Brain Change in Addiction as Learning, Not Disease

The new england journal of medicine

n engl j med 379;16 nejm.org October 18, 2018

1551

Review Article

uring the past 30 years, the assumption that addiction is a dis-

ease or pathology has crystallized into the “brain disease model of addic-

tion.”

This trend was driven by the convergence of 12-step thinking with

residential treatment approaches in the latter half of the 20th century,

the explo-

sion of neuroimaging technologies that began in the 1990s, and promotion by pro-

fessional organizations

and community groups.

According to the brain disease

model, addiction is a chronic disease brought about by changes in the brain sys-

tems that mediate the experience and anticipation of reward and in higher-order

systems that underlie judgment and cognitive control.

1,5

The proponents of the

model propose that these changes are driven by exposure to drugs of abuse or

alcohol, though links with behavioral addictions have also been explored.

The brain disease model is the most prevalent model of addiction in the western

world. Particularly in the United States, it dominates professional and public dis-

course on prevention, treatment, research agendas, and policy issues. Because the

disease model focuses on brain change, it has helped explain why persons with

addictions find it difficult to change their thoughts and behaviors quickly or easily.

Because it focuses on biologic factors rather than moral arguments, it has helped

reduce the stigma faced by those with addictions and their families, at least in

some respects. (See Table 1 for a broader discussion of stigma.) The brain disease

model has also legitimized the role of doctors and other medical professionals in

addiction treatment and driven research on new drugs to combat addiction, and it

has been used to advocate for access to treatment and care rather than segregation

and punishment.

These aims and outcomes are well intended, and they have been beneficial in

some contexts, but the narrow focus of the disease model on the neurobiologic

substrates of addiction has diverted attention (and research funding) from other

models.

Alternatives to the brain disease model often highlight the social and

environmental factors that contribute to addiction, as well as the learning pro-

cesses that translate these factors into negative outcomes.

11-15

For example, it has

been shown repeatedly that adverse experiences in childhood and adolescence in-

crease the probability of later addiction.

13,14

Also, exposure to physical, economic,

or psychological trauma greatly increases susceptibility to addiction.

14-17

Learning

models propose that addiction, though obviously disadvantageous, is a natural,

context-sensitive response to challenging environmental contingencies, not a dis-

ease.

18,19

Yet the brain disease model construes addictive learning in terms of patho-

logic brain changes triggered mainly by substance abuse. Learning models also

favor individual solutions for overcoming addiction, facilitated by cognitive modi-

fications and personal agency. (See Table 2 for a discussion of empowerment.)

Learning models can include multiple levels of analysis: societal, social, psycho-

logical, and biologic. According to experts both inside and outside the medical

From the University of Toronto, Toronto.

Address reprint requests to Dr. Lewis at

Klingelbeekseweg 24, 6812DH Arnhem, the

Netherlands, or at m . lewis@ psych . ru . nl.

N Engl J Med 2018;379:1551-60.

DOI: 10.1056/NEJMra1602872

Dan L. Longo, M.D., Editor

Brain Change in Addiction as Learning,

Not Disease

Marc Lewis, Ph.D.

n engl j med 379;16 nejm.org October 18, 2018

1552

The new england journal of medicine

field,

these levels of analysis should ideally be

integrated for a comprehensive understanding of

addiction. Unfortunately, however, the neural level

of analysis is almost always ignored by nondis-

ease models that emphasize learning. (Work by

Szalavitz is a notable exception.

) Rather than

ignore (or dispute) evidence of brain change in

addiction, the current learning model reinterprets

such evidence. Psychological change, develop-

ment, and indeed all learning involve brain

change. It is therefore unnecessary and perhaps

unreasonable for a learning model of addiction

to dismiss neural findings.

In this review, I examine addiction within a

learning framework, informed by classic and

contemporary cognitive principles, which can

incorporate the brain changes seen in addiction

without reference to pathology or disease. In do-

ing so, I hope to connect neurobiologic and en-

vironmental accounts to make sense of addiction

with a degree of depth and precision that could

not be achieved by either one alone. I also inter-

pret key neurocognitive findings from both learn-

ing and disease perspectives to highlight their

parallels as well as their disparities (Table 3).

Addiction as Learning

Psychologists have historically divided learning

into operant conditioning, by which animals

work to receive rewards predicted by specific

cues, and Pavlovian conditioning, by which ani-

mals respond automatically to the stimulus prop-

erties of cues themselves. Advances in cognitive

psychology reveal that learning also involves

planning, decision making, inhibitory control,

and strings of cues that eventually lead to pre-

dicted rewards. The contemporary view from

cognitive science has extended this understand-

ing with models of “embodied cognition,” which

propose that all cognitive activity (including

learning) results from iterative, self-perpetuat-

ing interactions (i.e., feedback) between the ani-

mal and the environment.

From this perspec-

tive, learning occurs when the animal’s neural

capacities become entrained with an environ-

mental context. Thus, learning is not just a re-

sponse to stimuli but active engagement with

meaningful aspects of the environment.

The brain disease model does not dismiss the

importance of learning but views this learning

as pathologic. Addictive behaviors are proposed

to begin as impulsive bids for highly motivating

rewards, consolidated through operant condi-

tioning, but to end up as automatic (Pavlovian)

responses that bypass intention, augmented by

a loss of inhibitory control and a capacity for

choice. This observation is consistent with mod-

els of “delay discounting,” which propose that

immediate payoffs are inflated in their perceived

value, whereas longer-term rewards are “dis-

counted” (devalued).

31,32

Psychologists view delay

discounting as an intrinsic cognitive bias, not

only in humans but in other mammals as well.

Yet delay discounting seems to be augmented in

addiction, with long-term rewards falling off the

radar almost entirely. “Dual process” models of

addiction may help to explain this phenomenon,

in that a cognitive “overseer” loses the capacity

to override impulsive choices.

Although none

Proponents of the brain disease model of addiction have consistently claimed

that the disease definition has major social benefits for people with addic-

tion. Before addiction was defined as a disease, it was mostly viewed as a

moral failure, and “addicts” were reviled as self-indulgent, weak, dirty, or

malicious. But if addiction is viewed as a disease (like any other disease),

then the behaviors of people with addiction should not be seen as their

fault. In this way, the disease model was proposed to reduce stigma, blame,

and the assumption that people with addiction should be punished or re-

moved from society. The disease model should be commended for even

partial success in achieving these humanitarian goals.

