Is Drug Addiction a Brain Disease? This popular claim lacks evidence and leads to poor policy

Abstract

The concept of the addicted brain appeals to the intuitive idea that because recreational drugs exert their actions (for example, eliciting euphoria) through specific brain structures, there must exist discernible and meaningful brain differences between those who are afflicted with addiction and those who are not. These differences are then seen as the cause of the addiction and, therefore, as targets for treatment. Despite this seemingly solid scientific consensus, there are virtually no data in humans indicating that addiction
is a disease of the brain in the way that, for instance, Huntington’s or Parkinson’s are diseases of the brain. The existing paradigm is based on intuition and political necessity, not on data and useful clinical results. Yet the diseased-brain perspective has outsized
influence on research funding and drug policy.

Full text

160 American Scientist, Volume 106
The notion that drug addiction
is a brain disease has become
axiomatic. Around the globe
aspiring health professionals
treating substance abuse are indoctri-
nated with this belief, especially after
the idea became popular in the 1990s.
Its popularity extends far beyond the
hallowed halls of academia. Both the
May 1997 Time and the September 2017
National Geographic magazines were
dedicated to the brain science of ad-
diction. Numerous other popular mag-
azines have run similar cover stories
over the past two decades.
But after 20 years of research, one
of us (Hart) saw that paradigm yield-
ing dismal results. Meanwhile, be-
havioral research on outcomes after
providing both animals and humans
with attractive alternatives to drugs
has yielded positive results regarding
effective treatments, despite the lack
of mainstream attention. This observa-
tion prompted Hart to refocus his re-
search on these behavioral treatments.
So in 2016 we teamed up to reexamine
the prevailing assumptions supporting
the brain-disease model of addiction
and the data behind those assump-
tions. Like many other people in ad-
diction research, coauthor Grifell had
not directly questioned this paradigm
until teaming up with Hart and dig-
ging into the evidence. Brain-imaging
data from methamphetamine- addicted
users provide the strongest support for
the prevailing paradigm but still can
be interpreted in other ways.
The concept of the addicted brain
appeals to the intuitive idea that be-
cause recreational drugs exert their ac-
tions (for example, eliciting euphoria)
through specific brain structures, there
must exist discernible and meaningful
brain differences between those who
are afflicted with addiction and those
who are not. These differences are then
seen as the cause of the addiction and,
therefore, as targets for treatment. The
“diseased brain” perspective is appeal-
ing also because it offers treatment ap-
proaches that seem straightforward.
This view was consolidated and per-
suasively expressed in 1997 when the
director of the U.S. National Institute
on Drug Abuse (NIDA), Alan Leshner,
published an influential editorial in
the journal Science titled “Addiction
is a brain disease, and it matters.” He
explained, “That addiction is tied to
changes in brain structure and func-
tion is what makes it, fundamentally,
a brain disease.” Multiple proponents,
including current NIDA Director, Nora
Volkow, neuroscience researchers, ad-
diction experts, and even politicians,
have echoed these sentiments.
Despite this seemingly solid scien-
tific consensus, there are virtually no
data in humans indicating that ad-
diction is a disease of the brain in the
way that, for instance, Huntington’s or
Parkinson’s are diseases of the brain.
The existing paradigm is based on in-
tuition and political necessity, not on
data and useful clinical results. Yet the
diseased-brain perspective has out-
sized influence on research funding
and direction, as well as on how drug
use and addiction are viewed around
the globe. This situation contributes to
unrealistic, costly, and harmful drug
policies: If the problem is a person’s
neurobiological state after exposure to
a drug, then either the drug must be
eradicated from society through law
enforcement or an individual’s brain
must be treated. In such a myopic ap-
proach, the socioeconomic and societal
factors that contribute to drug addic-
tion are considered a footnote in re-
search, clinical practices, and policy,
despite their apparent importance.
What Is Addiction?
It’s easy to fall into circular thinking
by defining addiction as a brain disor-
der and then researching it as a brain
disorder. Knowing that someone uses
a drug, even regularly, does not tell us
whether that person is “addicted.” It
doesn’t even mean that the person has
a drug problem.
To meet the most widely accepted
definition of addiction—the one in
psychiatry’s Diagnostic and Statistical
Manual of Mental Disorders, 5th Edition
(DSM-5)—a person’s drug use must
interfere with important life functions
such as parenting, work, and intimate
relationships. This use must continue
despite ongoing negative consequences,
taking up a great deal of time and men-
tal energy and persisting in the face of
repeated attempts to stop or cut back. It
also may include the experience of need-
ing more of the drug to get the same ef-
fect (tolerance) and suffering withdraw-
al symptoms if use suddenly ceases.
But 75 percent or more of drug
users— whether they use alcohol, tobac-
co, prescription medications, or other
drugs—do not meet criteria for drug
addiction, according to numerous stud-
ies, including those by James C. An-
thony of Michigan State University and
his colleagues. Indeed, research shows
Marc Grifell is a physician completing his psychia-
try residency in May 2018. He is also completing
his PhD thesis under the mentorship of Carl Hart
at Columbia University and New York State Psy-
chiatric Institute. Dr. Grifell’s research focuses on
understanding the neurobiological and environ-
mental factors that mediate recreational drug effects.
