Obesity and the brain: How convincing is the addiction model?

Abstract

An increasingly influential perspective conceptualizes both obesity and overeating as a food addiction accompanied by corresponding brain changes. Because there are far-reaching implications for clinical practice and social policy if it becomes widely accepted, a critical evaluation of this model is important. We examine the current evidence for the link between addiction and obesity, identifying several fundamental shortcomings in the model, as well as weaknesses and inconsistencies in the empirical support for it from human neuroscientific research.

Full text

Obesity, which has a profound impact on
personal well-being and on the demand for
health care, is at pandemic levels1. Central
to weight gain is the development of an
energy imbalance, a situation that arises as
a result of complex interactions between an
individual’s biology and environmental fac-
tors1,2. Clinicians, researchers and politicians
recognize the importance of understanding
how the brain interacts with an obesogenic
environment and the corresponding poten-
tial for neuroscience to develop our under-
standing of the causes and consequences of
obesity. The messages now emerging from
the neuroscientific research community may
therefore have an unprecedented impact on
policy development.
A fast-growing consensus is that obesity
might be understood within the same
neuro biological framework as addiction and
that research, investigations, treatments
and policy should be shaped accordingly3.
Essentially, the view is that obesity results
from an addiction to food that strongly
resembles addiction to drugs, both behav-
iourally and in terms of underlying neural
processes. This idea is exerting a tremendous
influence on the field of obesity research and
has driven cogent, although unsuccessful,
arguments for the inclusion of obesity or
overeating as a category in the fifth edition
of the Diagnostic and Statistical Manual of
Mental Disorders (DSM-V)4,5. Although
there has been debate about the validity of
arguments for phenotypic similarity between
overeating and addiction — and questions
over whether such a model can generate real-
istic goals for policymakers3 — one area that
has not yet been critically scrutinized is the
human neuroscience work that is often cited
in support of the addiction model and that
provides a pervasive framework for design
and inference in human studies of overeating.
In this Perspective article, we describe
how the addiction model has been applied
to obesity and overeating and critically
review each of the five main lines of research
that are usually invoked to support this
conflation. At the outset, it is important to
acknowledge that the food-addiction litera-
ture has largely adopted the clinical model
of addiction as defined by the DSM-IV.
Although this model has clinical validity, in
the addiction research literature it has been
supplemented, and to an extent superseded,
by powerful neurobiological models that
have decomposed the clinical syndrome in
terms of its core cognitive processes and
their possible neural substrates (BOX 1).
This approach, which is based on a growing
understanding of the neurobiology of
addiction, is welcome and — as we discuss
— may offer new ways of identifying overlap
between obesity and addiction. However,
this article is primarily concerned with the
existing arguments in favour of addiction as
a model for obesity, arguments that draw on
clinical definitions.
Obesity and addiction: two views
The addiction model has been applied
to obesity in a number of ways. Central to
each is the idea that someone can become
a ‘food addict’. What might this mean? Two
broad ideas have been discussed. The first
is that certain foods (those high in fat, salt
and sugar6–8) are akin to addictive sub-
stances insofar as they engage brain sys-
tems and produce behavioural adaptations
comparable to those engendered by drugs
of abuse. This in itself is not surprising,
given that current addiction models sug-
gest that addictive drugs hijack the brain
circuitry subserving the motivation for and
enjoyment of, among other things, food9,10.
What the putatively addictive foods are has
yet to be fully defined. The case has been
made that processed foods — as opposed
to unrefined foods — are addictive because
they have nutrient profiles, such as very
high sugar content or combinations of high
sugar and high fat, that are not found in
naturally occurring foods3,6. However, this
classification (processed versus unrefined
foods) is very broad and imprecise, and it
would ultimately be important to specify in
more detail a particular substance or a level
of nutrient (for example, a fat percentage)
that would distinguish an addictive food
from a non-addictive one. Sugar addic-
tion, for example, has been demonstrated
in animals, but not in humans. Indeed,
the validity of sugar addiction as a con-
cept that could apply to humans has been
criticized11.
A second view is that food addiction is
a behavioural phenotype that is seen in a
subgroup of people with obesity and resem-
bles drug addiction. This view draws on
the parallels between the DSM-IV criteria
for a substance-dependence syndrome and
observed patterns of overeating (TABLE 1).
A quantitative measure of the features of
the syndrome has recently been developed
in the form of the Yale Food Addiction
Scale (YFAS)5,12–14. However, although
there seem to be some similarities between
these two phenotypes, the overlap is only
partial (TABLE1). A related, but narrower,
view asserts that a food-addiction pheno-
type is most apparent in individuals with
binge-eating disorder (BED), which is
OPINION
Obesity and the brain: how
convincing is the addiction model?
Hisham Ziauddeen, I.Sadaf Farooqi and Paul C.Fletcher
Abstract | An increasingly influential perspective conceptualizes both obesity and
overeating as a food addiction accompanied by corresponding brain changes.
Because there are far-reaching implications for clinical practice and social policy if it
becomes widely accepted, a critical evaluation of this model is important. We
examine the current evidence for the link between addiction and obesity, identifying
several fundamental shortcomings in the model, as well as weaknesses and
inconsistencies in the empirical support for it from human neuroscientific research.
PERSPECTIVES
NATURE REVIEWS
|
NEUROSCIENCE VOLUME 13
|
APRIL 2012
|
279
© 2012 Macmillan Publishers Limited. All rights reserved

characterized by recurrent episodes (binges)
of uncontrolled, often rapid consumption of
large amounts of food, usually in isolation,
even in the absence of hunger. This eating
persists despite physical discomfort, and
binges are associated with marked distress
and feelings of guilt and disgust15. Once
again, there is an important caveat: although
BED is associated with obesity16, a substantial
number of people who show binge-eating
behaviour are not obese and most obese
people do not haveBED.
A closer look at the evidence
At a population level, one of the main
drivers of the rise in prevalence of obesity
seems to be increased availability of food,
with a consequent imbalance between
energy intake and expenditure1. A modest
energy imbalance over a sustained period
of time can account for the observed
changes in the body mass index (BMI)
distributions of populations2,17. This sug-
gests that any loss of control of eating,
which is important to the idea of obesity
as addiction, is very subtle in most of
the obese population. Moreover, in con-
sidering unhealthy food choices and
consumption, we cannot ignore social
circumstances. For example, limited fam-
ily budgets direct choice to more obeso-
genic foods18. However, as we note above,
although obesity perse is often linked to
addiction, a more nuanced view suggests
that if food addiction produces obesity, it
is likely to do so only in certain individu-
als with disordered eating behaviours
such as BED15,19,20. Here, we consider both
perspectives.
There are five key pieces of evidence
cited in support of the addiction model:
first, a clinical overlap between obesity
(or, more specifically, BED) and drug
addiction15; second, evidence of shared
vulnerability to both obesity and substance
addiction; third, evidence of tolerance,
withdrawal and compulsive food-seeking
in animal models of overexposure to high-
sugar and/or high-fat diets21; fourth, evi-
dence of lower levels of striatal dopamine
receptors (similar to findings in patients
with drug addiction) in obese humans22; and
fifth, evidence of altered brain responses to
food-related stimuli in obese individuals
compared with non-obese controls in func-
tional imaging studies. Below, we consider
each of these inturn.