Yet the disease definition can replace one kind of stigma with another. The

notion of a mental illness or disease can hurt more than help those with

behavioral problems such as addiction, because it fuels discrimination

and alienation of another sort. The disease designation can reinforce the

belief that an inviolable or essentialist “badness” is built in and perma-

nent, resulting in a sense that one is fundamentally different from “normal”

people, with concomitant feelings of inferiority and shame.

The label can

also curtail attempts to improve one’s functioning without medical care.

Biogenetic explanations carry the implication that people with addictions

are not really trustworthy, now or in the future, because of a biologic pro-

clivity they cannot control.

Not only does this fuel one kind of stigmatiza-

tion; it also helps rationalize a long-standing policy of withholding employ-

ment benefits and positions of authority from anyone who has ever been

labeled an addict.

It is true that some people with addiction feel consoled by the disease label.

In fact, psychiatric classifications have provided people who have diverse

emotional and mental problems with a label and (sometimes) a hypotheti-

cal explanation for adversities that can otherwise seem indefinable, amor-

phous, and yet blameworthy. Distinct categories with concrete labels can

help provide closure, context, and even a sense of belonging (to a particu-

lar group).

Yet many people with addiction recoil from the disease label. Especially when

they are successful in galvanizing their willpower and rejigging their habits

(i.e., recovering), they often find it confusing and debilitating to be told they

are chronically ill. People with previous addictions (“recovered addicts”)

usually want to feel that they have developed beyond their addiction and

become better people as a result. Many would prefer respect for that

achievement over the pity bequeathed by the disease definition.

Table 1. Brain Disease Model and Stigma.

n engl j med 379;16 nejm.org October 18, 2018

1553

Brain Change in Addiction as Learning, Not Disease

of these learning mechanisms are necessarily

unique, the brain disease model of addiction

views the progression of decreasing control as a

reflection of pathologic brain changes.

Addiction neuroscience explores these brain

changes. The shift from impulsive (operant,

reward-driven) actions to compulsive (automatic,

Pavlovian) associations is a case in point. When

drug taking is found to be highly rewarding, the

ventral striatum (including the nucleus accum-

bens) focuses attention on the desired goal, acti-

vates a behavioral sequence to achieve that goal,

and produces a motivational urge to energize

that behavior.

Over time, however, as behavior

becomes more compulsive and less impulsive

(less reward-driven), activation increases in the

dorsal striatum, the region most associated with

automatic responses.

10,33,36,37

This progression is

thought to eradicate willpower,

because con-

scious choice is no longer driving the behavior.

The neurotransmitter dopamine has often

been the focus of neural models of addiction.

But dopamine has many functions, both in the

striatum and in the prefrontal cortex, depending

partly on the receptor type absorbing it. For the

purposes of this discussion, we can think of

dopamine as activating synaptic activity and,

over time, synaptic change, both in the ventral

and dorsal striatum and in the prefrontal cortex

(partly through its effect on glutamate transmis-

sion). The release of dopamine to these and

other systems is triggered by the perception of

cues paired with anticipated rewards (in the case

of operant learning) or with automatic responses

(in the case of Pavlovian conditioning). Yet dopa-

mine metabolism also responds to the experi-

ence of rewards, increasing when rewards ex-

ceed expectations and decreasing when they fall

short. Addiction neuroscientists highlight the

long-lasting sensitization of the dopamine sys-

tem to addictive rewards or the cues that predict

them, resulting in craving and narrowed atten-

tion

6,37

as well as the subsequent blunting of the

dopamine system over time.

Striatal systems engage in constant cross-talk

with regions of the prefrontal cortex. Prefrontal

activation (in the orbitofrontal cortex) determines

the attractiveness of potential rewards and also

(in the dorsolateral prefrontal cortex) the exer-

cise of judgment and perspective shifting. In fact,

disrupted activation of the lateral prefrontal cor-

tex has been shown to increase delay discount-

ing (i.e., the proportion of impulsive choices).

A key finding in support of the brain disease

model is that drug use reduces connectivity be-

tween the prefrontal cortex and striatum, and

long-term addiction corresponds with reduced

gray-matter density (synaptic loss) in several

prefrontal and related regions. Such changes are

hypothesized to underlie diminished capacities

for judgment and self-control, or “impaired re-

sponse inhibition,” in people with addictions.

5,40

According to the brain disease model, the

cognitive and neural changes characterizing ad-

Viewing addiction in terms of learning rather than disease may have direct

advantages for those who are struggling. If people think that their addic-

tion results from an underlying pathology, as implied by the brain disease

model, and that the pathology is chronic, as highlighted both by profes-

sional bodies and by the 12-step movement, then they are less likely to

believe they will ever be free of it, especially as a result of their personal ef-

forts.

This characterization of addiction flies in the face of research show-

ing that a majority of persons with addictions recover without professional

treatment.

In fact, addiction workers generally agree that personal mo-

tivation, a sense of empowerment, and belief in one’s own agency are the

most important psychological resources for overcoming addiction. These

qualities would seem peripheral rather than mandatory if addiction were

indeed a disease.

In response to this argument, proponents of the brain disease model have

pointed out that defining something as a disease does not exempt patients

from responsibility for self-care (e.g., making lifestyle choices that improve

their prognosis). There is some truth to this counterargument; a sense

of empowerment can bolster self-care for patients with various medical

problems.

Yet viewing oneself as a patient implies that one’s primary duty is to follow

the instructions of knowledgeable professionals rather than examine one’s

own motivations, beliefs, and intuitions. Taking on the role of a patient

may be especially counterproductive in institutional settings, where people

with addictions tend to offload responsibility to treatment staff.

More-

over, biogenetic explanations for psychological problems induce “prog-

nostic pessimism.”

People dealing with addiction will try to change only

that which they feel is within their power to change.

Thus, their own faith

in their recovery and the confidence of those around them are hampered

by the disease definition.