Dr. Hart is chair of the Department of Psychology
atColumbia University. He is also the Dirk Ziff
Professor of Psychology in the Departments of Psy-
chology and Psychiatry and a research scientist at
the New York State Psychiatric Institute. He is the
author of High Price: A Neuroscientist’s Journey
of Self-Discovery That Challenges Everything
You Know about Drugs and Society. Email for
Dr. Hart: clh42@columbia.edu.
Is Drug Addiction a
Brain Disease?
This popular claim lacks evidence and leads to poor policy.
Marc Grifell and Carl L. Hart

2018 May–June 161www.americanscientist.org
repeatedly that such issues affect only
10 percent to 25 percent of those who
use even the most stigmatized drugs,
such as heroin, methamphetamine, and
crack cocaine. Our use of the term addic-
tion here is inter changeable with DSM-
5’s substance use disorder, which always
means problematic use of the sort that
interferes with functioning—not just
regularly ingesting a substance.
Brain-Disease Perspective in Action
Even under the DSM-5’s more subtle
definition of addiction, drug use could
fit under the brain-disease theory. To
investigate whether that is the case, we
must look at the studies of substance
abusers’ brains. The utility of any theory
is measured by how well it can account
for, predict, and even suggest strategies
to control or treat specific phenomena.
In Parkinson’s disease, for example, the
dominant theory asserts that after the
loss of more than 30 percent of dopa-
mine neurons in the midbrain, clinical
manifestation of the illness will com-
mence. Although Parkinson’s disease
progresses inexorably— that is to say,
it is irreversible and fatal—the brain-
disease theory used to explain the
mechanism underlying the disease’s
symptoms has led to the development
of treatments that control the telltale
gait, rigidity, and slow movement. With-
out dopamine replacement, currently
considered the gold-standard treatment,
advanced-stage Parkinson’s patients are
unable to move.
Viewing this illness from a neurobio-
logical perspective provides clear, in-
disputable evidence of the superiority
of this approach compared with a psy-
chological or behavioral theory. Such a
theory might, for example, emphasize
an environmental or behavioral modi-
fication to lessen Parkinsonian symp-
toms. This approach, of course, would
be inappropriate and less effective than
neuropharmacological manipulation
consisting of dopamine replacement.
In the case of addiction, however,
the diseased-brain theory does not
have such clear explanatory power, es-
pecially when informing guidelines to
treat the disorder. Behavioral and psy-
chosocial therapies, such as cognitive
behavioral therapy, contingency man-
agement, or motivational interview-
ing, remain the predominant treat-
ments for substance-use disorders.
It is true that over the past several
decades, data from basic research have
contributed to an increased understand-
ing of neural mechanisms involved in
many effects produced by recreational
drugs. Indeed, much of this evidence is
the bedrock on which the brain-disease
theory is built. But simply knowing that
a drug causes, for instance, an increase
in dopamine transmission does not nec-
essarily provide any information about
addiction to that drug.
We need to also make clear that
specific neural elements underlying
addiction have yet to be identified. De-
spite this empirical void, a key sup-
position of this theory purports that
detectable brain dysfunctions in indi-
viduals afflicted with addiction are the
source of the problem and that focusing
on the brain should be paramount in
our search for a solution.
A Lack of Supporting Data
Some of the most compelling evidence
that seems to support the diseased-
The prevailing paradigm among neuroscience researchers that addiction is a brain disease is not
supported by evidence and contributes to social injustice, contend authors Marc Grifell and Carl
Hart. The idea has outsized influence on research funding and direction, and on how drug use is
viewed around the globe. This outlook ignores the ways that socioeconomic factors contribute to
problems with substance abuse and contribute to unrealistic, costly, and harmful drug policies.
Taken to extremes, it can lead to situations such as that seen in the Philippines, where thousands of
people accused of using illegal drugs have been killed at the order of President Rodrigo Duterte.
Brian Hubble

162 American Scientist, Volume 106
brain view of addiction comes from
studies, mostly published in the 1980s
and 1990s, of laboratory animals that
were administered amphetamine. Dur-
ing these studies, the short- and long-
term effects of amphetamines on neu-
ral structure and functioning as well as
behavior were investigated. One of the
most consistent findings is that a single
large amphetamine dose administered
to nontolerant animals produces ex-
tensive damage to dopamine neurons,
meaning neuro toxicity. Areas of the
brain that are rich in dopamine serve a
wide range of important human func-
tions, from mood regulation to move-
ment to learning and memory. Indeed,
a substantial database collected in labo-
ratory animals indicates that large am-
phetamine doses produce disruptive
effects in multiple behavioral domains,
including learning and memory.