Box 1 | The addiction model for drugs of abuse
Influential models of drug dependence have divided the Diagnostic and Statistical Manual of
Mental Disorders IV (DSM-IV) behavioural syndrome into several core processes that are involved
in the transition from drug-taking to drug dependence in the subgroup of individuals who
develop the syndrome. This transition involves a shift from voluntary drug-taking, under ventral
striatal and prefrontal control, to habitual and compulsive drug-seeking, driven predominantly by
the dorsal striatum, with loss of executive control over this behaviour75. Trait impulsivity, which
relates to lower levels of striatal D2 dopamine receptors (D2Rs), has been shown to increase the
vulnerability to this process44,76. Lower levels of striatal D2Rs may indicate a reward-deficiency
state that leads to greater drug-taking in an attempt to achieve the same level of reward. The
transition from initial impulsivity to later compulsivity has been proposed to progress through
a three-stage model of anticipation and/or preoccupation; binge and/or intoxication; and
withdrawal and/or negative effect77. Furthermore, drugs of abuse are also thought to sensitize
the mesolimbic dopaminergic systems, leading to an enhanced salience of, and consequent
motivation towards, drug-related cues as well as to cravings induced by such cues78. Increasing
drug intake leads to neural adaptations in the striatum (further decrease of D2Rs) that promote
compulsive drug-seeking and impaired inhibitory control79, whereas adaptations in the amygdala
counter the negative states of dysphoria and withdrawal related to drug use77. These adaptations
serve to perpetuate the syndrome.
Table 1 | Modelling food addiction on substance dependence
DSM-IV criteria for substance dependence Proposed food-addiction equivalent* Comment
Tolerance: increasing amounts of drug are required to
reach intoxication
Tolerance: increasing amounts of food are
required to reach satiety
Not a convincing equivalent to
drug tolerance because it assumes
an equivalence between satiety
and intoxication. In addition, key
characteristics of binges are eating in
the absence of hunger and to the point
of physical discomfort (beyond satiety)
Withdrawal symptoms on drug discontinuation,
including dysphoria and autonomic symptoms such as
shakes and sweats
Distress and dysphoria during dieting No convincing evidence of a human
withdrawal syndrome for foods
Persistent desire for and unsuccessful attempts to cut
drug use
Persistent desire for food and unsuccessful
attempts to curtail the amount of food eaten
This criterion requires the application
of severity and impairment thresholds
to be meaningful
Larger amounts of drug taken than intended Larger amounts of food eaten than intended This criterion requires the application
of severity and impairment thresholds
to be meaningful
A great deal of time is spent on getting the drug, using
the substance or recovering from it
A great deal of time is spent eating It is difficult to apply this criterion
because of the easy availability of foods
in most developed societies
Important social, occupational or recreational activities
are given up or reduced because of substance abuse
Activities are given up through fear of rejection
because of obesity
A strict equivalence would require
engagement in eating to the exclusion
of other activities
Substance use is continued despite knowledge of having
a persistent or recurrent physical or psychological
problem caused or exacerbated by the drug
Overeating is maintained despite knowledge of
adverse physical and psychological consequences
caused by excessive food consumption
This criterion requires the application
of severity and impairment thresholds
to be meaningful
*Data in second column are taken from REFS 5,14. DSM-IV, Diagnostic and Statistical Manual of Mental Disorders IV.
PERSPECTIVES
280
|
APRIL 2012
|
VOLUME 13 www.nature.com/reviews/neuro
© 2012 Macmillan Publishers Limited. All rights reserved

Clinical overlap. Substance dependence is
defined in the DSM-IV by the presence of
characteristic patterns of behaviour (TABLE1),
and it has been suggested that similar patterns
characterize obesity5,6,14. Although some fea-
tures (persistent desire, unsuccessful attempts
to cut down and continued use despite nega-
tive consequences) translate reasonably well
from substance abuse to overeating6,14, others
do not. Tolerance and withdrawal are not
convincingly observed in the human eating
literature14. Furthermore, food, unlike drugs,
is necessary for survival, is easy to obtain
openly and does not (generally) provoke
social opprobrium. As a result, it is difficult to
apply criteria that relate to efforts expended
in acquiring and consuming: such criteria are
useful in addiction to separate use from abuse
for purposes of a DSM-IV diagnosis, but have
little value with respect tofood.
As shown in TABLE 1, three criteria translate
reasonably well from substance dependence
to overeating. Crucially, drug dependence
can be diagnosed if any three criteria are met.
Extending this to food, an individual who ate
more than intended (loss of control), dieted
frequently and unsuccessfully (persistent
attempts to cut down) and continued eating
despite significant weight gain (continued
use despite negative consequences) would
meet the requisite criteria and be deemed a
food addict. The YFAS has applied severity
and impairment thresholds that must be met
to satisfy the criteria13. Although this cer-
tainly may capture a pattern of eating behav-
iour that is abnormal, we question whether
such an approach is sufficiently rigorous to
constitute good grounds for assuming an
addictive basis for overeating in research
studies and in clinical policy decisions.
If we narrow our application of the con-
cept to individuals with BED15, who clearly
have abnormal eating behaviour and a
high prevalence of obesity16, the argument
becomes more convincing. We can recognize
a behavioural syndrome more convincingly
like that of drug addiction, entailing loss of
control of eating, escalating consumption,
compulsivity, restriction of activities, time
spent in pursuing behaviour, and possibly
consuming to ameliorate dysphoric and
negative effects23. It seems that the face value
of the food-addiction construct is strongest
when it is applied to certain (although not
all) individuals with BED19. Perhaps this
highlights a key limitation of the current and
pervasive DSM-IV-based model relating
over-consumption to addiction. The syn-
drome as it is defined and measured captures
a phenotype that may be too imprecise to
evaluate rigorously (BOX 2).
Shared vulnerabilities. Another observa-
tion linking obesity to drug addiction
comes from family studies indicating
that there may be shared genetic suscep-
tibilities to the two conditions. A family
history of alcoholism is associated with
an increased risk of obesity24, and BED
is associated with increased levels of
substance-use disorder in relatives25. The
possible contribution of specific genetic
variants has been explored26–28. The most
widely studied of these has been the Taq 1A
minor (A1) allele of the dopamine receptor
D2 (DRD2) gene, which has been associ-
ated with alcoholism29; substance-misuse
disorders, including cocaine30, smoking31
and opioid dependence32; and obesity33.
However, many studies, including large
meta-analyses that addressed concerns
about population stratification and sample
size, have failed to replicate these find-
ings34–36. Moreover, this polymorphism is
located 10kilobases downstream of the
DRD2 gene, and convincing evidence of an
effect on the expression or function of the
receptor is lacking, although an association
with lower levels of D2 dopamine receptors
(D2Rs) in the striatum, measured by posi-
tron emission tomography (PET), has been
reported37,38.
Obese individuals with BED have also
been reported to have a higher prevalence
of a gain-of-function allele (A118G) of the
μ-opioid receptor (OPRM1)33 that has been
associated with increased sensitivity to
reward, greater preference for sweet and fatty
foods39 and substance addiction40,41. Indeed,
sensitivity to reward is a personality trait that
has been associated with obesity and drug
addiction. It has been argued that, as
in drug addiction, obese individuals have
lower reward sensitivity (the reward-
deficiency hypothesis42), resulting in a
compensatory overconsumption. However,
the relationship between BMI and reward
sensitivity is not straightforward, and in
some people overeating occurs in the setting
of an apparently enhanced sensitivity to the
hedonic aspect of food43. Reward sensitivity
may be mediated by the OPRM1 and Taq 1A
allele polymorphisms mentionedabove.