The choice of terminology suggests specific guidelines for treatment. If replac-

ing the disease nomenclature with an emphasis on motivation and self-

direction increases the probability of successful outcomes, then treatment

professionals (including doctors) should advise those seeking help that

they do not have a chronic disease. They should encourage people with ad-

diction not to strive for obedience to a set of rules or pharmaceutical sub-

stitutes (unless heroin use prioritizes the need for medication-assisted

treatment) but instead to seek counseling or psychotherapy to help them

organize and modify their own attentional and motivational habits. For ex-

ample, a psychotherapeutic technique called motivational interviewing has

been developed in which nonconfrontational counseling by the clinician

encourages increased awareness of one’s own motives, conscious choices

that are consistent with one’s long-term goals, and reduced ambivalence;

this approach is best known for its success in reducing substance use.

More conventional psychotherapies such as cognitive behavioral therapy

also show efficacy in overcoming addiction,

and cognitively oriented

group interventions such as Self-Management and Recovery Training

(SMART Recovery) are quickly gaining recognition.

Table 2. Learning Models and Empowerment.

n engl j med 379;16 nejm.org October 18, 2018

1554

The new england journal of medicine

diction are unique and pathologic. Some theories

highlight distinct phases or stages: drug taking

is driven by positive reinforcement at first, then

by negative reinforcement (underpinned by re-

duced dopamine signaling and blunted receptor

responses), and finally by the loss of prefrontal

control.

1,41

A closely related theory suggests that

addictive urges are increasingly driven by the

brain’s rebound from drug stimulation — an

“antireward” effect resulting from an overactive

stress-response system, dopamine blunting, and

physical withdrawal symptoms.

These theories

emphasize repeated episodes of negative reinforce-

ment (learning to avoid an aversive outcome) and

positive reinforcement, plus changes in neuro-

chemistry and circuitry.

But are the neurocognitive processes that give

rise to addiction actually pathologic, or are they

constituents of normal learning with detrimen-

tal consequences? To help resolve this question,

I examine four neurocognitive changes central to

brain disease models. The first is the hypothe-

sized shift from impulsive behavior mediated by

the ventral striatum to compulsive responses

mediated by the dorsal striatum.

The second

change, which also supports the presumption of

involuntary behavior, is a reduction in functional

and structural connectivity between the striatum

and prefrontal cortex.

1,5

The third change is in-

creased and enduring sensitivity (i.e., sensitiza-

tion) to cues predicting addictive rewards, under-

pinned by mesolimbic dopamine.

The fourth

change is a decrease in sensitivity, not only to

alternative rewards but even to addictive rewards

themselves.

I argue that these four neurocogni-

tive changes are not specific to addiction and do

not indicate a disease process.

Reinterpreting

the Neurocognitive Data

Role of Compulsive or Automatic Responses

According to the brain disease model, impulsive

drug seeking and use are linked with activation

of the ventral striatum or nucleus accumbens at

first, but these behaviors become compulsive and

automatic with activation of the dorsal striatum

over time.

35,43

Yet behavior generally becomes

more automatic with practice, as novelty is re-

placed by familiarity, and dorsal striatal (includ-

ing globus pallidus) involvement underlies this

automatization even in a simple finger-tapping

task.

As Everitt and Robbins, acknowledged

experts on the ventral-to-dorsal shift, state, “There

is nothing aberrant or unusual about devolving

behavioural control to a dorsal striatal S-R [stimu-

lus–response or Pavlovian] ‘habit’ mechanism.”

They assert that this shift is to be expected in

Disease Model Learning Model Evidence for Learning

Addiction is characterized by a shift from

impulsive to compulsive processing,

loss of free will, and a shift of activation

to dorsal striatum.

All behavioral habits devolve to stimulus–

response mechanisms; automatization

is a normal outcome of learning.

Dorsal striatal activation or behavioral automati-

zation is seen with practice of even simple

(e.g., motor) tasks; for people with addiction,

operant contingencies facilitate the choice to

abstain from using drugs.

Functional connectivity between striatum

and PFC is lost, with reduced synaptic

density in specific PFC regions.

When planning and decision making are

bypassed, PFC demand is reduced; ex-

tended plasticity is normal; underused

synapses may be pruned.

Immediate or valued rewards lead to increased

striatal activation and decreased dorsolateral

PFC activation and cognitive control; synaptic

density in the PFC has been shown to rebound

with recovery.

Sensitization to drug cues is increased

(and enduring), mediated by increased

mesolimbic dopamine uptake.

Sensitization to valued rewards is normal;

an ongoing need or desire leads to on-

going sensitization (e.g., love, attachment,

wealth acquisition, religious practice).

Motivated goal pursuit leads to increased dopa-

mine, cue sensitization, and learning; high

emotional salience facilitates lasting synaptic

alterations (e.g., after trauma).

Ongoing drug use leads to loss of receptor

availability or sensitivity and reduced

pleasure (dopaminergic blunting).

Adversity, trauma (with or without drug

use), isolation, and overstimulation

lead to reduced dopamine-receptor

response or pleasure.

Loss of social status or trauma leads to reduced

D2 or D3 receptor availability; high levels of

mating behavior, eating, engagement with

pornography, and Internet use lead to a hypo-

dopaminergic system.

* PFC denotes prefrontal cortex.

Table 3. Comparison of Claims Made by Disease and Learning Models of Addiction and Sample Evidence for Learning.*

n engl j med 379;16 nejm.org October 18, 2018

1555

Brain Change in Addiction as Learning, Not Disease

many aspects of our lives, including eating and

other habitual activities. “Automatisation of be-

haviour frees up cognitive processes,” Everitt and

Robbins continue, which explains why we can

talk, eat, and drive at the same time.

Not only is normal behavior partly automatic,

but also addictive behavior, even in its later stages,

remains partly operant (reward-driven).

Support-

ing evidence comes from numerous studies in

which the reward value of the addictive goal

(e.g., the amount of drug offered) shifts in rela-

tion to the reward value of an alternative goal

(e.g., money).

45-49

In fact, these studies show that

the probability of abstaining is proportional to

the relative reward value of the two choices; this

sensitivity to environmental contingencies is the

hallmark of operant learning. Contingency man-

agement programs, based on these principles,

have shown a consistent effect in the reduction

of drug use.