Because d-amphetamine and meth-
amphetamine are used in several
countries, including the United States,
to treat a variety of disorders, such as
attention-deficit hyperactivity disorder
(ADHD), narcolepsy, and obesity, it is
not difficult to see how the possibil-
ity of amphetamine-induced neuro-
toxicity might cause alarm. This in-
formation also raises concerns about
the potential harmful consequences of
methamphetamine addiction on the
human brain and behavior.
Still, an important question remains:
To what extent do the neuro toxicity
data collected in animals extrapolate
to people who may use amphetamine
medically or who may be addicted to
this class of drug? In other words, is
this research relevant to humans who
use amphetamines? Furthermore,
amphetamine-induced neuro toxicity
in animals can be prevented with
previous exposure to several days of
escalating amphetamine doses. Put
more simply: Dosing regimens used
in animal studies do not correspond
with those typically used by humans;
and the development of tolerance seen
in long-term use can be protective
against amphetamine neurotoxicity.
These caveats are critical because
human recreational drug users—
as well as patients on prescription
amphetamines— usually increase (or
decrease) their doses gradually as they
gain more experience with a particular
substance. This observation suggests
that deleterious neurobiological and
behavioral changes observed in many
animal studies may not be apparent
in human amphetamine users. It also
highlights the importance of employ-
ing relevant models in future investi-
gations of drug-related effects on brain
functioning and behavior in animals.
Over the past two decades, research
investigating the direct effects of am-
phetamine use on human functioning
has steadily increased. The immediate
effects of low to moderate doses of am-
phetamine or methamphetamine are
clear and consistent: increased energy,
enhanced ability to focus and concen-
trate, reduced subjective feelings of
tiredness, and attenuation of cognitive
disruptions brought about by fatigue
or sleep deprivation. It is precisely be-
cause of these effects that several na-
tions’ militaries, including the United
States, have used (and continue to use)
amphetamine in specific situations
since World War II. Amphetamines
help specialized professionals work
better and longer in critical situations.
The impact of long-term adminis-
tration of large drug doses in humans
is less clear, owing to ethical consid-
erations. An alternative approach to
determining the extent to which large
methamphetamine doses might pro-
duce brain or behavioral detrimental
effects is to compare the brains and
cognitive performance of healthy con-
trol participants with abstinent indi-
viduals whose drug use meets criteria
for methamphetamine addiction. The
idea is that regular use of illicit meth-
amphetamine over several years may
neuronneuron
neuronneuron
neuronneuron
receptor
postsynaptic neuronpostsynaptic neuron
postsynaptic neuron
postsynaptic neuron
postsynaptic neuron
neurotransmitters
receptor
receptor receptor
receptor
neurotransmitters
neurotransmitters
1 2
1 2
1 2
NORMAL
Two normally
functioning neurons
(1 and 2). Binding
potential of 100
percent with 12
receptors
TOLERANCE
There is a reduc-
tion of receptors for
both neuron 1 and
2. This is an
example of
receptor down-
regulation of 50
percent with 6
receptors.
NEUROTOXICITY
Neuron 1 remains
unchanged, and
postsynaptic
neuron 2 is dead.
Binding potential
is 50 percent with
6 receptors.
Positron emission tomography (PET) imaging is commonly used to measure dopamine binding po-
tentials in the brain, which is relevant because in laboratory animals, repeated methampehtamine
doses decrease dopamine transporter density and dopamine receptor availability. There are two
potential conclusions, however, from observing lower binding potentials in drug users compared
with controls: They could be building up tolerance to the drug, or they could be damaging neurons.
Barbara Aulicino

2018 May–June 163www.americanscientist.org
result in neurotoxic effects to specific
neurons, including dopamine neurons.
Damage to these neurons, of course,
can have disruptive consequences on
specific behaviors such as cognitive
functioning and movement.
Typically, these studies combine
brain-imaging techniques with cog-
nitive testing so that brain structure
integrity or activity can be correlated
with a relevant behavior. Here, we will
focus on studies using positron emis-
sion tomography (PET) imaging tech-
niques for several reasons. First, PET
studies provide the most consistent
findings to date regarding neurobio-
logical differences between individuals
with substance use disorders and con-
trol participants. Second, PET imaging
is currently the most suitable technique
to detect dopamine transporter density
and dopamine receptor availability.
This function is relevant because an
ample amount of evidence collected in
laboratory animals demonstrates that
large, repeated methamphetamine dos-
es decrease these dopamine markers.
Consequently, some researchers
have reasoned that long-term abuse of
methamphetamine by humans should
produce a reduction in dopamine trans-
porter density and dopamine receptor
availability. If this reasoning were cor-
rect, PET imaging studies would show
these brain differences, which then may
be interpreted as pathological changes
in the brain produced by methamphet-
amine. Finally, this reasoning in turn
would lead to the conclusion that such
changes are the cause of addiction.
