Food for Thought? Trust your unconscious when energy is low
Maarten W. Bos
Rick van Baaren
Radboud University of Nijmegen
Harvard Business School
Number of words main text (including references): 2482 (3803)
Number of words abstract: 125
Running head: Food for Thought?
Address for correspondence: Maarten Bos, Harvard Business School, Baker Library 467,
Boston, MA 02163, USA, firstname.lastname@example.org.
Recent studies showed that a period of unconscious thought can help when
making complex decisions. Under some circumstances, unconscious thought improves
decisions even more than conscious thought. Executive functioning depends on energy
provided by glucose, and we know from previous research that the performance of
various conscious processes deteriorates when energy is low. Unconscious processes
require less energy and may operate unhampered when energy is low. Therefore we
propose that whereas low blood glucose levels impair conscious thought, this is not the
same for unconscious thought. An experiment, where we manipulated blood glucose
levels, indicated that indeed, when making decisions, the unconscious can best be trusted
when blood glucose levels are low, whereas conscious deliberation yields the best results
when blood glucose levels are elevated.
Common knowledge dictates that when we have a complex decision to make, we
should consciously weight all relevant information and think hard. However, research
shows that this may be ill-advised, as conscious thought is – at least under some
circumstances - sensitive to various pitfalls and biases (Bos & Dijkterhuis, 2011; Bos &
Dijksterhuis, in press; Dijksterhuis & Nordgren, 2006; Wilson & Schooler, 1991;
Dijksterhuis, Van Baaren, Bongers, Bos, Van Leeuwen, & Van der Leij, 2010). So what
is the best way to come to optimal decisions? A growing body of literature says that
under some circumstances, a period of unconscious thought can be beneficial for decision
Unconscious thought (UT) can be defined as a deliberative cognitive process in
the absence of conscious attention (Dijksterhuis & Nordgren, 2006). In a typical
experiment, participants were given information about a decision problem. Some
participants were asked to consciously think about the information, whereas others were
given a distracting task and engaged in unconscious thought. The common finding is that
unconscious thought leads to good decisions (Bos, Dijksterhuis, & Van Baaren, 2011),
and sometimes even better decisions than conscious thought (Bos, Dijksterhuis, & Van
Baaren, 2008; Dijksterhuis, 2004; Dijksterhuis, Bos, Nordgren, & Van Baaren, 2006a;
Dijksterhuis, Bos, Van der Leij, & Van Baaren, 2009; Ham, Van den Bos, & Van Doorn,
2009; Lerouge, 2009).
The UT effect is controversial (e.g., Gonzalez-Vallejo, Lassiter, Bellezza, &
Lindberg, 2008; Payne, Samper, Bettman, & Luce, 2008) and the controversy is fed by
the fact that the UT effect is far from robust. There are also a growing number of
researchers who reported failures to replicate the UT effect (Acker, 2008; Calvillo, &
Penaloza 2009; Newell, Wong, Cheung, & Rakow, 2009; Thorsteinson, & Withrow,
2009; Waroquier, Marchiori, Klein, & Cleeremans, 2009, 2010). However, given that the
number of successful replications is also increasing rapidly (for a meta-analysis see
Strick, Dijksterhuis, Bos, Sjoerdsma, Van Baaren, & Nordgren, 2011), investigating
circumstances under which unconscious thought is or is not beneficial can be helpful.
Various people have asked the question to what extent the UT effect is
generalizable to other paradigms or situations (see also Bekker, 2006; Payne et al., 2008;
Shanks, 2006, but see Dijksterhuis, Bos, Nordgren, & Van Baaren, 2006b; 2006c; Krans
& Bos, in press). By now, the UT effect has been obtained for many different choices,
including hypothetical houses, cars, and computers, posters that people could take home
(whereby quality of choice was operationalized as post-choice satisfaction), roommates,
and even for the predictions of sport games (Dijksterhuis, et al., 2009). In addition, the
UT effect extends to situations whereby people make moral or justice judgments (Ham et
al., 2009; Ham & Van den Bos, 2010), but it seems to not generalize to decisions
whereby people have to rely on numerical information, such as used in many games
(Payne et al, 2008). It seems a logical step to find more theoretical underpinnings of
when either conscious or unconscious thought would be more beneficial, depending on
the situation. Why do some decisions benefit from thinking consciously and do other
decisions benefit from thinking unconsciously?