26,49

The ventral striatum continues

to be involved in reward seeking in later-stage

addiction, even when the dorsal striatum domi-

nates behavior control.

In sum, a combination

of deliberate and automatic neurobehavioral

mechanisms characterizes both addiction and

“normal” habitual behavior.

Loss of Prefrontal Connectivity

and Synaptic Pruning

Evidence of a functional and (in some studies)

structural disconnection between the prefrontal

cortex and striatum has been pivotal for defin-

ing addiction as a brain disease.

Unfortunately,

these findings come from cross-sectional, not

longitudinal, research, so some cortical differ-

ences must precede rather than follow addiction,

as acknowledged by the researchers. Yet even

cortical changes that arise from (or with) addic-

tive drug use need not be considered pathologic.

When skills become streamlined with prac-

tice, they no longer engage conscious, reflective,

or effortful control. In fact, higher-order cogni-

tion is unnecessary once behavior becomes habit-

ual, as any professional musician or athlete can

demonstrate. Also, rewards perceived as both

immediate and valuable often bypass cognitive

control, as seen in the reduction of planning,

decision making, and concomitant prefrontal in-

volvement when it comes to sex, gambling, and

eating fast food.

50-55

Research points to an inverse

correlation between striatal activation and dorso-

lateral prefrontal engagement, both in delay dis-

counting

and more generally in effortful reward

seeking.

But would this loss of functional con-

nectivity normally lead to structural changes?

Indeed, the elimination, or “pruning,” of under-

used synapses is considered a key mechanism of

learning.

57,58

Massive cortical pruning has tradi-

tionally been associated with adolescence,

when

most addictions develop. However, since pruning

makes the brain more efficient when new skills

are practiced and consolidated, it is now thought

to underpin learning over the lifespan.

57,6 0

Synaptic density in certain prefrontal regions

decreases with the duration of drug use, but a

contrasting increase in synaptic density (in simi-

lar but not identical regions) correlates with the

number of weeks of abstinence.

In studies using

functional magnetic resonance imaging (fMRI),

“cocaine-dependent” participants who became

abstinent no longer differed from controls with

respect to the activation of inhibitory control

networks in the prefrontal cortex or the perfor-

mance of motor-inhibition tasks.

Thus, reduc-

tions in prefrontal involvement and synaptic

density appear to be restricted to the period of

habitual drug use, which may be followed by a

period of synaptic growth when a new skill —

abstinence — is learned. This two-way street in

frontal neuroplasticity is consistent with evidence

that most people with addiction recover,

21,22

and

most of those who recover do so without treat-

ment.

21,63

This finding would seem to be impos-

sible if prefrontal changes were permanent and

therefore pathologic.

Sensitization to Cues

People with drug addiction are highly sensitive

to drug-related cues, even after they quit using

drugs. To account for this sensitization, the brain

disease model points to a sharp rise in mesolim-

bic (reward-related) dopamine uptake.

The moti-

vational drive provided by mesolimbic dopamine

is essential for survival, because it ensures that

we prioritize eating, social relationships, and pro-

creation. Addiction neuroscientists acknowledge

that the levels of cue-triggered dopamine seen in

addiction can parallel those related to “natural”

rewards.

Indeed, romantic relationships de-

pend on motivational dopamine uptake,

50,65

and

desire after romantic rejection matches the crav-

ing for cocaine.

Motivated pursuits (natural or

otherwise), including shopping, sports, religious

practice, wealth acquisition, gambling, binge eat-

n engl j med 379;16 nejm.org October 18, 2018

1556

The new england journal of medicine

ing, romantic love, and pornography, correspond

with cue sensitization and increased activation

of striatal dopamine.

35,50-55,64-66

Even a simple in-

crease in reward availability on a computer screen

is sufficient to increase mesolimbic dopamine,

with a concomitant increase in effort.

Proponents of the brain disease model empha-

size that cue sensitivity in addiction is not only

extreme but also prolonged, whereas cue sensitiv-

ity returns to normal levels in relation to natural

reinforcers, once the need has been met.

This

prolonged sensitization is seen as the cause of

relapse.

37,68

Yet prolonged sensitization also results

from normal learning of emotionally salient

associations, through synaptic alterations in re-

gions that process emotion, such as the amyg-

dala.

69,70

Stimuli associated with past triumphs or

traumas or even a once-loved song will reliably

trigger strong feelings. Because these cues refer

to still-meaningful experiences, dopamine uptake

remains adaptive (rather than pathologic) for

ongoing behavioral adaptations.

Perhaps the most parsimonious explanation

for enduring cue sensitivity is that, in addiction,

goal seeking remains unfulfilled. The drug or

activity that was pursued to satisfy emotional

needs may have lost its effect because of a short

duration of action, chemical tolerance, or habitu-

ation. The value of addictive rewards is always

determined by context, including both the strength

of aversive feelings and the effectiveness of drugs,

for example, in quelling them. Unresolved needs

can make drug taking relevant indefinitely.

Desensitization to Drug-Related

and Natural Rewards

In parallel with cue sensitization and increased

levels of dopamine release, there is an appar-

ently paradoxical decrease in sensitivity to alter-

native rewards and even to drugs themselves.

1,68

This reward desensitization is thought to con-

tribute to increasing drug consumption. Brain

disease models ascribe this blunting to the down-

regulation (reduced availability or responsiveness)

of dopamine receptors (e.g., D2 and D3 recep-

tors), a pathologic process that may be mani-

fested as tolerance or withdrawal effects.

37,42

Yet

many studies of addiction use psychostimulants

(e.g., cocaine and methamphetamine), serious-

ly confounding this observation.

The buildup

(e.g., delayed reuptake) of dopamine resulting

from psychostimulants may directly trigger a

chemical rebound effect, independent of addictive

learning.

But even if addictive learning results in dopa-

minergic blunting, it need not denote pathologic

brain change. Poverty, trauma, and diminished

social status reduce the availability of the D2 and

D3 dopamine receptors in humans and nonhu-

man primates.

In fact, a reduction in D2 or D3

receptor availability has been shown to corre-

spond with reduced social dominance or isola-

tion, driving drug or alcohol use as a means of

countering anxiety or distress.