With PET techniques, a radioactive-
ly labeled chemical is injected into the
bloodstream, and then a computerized
scanning device maps out the relative
amounts of the chemical in various
brain regions. For our purposes, ra-
dioactively labeled drugs that bind to
specific receptors or transporters are
used, and so it is possible to see the
extent to which binding occurs in the
living human brain. In case of dam-
age or death of dopamine neurons we
would expect to see less binding of the
radioactive dopamine-binding drug,
because of lower availability of dopa-
mine receptors.
In this way, PET provides an indi-
rect measure of toxicity in the living
human brain. This measure is indirect
because neuron toxicity cannot be de-
finitively determined in PET studies
alone. Additional procedures or tech-
niques are needed to definitively mea-
sure toxicity, because changes in radio-
active binding may reflect adaptation
(such as downregulation in tolerance
development) of the neuron and not
toxicity (see the opposite figure and
the one on the following page).
Studies of Methamphetamine Users
Many comparative studies have been
conducted over the past several de-
cades that assess how well observa-
tions of methamphetamine’s effects in
animals are consistent with those in
people, but these comparisons are of-
ten inconclusive about cause and ef-
fect because of their necessary experi-
mental design. In the human studies,
methamphetamine users and healthy
control participants are recruited. Then
each individual undergoes a single
brain scan and completes a cognitive
task battery. An advantage of this ap-
proach is that it allows researchers to
determine brain or cognitive differ-
ences between the groups at the time
of testing. Brain images, however, are
typically collected at only a single time
point for both groups of participants.
This methodology makes it virtually
impossible to determine whether meth-
amphetamine use (or any other drug
use) caused any observed differences,
because preexisting differences be-
tween the two groups cannot be ruled
out. Usually authors tend to adhere to
the brain-disease model of addiction
by attributing these differences to pro-
longed methamphetamine use, using
words such as “changes,” “reductions,”
“deterioration,” or “atrophy,” among
others. This implication is clearly inap-
propriate, because these words imply
a temporality that is not evaluated in
these cross-sectional studies.
A caveat to the above is that a few re-
cent studies have scanned participants’
brains over multiple time points. These
studies have focused on how these
neuromarkers differ between the meth-
amphetamine users who are able to
remain abstinent and those who keep
consuming. Since these findings do not
add new information in the compari-
son between methamphetamine users
and healthy controls, we have focused
primarily on studies when abstinent
methamphetamine users were com-
pared with control participants.
One consistent finding in such stud-
ies was lower striatal dopamine trans-
port density in methamphetamine us-
ers compared with control participants.
Because the striatum is a critical com-
ponent of the brain’s circuitry related
to reward and motivation, some have
argued that this difference is indicative
of a diseased brain following drug ad-
diction, while others caution that neural
differences do not necessarily equate
to deficits or pathology. Because acute
administration of methamphetamine
seems to increase motivation through
dopaminergic activity in the striatum,
a reduction in its transport would be
consistent with the development of tol-
erance to this particular effect.
Nearly 20 years ago, a study
showed that taxi drivers in London
had larger hippocampal volume than
control participants who were not taxi
drivers. No one would conclude, how-
ever, that either group was impaired or
pathological.
One of the most rigorous studies,
published in 2006, reporting dopamine
transport density differences between
the groups illustrates how specific
neural data might be over interpreted
in these studies. Chris-Ellyn Johanson,
now retired from Wayne State Uni-
versity, and her colleagues used PET
imaging procedures and a comprehen-
sive cognitive battery to assess brain
functioning. The researchers found
that dopamine binding potentials at a
specific area of the brain were 10 per-
cent to 15 percent lower in metham-
phetamine users.
There was considerable overlap,
however, in the binding potentials of
methamphetamine participants and
those of the control group. As is shown
in the figure on page 165, monoamine
binding potential values for some meth-
amphetamine users were equal to or
higher than those of some individuals
The existing paradigm is based on
intuition and political necessity, not on
data and useful clinical results.

164 American Scientist, Volume 106
in the control group, suggesting that the
clinical relevance of the observed statis-
tical differences might be limited.
In practical terms, that result means
that if these brain images were shuffled
into a single set, experts would not be
able to distinguish between the brains
of controls and those of abstinent us-
ers. Furthermore, if one subscribes to
the brain-disease theory, one would
anticipate that individuals with the
largest methamphetamine addiction
histories would have the lowest dopa-
mine transporter binding potentials.
This was not the case. Not only do im-
aging studies fail to show the expected
drug-related changes in the brain, but
clinical studies of addicts don’t show
the expected functional changes, either.
Some proponents of the brain-
disease theory have implied that as
more sensitive methodologies are de-
veloped, clear brain differences will be
revealed. For example, in the supple-
ment to her 2016 New England Journal of
Medicine paper, Volkow argued that the
“overlap [in binding potentials of drug
users and healthy controls] is likely
to reflect the limitation of currently
available brain imaging techniques.”
