel, and subsequently revised this with the nuclear fusion model. Thus providing an example of an obvious, intuitive model (similar to the glucose willpower model) that required revision and thus became a less intuitive but accurate model (similar to the EP willpower model).

          • http://vertebratesocialbehavior.blogspot.com Clara B. Jones

            Thank you for the correction. I was talking about “cold fusion”.

  • http://twitter.com/#!/robsica Rob

    What about the recent research by Heatherton, mentioned in the NY Times piece?

    Heatherton reported that administering glucose completely reversed the brain changes wrought by depletion — a finding, he said, that thoroughly surprised him. Heatherton’s results did much more than provide additional confirmation that glucose is a vital part of willpower; they helped solve the puzzle over how glucose could work without global changes in the brain’s total energy use. Apparently ego depletion causes activity to rise in some parts of the brain and to decline in others. Your brain does not stop working when glucose is low. It stops doing some things and starts doing others. It responds more strongly to immediate rewards and pays less attention to long-term prospects. (Perhaps appearing here?)

    • Robert Kurzban

      I’m not familiar with the study discussed in the NYT article. (I don’t think it’s the TiCS piece that you linked to, is it?) However, I want to distinguish two claims. The first is that “self control” tasks drain glucose (on the margin). I dispute this. The second (which I don’t dispute) is that glucose levels affect decision making. I would expect that, as the Times articles says, reward centers in hungry people would respond more strongly to food and probably other immediate rewards. I would also expect hungry people to discount the future more. Can anyone confirm which article the NYT article is referring to?

      • http://twitter.com/#!/robsica Rob

        The study I parenthically linked to does not appear to be the one.

    • Ian

      You know, I think they describing something more interesting than just glucose depletion or hunger in that NYT article. In most of the examples given, people started to become much more conservative in their decision making after they had been asked to make many decisions, one after the other.

      It seems to me that this would be far more effectively explained not in physiological terms, but in terms of a protective mechanism. If you’re forced to make 50 decisions in a row, at some point the anxiety of increasing uncertainty in the accuracy of your decision-making model should creep up. After all, that’s a lot of information to synthesize. All one needs to do then is posit that we have some sort of uncertainty anxiety, and that after a while it becomes advantageous to become far more conservative in our decision-making, or just disengage from the environment that is forcing us to make these decisions (ie become fatigued).

      Thus, judges aren’t denying parole towards the end of the day because they’re hungry, they’re denying parole because their uncertainty anxiety is becoming too high. Furthermore, the switch that they describe when fatigued away from a conscious decision-making style towards an instinctive, unconsciously decided decision-making style is unlikely to occur because it’s less glucose-intensive. Far more likely that it occurs because our unconscious modules are much better at dealing with partial information and uncertainty.

    • Ian

      Oh, and obviously I’m not trying to argue that the anxiety is the cause of this effect. Rather, I’m saying that that an uncertain, difficult to model environment is the cause. Anxiety is the emotion that then organizes the brain into a state which helps it produce a coordinated response to this environment.

      Thus, if an environment is changing quickly, or demanding frequent decisions be made, the result is a more instinctual cognitive style (so as to use parts of the brain better equipped to dealing with uncertainty), and fatigue in order to cause the individual to withdraw to a more predictable environment.

    • Ian

      Also, that’s not to say that a safe, controllable environment is always better, only that it is less fatiguing. See this article for very interesting examples, albeit somewhat suspect interpretation on the journalist’s part: http://www.wired.com/wiredscience/2011/08/the-creativity-of-anger/

  • Jesse Marczyk

    Perhaps I don’t know just because I’m not up on the relevant literature, but is there some rationale given as to why not doing something increases glucose demands in the brain?

    • Ian

      You probably know more about this than I do, but physiology is complicated. For all we know, the neurons are too busy during periods of high demand to process glucose, and then their uptake subsequently increases when they have time to process it, ie during slow periods. I mean, I don’t get hungry during hard exercise, yet that’s when my energy demands are the highest.