Another personality trait, impulsivity —
the tendency to initiate behaviour without
adequate forethought of its consequences
— has been identified as a risk factor for
substance addictions44 (BOX 1). This trait has
shown a modest association with the Taq 1A
polymorphism45,46 and has been shown to
be higher in obese and BED individuals,
correlating with food intake47–49.
Box 2 | Towards a food-addiction model?
We have argued that attempts to develop the food-addiction model by relating obesity to the
current clinical definition of addiction (in the Diagnostic and Statistical Manual of Mental Disorders
IV (DSM-IV)) have been unconvincing. A possible future direction that we feel offers more hope of
identifying a convincing, useful clinical entity is to separate the consideration of a putative
food-addiction model from both obesity and the DSM-IV criteria of substance dependence. This
separation must be made for two reasons. First, food addiction, if it exists, may be a cause, a
co-morbidity or possibly a consequence of obesity. Accordingly, food addiction may prevail in
non-obese and not-yet-obese individuals. Therefore, obesity, particularly when assessed solely
cross-sectionally by body-mass index (BMI), will be an unsatisfactory phenotype for food addiction.
Second, the DSM-IV criteria for substance dependence translate poorly to food-related
behaviours (TABLE 1) and, more importantly, these criteria aggregate core features (such as
maintained use despite negative consequences) with markers of long-term use (such as tolerance)
and severity of impairment (such as time spent in acquiring substance).
Future research into the possibility of food addiction would gain by becoming more focused and
neuroscientifically driven in the following ways:
•By creating a more precise neurobehavioural definition of food addiction in which a core set of
measurable behaviours is clearly defined (inability to control consumption, increased motivation
to consume and persistent consumption despite negative consequences75,80). This would capture
a range of problem-eating behaviours, including, but not restricted to, binge eating.
•By incorporating impulsivity, compulsivity and specific patterns of cognitive response as markers
of vulnerability to and endophenotypes of the addiction81.
•By applying current models of addiction that are based on recent empirical neuroscientific work.
For example, demonstrating a transition from goal-directed food-seeking under voluntary
control to compulsive habitual seeking and consumption driven by environmental cues75.
•By relating more precise behavioural and cognitive phenotypes, rather than BMI, to
neuroimaging findings and outcomes.
With these principles in mind, we believe that future work on food addiction could obviate the
problems that have so far led to an inconsistent and contradictory literature.
PERSPECTIVES
NATURE REVIEWS
|
NEUROSCIENCE VOLUME 13
|
APRIL 2012
|
281
© 2012 Macmillan Publishers Limited. All rights reserved

It is possible, therefore, that there are
some shared vulnerabilities between drug
addiction and obesity. However, this does not
in itself strongly support an argument that
the same processes occur in each condition.
Evidence from animal models. By far the
strongest evidence for a food-addiction syn-
drome comes from animal models50. Using
highly palatable foods and highly structured
intermittent-access regimes, it has been
possible to induce an addiction-like pheno-
type in rats. Rats with intermittent access
to high-sugar and high-fat foods develop
escalating, binge-like eating behaviours21,51,
a phenomenon that seems to be related to
the palatability of the foods rather than their
macronutrient composition52. However, this
escalation of sugar and fat intake is offset
by decreases in intake of their normal food
supply, so although these animals become
‘addicted’, they do not become obese53.
Adifferent picture is seen when fat and
sugar are combined (as in ‘cafeteria’ diets,
in which animals are fed on foods such as
bacon, cheesecake and chocolate), where-
upon increased consumption and weight
gain occur in the context of eating that
appears more compulsive54.
In the case of sugar ‘addiction’, enforced
abstinence is associated with enhanced
motivation towards food55. Moreover, a
withdrawal syndrome, which can be induced
by challenge with the opioid antagonist
naloxone or by enforced abstinence, has
also been demonstrated56. The features of
the syndrome — including teeth chattering,
forepaw tremor and head shakes — along
with their induction by administration of
an opioid antagonist, indicate an opioid-
mediated effect of the high-sugar diet. In
these withdrawal states, levels of dopamine
in the accumbens fall and acetylcholine
levels rise56. However, such a withdrawal
syndrome has not been demonstrated with
high-fat and cafeteriadiets51.
How do these behavioural changes relate
to altered neural substrates? In animals
binge-eating on high-sugar diets, the dopa-
mine release that occurs with food exposure
fails to habituate with loss of novelty, even in
those that are sham fed (food is consumed
orally but not digested because it is removed
immediately by a gastric cannula)52,57. In
animals binge-eating on sugar and those fed
a cafeteria diet, striatal D2R levels fall54,58.
Moreover, in the animals of the latter group,
brain self-stimulation thresholds (the mini-
mum intensity of electrical stimulation in
the lateral hypothalamus that will maintain
self-administration of the stimulation by
the animal) increase and remain elevated
2weeks after cessation of the diet, indicat-
ing early and persistent alteration of reward
thresholds54. These findings suggest the
development of a reward-deficiency state
similar to that seen with drugs of abuse59,60.
Reductions in presynaptic dopamine have
also been shown in animals on cafeteria
diets, and their dopamine activity is reduced
in response to standard chow but not palat-
able food61. A complementary finding is that
obesity-prone animals have been shown to
have lower baseline levels of dopamine62,63.
In summary, highly controlled condi-
tions for short periods of time can produce
sugar dependence in rats, although this is
not associated with obesity. Conversely, the
combination of high fat and high sugar can
produce a compulsive overeating syndrome,
accompanied by obesity and the develop-
ment of a negative anhedonic state. In both
situations, there is a corresponding reduc-
tion in D2Rs. Notably, researchers who have
carried out experiments evaluating food
addiction in animals are at pains to point out
that there are important differences between
the effects of foods and drugs (for example,
dopamine release in response to drugs per-
sists across multiple administrations, whereas
dopamine release induced by palatable foods
ceases when the food is no longer novel or
the animal is no longer hungry21). The neces-
sity for highly specific food presentation in
order to engender addictive behaviours is
also an important consideration64. Given that
the environments of humans are much more
variable than those of laboratory animals,
the degree to which models of food addiction
in animals may extend to human obesity has
yet to be explored.
Dopamine receptor studies in human obesity.
In 2001, a landmark PET study demon-
strated reduced striatal D2R binding in a
group of obese individuals22. Importantly,
D2R levels were negatively correlated with
BMI. The ensuing inference, that obesity is
characterized by striatal hypofunction, is
consistent with a reward-deficiency account
of overeating22. The idea is that overeating
arises because there is less hedonic value
in food, leading to compensatory over-
consumption. Complementing this was the
observation that D2R binding correlated
with prefrontal metabolism65, suggesting
that striatal hypofunction is compounded
by reduced inhibitory control. This work
has been important in developing the
addiction model of obesity, although such
correlative, cross-sectional observations
do not tell us whether the receptor changes
occur as a consequence of, rather than a
cause of, increased BMI. More importantly,
subsequent PET studies have not produced
consistent findings.