Decisions based on conscious thought depend on declarative memory
(Dijksterhuis & Nordgren, 2006; Schooler, Ohlsson, & Brooks, 1993), so circumstances
under which declarative memory is facilitated may prove beneficial for conscious
thought. Administering glucose improves verbal memory in difficult tasks (Sünram-Lea,
Foster, Durlach, & Perez, 2002a; Meikle, Riby, & Stollery, 2005) and supports episodic
memory (Riby, McMurtrie, Smallwood, Ballantyne, Meikle, & Smith, 2006). Could a
low blood sugar level negatively impact conscious thought, but leave unconscious
thought unhampered? Let us first turn to the influence of blood glucose level on cognitive
The brain relies on glucose for almost all of its operations (Scholey, Harper, &
Kennedy, 2001; Siesjo, 1978; Taylor & Rachman, 1988). Cerebral stores of glucose and
glycogen run out in a matter of minutes (Marks & Rose, 1981). This makes the brain, in
contrast to other organs and tissues, dependent upon a continuous supply of glucose. So
glucose furnishes energy for the brain and indeed, various conscious processes deteriorate
when energy is low (Donohoe & Benton, 1999). For instance, low blood glucose levels
impair memory, attention and problem solving abilities (e.g. Benton, Owens, & Parker,
1994; Benton & Sargent, 1992; Donohoe & Benton, 1999; Kennedy & Scholey, 2000;
Morris & Sarll, 2001; Owens & Benton, 1994; Pollitt, Jacoby, & Cueto, 1996; Pollitt,
Leibel, & Greenfield, 1981; Pollitt, Lewis, Garza, & Shulman, 1983). Participants who
were asked to remember words from a list show an increase in memory performance
when blood glucose levels were elevated (Benton et al., 1994) and problem solving
abilities were increased in students after eating a snack containing sugar (Kanarek &
Not only do blood glucose levels affect performance, blood glucose levels also
decrease as a result of cognitively demanding tasks (e.g. Fairclough & Houston, 2004).
Brain glucose and glycogen levels run low after a demanding task (Gibson & Green,
2002) and it follows naturally that performing a demanding task negatively influences
performance on a subsequent demanding task.
Coming back to decisions: Given that various brain processes are dependent on
glucose, it seems likely that consciously thinking about decisions also suffers when blood
glucose is low. As said, decisions based on conscious thought are dependent on
declarative memory, and declarative memory is dependent on glucose (Sünram-Lea,
Foster, Durlach, & Perez, 2001, 2002b). So, how are we to make good decisions when
glucose levels are low?
Decisions based on unconscious thought are probably less dependent on
declarative memory (Dijksterhuis & Nordgren, 2006), so unconscious thought may suffer
less from low blood glucose. With an experiment, we examined the effects of
manipulating blood sugar levels on decisions made after conscious and unconscious
thought. Fasted participants were given either a drink containing sugar or a drink without
sugar, after which they were presented with a decision problem. Some participants were
asked to make a decision after a period of conscious thought, whereas other were given a
distracting task and engaged in unconscious thought. With this setup, we tried to answer
the question: Under what blood glucose conditions should we trust conscious deliberation
to provide the best results and when can we trust unconscious thought?
Participants and Design
Of the 156 Dutch undergraduate students of the University of Amsterdam with an
average age of 21.18 (SD = 3.98) 35 were male. They were randomly assigned to the
cells of a 2 (Mode of Processing: Conscious versus Unconscious) x 2 (Blood sugar level:
Low versus Increased) between-participants design. They received course credits or 7
Euros for their participation.