73-76

As noted above,

early adversity and trauma are reliable predictors

of subsequent drug use.

13,16,17,77

However, social

adversity may also result from drug use itself.

Society responds to illicit drug use by excluding

or punishing users, which in turn leads to bro-

ken relationships and erosion of self-esteem.

Thus, social and psychological hardships may

result in dopaminergic blunting, which then en-

courages addictive activities, amplifying these

hardships.

Dopaminergic blunting can also result from

nondrug rewards. Mating behavior in rats reduces

dopamine output in mesolimbic dopamine cir-

cuitry, leading to “a hypodopaminergic system,”

and identical changes result from prolonged ex-

posure to opiates.

In addition, obesity has been

linked to reduced dopamine receptivity, with the

hypothetical explanation that dopaminergic blunt-

ing leads to increased food consumption.

79,80

Ex-

posure to other potentially habit-forming plea-

surable activities also leads to dopaminergic

blunting, as shown with pornography use

and

extensive Internet use.

Thus, it seems that dopa-

minergic blunting can result from frequent acti-

vation of the mesolimbic dopamine system by

any repetitive reward-seeking behavior rather than

by drug exposure itself. Kent Berridge, a renowned

addiction neuroscientist, views dopaminergic

suppression as a temporary effect of overstimu-

lation, which may result from drug addiction but

does not cause it.

Addiction as Organism–

Environment Entrainment

Most alternatives to the brain disease model

of addiction share the view that explanations of

addiction should include societal, social, and fa-

milial factors that predict drug misuse. The brain

disease model has acknowledged these factors,

n engl j med 379;16 nejm.org October 18, 2018

1557

Brain Change in Addiction as Learning, Not Disease

but its emphasis on brain pathology sidelines

their causal status and their relevance to preven-

tion and treatment efforts. Yet viewing addiction

solely as the product of environmental forces

tends to ignore the properties of the organism,

its nervous system, and its response proclivities.

A comprehensive, balanced model of addiction

needs to recognize that the organism and its

environment are connected at every level, from

perception to cognition to behavior, and interact

continuously as an open system.

I have presented arguments and evidence that

automatization, reduced neural flexibility, endur-

ing cue sensitization, and reward desensitization

are normal features of learning highly motivat-

ing, repetitive, and habitual behavioral patterns.

Thus, I dispute the idea that addiction is patho-

logic. Nevertheless, there is considerable poten-

tial for reconciliation between aspects of the brain

disease model and an environmental model of

addiction, given that both view a rigidified be-

havioral pattern as learned, and learned deeply.

Classic learning models have limited value for

this synthesis, since they view the learner as an

independent agent responding to a static environ-

ment. In contrast, principles of embodied cogni-

tion construe learning as a process of reciprocal

adjustments between the activities of the organ-

ism and meaningful features of the environment.

What is meaningful is assumed to be con-

strained by biologic antecedents and emerging

biologic sensitivities, as well as the stimulus

properties of the animal’s environment (i.e.,

features of the environment that afford or invite

specific actions, known as affordances).

For a young human, the range of potentially

meaningful environmental features can be vast,

at least until social, familial, and psychological

setbacks narrow it down to a small subset of

suboptimal rewards. For example, many children

grow up with an unpredictable, disengaged, or

violent parent. As adolescents, they may face

disruptions in education, employment, or rela-

tionships as a result of financial or other disad-

vantages. These persons tend to find increased

meaning in drugs that reduce stress or promote

feelings of security and well-being, especially

because these effects can be attained without

mediation by other people. As drug use pro-

gresses and becomes a more consistent focus of

attention and behavior, the properties of the in-

dividual and of the environment tend to become

synchronized through mutual adjustments. Be-

havioral outcomes continue to shape a social en-

vironment that progressively narrows behavioral

options. For example, the social environment may

become increasingly limited to people who can

supply drugs (dealers or doctors), people with

whom to take drugs, and “friends” who remain

apathetic and disengaged. Behavioral proclivities

will change accordingly. Besides the increasing

habit strength of drug pursuit itself, there is

likely to be increased lying to avoid rejection or

punishment, as well as disengagement from

romantic partners and family members, further

limiting the chance to feel connected and pro-

tected. These changes would be mediated by cog-

nitive modifications — changes in attentional

foci, belief systems, identity, and self-esteem — as

well as by immature habits of emotion regula-

tion (e.g., suppression or denial) more generally.

But how might this addiction spiral get

started? The embodied-cognition view encour-

ages us to look for biologic and environmental

vulnerabilities that amplify and reinforce each

other. The goal here is not to list organismic

(e.g., genetic) and environmental risk factors and

add them together, but instead to track the inter-

action of factors that reciprocally influence each

other. I suggest that the addiction spiral gets

started with early psychosocial adversity. First,

we already know that early adversity and trauma

are strong predictors of later addiction.

13,16,17,77

Second, developmental psychologists have shown

that early trauma (physical, emotional, or sexual)

leaves enduring effects on nervous system func-

tion, such as sympathetic or parasympathetic

overattunement (causing hyperreactivity or hypo-

reactivity), oversensitivity to threat based on ac-

celerated amygdala development, and hippocam-

pal damage resulting from excessive cortisol

levels. Third, in animal models, researchers have

pinpointed epigenetic changes (e.g., methylation

of a gene that tunes the glucocorticoid feedback

loop) that take place in utero or the first year of

life in response to inadequate nurturing. But

these neuropsychological insults do not emerge

in a vacuum. Both trauma and “stress methyla-

tion” can begin with overstressed parents and

even grandparents

83,84

in families challenged by

unemployment, marital discord, histories of abuse,

or alienation from the community, affecting the

stress response in childhood and throughout life.

From these beginnings, a narrowing spiral of

n engl j med 379;16 nejm.org October 18, 2018

1558

The new england journal of medicine

ineffective coregulation emerges between devel-

oping children and their caregivers, leading even-

tually to entrainment between drug seeking and

its environmental concomitants. From the Rat

Park studies of the 1970s and 1980s, in which

even addicted rats avoided ingesting morphine

when allowed to socialize and play,

to contem-

porary evidence of the adverse consequences of

socioeconomic fragmentation,

Bruce Alexander

has shown that addiction emerges universally

as a response to the disruption of normal social

interactions. Therefore, models of addiction pred-

icated on embodied cognition should focus on

environments in which social stressors affect

early neuropsychological development, as a gate-

way to ongoing reciprocal adjustments between

disadvantageous organismic adaptations and nar-

rowing environmental opportunities.