In studies on normal-weight participants,
the act of consuming food was initially shown
to be associated with a reduction in dopamine
binding in the dorsal striatum to a degree that
correlated with subjectively rated meal pleas-
antness66. However, in a subsequent study,
the presence of food in the mouth was not
associated with a significant change in striatal
dopamine binding, although high levels of
dietary restraint were associated with greater
food-induced alterations in dopamine-
receptor availability in the dorsal striatum67.
Furthermore, using an elegant combination
of drug challenge (methylphenidate com-
pared with placebo) and stimulus presenta-
tion (food and neutral non-food stimuli), it
was shown that food stimulation alone does
not always have an impact on D2R striatal
binding and that, although food stimulation
combined with a methylphenidate challenge
is associated with reduced dopamine binding,
the same is true for the combination of meth-
ylphenidate and a neutral non-food stimulus
(and, moreover, binding changes produced
by the food–methylphenidate combination
do not differ significantly from those found
with a food–placebo combination)68. In
short, PET data relating to dopamine bind-
ing and food consumption in normal-weight
people are inconsistent, although this may be
due, in part, to the different methodological
approaches used, such as consuming versus
tastingfood.
Given the variability in dopamine respon-
sivity to food stimuli in normal-weight
humans, it is perhaps unsurprising that the
picture in obesity is also inconsistent. Even in
the first study, which showed reduced D2R
availability in morbidly obese individuals
(BMI range 42–60), there was considerable
overlap with binding measures in healthy-
weight controls22. In a more recent study69, a
comparable striatum-based analysis showed
no difference in baseline dopamine-binding
measures between overweight or obese
individuals and normal-weight controls
(although a subsequent voxel-wise analysis
showed a thalamic difference that extended
into the striatum). The negative correlation
between BMI and striatal dopamine bind-
ing was not replicated. There are, of course,
numerous reasons why one might expect dif-
ferences between the original sample, which
consisted of a group of people with a BMI of
more than 40, and the more recent one, in
which mean BMI was much less. For exam-
ple, peripheral metabolic profiles might be
PERSPECTIVES
282
|
APRIL 2012
|
VOLUME 13 www.nature.com/reviews/neuro
© 2012 Macmillan Publishers Limited. All rights reserved

quite different, as might food intake. But the
fact remains that reduced D2R binding
is not a consistent correlate of BMI or
obesity and, as such, this does not, as is
usually claimed, provide consistent evidence
in favour of the addiction hypothesis.
Perhaps the inconsistency is a conse-
quence of the phenotypic complexity of
obesity. However, a study focusing specifi-
cally on differences between binge eaters
and BMI-matched controls20 demonstrated
neither a correlation between receptor
binding and BMI nor group differences
that accord with an addiction model. In
BED, the combination of a food stimulus
and methylphenidate was associated with
reduced dopamine binding in the caudate,
whereas in non-binge-eating obese indi-
viduals only the combination of a non-food
stimulus and methylphenidate produced a
significant change. Other studies examin-
ing the impact of bariatric surgery have
also produced conflicting results, suggest-
ing both decreases and increases in recep-
tor binding subsequent to surgery70,71.
In short, the message emerging from
PET ligand studies is rather more complex
than is frequently asserted. Although it has
been shown that dopamine ligand binding
is reduced in obese individuals, this finding
has not been replicated, and studies involv-
ing challenges with dopamine-stimulating
drugs and food-related stimuli produce
complex results that do not corroborate an
addiction model. Nor does a narrowing of
the phenotypic question to BED do anything
to clarifymatters.
Functional neuroimaging. Functional neuro-
imaging is an important tool in testing the
addiction model, which predicts that func-
tional responses to foods and food-related
stimuli in key reward-related brain regions
should be consistently perturbed. This is
not the case. Although studies exploring
brain responses to food and food-related
stimuli in normal-weight people have shown
largely consistent activation in reward cir-
cuitry (including the amygdala, insula and
striatum), the pattern emerging from studies
comparing obese individuals and binge-eaters
with controls is most remarkable for its vari-
ability and inconsistency (TABLE 2). A more
specific prediction, based on the reward-
deficiency hypothesis, is an enhancement
of anticipatory responses and a reduction of
consummatory responses to food rewards
in obese individuals72. However, studies that
explicitly distinguish between anticipation-
and consumption-related brain activity are
rare, and their results are equivocal.
TABLE 2 summarizes key findings from
functional neuroimaging studies of children,
adolescents and adults that explored brain
responses to food-related stimuli (typically
images) and to anticipation and consumption
Table 2 | Summary of the findings of studies exploring altered brain responses in people with obesity or altered eating patterns
Brain region
Response to presentation of food
images
Response to cues signalling
imminent presentation of food/
juice reward (anticipation) Response to consumption of reward
Obese BED BMI FA Obese BED BMI FA Obese BED BMI FA
Regions associated with the reward circuitry
Striatum 2↑83,84,
1↓85,
1↔86
2↔87,88 1↑89,
1↓90,
3↔85,91,92
NA 1↑93, 1↔94 NA NA 1↑95 5↔93,94,96–98 1↓99,
1↔100 1↓94 1↔95
Midbrain 4↔83–86 2↔87,88 5↔85,89–92 NA 2↔93,94 NA NA 1↔95 1↑96,
4↔93,94,97,98 2↔99,100 1↔94 1↔95
PFC (orbital) 1↑86,
3↔83–85 1↑87,
1↔88 3↑90–92,
1↓89,
1↔85
NA 2↔93,94 NA NA 1↑95 1↑96,
4↔93,94,97,98 1↓99,
1↔100 1↔94 1↓95
PFC (lateral) 3↑84–86,
1↔83 2↔87,88 1↑85,
1↓92,
3↔89–91
NA 1↑93, 1↔94 1↑101 NA 1↔95 1↑93,
2↓97,98,
2↔94,96
2↔99,100 1↔94 1↔95
PFC (medial) 2↑84,86,
1↓85,
1↔83
1↑87,
1↔88 1↓92,
4↔85,89–91
NA 1↑94, 1↔93 NA NA 1↑95 5↔93,94,96–98 2↔99,100 1↔94 1↔95
Amygdala 4↔83–86 2↔87,88 5↔85,89–92 NA 2↔93,94 NA NA 1↑95 1↑93,
4↔94,96–98 1↓99,
1↔100 1↔94 1↔95
Gustatory
cortex (AI/FO) 1↑83,
3↔84–86 1↑87,
1↓88 3↑89,90,92,
2↔85,91
NA 1↑94, 1↔93 NA NA 1↔95 3↑93,94,96,
2↔97,98 2↓99,100 1↔94 1↔95
Hippocampus/
PHG 2↑84,86,
1↓85,
1↔83
2↔87,88 1↓85,
4↔89–92
NA 1↑93, 1↔94 NA NA 1↔95 5↔93,94,96–98 2↔99,100 1↔94 1↔95
Brain regions not associated with the reward circuitry
Thalamus 1↓85,
3↔83,84,86 2↔87,88 5↔85,89–92 NA 2↔93,94 NA NA 1↔95 5↔93,94,96–98 2↔99,100 1↔94 1↔95
Rolandic
operculum 4↔83–86 2↔87,88 5↔85,89–92 NA 2↑93,94 NA NA 1↔95 2↑93,94,
3↔96–98 2↔99,100 1↔94 1↔95
The table shows responses that were elevated (↑) or reduced (↓) in groups of obese individuals or those with binge-eating disorder (BED) relative to controls.