Procedure and Materials
Participants were asked not to eat or drink anything (other than water) three hours
prior to coming into the lab. Self-reports indicated that participants had indeed followed
this instruction. The entire study lasted for a total of 30 minutes, and participants were
tested somewhere between 9AM and 4PM. Blood glucose levels naturally change during
the day and with circadian rhythm, but conditions were randomly divided over the day, to
eliminate circadian rhythm or blood glucose variability confounds. Upon coming into the
lab, half the participants were given a drink containing sugar (250 ml of 7-up ©,
containing roughly 28 grams of sugar), whereas the other half were given a drink without
sugar (250 ml of 7-up Free ©, sweetened with aspartame). Both drinks were pre-tested
and tested in our experiment and participants in the pre-test as well as in the experiment
reported no difference in taste (sweetness) or liking of the drinks (all Fs<1). It takes about
12-15 minutes for sugar in a liquid to have an effect on the brain (there is a range from
10-12 minutes, as described by Masicampo & Baumeister in 2008 to 20 minutes, as
described by Benton in 1990), so after imbibing the drink, participants were presented
with a movie about sea life for 9 minutes and some demographic questions before they
read information about four hypothetical choice options.
Some participants read information about four hypothetical cars (n=54); some
about four hypothetical jobs (n=102). The car materials had been used previously
(Dijksterhuis et al., 2006a) and the job materials were constructed in the same way the
car materials had been constructed. Each choice option (car or job) was described by
twelve attributes. The attributes were either positive or negative. The optimal choice
option was described by 75% positive attributes, two other choice options were described
by 50% positive attributes, and the last choice option by 25% positive attributes (see
Online supporting materials of Dijksterhuis et al., 2006a, appendix 1 for insight in the
construction of the materials). The attributes were presented one by one in random order
on a computer screen. Each attribute was presented for 4 seconds. After reading the
information, participants were either assigned to a conscious thought condition, or to an
unconscious thought condition. The timing was such that at the time participants were
assigned to either a conscious or unconscious thought condition, the glucose should have
its effect. In the conscious thought condition, participants were presented with a computer
screen with instructions asking them to think about the information carefully for four
minutes, so they could later make a decision. After that they rated the choice options. In
the unconscious thought condition, participants were distracted for four minutes (they
were given another, unrelated movie clip about sea life). Before watching the movie they
were told that after the period of distraction they would be asked their opinion of the
choice options. Participants in the unconscious thought condition were asked questions
about the second movie clip, which they received to distract them, to assess whether they
had paid attention, and were therefore sufficiently distracted. In general, participants in
both drink conditions performed equally well (F<1), scoring an average of 1.33
(SD=1.53) out of 4 in-depth questions about the movie.
Participants gave their attitude towards each of the four individual choice options.
The attitudes towards the individual choice options were measured by having participants
click a score between 1 (“very negative”) to 20 (“very positive”). The attitudes towards
the four choice options were measured in random order. A higher score on the attitude
difference measure indicates a greater ability to differentiate between choice options.
Participants were asked some questions about the drinks (taste, sweetness), after which a
subset of the participants received a mood questionnaire (n = 54). Further analyses
revealed no effect of the drinks on experienced mood (F<1). Mood was measured after
all other dependent variables, by using a subset of the PANAS mood questionnaire items
(Watsen, Clark, & Tellegen, 1988).
We calculated a difference score by subtracting participants’ rating (scored 1-20)
of the worst car from participants’ rating (scored 1-20) of the best car (as used by
Dijksterhuis et al., 2006a; Bos et al., 2008). This resulted in a scale ranging from -19 to
19. A higher score indicates a greater ability to correctly differentiate between the best
option and the worst option. We compared the difference scores between conditions
using a 2 (Mode of processing: Conscious versus Unconscious) x 2 (Blood sugar level:
Low versus Increased) analysis of variance.
We found no main effect of either mode of processing condition (F<1) or blood
sugar level condition (F<1). The two-way interaction is significant however, F(1,152) =
10.19, p < .01, R
= .07. As depicted in Figure 1, conscious deliberation improved (from
M=3.439, SE=.741 to M=5,220, SE=.741) when blood sugar level was increased, F(1, 82)
= 4.05, p < .05, η
= .05. However, the reverse is true for unconscious processing F(1, 74)
= 5.90, p < .05, η
= .05, (from M= 5.444, SE=.790 for low to M=2.368, SE=.769 for
increased blood sugar). Performance after unconscious processing actually declined when
blood glucose level was increased. A possible explanation is that a change of blood
glucose level disrupts unconscious processes. Such a disruptive influence of blood
glucose changes on unconscious, automatic processes has indeed been found before
(Holmes, Hayford, Gonzalez, & Weydert, 1983), although this was shown in diabetes
patients, who have a somewhat different glucose metabolism and tolerance for low and
high blood glucose.