In summary, the embodied-cognition frame-

work can help model the interaction between

neurobiologic and social-environmental contrib-

utors to addiction. Addictive activities are deter-

mined neither solely by brain changes nor solely

by social conditions. Although they indeed result

from and contribute to brain changes, addictive

activities also feed back to the social environ-

ment, further narrowing what are often already

limited opportunities for well-being, which in

turn further narrows cognitive and neural flex-

ibility. It follows that the narrowing seen in ad-

diction takes place within the behavioral reper-

toire, the social surround, and the brain — all

at the same time. It also follows that growth

beyond addiction can be facilitated by improved

social support, extended behavioral opportuni-

ties, targeted pharmacologic interventions, or

some combination of these strategies.

Disclosure forms provided by the author are available with the

full text of this article at NEJM.org.

I thank Shaun Shelly (University of Pretoria, Department of

Family Medicine) for providing information and references sup-

porting the arguments reviewed here, as well as feedback on

previous versions of the manuscript.

References

1. Volkow ND, Koob GF, McLellan AT.

Neurobiologic advances from the brain

disease model of addiction. N Engl J Med

2016; 374: 363-71.

2. Fletcher AM. Inside rehab: the sur-

prising truth about addiction treatment

— and how to get help that works. New

York: Penguin Books, 2013.

3. American Society of Addiction Medi-

cine. Definition of addiction. April 12,

2011 (http://www .asam .org/ quality - practice/

definition - of - addiction).

4. Narcotics Anonymous World Servic-

es. What is addiction? Narcotics Anony-

mous Bulletin no. 16. 2017 (https://na .org/

?ID=bulletins - bull17 - r).

5. Goldstein RZ, Volkow ND. Drug ad-

diction and its underlying neurobiologi-

cal basis: neuroimaging evidence for the

involvement of the frontal cortex. Am J

Psychiatry 2002; 159: 1642-52.

6. Volkow ND, Baler RD, Goldstein RZ.

Addiction: pulling at the neural threads of

social behaviors. Neuron 2011; 69: 599-602.

7. Buchman D, Reiner PB. Stigma and

addiction: being and becoming. Am J Bio-

eth 2009; 9: 18-9.

8. Haslam N. Genetic essentialism, neu-

roessentialism, and stigma: commentary

on Dar-Nimrod and Heine (2011). Psychol

Bull 2011; 137: 819-24.

9. Kvaale EP, Haslam N, Gottdiener WH.

The ‘side effects’ of medicalization: a meta-

analytic review of how biogenetic expla-

nations affect stigma. Clin Psychol Rev

2013; 33: 782-94.

10. Hall W, Carter A, Forlini C. The brain

disease model of addiction: is it support-

ed by the evidence and has it delivered

on its promises? Lancet Psychiatry 2015; 2:

105-10.

11. Alexander BK. The globalization of

addiction: a study in the poverty of spirit.

Oxford, England: Oxford University Press,

2008.

12. Heilig M, Epstein DH, Nader MA,

Shaham Y. Time to connect: bringing so-

cial context into addiction neuroscience.

Nat Rev Neurosci 2016; 17: 592-9.

13. Dube SR, Felitti VJ, Dong M, Chap-

man DP, Giles WH, Anda RF. Childhood

abuse, neglect, and household dysfunction

and the risk of illicit drug use: the adverse

childhood experiences study. Pediatrics

2003; 111: 564-72.

14. Enoch MA. The role of early life stress

as a predictor for alcohol and drug depen-

dence. Psychopharmacology (Berl) 2011;

214: 17-31.

15. Buchanan J. Understanding problem-

atic drug use: a medical matter or a social

issue? Br J Community Justice 2006; 4: 47-61

(http://epubs .glyndwr .ac .uk/ siru/ 18).

16. Valentine K, Fraser S. Trauma, dam-

age and pleasure: rethinking problematic

drug use. Int J Drug Policy 2008; 19: 410-6.

17. McDevitt-Murphy ME, Murphy JG,

Monahan CM, Flood AM, Weathers FW.

Unique patterns of substance misuse asso-

ciated with PTSD, depression, and social

phobia. J Dual Diagn 2010; 6: 94-110.

18. Lewis M. Addiction and the brain: de-

velopment, not disease. Neuroethics 2017;

10: 7-18.

19. Heather N, Best D, Kawalek A, et al.

Challenging the brain disease model of

addiction: European launch of the Addic-

tion Theory Network. Addict Res Theory

2018; 26: 249-55.

20. Wiens TK, Walker LJ. The chronic dis-

ease concept of addiction: helpful or

harmful? Addict Res Theory 2015; 23: 309-

21.

21. Heyman GM. Quitting drugs: quanti-

tative and qualitative features. Annu Rev

Clin Psychol 2013; 9: 29-59.

22. Lopez-Quintero C, Hasin DS, de Los

Cobos JP, et al. Probability and predictors

of remission from life-time nicotine, alco-

hol, cannabis or cocaine dependence: re-

sults from the National Epidemiologic

Survey on Alcohol and Related Conditions.

Addiction 2011; 106: 657-69.

23. Bonnie RJ. Responsibility for addic-

tion. J Am Acad Psychiatry Law 2002; 30:

405-13.

24. Pickard H, Ward L. Responsibility with-

out blame: philosophical reflections on

clinical practice. In: Fulford KWM, Davies

M, Gipps RGT, et al., eds. The Oxford

handbook of philosophy and psychiatry.

Oxford, England: Oxford University Press,

2013 (http://oxfordhandbooks .com/ view/

10 .1093/ oxfordhb/ 9780199579563 .001

.0001/ oxfordhb - 9780199579563 - e - 066).

25. Dunn C, Deroo L, Rivara FP. The use

of brief interventions adapted from moti-

vational interviewing across behavioral

domains: a systematic review. Addiction

2001; 96: 1725-42.