Nogroup difference is signified by ‘↔’. Numbers before the arrows indicate the number of studies. The table also shows studies reporting positive (↑), negative (↓)
or no (↔) reported group difference between neural activity and body mass index (BMI) or food addiction (FA) scores. AI, anterior insula; FO, frontal operculum;
NA,no reports available (at the time of writing); PFC, prefrontal cortex; PHG, parahippocampal gyrus.
PERSPECTIVES
NATURE REVIEWS
|
NEUROSCIENCE VOLUME 13
|
APRIL 2012
|
283
© 2012 Macmillan Publishers Limited. All rights reserved

Nature Reviews | Neuroscience
Energy balance
Energy
expenditure
Food environment
Appetite
and satiety
Individual
predispositions
Environmental
influences
Personality and
reward circuitry
Physical
environment
• Fat-free mass
• Physical activity
• Availability
• Palatability
• Energy density
• Portion size
• Food access
• Access to physical
activity
• Advertising
• TV watching
• Parental and societal
influences
• Media
• Impaired satiety
signalling
• Insensitivity to
hunger and fullness
• Eating rate
• Early developmental
programming
• Genetic and epigenetic
factors
• Reward sensitivity
• Impulsivity
• Abnormal eating
• Food addiction
of actual food stimuli (typically milkshake).
A number of approaches have been used to
explore obesity and altered eating patterns.
Case–control studies comparing obese indi-
viduals with normal-weight controls are typi-
cal and are complemented by analyses of the
extent to which activity correlates with BMI
and, in one study, with food-addiction score.
Studies of binge eating (with bulimia nervosa
or BED) have also been carried out. The
findings shown in TABLE 2 indicate a striking
lack of consistency acrossstudies.
Of course, there are differences in tasks
and stimuli across the studies and there are
age and gender differences across the groups
studied. But, given that the striatum, mid-
brain and prefrontal cortex are core compo-
nents of the dopaminergic-reinforcement
circuitry, the lack of consistent findings across
a large set of studies militates strongly against
the addiction model. If we consider the region
of the anterior insula and frontal operculum
that is sometimes referred to as the gusta-
tory cortex, the inconsistency remains. Nor
is observation of responses in the amygdala
helpful in distinguishing obese individuals
from normal-weight controls. The over-
whelming message emerging from TABLE 2,
even allowing for technical and participant
differences, is that functional neuroimaging
does not support the addictionmodel.
Functional neuroimaging allows us to
measure not only regional responses but
also inter-regional relationships. Alterations
in these system-wide patterns have been
assessed in association with external food
sensitivity — the extent to which external
food cues evoke the desire to eat73 — and
obesity74. Although intriguing observations
have been made, particularly with respect
to the regions described above (which
constitute the ‘reward circuitry’), it is too
soon to judge whether connectivity studies
will show a consistency that eludes regional
measures.
There are two clear messages emerging
from the functional neuroimaging literature
on obesity and overeating. First, a growing
body of work has not supported any single
view of obesity and overeating. Second,
even when analysis is confined to sub-
groups showing binge-eating behaviour,
there has been no convincing or consist-
ent pattern of abnormal responding in the
reward circuitry. If the addiction model
of overeating has currency beyond phe-
notypic similarities (which, as we argue
above, are themselves weak), we would
expect functional neuroimaging studies to
identify core similarities. Why have they
failed to provide any consistent insight into
the behaviour of brain reward circuitry in
overeating, let alone support for the addic-
tion model? We find it hard to believe that
such circuitry is unaltered. One possibility
is that overeating and its consequences are
just too complex to expect consistency when
individuals are grouped simply according to
BMI, or to binge-eating or food-addiction
scores. Given that obesity and binge eating
are complex phenotypes emerging for a
host of genetic and environmental reasons,
in failing to account for this complexity
our capacity to identify group or factor-
related differences is markedly reduced.
Furthermore, both of these phenotypes
have often been measured cross-sectionally,
without taking into account the natural his-
tory of these conditions (BOX 2). We clearly
need more precise behavioural, temporal,
metabolic, genetic and cognitive profiling in
such investigations. Moreover, the growing
sophistication of cognitive neuroscientific
models of addictive behaviours points
to crucial process-specific alterations in
regional responding. Dissecting out these
processes will require more complex task-
dependent measurements than are typically
applied in overeating and obese individuals.
In the future, those imaging studies that
attempt to distinguish subtle processes and
simultaneously take into account individual
variability27 will prove useful and important.
Figure 1 | Mediators of energy balance and body weight. The outer
ring represents the major classes of mediators, the inner ring some of the
individual mediators in each class. We suggest that food addiction is one
of many factors in a more complex model of the obesity epidemic that
require further exploration and refinement. The data on which the figure
is based come from the Obesity Systems Map introduced by the UK
Foresight programme 2007, a multidisciplinary effort to plan the
UK response to obesity82.
PERSPECTIVES
284
|
APRIL 2012
|
VOLUME 13 www.nature.com/reviews/neuro
© 2012 Macmillan Publishers Limited. All rights reserved

Conclusions and future directions
The view that overeating and obesity are
directly related to addiction has provided
impetus to a series of elegant studies testing
this proposed link. Somehow, the view has
emerged that, overall, these studies sup-
port the link. We challenge this view and
argue that the work tells us three important
things. First, the vast majority of overweight
individuals have not shown a convincing
behavioural or neurobiological profile that
resembles addiction. Indeed, the enormous
inconsistency emerging from a review of
the neuroimaging literature tells us that
in this highly heterogenous disorder, the
application of a single model is likely to be
more of a hindrance than a help to future
research. Second, even when we refine the
phenotype to characterize individuals who
show obesity caused by BED, the evidence
for an overlap with addiction is inconsist-
ent and weak. Third, given the absence of
good evidence, the ubiquitous influence
of the addiction model of overeating and
consequent obesity is remarkable. Now is a
good time to question it and to acknowledge
that adherence to it in the face of data that
do not fit will lead to research that is too
narrowly focused and, ultimately, mislead-
ing. Given the attention that is rightly paid
to potential insights offered by neurosci-
ence, there is an associated danger that
clinical and policy recommendations will
be misguided. We suggest that alternative
approaches to exploring the brain’s con-
tributions to obesity be explored. Central
to these is an explicit acknowledgement of
the enormous heterogeneity of the condi-
tion, which requires further exploration
and characterization. This characterization
will, we anticipate, entail the use of cogni-
tive neuroscience to provide useful pheno-
typic markers of the numerous pathways
toobesity.
Our intention in this Perspective has
been to urge caution against the hasty adop-
tion of a model with limited applicability and
supporting evidence. We do not deny that
there may well be a place for an addiction
model in the understanding of overeating
and the spectrum of the obesity syndrome
(FIG1). However, successful development
of such a model will demand a progres-
sion beyond existing clinical definitions of
addiction to ideas that are guided by the
developing neuroscientific literature (BOX 2).
It will also demand sophisticated and precise
delineations of altered eating behaviour in
humans, and phenotypic markers that go
well beyond simple cross-sectional measures
such asBMI.
