The Neurobiology of Addiction: Implications for Voluntary Control of Behavior
Abstract and Keywords
This article focuses on addiction to examine the issues of decision-making and behavioral control and their ethical implications. Addicted people habitually engage in apparently voluntary behaviors, such as drug seeking and drug use, that are by standard definitions of addiction, compulsive or beyond the person's control. Addiction also provides a useful window because there has been made substantial progress toward understanding its neural mechanisms, even if somewhat divergent perspectives remain. For both ethical and technological reasons, most current experiments in human neurobiology, including many based on functional magnetic resonance imaging; yield correlative information rather than direct tests of causal mechanisms. The focus on compulsive drug use as the cardinal feature of addiction improves upon older views that had focused on dependence and withdrawal. The shaping of behavior to maximize future reward is dependent on the precise pattern of dopamine release. One of the most significant features of drug addiction discussed is its persistence.
A foundational assumption of most modern systems of morality and law is that human beings can voluntarily regulate their behavior in accord with freely chosen goals, and, when necessary, inhibit unbidden (prepotent) behavioral responses. In judicial proceedings there is a high bar for excusing illegal acts due to the perpetrator’s mental state—perhaps advanced dementia, some (but not all) cases of severe psychosis, or a powerfully compelling external force such as acting with a gun pointed at one’s head. It is widely understood that people are influenced by prior experience, by their biological makeup, and by the context in which they find themselves. Under most circumstances, however, such influences are not thought to extirpate free choice and behavioral control. Yet neurobiology is beginning to close the gap between diverse forms of influence and behavioral outputs with ever tighter mechanistic explanations. As a result, a central problem at the intersection of neuroscience with ethics, law, and policy is whether scientific progress has begun to erode the basis for believing that, for the most part, human behavior is under free and voluntary control, and thus whether new thinking is warranted concerning justifications for moral outrage, punishment, and policies directed toward diverse problematic behaviors. I believe that progress can be made on such ethical and policy implications of neuroscience independently of age-old philosophical discussions of free will and determinism. New findings concerning the neural mechanisms that underpin human behavior can be interpreted and applied independently of questions concerning chance and determinism in our universe. At a metaphysical level, neuroscience is less of a challenge to the concept of free will than that of an omnipotent and omniscient God, which had been at the heart of such discussions for centuries (p. 204) (Roskies 2006). What is currently at stake here is a more worldly set of issues related to the moral status of individuals who transgress social conventions and laws as well as significant questions about justice (Green and Cohen 2004).
In recent years, theories of decision-making and behavioral control derived from cognitive neuroscience and other branches of neurobiology have become increasingly prominent in policy debates. The influence of such theories can be seen in diverse discussions ranging from the possible utility of mildly paternalistic applications of behavioral economics (Thaler and Sunstein 2008), to debates over involuntary treatment for drug addiction (Caplan 2008; Hall et al. 2008), to disagreements about the moral underpinnings of retributive justice (Greene and Cohen 2004; Snead 2008). A salient example is found in arguments made in a US Supreme Court case, Roper v. Simmons (2005). The court found it unconstitutional to impose capital punishment if a defendant was under the age of 18 at the time of the crime because that would represent a cruel and unusual punishment that is impermissible under the 8th amendment to the Constitution. The American Psychological Association and the American Medical Association argued in influential amicus curiae briefs, based largely on neuroimaging evidence, that the adolescent prefrontal cortex is anatomically and functionally immature, thus limiting the capacity for decision-making and impulse control. As the broad sweep of these several examples suggest, it is timely to discuss the long-term implications of research on decision-making and behavioral control. It is also important, given the excitement that often surrounds new technologies and scientific findings, to forestall a premature embrace of interim conclusions.
In this chapter I have chosen addiction as the lens through which to examine the issues of decision-making and behavioral control and their ethical implications. I have chosen to focus on addiction for several reasons. Addicted people habitually engage in apparently voluntary behaviors, such as drug seeking and drug use, that are by standard definitions of addiction, compulsive or beyond the person’s control (World Health Organization 1992; American Psychiatric Association 2000). It is useful to consider what it means to engage in voluntary behaviors that one might not have intended or that one cannot control. In the case of putative behavioral addictions such as compulsive gambling, shopping, eating, or Internet use (Potenza 2006; Volkow et al. 2008), the question of what is meant by loss of control is even more pointed: lacking even the action of drugs, it is especially unlikely that such behavior is caused by a brain lesion (such as frontal lobe damage) or by biochemical toxicity. Such compulsions are more likely to result from the effects of normal brain mechanisms, such as experience-dependent neural plasticity, taken to an extreme.