In other words, she suggests that our
faith in future technologies will be rein-
forced by the demonstration of meth-
amphetamine- or other drug- induced
neurotoxicity. It seems, much to our
chagrin, that some addiction- as-a-
brain-disease scientists are encourag-
ing faith-based thinking rather than
applying evidence- based methods to
inform drug addiction theories.
Nobel Prize winner and neuropa-
thologist and -anatomist Santiago
Ramón y Cajal perhaps stated it best
when he observed in 1906: “Present-
day science, in spite of its well-founded
conclusions, has not the right to foretell
the future. Our assertion can go no fur-
ther than the revelations of contempo-
rary methods.” The present data indi-
cate that we do not have the evidence
to support claims that the brains of ad-
dicted persons can be distinguished
from those of nonaddicted individuals.
Regarding the cognitive findings in
the study by Johanson and colleagues,
the performance of methamphetamine
users and control group participants
did not significantly differ on most
tasks. Methamphetamine users, how-
ever, performed more poorly than con-
trols on measurements of sustained at-
tention and immediate and long-term
memory. Still, the methamphetamine
users’ performance remained within
the normal range for their age and ed-
ucational group. In other words, the
methamphetamine users were cogni-
tively intact.
Although there are brain differences
between addicts and non addicts, there
are no data to support the idea that
these differences qualify as a brain
“disease” or even that some addictive
drugs lead to a loss of brain function.
Even though some drugs produce
neurotoxicity when administered at
high doses during long periods of
time, it is not possible to generalize
such effects to all drugs. To determine
clinical relevance, cognitive scores
should be compared against a norma-
tive database. Normative data, which
are obtained from a large, randomly
selected representative sample, incor-
porate important variables such as age
and education, and establish a baseline
distribution for a measurement. Un-
fortunately, this basic requirement is
often ignored; as a result, the addiction
literature is replete with a tendency to
interpret any difference as deficits rep-
resenting substantial loss of function.
This point is highlighted in the con-
clusions drawn from multiple studies
that compare the cognitive or brain
functioning of people with drug ad-
diction with that of nonaddicted con-
trol participants. For example, Sara L.
Simon and others at the University of
California, Los Angeles, warned the
following in a 2002 article in the Jour-
nal of Addictive Diseases:
“The national campaign against
drugs should incorporate infor-
mation about the cognitive defi-
cits associated with methamphet-
amine….Law enforcement officers
and treatment providers should be
aware that impairments in memo-
ry and in the ability to manipulate
information and change points of
view (set) underlie comprehen-
sion….Methamphetamine abusers
will not only have difficulty with
inferences…but…they also may
have comprehension deficits….
The cognitive impairment asso-
ciated with [methamphetamine
abuse] should be publicized.”
Such warnings were based on mea-
sures that revealed statistically signifi-
cant differences between methamphet-
amine users and controls, which alone
are insufficient to determine true cog-
nitive dysfunctions. Despite such inap-
propriately dire conclusions, findings
from the bulk of the brain- imaging
and cognitive literature assessing indi-
viduals who meet criteria for drug ad-
diction indicate that they are virtually
indistinguishable from their age- and
education-matched non-drug-using
counterparts. As methamphetamine
neurotoxicity in animals has been the
most widely demonstrated, the failure
to replicate these results in humans
leads us to conclude that it is a tremen-
dous stretch to argue that the scientific
data show that drug use causes brain
disease.
Higher dopamine receptor availability in PET
scans from nonaddicted individuals com-
pared with drug-addicted individuals can be
interpreted in several ways: It could indicate
drug-induced changes in brain function or
structure; reversible brain dysfunction caused
by drug use; irreversible brain damage caused
by drug use; or differences in ligand bind-
ing for some other reason. Even though such
comparisons are not conclusive about cause
and effect, most researchers intepret these dif-
ferences as changes that result from drug use.
CONTROL
cocainemethamphetaminealcoholheroin
DA D2 receptor availability
ADDICTED
Science Source/Science Source

2018 May–June 165www.americanscientist.org
An Influential
Unsupported Theory
Despite this empirical re-
ality, the diseased- brain
perspective has outsized
influence on research
funding and direction, as
well as on how drug use
and addiction are viewed
in society. For example, the
U.S. National Institutes of
Health recently initiated
a study that will cost tax-
payers more than $300 mil-
lion: The Adolescent Brain
Cognitive Development
longitudinal study primar-
ily seeks to gather neu-
roimaging data to better
understand the neural un-
derpinnings of drug addic-
tion among young people.
Researchers will follow
more than 10,000 people
between the ages of 9 and
10 for a decade, collecting their genetic
information and assessing their drug
use and academic achievement.
This research endeavor will give less
consideration to important social fac-
tors, such as parental income, neighbor-
hood, or family structure. This oversight
is to be expected, in part, because most
of the lead investigators are neuroim-
aging researchers. Notably, there has
never been such an ambitious funding
effort focused on psychosocial deter-
minants or consequences (for example,
employment status, racial discrimina-
tion, neighborhood characteristics, or
policing) of drug use or addiction.