Hisham Ziauddeen and Paul C. Fletcher are in the
Department of Psychiatry, University of Cambridge,
Herchel Smith Building, Addenbrooke’s Hospital,
Cambridge CB2 0SZ, UK; and at the Cambridgeshire
and Peterborough NHS Foundation Trust, Fulbourn
Hospital, Cambridge CB21 5EF, UK.
Hisham Ziauddeen and I. Sadaf Farooqi are at the
University of Cambridge Metabolic Research
Laboratories, Institute of Metabolic Science,
Addenbrooke’s Hospital, Cambridge CB2 0QQ, UK.
Correspondence to P.C.F.
e-mail: pcf22@cam.ac.uk
doi:10.1038/nrn3212
Published online 14 March 2012
1. Swinburn, B.A. etal. The global obesity pandemic:
shaped by global drivers and local environments.
Lancet 378, 804–814 (2011).
2. Hall, K.D. etal. Quantification of the effect of energy
imbalance on bodyweight. Lancet 378, 826–837
(2011).
3. Gearhardt, A.N., Grilo, C.M., Dileone, R.J.,
Brownell, K.D. & Potenza, M.N. Can food be
addictive? Public health and policy implications.
Addiction 106, 1208–1212 (2011).
4. Devlin, M.J. Is there a place for obesity in DSM‑V?
Int.J.Eat. Disord. 40, S83–S88 (2007).
5. Volkow, N.D. & O’Brien, C.P. Issues for DSM‑V:
should obesity be included as a brain disorder?
Am.J.Psychiatry 164, 708–710 (2007).
6. Ifland, J.R. etal. Refined food addiction: a classic
substance use disorder. Med. Hypotheses 72,
518–526 (2009).
7. Cocores, J.A. & Gold, M.S. The Salted Food Addiction
Hypothesis may explain overeating and the obesity
epidemic. Med. Hypotheses 73, 892–899 (2009).
8. Volkow, N.D., Wang, G.‑J. & Baler, R.D.
Reward,dopamine and the control of food intake:
implications for obesity. Trends Cogn. Sci. 15, 37–46
(2011).
9. Volkow, N.D. & Wise, R.A. How can drug addiction
help us understand obesity? Nature Neurosci. 8,
555–560 (2005).
10. Kelley, A.E. & Berridge, K.C. The neuroscience of
natural rewards: relevance to addictive drugs.
J.Neurosci. 22, 3306–3311 (2002).
11. Benton, D. The plausibility of sugar addiction and its
role in obesity and eating disorders. Clin. Nutr. 29,
288–303 (2010).
12. American Psychiatric Association. Diagnostic and
Statistical Manual of Mental Disorders 4th edn
(American Psychiatric Association, Washington, DC,
2000).
13. Gearhardt, A.N., Corbin, W.R. & Brownell, K.D.
Preliminary validation of the Yale Food Addiction
Scale. Appetite 52, 430–436 (2009).
14. Gearhardt, A.N., Corbin, W.R. & Brownell, K.D.
Food addiction: an examination of the diagnostic
criteria for dependence. J.Addict. Med. 3, 1–7
(2009).
15. Davis, C. & Carter, J.C. Compulsive overeating as an
addiction disorder. A review of theory and evidence.
Appetite 53, 1–8 (2009).
16. Striegel‑Moore, R.H. & Franko, D.L. Epidemiology
of binge eating disorder. Int. J.Eat. Disord. 34,
S19–S29 (2003).
17. O’Rahilly, S. Human genetics illuminates the paths to
metabolic disease. Nature 462, 307–314 (2009).
18. Drewnowski, A. Obesity, diets, and social inequalities.
Nutr. Rev. 67, S36–S39 (2009).
19. Davis, C. etal. Evidence that ‘food addiction’ is a valid
phenotype of obesity. Appetite 57, 711–717 (2011).
20. Wang, G.J. etal. Enhanced striatal dopamine release
during food stimulation in binge eating disorder.
Obesity 19, 1601–1608 (2011).
21. Avena, N.M., Rada, P. & Hoebel, B.G. Evidence for
sugar addiction: behavioral and neurochemical
effects of intermittent, excessive sugar intake.
Neurosci. Biobehav. Rev. 32, 20–39 (2008).
22. Wang, G.J. etal. Brain dopamine and obesity.
Lancet 357, 354–357 (2001).
23. Cassin, S.E. & von Ranson, K.M. Is binge eating
experienced as an addiction? Appetite 49, 687–690
(2007).
24. Grucza, R.A. etal. The emerging link between
alcoholism risk and obesity in the United States.
Arch. Gen. Psychiatry 67, 1301–1308 (2010).
25. Lilenfeld, L.R.R., Ringham, R., Kalarchian, M.A. &
Marcus, M.D. A family history study of binge‑eating
disorder. Compr. Psychiatry 49, 247–254 (2008).
26. Stice, E., Spoor, S., Bohon, C. & Small, D.M.
Relation between obesity and blunted striatal
response to food is moderated by Taq IA A1 allele.
Science 322, 449–452 (2008).
27. Stice, E., Yokum, S., Bohon, C., Marti, N. & Smolen, A.
Reward circuitry responsivity to food predicts
future increases in body mass: moderating effects
of DRD2 and DRD4. Neuroimage 50, 1618–1625
(2010).
28. Felsted, J.A., Ren, X., Chouinard‑Decorte, F. &
Small, D.M. Genetically determined differences in
brain response to a primary food reward. J.Neurosci.
30, 2428–2432 (2011).
29. Munafo, M.R., Matheson, I.J. & Flint, J. Association
of the DRD2 gene Taq1A polymorphism and
alcoholism: a meta‑analysis of case‑control studies
and evidence of publication bias. Mol. Psychiatry 12,
454–461 (2007).
30. Noble, E.P. etal. Allelic association of the D2
dopamine receptor gene with cocaine dependence.
Drug Alcohol Depend. 33, 271–285 (1993).
31. Zuo, Y. etal. DRD2‑related TaqIA polymorphism
modulates motivation to smoke. Nicotine Tob. Res. 11,
1321–1329 (2009).
32. Doehring, A. etal. Genetic variants altering dopamine
D2 receptor expression or function modulate the risk
of opiate addiction and the dosage requirements of
methadone substitution. Pharmacogenet. Genomics
19, 407–414 (2009).
33. Davis, C.A. etal. Dopamine for “wanting” and opioids
for “liking”: a comparison of obese adults with and
without binge eating. Obesity 17, 1220–1225
(2009).
34. Smith, L., Watson, M., Gates, S., Ball, D. & Foxcroft, D.
Meta‑analysis of the association of the Taq1A
polymorphism with the risk of alcohol dependency:
aHuGE gene‑disease association review.
Am.J.Epidemiol. 167, 125–138 (2008).
35. Munafo, M.R., Timpson, N.J., David, S.P., Ebrahim, S.
& Lawlor, D.A. Association of the DRD2 gene Taq1A
polymorphism and smoking behavior: a meta‑analysis
and new data. Nicotine Tob. Res. 11, 64–76
(2009).
36. Fernandez‑Castillo, N. etal. Association study
between the DAT 1, DBH and DRD2 genes and cocaine
dependence in a Spanish sample. Psychiatr. Genet.