In line with our hypotheses, we found that when we have enough energy,
conscious deliberation enables us to make good decisions. The unconscious on the other
hand seems to operate fine with low energy. We suggest the reason why conscious
deliberation is aided by an increase in blood sugar level is through declarative memory,
which improves with increases in blood sugar level. We did not directly measure
declarative memory however, the actual process behind our findings remain therefore
There are several other shortcomings in the study, that need to be addressed in
future. One important drawback of the method used, is that participants did not make a
choice in this task, but instead only rated the choice options. Although this gives an
indication of the extent to which participants are able to differentiate between choice
options, it could be that participants with a low score on this measure would still make
good decisions. Another shortcoming is that the used task is artificial (as argued by
Newell et al., 2009). This goes for the distraction task less so, as watching a movie is
more common than solving anagrams or doing an n-back task, which has been used in
various other unconscious thought tasks. However, it could well be that watching a movie
does not fully capture conscious attention. Therefore, conscious processing of the
information in the ‘unconscious processing’ condition can not be ruled out completely.
There are conceptual concerns with regard to the influence of blood sugar level on
performance measures in general. There is debate about the ability of blood glucose to
have a direct effect on performance. That is, the amount of glucose necessary to perform
mental operations is minimal. However, arguments have been made that there is a role for
peripheral detection (outside of the brain) of the glucose supply in the body. Several
receptors throughout the body feed information to the brain about the amount of glucose
in the body (Langhans 2000). In this way, the ingestion of glucose could change mental
function indirectly, by giving off signals that the body is ‘safe’ in terms of glucose stores.
Clearly, neither solely conscious nor solely unconscious thought is always
beneficial. Previous papers have argued that the complexity of a decision is one of the
determinants for which mode of thought is beneficial (Dijktershuis et al., 2006a). Another
recent paper has argued that, under certain circumstances, the combination of both
conscious and unconscious thought leads to optimal decisions (Norgren, Bos, &
Dijksterhuis 2010). This paper indicates another potential boundary for when to use
either conscious or unconscious thought. Our unconscious does not seem to require the
same energy needed for conscious processes. Our data show that when we are low on
energy we can employ another decision strategy than thinking consciously: We can trust
Acker, F. (2008). New findings on unconscious versus conscious thought in
decision making: Additional empirical data and meta-analysis. Judgment and Decision
Making, 3, 292-303.
Bekker, H. L. (2006). Making choices without deliberating. Science, 312, 1472.
Benton, D. (1990). The impact of increasing blood glucose on psychological
functioning. Biological Psychology, 30, 13–19.
Benton, D., Owens, D.S., & Parker, P.Y. (1994). Blood glucose influences
memory and attention in young adults. Neuropsychologia, 32, 595-607.
Benton, D., & Sargent, J. (1992). Breakfast, blood glucose and memory.
Biological Psychology, 33, 207–210.
Bos, M.W. & Dijksterhuis, A. (in press). Self-knowledge, Unconscious thought
and decision making. In S. Vazire & T. Wilson (Eds.). Handbook of Self-knowledge.
Bos, M.W. & Dijksterhuis, A. (2011). Unconscious thought works bottom-up and
conscious thought works top-down when forming an impression. Social Cognition 29(6),
Bos, M.W., Dijksterhuis, A., & Van Baaren, R.B. (2008). On the goal-
dependency of unconscious thought. Journal of Experimental Social Psychology, 44,
Bos, M.W., Dijksterhuis, A., & Van Baaren, R.B. (2011). The benefits of
“sleeping on things”: Unconscious thought leads to automatic weighting. Journal of
consumer psychology, 21, pp. 4-8.
Calvillo, D.P., & Penaloza, A. (2009). Are complex decisions better left to the
unconscious? Further failed replications of the deliberation-without-attention effect.
Judgment and Decision Making, 4, 509–517.
Dijksterhuis, A. (2004). Think different: The merits of unconscious thought in
preference development and decision making. Journal of Personality and Social
Psychology, 87, 586–598.
Dijksterhuis, A., Bos, M.W., Nordgren, L.F., & Van Baaren, R.B. (2006a). On
making the right choice: The deliberation-without-attention effect. Science, 311(5763),
Dijksterhuis, A., Bos, M.W., Nordgren, L.F., & Van Baaren, R.B. (2006b).