Addiction also provides a useful window because there has been made substantial progress toward understanding its neural mechanisms, even if somewhat divergent perspectives remain (Berke and Hyman 2000; Everitt and Robbins 2005; Koob and LeMoal 2005; Hyman et al. 2006). At the same time, the topic of addiction directly raises significant ethical and policy issues related to autonomy and personal responsibility. If, for example, addicted people have lost control of their behavior, and thus put themselves and others at risk of harm, might involuntary treatment be ethically acceptable (Caplan 2008; Hall et al. 2008)? If loss of control is so severe as to be considered a symptom of a disorder or disease (World Health Organization 1992; American Psychiatric Association 2000), is punishment still justifiable for violations of the law, such as drug possession, that are direct consequences of the disorder? Do mechanistic understandings or acceptance of a disease model (p. 205) (Leshner 1997) diminish the moral opprobrium that has historically attached to addicted individuals? If so, is this a societally beneficial or damaging result (Satel 1999)?
Challenges to Neuroscience
It will prove extraordinarily difficult to discover the precise mechanisms by which human beings make choices, use those choices to regulate their behavior, and conform to laws and to social norms. Without claiming too much, progress is being made. The basis for this progress includes new technologies that make it possible to investigate the human brain at work and the integration of results from multiple subdisciplines of neuroscience. Relevant neurobiological approaches include invasive physiological experiments in animals (Sugrue et al. 2005; Kable and Glimcher 2009), neuropsychological analysis of humans who have suffered brain lesions (Wallis 2007), cognitive neuroscience (Koechlin and Hyafil 2007; Haggard 2008; Soon et al. 2008), pharmacology (Robbins and Arnsten 2009), and computational neuroscience (Gold and Shadlen 2007). Nonetheless significant obstacles remain, especially for those aspects of human brain function, such as “volition” or “will,” that are difficult to model in animals. Indeed agreed scientific definitions for slippery concepts such as “autonomy,” “volition,” and “will” remain elusive (Roskies 2010).
For both ethical and technological reasons, most current experiments in human neurobiology, including many based on functional magnetic resonance imaging (fMRI), yield correlative information rather than direct tests of causal mechanisms. That said, current methods have yielded many important new observations, and can test hypotheses concerning the spatiotemporal substrates of neural processes. For example, neuroimaging can determine which circuits are, or are not activated by a certain stimulus or cognitive task. Based on pattern analysis, it is possible, within significant constraints, to predict actions from brain imaging, even before the subject is conscious of his intentions (Haynes and Rees 2006; Norman et al. 2006). That said, until ethically acceptable experiments can be performed that directly and precisely examine causal mechanisms underlying cognition, we will have only limited understandings of human decision-making and behavioral control, among other aspects of cognition. Current tools to activate or inactivate human neural circuits, such as deep brain stimulation (DBS) (Mayberg et al. 2005) or repetitive transcranial magnetic stimulation (rTMS) (Knoch et al. 2009) are either limited to particular patient populations (DBS) for ethical reasons or to relatively broad swaths of cerebral cortex (rTMS) for technical reasons. In addition, DBS and rTMS may activate or inhibit circuits several synapses away from the pathways that are the direct targets of investigation. Pharmacologic tools to activate or block specific neurotransmitter or hormone receptors (Kosfeld et al. 2005) or other molecular targets are quite useful. Their precision is limited, however, by issues of selectivity, toxicity, and perhaps most importantly, by the distribution of the molecular target in the nervous system. Particular receptors have, as a general rule, been used by evolution for divergent functions on diverse cells and circuits. Thus it is likely that deep mechanistic understandings will only be achieved slowly and iteratively as new technologies emerge that can be applied safely to study the neural representations of human cognition and action.
(p. 206) Despite these current limitations, progress has been made in understanding how internally represented goals and both external and interoceptive stimuli regulate human behavior (Schultz et al. 1997; Miller and Cohen 2001; Montague et al. 2004; Wallis 2007). Much data converges on the idea that even healthy people under ordinary conditions have less control of their choices and actions than is generally believed based on introspection. Insight is also being gained into how neuropsychiatric disorders such as schizophrenia or attention deficit hyperactivity disorder degrade “top down” or “cognitive” control of behavior (Barch 2005; Gilbert and Sigman 2007; Vaidya et al. 2005). Perhaps the largest such body of research has focused on addiction. As described earlier, given that obtaining and taking drugs represents a series of voluntary acts, this research has not only contributed to understandings of pathogenesis and treatment, but has fueled a vibrant discussion of the degree to which addicted people can be treated as moral agents, responsible for their drug-taking behavior (Morse 2004, 2007; Hyman 2007; Caplan 2008; Hall et al. 2008).