20, 317–320 (2010).
37. Pohjalainen, T. etal. The A1 allele of the human D2
dopamine receptor gene predicts low D2 receptor
availability in healthy volunteers. Mol. Psychiatry 3,
256–260 (1998).
38. Jonsson, E.G. etal. Polymorphisms in the dopamine
D2 receptor gene and their relationships to striatal
dopamine receptor density of healthy volunteers.
Mol. Psychiatry 4, 290–296 (1999).
39. Davis, C. etal. Opiates, overeating and obesity: a
psychogenetic analysis. Int. J.Obes. 35, 1347–1354
(2011).
40. Miranda, R. etal. Initial evidence of an association
between OPRM1 and adolescent alcohol misuse.
Alcohol. Clin. Exp. Res. 34, 112–122 (2010).
41. Ramchandani, V.A. etal. A genetic determinant of
the striatal dopamine response to alcohol in men.
Mol. Psychiatry 16, 809–817 (2011).
42. Comings, D.E. & Blum, K. Reward deficiency
syndrome: genetic aspects of behavioral disorders.
Prog. Brain Res. 126, 325–341 (2000).
43. Davis, C. & Fox, J. Sensitivity to reward and body
mass index (BMI): evidence for a non‑linear
relationship. Appetite 50, 43–49 (2008).
44. Verdejo‑Garcia, A., Lawrence, A.J. & Clark, L.
Impulsivity as a vulnerability marker for
substance‑use disorders: review of findings from
high‑risk research, problem gamblers and genetic
association studies. Neurosci. Biobehav. Rev. 32,
777–810 (2008).
45. Limosin, F. etal. Impulsiveness as the intermediate
link between the dopamine receptor D2 gene and
alcohol dependence. Psychiatr. Genet. 13, 127–129
(2003).
46. Eisenberg, D.T. etal. Examining impulsivity as an
endophenotype using a behavioral approach: a DRD2
TaqI A and DRD4 48-bp VNTR association study.
Behav. Brain Funct. 3, 2 (2007).
47. Guerrieri, R. etal. The influence of trait and
induced state impulsivity on food intake in
normal‑weight healthy women. Appetite 49, 66–73
(2007).
PERSPECTIVES
NATURE REVIEWS
|
NEUROSCIENCE VOLUME 13
|
APRIL 2012
|
285
© 2012 Macmillan Publishers Limited. All rights reserved

48. Galanti, K., Gluck, M.E. & Geliebter, A. Test meal
intake in obese binge eaters in relation to impulsivity
and compulsivity. Int. J.Eat. Disord. 40, 727–732
(2007).
49. Davis, C. etal. Personality and eating behaviors:
a case–control study of binge eating disorder.
Int.J.Eat. Disord. 41, 243–250 (2008).
50. Kenny, P.J. Common cellular and molecular
mechanisms in obesity and drug addiction.
Nature Rev. Neurosci. 12, 638–651 (2011).
51. Bocarsly, M.E., Berner, L.A., Hoebel, B.G. &
Avena, N.M. Rats that binge eat fat‑rich food do not
show somatic signs or anxiety associated with opiate‑
like withdrawal: implications for nutrient‑specific food
addiction behaviors. Physiol. Behav. 104, 865–872
(2011).
52. Avena, N.M., Rada, P., Moise, N. & Hoebel, B.G.
Sucrose sham feeding on a binge schedule releases
accumbens dopamine repeatedly and eliminates the
acetylcholine satiety response. Neuroscience 139,
813–820 (2006).
53. Avena, N.M., Rada, P. & Hoebel, B.G. Sugar and fat
bingeing have notable differences in addictive‑like
behavior. J.Nutr. 139, 623–628 (2009).
54. Johnson, P.M. & Kenny, P.J. Dopamine D2 receptors
in addiction‑like reward dysfunction and compulsive
eating in obese rats. Nature Neurosci. 13, 635–641
(2010).
55. Avena, N.M., Long, K.A. & Hoebel, B.G.
Sugar‑dependent rats show enhanced responding
for sugar after abstinence: evidence of a sugar
deprivation effect. Physiol. Behav. 84, 359–362
(2005).
56. Colantuoni, C. etal. Evidence that intermittent,
excessive sugar intake causes endogenous opioid
dependence. Obes Res. 10, 478–488 (2002).
57. Rada, P., Avena, N.M. & Hoebel, B.G. Daily bingeing
on sugar repeatedly releases dopamine in the
accumbens shell. Neuroscience 134, 737–744
(2005).
58. Colantuoni, C. etal. Excessive sugar intake alters
binding to dopamine and mu‑opioid receptors in the
brain. Neuroreport 12, 3549–3552 (2001).
59. Ahmed, S.H., Kenny, P.J., Koob, G.F. & Markou, A.
Neurobiological evidence for hedonic allostasis
associated with escalating cocaine use. Nature
Neurosci. 5, 625–626 (2002).
60. Kenny, P.J., Chen, S.A., Kitamura, O., Markou, A. &
Koob, G.F. Conditioned withdrawal drives heroin
consumption and decreases reward sensitivity.
J.Neurosci. 26, 5894–5900 (2006).
61. Geiger, B.M. etal. Deficits of mesolimbic dopamine
neurotransmission in rat dietary obesity. Neuroscience
159, 1193–1199 (2009).
62. Geiger, B.M. etal. Evidence for defective mesolimbic
dopamine exocytosis in obesity‑prone rats. FASEB J.
22, 2740–2746 (2008).
63. Rada, P., Bocarsly, M.E., Barson, J.R., Hoebel, B.G.
& Leibowitz, S.F. Reduced accumbens dopamine in
Sprague‑Dawley rats prone to overeating a fat‑rich
diet. Physiol. Behav. 101, 394–400 (2010).
64. Corwin, R.L. & Grigson, P.S. Symposium overview‑
food addiction: fact or fiction? J.Nutr. 139, 617–619
(2009).
65. Volkow, N.D. etal. Low dopamine striatal D2
receptors are associated with prefrontal metabolism
in obese subjects: possible contributing factors.
Neuroimage 42, 1537–1543 (2008).
66. Small, D.M., Jones‑Gotman, M. & Dagher, A.
Feeding‑induced dopamine release in dorsal striatum
correlates with meal pleasantness ratings in healthy
human volunteers. Neuroimage 19, 1709–1715
(2003).
67. Volkow, N.D. etal. Brain dopamine is associated with
eating behaviors in humans. Int. J.Eat. Disord. 33,
136–142 (2003).
68. Volkow, N.D. etal. “Nonhedonic” food motivation in
humans involves dopamine in the dorsal striatum and
methylphenidate amplifies this effect. Synapse 44,
175–180 (2002).
69. Haltia, L.T. etal. Effects of intravenous glucose on
dopaminergic function in the human brain invivo.
Synapse 61, 748–756 (2007).
70. Steele, K.E. etal. Alterations of central dopamine
receptors before and after gastric bypass surgery.
Obes. Surg. 20, 369–374 (2010).
71. Dunn, J.P. etal. Decreased dopamine type2 receptor
availability after bariatric surgery: preliminary
findings. Brain Res. 1350, 123–130 (2010).
72. Volkow, N.D., Wang, G.J., Fowler, J.S. & Tomasi, D.
Addiction circuitry in the human brain. Annu. Rev.