Making choices without deliberating. Science, 312(5779), 1472.
Dijksterhuis, A., Bos, M.W., Nordgren, L.F., & Van Baaren, R.B. (2006c).
Complex choices better made unconsciously? Science, 313(5788), 760–761.
Dijksterhuis, A., Bos, M.W., Van der Leij, A., & Van Baaren, R.B. (2009).
Predicting soccer matches after unconscious and conscious thought as a function of
expertise. Psychological Science, 20(11), 1381-1387.
Dijksterhuis, A., & Nordgren, L.F. (2006). A theory of Unconscious Thought.
Perspectives on Psychological Science, 1, 95–109.
Dijksterhuis, A., Van Baaren, R.B., Bongers, K.C.A., Bos, M.W., Van Leeuwen,
M.L., & Van der Leij, A. (2010). The rational unconscious: Conscious versus
unconscious thought in complex consumer choice. In M. Wanke (Ed.,). The social
psychology of consumer behavior, 89-108. New York: Psychology Press.
Donohoe, R.T., & Benton, D. (1999). Cognitive functioning is susceptible to the
level of blood glucose. Psychopharmacology, 145, 378-385.
Fairclough, S.H., & Houston, K. (2004). A metabolic measure of mental effort.
Biological Psychology, 66, 177-190.
Gibson, E.L., & Green, M.W. (2002). Nutritional influences on cognitive
function: mechanisms of susceptibility. Nutrition Research Reviews, 15, 169–206.
Gonzalez-Vallejo, C., Lassiter, G.D., Belezza, F.S., & Lindberg, M. (2008). “Save
angels perhaps”: A critical examination of unconscious thought theory and the
deliberation-without-attention effect. Review of General Psychology, 12(3), 282–296.
Ham, J., & Van den Bos, K. (2010). The merits of unconscious processing of
directly and indirectly obtained information about social justice. Social Cognition, 28(2),
Ham, J., Van den Bos, K., & Van Doorn, E. (2009). Lady Justice thinks
unconsciously: Unconscious thought can lead to more accurate justice judgments. Social
Cognition 27 (4), 509-521.
Holmes, C., Hayford, J.T., Gonzalez, J.L., & Weydert, J.A. (1983). A survey of
cognitive functioning at different glucose levels in diabetic persons. Diabetes Care, 6(2),
Kanarek, R.B. & Swinney, D. (1990). Effects of food snacks on cognitive
performance in male college students. Appetite, 14, 15–27.
Kennedy, D.O., & Scholey, A.B. (2000). Glucose administration, heart rate and
cognitive performance: Effects of increasing mental effort. Psychopharmacology, 149,
Krans, J., & Bos, M. W. (in press). To think or not to think about trauma? An
experimental investigation into unconscious thought and intrusion development." Journal
of Experimental Psychopathology.
Langhans, W. (2000). Portal-hepatic sensors for glucose, amino acids, fatty acids,
and availability of oxidative products. In: Neural and Metabolic Control of
Macronutrient Intake, (HR Berthoud & RJ Seeley eds), pp. 309–24. CRC Press: Boca
Lerouge, D. (2009). Evaluating the benefits of distraction on product evaluations:
The mindset effect. Journal of Consumer Research, 36(3), 367-379.
Newell, B.R., Wong, K.Y., Cheung, J.C.H., & Rakow, T. (2009). Think, blink or
sleep on it? The impact of modes of thought on complex decision making. Quarterly
Journal of Experimental Psychology, 62, 707–732.
Marks, V., & Rose, F.G. (1981). Hypoglycaemia. Blackwell Scientific
Masicampo, E.J., & Baumeister, R.F. (2008). Toward a physiology of dual-
process reasoning and judgment: Lemonade, willpower, and expensive rule-based
analysis. Psychological Science, 19, 255–260.
Meikle, A., Riby, L.M., & Stollery, B. (2005). Memory processing and the
glucose facilitation effect: the effects of stimulus difficulty and memory load. Nutritional
Neuroscience, 8, 227–32.
Morris, N., & Sarll, P. (2001). Drinking glucose improves listening span in
students who miss breakfast. Educational Research, 43, 201–207.