    • http://vertebratesocialbehavior.blogspot.com Clara B. Jones

      Unless I am mistaken, the relevant idea is that the metabolic demands of neural processing places energetic constraints on brains, particularly where metabolism increases above some baseline and/or homeostatic level(s). My review of this literature for my 2005 book and recent reading of a few papers on this topic discuss these varying constraints as “costs”, again, when they stress neural systems beyond their standard range of flexibility.

      • Jesse Marczyk

        I don’t see where the “increase relative to baseline comes in” part comes in in the study though. The self-control task in the study Rob mentions (don’t look at the words) is supposed to use up more glucose than the control task (look at the words).

        With the control task, one needs to process the words and their meanings – as well as whatever associated tasks come with it – in addition to watching the movie. I don’t get how not reading those words is supposed to be more costly, much less substantially so, in terms of metabolic activity, if reading them is considered the baseline.

        • Alex

          My guess for the rationale is that your attention would be naturally drawn to the words so you are using an inhibitory process in order not to have your attention captured by the words. This inhibitory process of trying to avoid your attention being captured is what is supposed to drain your brain of glucose.

          • Jesse Marczyk

            That was kind of the angle I was figured they’d try take on it. I just don’t know if inhibition, in the form of not looking down for five whole minutes, would actually put a greater metabolic strain – presumably a substantially greater one – on the system

  • http://scienceblogs.com/thepumphandle/ Liz

    Fascinating! These studies have been mentioned a lot recently, so thanks for taking the time to go into them in more detail.

    I’m wondering if there might be a distinction between how much energy the brain actually uses during difficult tasks and how much it signals that it need. Could a very slight uptick in metabolism spark a “need more glucose” signal, even if no glucose shortage is imminent?

    • Ian

      Yeah, I just noticed that myself. Perhaps blood glucose levels are not a good signal for cellular energy reserves (see my second comment). But the fact remains that it’s unlikely our behaviour is modulated too heavily by physiological inputs, and far more likely that it is modulated based on environmental inputs. Simply put, if you keep being lazy in the face of opportunity, somebody else is going to eat your lunch. (see my first comment)

  • http://vertebratesocialbehavior.blogspot.com Clara B. Jones

    Your focus on “willpower” isn’t clear to me; however, there is a very large and large-standing literature demonstating the high metabolic costs of brain processing and neural processing in general–in organisms from Drosophila to primates, including humans. Recent studies support conclusions of earlier ones that large brains (brain to body size ratio), within and between taxa are expecially costly to “run”. I attach the paper below only because it refers to highly-regarded work by primatologists (Van Schaik’s group) and clearly explicates methodology. I think that other papers (e.g., Sobrero et al., 2011) are, on whole, more precise about the relevant arguments and definitions.

    • Ian

      Forest for the trees. I don’t think anybody disputes that there are high fixed costs to maintaining a large and complex brain, however what Kurzban is trying to point out is that these costs are fixed. The variable cost of increased brain activity is extremely small, and is unlikely to have any effect on willpower.

      And yes, willpower is a high-level concept which is difficult to pin down in terms of lower-level concepts. I’m not going to try. But like the judge who stated that pornography can’t be defined, but I know it when I see it, I think we all know what is being discussed here. I’ll leave the specific definitions for someone smarter than me.

      If these are the trees, what is the forest? Put it this way – if our instinctual response to low blood glucose is to sit down and stop working, our ancestors may have found themselves with a problem. You try losing motivation every time you get hungry, and I’ll try using brain modules to calculate when it is most appropriate to use my limited energy (thus often working *harder* when I have low blood glucose), and we’ll see who survives longest in a paleolithic society.

  • Pingback: Diabetes and Aggression | The Trait-State Continuum

  • Pingback: The politics and psychology of self-help | Bjørn Stærk

  • Pingback: h+ Magazine | Covering technological, scientific, and cultural trends that are changing human beings in fundamental ways.

Copyright 2011 Robert Kurzban, all rights reserved.

Opinions expressed in this blog do not reflect the opinions of the editorial staff of the journal.