Pharmacol. Toxicol. 52, 321–336 (2012).
73. Passamonti, L. etal. Personality predicts the brain’s
response to viewing appetizing foods: the neural basis
of a risk factor for overeating. J.Neurosci. 29, 43–51
(2009).
74. Kullmann, S. etal. The obese brain: association of
body mass index and insulin sensitivity with resting
state network functional connectivity. Hum. Brain
Mapp. 21Apr 2011 (doi:10.1002/hbm.21268).
75. Everitt, B.J. etal. Review. Neural mechanisms
underlying the vulnerability to develop compulsive
drug‑seeking habits and addiction. Phil. Trans. R.Soc. B
363, 3125–3135 (2008).
76. Dalley, J.W. etal. Nucleus accumbens D2/3 receptors
predict trait impulsivity and cocaine reinforcement.
Science 315, 1267–1270 (2007).
77. Koob, G.F. & Le Moal, M. Review. Neurobiological
mechanisms for opponent motivational processes in
addiction. Phil. Trans. R.Soc. B 363, 3113–3123
(2008).
78. Robinson, T.E. & Berridge, K.C. The neural basis of
drug craving: an incentive‑sensitization theory of
addiction. Brain Res. Brain Res. Rev. 18, 247–291
(1993).
79. Volkow, N.D. etal. Low level of brain dopamine D2
receptors in methamphetamine abusers: association
with metabolism in the orbitofrontal cortex.
Am.J.Psychiatry 158, 2015–2021 (2001).
80. Deroche‑Gamonet, V., Belin, D. & Piazza, P.V.
Evidence for addiction‑like behavior in the rat.
Science 305, 1014–1017 (2004).
81. Robbins, T.W., Gillan, C.M., Smith, D.G., de Wit, S. &
Ersche, K.D. Neurocognitive endophenotypes of
impulsivity and compulsivity: towards dimensional
psychiatry. Trends Cogn. Sci. 16, 81–91 (2012).
82. Vandenbroeck, P., Goossens, J. & Clemens, M.
Tackling Obesities: Future Choices — Obesity System
Atlas (Government Office for Science, UK, 2007).
83. Rothemund, Y. etal. Differential activation of the
dorsal striatum by high‑calorie visual food stimuli in
obese individuals. Neuroimage 37, 410–421 (2007).
84. Martin, L.E. etal. Neural mechanisms associated with
food motivation in obese and healthy weight adults.
Obesity 18, 254–260 (2010).
85. Davids, S. etal. Increased dorsolateral prefrontal
cortex activation in obese children during observation
of food stimuli. Int. J.Obes. 34, 94–104 (2010).
86. Bruce, A.S. etal. Obese children show hyperactivation
to food pictures in brain networks linked to motivation,
reward and cognitive control. Int. J.Obes 34,
1494–1500 (2010).
87. Schienle, A., Schäfer, A., Hermann, A. & Vaitl, D.
Binge‑eating disorder: reward sensitivity and brain
activation to images of food. Biol. Psychiatry 65,
654–661 (2009).
88. Brooks, S.J. etal. Differential neural responses to
food images in women with bulimia versus anorexia
nervosa. PLoS ONE 6, e22259 (2011).
89. Stice, E., Yokum, S., Blum, K. & Bohon, C. Weight gain
is associated with reduced striatal response to
palatable food. J.Neurosci. 30, 13105–13109 (2010).
90. Yokum, S., Ng, J. & Stice, E. Attentional bias to food
images associated with elevated weight and future
weight gain: an FMRI study. Obesity 19, 1775–1783
(2011).
91. Killgore, W.D.S. & Yurgelun‑Todd, D.A. Body mass
predicts orbitofrontal activity during visual
presentations of high‑calorie foods. Neuroreport 16,
859–863 (2005).
92. Batterink, L., Yokum, S. & Stice, E. Body mass
correlates inversely with inhibitory control in response
to food among adolescent girls: an fMRI study.
Neuroimage 52, 1696–1703 (2010).
93. Ng, J., Stice, E., Yokum, S. & Bohon, C. An fMRI study
of obesity, food reward, and perceived caloric density.
Does a low‑fat label make food less appealing?
Appetite 57, 65–72 (2011).
94. Stice, E., Spoor, S., Bohon, C., Veldhuizen, M.G. &
Small, D.M. Relation of reward from food intake and
anticipated food intake to obesity: a functional
magnetic resonance imaging study. J.Abnorm.
Psychol. 117 , 924–935 (2008).
95. Gearhardt, A.N. etal. Neural correlates of food
addiction. Arch. Gen. Psychiatry 68, 808–816 (2011).
96. DelParigi, A., Chen, K., Salbe, A.D., Reiman, E.M. &
Tataranni, P.A. Sensory experience of food and
obesity: a positron emission tomography study of the
brain regions affected by tasting a liquid meal after a
prolonged fast. Neuroimage 24, 436–443 (2005).
97. Le, D.S. etal. Less activation of the left dorsolateral
prefrontal cortex in response to a meal: a feature of
obesity. Am. J.Clin. Nutr. 84, 725–731 (2006).
98. Le, D.S. etal. Less activation in the left dorsolateral
prefrontal cortex in the reanalysis of the response to a
meal in obese than in lean women and its association
with successful weight loss. Am. J.Clin. Nutr. 86,
573–579 (2007).
99. Frank, G.K.W., Reynolds, J.R., Shott, M.E. &
O’Reilly, R.C. Altered temporal difference learning in
bulimia nervosa. Biol. Psychiatry 70, 728–735
(2011).
100. Bohon, C. & Stice, E. Reward abnormalities among
women with full and subthreshold bulimia nervosa:
a functional magnetic resonance imaging study.
Int. J.Eat. Disord. 44, 585–595 (2010).
101. Karhunen, L.J. etal. Regional cerebral blood flow
during exposure to food in obese binge eating women.
Psychiatry Res. 99, 29–42 (2000).
Acknowledgements
We would like to thank B. Everitt for his comments, particu‑
larly on Box2. H.Z. is jointly funded by the Wellcome Trust
and GlaxoSmithKline. I.S.F. and P.C.F. are supported by the
Bernard Wolfe Health Neuroscience Fund. H.Z., I.S.F. and
P.C.F. are also supported by the Wellcome Trust, MRC
(Medical Research Council) Centre for Obesity and Related
Diseases, and the UK National Institute for Health Research
(NIHR) Cambridge Biomedical Research Centre. This work
was carried out at the Institute of Metabolic Science and the
Wellcome–MRC‑funded Behavioural and Clinical
Neuroscience Institute. It was inspired by discussions with
fellow members of the Behaviour and Health ResearchUnit.
Competing interests statement
The authors declare competing financial interests; see Web
version for details.
DATABASES
Pathway Interaction Database: http://pid.nci.nih.gov
FURTHER INFORMATION
Behaviour and Health Research Unit:
http://www.bhru.iph.cam.ac.uk
Metabolic Research Laboratories:
http://www.mrl.ims.cam.ac.uk
ALL LINKS ARE ACTIVE IN THE ONLINE PDF
PERSPECTIVES
286
|
APRIL 2012
|
VOLUME 13 www.nature.com/reviews/neuro
© 2012 Macmillan Publishers Limited. All rights reserved