Nordgren, L.F., Bos, M.W., & Dijksterhuis, A. (2010). The best of both worlds:
Integrating conscious and unconscious thought best solves complex decisions. Journal of
Experimental Social Psychology, 47(2), 509–511.
Owens, D.S., & Benton, D. (1994). The impact of raising blood glucose on
reaction times. Neuropsychobiology, 30, 106-113.
Payne, J.W., Samper, A., Bettman, J.R., & Luce, M.F. (2008). Boundary
condition on unconscious thought in complex decision making. Psychological Science,
Pollitt, E., Jacoby, E., & Cueto, S. (1996). School breakfast and cognition among
nutritionally at-risk children in the Peruvian Andes. Nutrition Reviews, 54, S22–S26.
Pollitt. E., Leibel, R.L., & Greenfield, D. (1981). Brief fasting, stress, and
cognition in children. American Journal of Clinical Nutrition, 34, 1526–1533.
Pollitt. E., Lewis, N.L., Garza, C., & Shulman, R.J. (1983). Fasting and cognitive
function. Journal of Psychiatric Research, 17, 169–174.
Riby, L.M., McMurtrie, H., Smallwood, J., Ballantyne, C., Meikle, A., & Smith,
E. (2006). The facilitative effects of glucose ingestion on memory retrieval in younger
and older adults: is task difficulty or task domain critical? British Journal of Nutrition,
Scholey, A.B., Harper, S., & Kennedy, D.O. (2001). Cognitive demand and blood
glucose. Physiology and Behavior, 73, 585–92.
Schooler, J. W., Ohlsson, S., & Brooks, K. (1993). Thoughts beyond words:
When language overshadows insight. Journal of Experimental Psychology: General, 122,
Shanks, D. R. (2006). Making choices without deliberating. Science, 313, 760.
Siesjo, B. K. (1978). Brain energy metabolism. Chichester, UK: Wiley.
Strick, M., Dijksterhuis, A., Bos, M. W., Sjoerdsma, A., Van Baaren, R. B., &
Nordgren, L. F. (2011). A meta-analysis on unconscious thought effects. Social
Cognition, 29(6), 738–762.
Sünram-Lea, S.I., Foster, J.K., Durlach, P., Perez, C. (2001). Examination of the
relation of fast-duration, time of day and pre-consumption baseline plasma glucose levels
on the glucose facilitation of cognitive performance effect. Psychopharmacology 157(1),
Sünram-Lea, S.I., Foster, J.K., Durlach, P., Perez, C. (2002a). Investigation into
the significance of task difficulty and divided allocation of resources on the glucose
memory facilitation effect. Psychopharmacology, 160, 387–397.
Sünram-Lea, S.I., Foster, J.K., Durlach, P., Perez, C. (2002b). The effect of
retrograde and anterograde glucose administration on memory performance in healthy
young adults. Behavioural Brain Research, 134, 505–516.
Taylor, L.A., & Rachman. S.J. (1988). The effects of blood sugar level changes
on cognitive function, affective state, and somatic symptoms. Journal of Behavioral
Medicine, 11, 279-291.
Thorsteinson, T.J., & Withrow, S. (2009). Does unconscious thought outperform
conscious thought on complex decisions? A further examination. Judgment and Decision
Making, 4, 235–247.
Waroquier, L., Marchiori, D., Klein, O., & Cleeremans, A. (2009).
Methodological pitfalls of the unconscious thought paradigm. Judgment and Decision
Making, 4, 601–610.
Waroquier, L., Marchiori, D., Klein, O., & Cleeremans, A. (2010). Is it better to
think unconsciously or to trust your first impression? A reassessment of unconscious
thought theory. Social Psychological & Personality Science, 1, 111–118.
Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of
brief measures of positive and negative affect: The PANAS scales. Journal of Personality
and Social Psychology, 54, 1063-1070.
Wilson, T.D., & Schooler, J.W. (1991). Thinking too much: Introspection can
reduce the quality of preferences and decisions. Journal of Personality and Social
Psychology, 60, 181–192.
Figure 1: Performance after conscious and unconscious processing of information as a
function of high energy (after sugar intake) and low energy (after no sugar intake).
Higher scores indicate greater ability to distinguish good from bad options.