To be clear, our goal here is not to
set up a dichotomy between biologi-
cal and social factors involved in drug
addiction, as if they are mutually ex-
clusive. They are not. In addition, we
recognize that many proponents of
the diseased-brain theory of addic-
tion habitually provide cursory and
pro forma statements attesting to the
importance of understanding the role
of psychosocial and environmental
factors in mediating drug addiction.
These statements are often unconvinc-
ing because they are not accompanied
with actions consistent with the claims.
Thus, our point is that there should
be greater parity in the funding of
drug addiction research and in how
drug addiction is viewed. The weight
of the evidence should drive research
direction and funding. Through re-
search published in 2016 that used
the National Comorbidity Survey–
Adolescent Supplement to study the
substance use and mental disorders of
more than 10,000 teenagers and also
through our research, we now know
that among the relatively small per-
centage of individuals who do become
addicted, co-occurring psychiatric dis-
orders and environmental and social
factors account for a substantial pro-
portion of these addictions.
This point indicates that far more
research resources and efforts should
be allocated to carefully studying these
factors. It also suggests avenues for ef-
fective substance-use disorder interven-
tions such as available mental health-
care and attractive alternatives such as
career opportunities or sports facilities.
The argument for a more pluralis-
tic view of drug addiction does not
exclude a role for neuroscience, as
long as data justify that role. At pres-
ent, the utility of explaining drug ad-
diction from an exclusive— or almost
exclusive— diseased-brain perspective
seems limited. A large proportion of
people who previously were addicted
managed to abstain without profession-
al help, according to a couple of 2011
studies by Catalina Lopez- Quintero,
Carlos Blanco, and their coauthors. For
other people struggling with addiction,
as noted above, the most effective treat-
ments are behaviorally based.
Despite the effectiveness of
evidenced- based behavioral therapies,
they are not widely used, according to
a 2016 review by Danielle
Davis of University of Ver-
mont and colleagues and
a 2014 review by Kathleen
M. Carroll of Yale Univer-
sity School of Medicine.
Even when medication-
assisted treatments are
used, such treatments are
far less effective without
adjunctive behavioral
therapies, as is argued in
a 2000 paper in Drug Safety
by James Bell and Deborah
Zador of the Langton Cen-
tre in Australia.
Viewing addiction as
a disease of the brain has
demonstrated, thus far,
limited utility for the de-
velopment of effective
strategies to deal with drug
addiction. Notably, the
most relevant biological
treatments, such as metha-
done, disulfiram, and others, were de-
veloped prior to the establishment of
the brain-disease model of addiction.
This Theory Promotes Social Injustice
Disproportionately viewing drug ad-
diction through the brain-disease lens
contributes to unrealistic, costly, and
harmful drug policies. If the real prob-
lem with drug addiction, for example,
is the interaction between the drug
itself and an individual’s brain, then
the solution to this problem lies in one
of two approaches: Either remove the
drug from society through policies and
law enforcement (for example, drug-
free societies), or focus exclusively
on the “addicted” individual’s brain
as the problem. In both cases, there
is neither genuine need for nor inter-
est in understanding the role of socio-
economic factors in maintaining drug
use or mediating drug addiction.
The entire removal of recreational
psychoactive substances from society is
both impractical and impossible. There
has never in history been a drug-free
society, and it is unlikely that there ever
will be one. In spite of this fact, law en-
forcement is charged with the unenvi-
able task of carrying out repressive
recreational drug-use policies that em-
phasize abstinence. Despite the claim
that viewing addiction as a brain disease
would lessen stigma and reduce drug-
related arrests, millions of people are ar-
rested annually for drug possession. In
the United States, for example, 2016 data
caudate
nucleus
anterior
putamen
posterior
putamen
dihydrotetrabenazine binding potential
4.5
4.0
3.5
3.0
2.5
2.0
1. 5
control group (n=16) methamphetamine group (n=15)
Clinical relevance of the observed statistical difference in monoamine
binding potential values between methamphetamine users and a control
group might be limited. If these brain images were shuffled into a single
set, experts would not be able to distinguish between the brains of con-
trols and those of abstinent users. (Graph adapted from Johanson, et al.
Psychopharmacology 186:620.)
Springer-Verlag 2006

166 American Scientist, Volume 106
from the U.S. Federal Bureau of Investi-
gation indicate there were 1.5 million an-
nual drug arrests, a number that hasn’t
appreciably changed since 1996.
Even worse, the abhorrent practice
of racism continues to flourish in drug
law enforcement. In the United States,
blacks are four times more likely than
whites to be arrested for cannabis pos-
session, even though both groups use
cannabis at similar rates.And more than
80 percent of those convicted of heroin
trafficking are black or Latino. There are
considerably more white heroin users
than black or Latino users, and most
drug users buy their drugs from dealers
within their own racial group.