26. Dutra L, Stathopoulou G, Basden SL,

n engl j med 379;16 nejm.org October 18, 2018

1559

Brain Change in Addiction as Learning, Not Disease

Leyro TM, Powers MB, Otto MW. A meta-

analytic review of psychosocial interven-

tions for substance use disorders. Am J

Psychiatry 2008; 165: 179-87.

27. Satel S, Lilienfeld SO. Addiction and

the brain-disease fallacy. Front Psychiatry

2014; 4: 141.

28. Szalavitz M. Squaring the circle: addic-

tion, disease and learning. Neuroethics

2017; 10: 83-6.

29. Cowart M. Embodied cognition. In-

ternet Encyclopedia of Philosophy, 2018

(https://www .iep .utm .edu/ embodcog/ ).

30. Wilson AD, Golonka S. Embodied cog-

nition is not what you think it is. Front

Psychol 2013; 4: 58.

31. Ainslie G. Précis of Breakdown of Will.

Behav Brain Sci 2005; 28: 635-50.

32. Bickel WK, Marsch LA. Toward a be-

havioral economic understanding of drug

dependence: delay discounting processes.

Addiction 2001; 96: 73-86.

33. Wiers RW, Bartholow BD, van den

Wildenberg E, et al. Automatic and con-

trolled processes and the development of

addictive behaviors in adolescents: a review

and a model. Pharmacol Biochem Behav

2007; 86: 263-83.

34. Stacy AW, Wiers RW. Implicit cogni-

tion and addiction: a tool for explaining

paradoxical behavior. Annu Rev Clin Psy-

chol 2010; 6: 551-75.

35. Everitt BJ, Robbins TW. From the ven-

tral to the dorsal striatum: devolving

views of their roles in drug addiction.

Neurosci Biobehav Rev 2013; 37: 9 Pt A:

1946-54.

36. Volkow ND, Fowler JS, Wang G-J,

Swanson JM, Telang F. Dopamine in drug

abuse and addiction: results of imaging

studies and treatment implications. Arch

Neurol 2007; 64: 1575-9.

37. Robinson TE, Berridge KC. The neural

basis of drug craving: an incentive-sensi-

tization theory of addiction. Brain Res

Brain Res Rev 1993; 18: 247-91.

38. Volkow ND. Addiction is a disease of

free will. Bethesda, MD: National Insti-

tute on Drug Abuse, 2015 (https://www

.drugabuse .gov/ about - nida/ noras - blog/ 2015/

06/ addiction - disease - free - will).

39. Figner B, Knoch D, Johnson EJ, et al.

Lateral prefrontal cortex and self-control

in intertemporal choice. Nat Neurosci

2010; 13: 538-9.

40. Volkow ND, Fowler JS. Addiction, a dis-

ease of compulsion and drive: involve-

ment of the orbitofrontal cortex. Cereb

Cortex 2000; 10: 318-25.

41. Volkow ND, Morales M. The brain on

drugs: from reward to addiction. Cell 2015;

162: 712-25.

42. Koob GF. Neurobiological substrates

for the dark side of compulsivity in addic-

tion. Neuropharmacology 2009; 56: Suppl 1:

18-31

43. Everitt BJ. Neural and psychological

mechanisms underlying compulsive drug

seeking habits and drug memories — in-

dications for novel treatments of addic-

tion. Eur J Neurosci 2014; 40: 2163-82.

44. Lehéricy S, Benali H, Van de Moortele

P-F, et al. Distinct basal ganglia territories

are engaged in early and advanced motor

sequence learning. Proc Natl Acad Sci U S A

2005; 102: 12566-71.

45. Heather N. Is the concept of compul-

sion useful in the explanation or descrip-

tion of addictive behaviour and experi-

ence? Addict Behav Rep 2017; 6: 15-38.

46. Cohen M, Liebson IA, Faillace LA. The

role of reinforcement contingencies in

chronic alcoholism: an experimental analy-

sis of one case. Behav Res Ther 1971; 9:

375-9.

47. Hart C. High price: a neuroscientist’s

journey of self-discovery that challenges

everything you know about drugs and so-

ciety. New York: Harper, 2013.

48. Hart CL, Haney M, Foltin RW, Fisch-

man MW. Alternative reinforcers differ-

entially modify cocaine self-administra-

tion by humans. Behav Pharmacol 2000;

11: 87-91.

49. Menza TW, Jameson DR, Hughes JP,

Colfax GN, Shoptaw S, Golden MR. Con-

tingency management to reduce metham-

phetamine use and sexual risk among

men who have sex with men: a random-

ized controlled trial. BMC Public Health

2010; 10: 774.

50. Fisher HE, Brown LL, Aron A, Strong

G, Mashek D. Reward, addiction, and emo-

tion regulation systems associated with

rejection in love. J Neurophysiol 2010; 104:

51-60.

51. Noori HR, Cosa Linan A, Spanagel R.

Largely overlapping neuronal substrates

of reactivity to drug, gambling, food and

sexual cues: a comprehensive meta-analy-

sis. Eur Neuropsychopharmacol 2016; 26:

1419-30.

52. Rømer Thomsen K, Fjorback LO,

Møller A, Lou HC. Applying incentive sen-

sitization models to behavioral addiction.

Neurosci Biobehav Rev 2014; 45: 343-9.

53. Joyner MA, Kim S, Gearhardt AN.

Investigating an incentive-sensitization

model of eating behavior: impact of a

simulated fast-food laboratory. Clin Psy-

chol Sci 2017; 5: 1014-26.

54. Balodis IM, Kober H, Worhunsky PD,

Stevens MC, Pearlson GD, Potenza MN.

Diminished frontostriatal activity during

processing of monetary rewards and loss-

es in pathological gambling. Biol Psychia-

try 2012; 71: 749-57.

55. Robinson MJF, Fischer AM, Ahuja A,

Lesser EN, Maniates H. Roles of “want-

ing” and “liking” in motivating behavior:

gambling, food, and drug addictions. Curr

Top Behav Neurosci 2016; 27: 105-36.

56. Westbrook A, Braver TS. Dopamine

does double duty in motivating cognitive

effort. Neuron 2016; 89: 695-710.

57. Craik FIM, Bialystok E. Cognition

through the lifespan: mechanisms of

change. Trends Cogn Sci 2006; 10: 131-8.