An insidious assumption of the
diseased- brain theory is that any use
of certain drugs (for example, crack
cocaine, heroin, or methamphetamine)
is considered pathological, even the
nonproblematic, recreational use that
characterizes the experiences of the
overwhelming majority who partake of
these drugs. For example, in the 1980s
crack cocaine addiction was said to
occur after only one hit. Drug experts
with neuroscience leanings weighed
in. “The best way to reduce demand,”
Yale University psychiatry professor
Frank Gawin told Newsweek in 1986,
“would be to have God redesign
the human brain to change the way
cocaine reacts with certain neurons.”
For a more recent example, a popular
U.S. antidrug campaign implies that
one hit of methamphetamine is enough
to cause irrevocable brain damage
(http://www.methproject.org/ads/
tv/deep-end.html).
“Neuro” remarks made about drugs,
even though they have no foundation
in evidence, can be pernicious: They can
help shape an environment in which
there is an unwarranted and unrealistic
goal of eliminating certain types of drug
use at any cost to marginalized citizens.
In 1986, the U.S. Congress passed leg-
islation setting penalties that were 100
times harsher for crack than for powder
cocaine violations. More than 80 percent
of those sentenced for crack cocaine of-
fenses are black, despite the fact that the
majority of the drug’s users are white.
Even today, some politicians enact
misguided drug policy based on these
neuroexaggerations. The recent ac-
tions of Philippine President Rodrigo
Duterte represent but one example.
A year into his presidency, more than
4,000 people accused of using or sell-
ing illegal drugs have been killed. This
tragedy results from Duterte’s view
of methamphetamine’s effects on the
brain. Duterte justifies his actions by
stating that methamphetamine shrinks
the brains of users, and as a result, these
individuals are no longer capable of re-
habilitation. It is not difficult to see how
Duterte’s misguided belief that meth-
amphetamine causes brain damage
might have been shaped by exagger-
ated claims drawn by researchers who
tout that addiction is a brain disease.
A Behavioral Lens on Addiction
Many addiction researchers begin with
the assumption that this condition is a
brain disease. Yet there are virtually no
data in humans indicating that addic-
tion is a brain disease in the way that, for
instance, Huntington’s or Parkinson’s
are brain diseases. The present evidence
indicates that this assumption should
be reevaluated to formulate a more
accurate view of drug addiction. An
evidence- informed view would be more
inclusive, would emphasize a promi-
nent role for psycho social and environ-
mental factors, and would focus on of-
fering alternative reinforcers— nondrug
alternatives that decrease problematic
drug-taking. From a practical or clinical
perspective, this approach means it is
unacceptable to tell substance-use dis-
order patients that they suffer from a
diseased brain. Instead, a comprehen-
sive psychosocial assessment should be
employed, and the resulting findings
should dictate intervention strategies.
Research has now shown repeatedly
that alternative reinforcers can be used
effectively to treat substance-use dis-
orders. This kind of treatment is called
contingency management. The idea
comes from basic behaviorism: Our
actions are governed to a large extent
by what we are rewarded for in our
environment. These cause-and- effect
relationships, where a reward is de-
pendent (contingent) upon the person
either doing or (in the case of drugs)
not doing a particular behavior, can be
used to help change all types of habits.
A body of literature now shows that
providing alternative reinforcers im-
proves addiction treatment outcomes. It
is far more effective than using punitive
measures such as incarceration, which
often is less useful in the long run. Al-
though many people stop or at least re-
duce their drug use while incarcerated,
jail and prison don’t provide positive al-
ternatives to replace drug habits. When
heavy drug users return to their com-
munities, they are not better equipped to
find work and support themselves and
their families; instead, having a criminal
record and a gap in their résumé makes
finding work even harder.
baseline
study week
voucher
1 2 3 4 5 1 2 3 4 5 6 7 8 9 10 11 12
percent of subjects cocaine-abstinent
100
80
60
40
20
0
abstinence reinforcement group
signicant difference
control group
Elsevier Science Inc.
Many studies show that offering alternative reinforcers—nondrug alternatives that decrease
problematic drug-taking—is effective in treating substance-use disorders. This study fol-
lowed 37 patients who regularly used cocaine and were enrolled in a methadone counseling
program. One group received a voucher redeemable for retail items when they abstained from
cocaine; the control group received vouchers that were not contingent on their cocaine avoid-
ance. This and subsequent studies indicate that people with drug addictions can and do make
rational decisions, providing an argument against the notion that they have diseased brains.
(Higgins, S. T. 1997. The influence of alternate reinforcers on cocaine use and abuse: a brief
review. Pharmacology Biochemistry and Behavior 57:419–427.)

2018 May–June 167www.americanscientist.org
A 1993 study led by Stephen Higgins
of the University of Vermont randomly
assigned treatment- seeking cocaine users
to either contingency management plus
behavioral counseling or to a traditional
12-step-focused counseling treatment.