58. Petanjek Z, Judaš M, Šimic G, et al.

Extraordinary neoteny of synaptic spines

in the human prefrontal cortex. Proc Natl

Acad Sci U S A 2011; 108: 13281-6.

59. Selemon LD. A role for synaptic plas-

ticity in the adolescent development of

executive function. Transl Psychiatry 2013;

3: e238.

60. Segal I, Korkotian I, Murphy DD. Den-

dritic spine formation and pruning: com-

mon cellular mechanisms? Trends Neuro-

sci 2000; 23: 53-7.

61. Connolly CG, Bell RP, Foxe JJ, Gara-

van H. Dissociated grey matter changes

with prolonged addiction and extended

abstinence in cocaine users. PLoS One

2013; 8(3): e59645.

62. Bell RP, Garavan H, Foxe JJ. Neural

correlates of craving and impulsivity in

abstinent former cocaine users: towards

biomarkers of relapse risk. Neuropharma-

cology 2014; 85: 461-70.

63. Dawson DA, Grant BF, Stinson FS,

Chou PS. Maturing out of alcohol depen-

dence: the impact of transitional life

events. J Stud Alcohol 2006; 67: 195-203.

64. Olsen CM. Natural rewards, neuroplas-

ticity, and non-drug addictions. Neuro-

pharmacology 2011; 61: 1109-22.

65. Fisher HE, Xu X, Aron A, Brown LL.

Intense, passionate, romantic love: a nat-

ural addiction? How the fields that inves-

tigate romance and substance abuse can

inform each other. Front Psychol 2016; 7:

687.

66. Sussman S, Lisha N, Griffiths M.

Prevalence of the addictions: a problem of

the majority or the minority? Eval Health

Prof 2011; 34: 3-56.

67. Hamid AA, Pettibone JR, Mabrouk

OS, et al. Mesolimbic dopamine signals

the value of work. Nat Neurosci 2016; 19:

117-26.

68. Berridge KC. Is addiction a brain dis-

ease? Neuroethics 2017; 10: 29-33.

69. Phelps EA, Lempert KM, Sokol-Hess-

ner P. Emotion and decision making: mul-

tiple modulatory neural circuits. Annu

Rev Neurosci 2014; 37: 263-87.

70. Panno A, Lauriola M, Figner B. Emo-

tion regulation and risk taking: predict-

ing risky choice in deliberative decision

making. Cogn Emot 2013; 27: 326-34.

71. Nutt DJ, Lingford-Hughes A, Erritzoe

D, Stokes PRA. The dopamine theory of

addiction: 40 years of highs and lows. Nat

Rev Neurosci 2015; 16: 305-12.

72. Masten CL, Eisenberger NI, Borofsky

LA, et al. Neural correlates of social exclu-

sion during adolescence: understanding

the distress of peer rejection. Soc Cogn

Affect Neurosci 2009; 4: 143-57.

73. Czoty PW, Gage HD, Nader MA. Dif-

ferences in D2 dopamine receptor avail-

ability and reaction to novelty in socially

housed male monkeys during abstinence

n engl j med 379;16 nejm.org October 18, 2018

1560

Brain Change in Addiction as Learning, Not Disease

images

clinical

medicine

The Journal welcomes consideration of new submissions for Images in Clinical

Medicine. Instructions for authors and procedures for submissions can be found

on the Journal’s website at NEJM.org. At the discretion of the editor, images that

are accepted for publication may appear in the print version of the Journal,

the electronic version, or both.

from cocaine. Psychopharmacology (Berl)

2010; 208: 585-92.

74. Czoty PW, Nader MA. Effects of dopa-

mine D2/D3 receptor ligands on food-

cocaine choice in socially housed male

cynomolgus monkeys. J Pharmacol Exp

Ther 2013; 344: 329-38.

75. Morgan D, Grant KA, Gage HD, et al.

Social dominance in monkeys: dopamine

D2 receptors and cocaine self-administra-

tion. Nat Neurosci 2002; 5: 169-74.

76. Blum K, Chen ALC, Chen TJH, et al.

Putative targeting of dopamine D2 recep-

tor function in reward deficiency syn-

drome (RDS) by synaptamine complex

variant (KB220): clinical trial showing

anti-anxiety effects. Gene Ther Mol Biol

2009; 13: 214-30.

77. Stein MD, Conti MT, Kenney S, et al.

Adverse childhood experience effects on

opioid use initiation, injection drug use,

and overdose among persons with opioid

use disorder. Drug Alcohol Depend 2017;

179: 325-9.

78. Pitchers KK, Coppens CM, Beloate LN,

et al. Endogenous opioid-induced neuro-

plasticity of dopaminergic neurons in the

ventral tegmental area inf luences natural

and opiate reward. J Neurosci 2014; 34:

8825-36.

79. Volkow ND, Wang GJ, Telang F, et al.

Low dopamine striatal D2 receptors are

associated with prefrontal metabolism in

obese subjects: possible contributing fac-

tors. Neuroimage 2008; 42: 1537-43.

80. Stice E, Spoor S, Bohon C, Small DM.

Relation between obesity and blunted

striatal response to food is moderated by

TaqIA A1 allele. Science 2008; 322: 449-

52.

81. Kühn S, Gallinat J. Brain structure and

functional connectivity associated with

pornography consumption: the brain on

porn. JAMA Psychiatry 2014; 71: 827-34.

82. Kim SH, Baik SH, Park CS, Kim SJ,

Choi SW, Kim SE. Reduced striatal dopa-

mine D2 receptors in people with Internet

addiction. Neuroreport 2011; 22: 407-11.

83. Meaney MJ. Maternal care, gene ex-

pression, and the transmission of individ-

ual differences in stress reactivity across

generations. Annu Rev Neurosci 2001; 24:

1161-92.

84. Keating DP. Born anxious: the life-

long impact of early life adversity — and

how to break the cycle. New York: St.

Martin’s Press, 2017.

85. Alexander BK, Beyerstein BL, Hada-

way PF, Coambs RB. Effect of early and

later colony housing on oral ingestion of

morphine in rats. Pharmacol Biochem

Behav 1981; 15: 571-6.