Patients in the contingency- management
arm of the study received vouchers for
merchandise whenever they had drug-
free urine. Fifty-eight percent of partici-
pants in the contingency-management
group completed the 24-week outpatient
treatment— compared with 11 percent in
the 12-step group. In terms of abstinence,
68 percent in the contingency-manage-
ment group achieved at least eight weeks
cocaine-free, versus just 11 percent in the
12-step group. And after the rewards are
stopped, people in contingency manage-
ment are no more likely to relapse than
other treatment graduates. Because more
people complete treatment with contin-
gency management, this outcome makes
for an overall reduction in relapse.
More than three dozen studies have
now been conducted on contingency
management—used in the treatment of
opioid, cocaine, alcohol, and multiple-
drug addiction. They show that contin-
gency management typically does bet-
ter than treatment without it—and that
larger, faster rewards are more effective
than smaller, less quickly received incen-
tives. This result is exactly what research
on other types of behavior predicts.
Recently, we have asked our col-
leagues who are proponents of the
brain-disease model of addiction to pro-
vide us with practical information that
might be helpful to patients. Invariably,
we are directed toward the future with
promises of new and effective therapies.
Multiple purported therapies (including
the cocaine vaccine, anticraving medica-
tions, genetic therapy, and many others)
have proven unsuccessful over the past
two decades. In general, the response to
such failure has not been to question the
theory guiding the search for treatments,
but instead to double down, making the
theory even more complex. Meanwhile,
addicted patients continue to suffer.
It is incumbent on us all to be on
guard against neuroexaggeration and
neurocentrism, because they have in-
appropriately and disproportionately
influenced our research funding priori-
ties and drug policies. The stakes are too
high, and the human cost is incalculable.
Bibliography
American Civil Liberties Union. 2013. The War
on Marijuana in Black and White, 1–190.
Accessed March 2018. https://www.aclu.
org/report/report-war-marijuana-black-
and-white.
Anthony, J. C., L. A. Warner, and R. C. Kessler.
1994. Comparative epidemiology of depen-
dence on tobacco, alcohol, controlled sub-
stances, and inhalants: Basic findings from
the National Comorbidity Survey. Experimen-
tal and Clinical Psychopharmacology 2:244–268.
Caldwell, J. A., and J. L. Caldwell. 2005. Fa-
tigue in military aviation: An overview of
U.S. military-approved pharmacological
countermeasures. Aviation Space and Envi-
ronmental Medicine 76(7 II):39–51.
Carroll, K. M. 2014. Lost in translation? Moving
contingency management and cognitive be-
havioral therapy into clinical practice. Annals
of the New York Academy of Sciences 1327:94–111.
Csete, J., et al. 2016. Public health and interna-
tional drug policy. The Lancet 387:1427–1480.
Davis, D. R., A. N. Kurti, J. M. Skelly, R. Redner,
T. J. White, and S. T. Higgins. 2016. A review
of the literature on contingency management
in the treatment of substance use disorders,
2009–2014. Preventive Medicine 92:36–46.
Dole, V. P., and M. Nyswander. 1966. A medi-
cal treatment for diacetylmorphine (heroin)
addiction. Journal of Occupational and Envi-
ronmental Medicine 8(4):241.
Hart, C. L. 2013. High Price: A Neuroscientist’s
Journey of Self-Discovery That Challenges Ev-
erything You Know About Drugs and Society.
New York: Harper-Collins.
Hart C. L., C. B. Marvin, R. Silver, and E. E.
Smith. 2012. Is cognitive functioning im-
paired in methamphetamine users? A critical
review. Neuropsychopharmacology 37:586–608.
Johanson, C.-E., et al. 2006. Cognitive func-
tion and nigrostriatal markers in abstinent
methamphetamine abusers. Psychopharma-
cology 185(3):327–338.
Leshner, A. I. 1997. Addiction is a brain dis-
ease, and it matters. Science 278(5335):45–47.
Maguire, E., et al. 2000. Navigation-related struc-
tural change in the hippocampi of taxi drivers.
Proceedings of the National Academy of Sciences
of the United States of America 97(8):4398–4403.
Simon, S. L., et al. 2002. A comparison of pat-
terns of methamphetamine and cocaine
use. Journal of Addictive Diseases 21(1):35–44.
U.S. Sentencing Commission. 2016. Quick Facts:
Crack cocaine trafficking. Accessed March
2018. https://www.ussc.gov/sites/default/
files/pdf/research-and-publications/quick-
facts/Crack_Cocaine_FY16.pdf.
Volkow, N. D., G. F. Koob, and A. T. McLel-
lan. 2016. Neurobiologic advances from the
brain disease model of addiction. The New
England Journal of Medicine 374(4):363–371.
For relevant Web links, consult this
issue of American Scientist Online:
www.amsci.org/magazine/issues/2018/
may-june
Brian Hubble
Despite the claim that viewing addiction as a brain disease would lessen stigma and reduce
drug-related arrests, in 2016 in the United States there were 1.5 million drug-related arrests, a
number that hasn’t appreciably changed since 1996. There has never been an ambitious fund-
ing effort focused on psychosocial determinants or consequences of drug use—for example,
employment status, racial discrimination, neighborhood characteristics, or policing.