Definition of Addiction
The core feature of current mainstream definitions of drug addiction (World Health Organization 1992; American Psychiatric Association 2000) is compulsive drug or substance use, despite serious negative consequences. Compulsive drug use means that the affected person cannot control use for a significant period of time, despite powerful reasons to do so, such as significant drug-related health problems, family disruption, threatened job loss, or arrest. The focus on compulsive drug use as the cardinal feature of addiction improves upon older views that had focused on dependence and withdrawal. This older view proved inadequate because some highly addictive drugs such as the psychostimulants, cocaine and amphetamine, may produce mild withdrawal symptoms if any. In addition, a focus on dependence and withdrawal fails to explain why late relapses may occur long after detoxification, and may be initiated by specific drug-associated cues (O’Brien et al. 1998; Hyman et al. 2006) or by stress (Koob 2008). Finally, the focus on compulsion leaves open the possibility that behavioral states, such as compulsive gambling, might plausibly share mechanisms with drug addiction. While some severe problem gamblers describe withdrawal-like symptoms, these are far from universal; what is most salient is loss of control characterized by continued gambling despite significant debt and other negative consequences (Potenza 2006).
Stages of Drug Use and Risk Factors
Addiction is, perforce, preceded by drug experimentation, followed by a period of variable length in which use becomes regular, but is not yet compulsive. During this period of regular use, tolerance, dependence, and withdrawal symptoms may occur. With continued regular use, a subset of individuals find that they can no longer cut back without feeling intense drug urges and discover that drug seeking and drug taking are now beyond their control. Not everyone who tries drugs, whether tobacco, cocaine, or heroin, ultimately becomes a (p. 207) regular user or becomes addicted. Of those who become addicted, some can cease drug use without outside help, perhaps because of a change in life circumstance or perhaps as an act of will. Some who eventually find their way to treatment respond rapidly. Yet many others continue to relapse despite many attempts at treatment and remain compulsive users for decades (Hser et al. 2001).
Risk factors both for initiating drug use and for becoming addicted include male sex (across countries and cultures), family history, and availability of drugs in the person’s environment and culture. Twin and adoption studies have demonstrated that familial risk is explained by genes rather than by shared environment. Twin studies consistently show higher rates of concordance for heavy drug use and addiction within monozygotic twin pairs than within dizygotic twin pairs (Tsuang et al. 1996; Merikangas et al. 1998). Adoption studies that have been performed in several Scandinavian countries and in the United States have focused mostly on alcoholism (Sigvardsson et al. 1996). These studies demonstrate that individuals adopted early in life resemble their biological rather than their adoptive parents with respect to patterns of alcohol use.
Although genes play a substantial role in vulnerability to addiction, few of the specific genetic variants that confer risk have been identified to date. Like all common neuropsychiatric disorders, addiction risk is highly genetically complex (Goldman et al. 2005); evidence from linkage and association studies is consistent with contributions to both risk and resilience from a very large number of allelic variants. Moreover, based on family and twin studies, there appear to be both shared and unshared genetic risk factors underlying a propensity to addiction in general, and underlying preferences for specific drugs.
It is often argued out that even if attributions of personal responsibility for drug use might be relaxed once a state of addiction has set in, individuals must be accounted responsible during earlier periods of experimentation and use. At one level this is straightforwardly the case. No matter what the temptations, the early stage drug user must have some notion of the risk, including illegality for many drugs, and should retain the capacity for impulse control. In some contexts, however, the case for simple attribution of responsibility and moral opprobrium becomes somewhat murkier. For example, within some peer cultures, for example in some colleges, approximately 80% of the students may use alcohol. After a period of years that may include periods of heavy drinking, the majority of these young people emerge as social drinkers. Perhaps one in ten of those in the initial cohort may find themselves addicted to alcohol. It would not be an easy argument to make that the majority, who do not become alcoholic, are morally superior to those who become alcoholic. They may have fewer risk factors or better luck than those who become addicted. While more of those who develop alcoholism may have alcoholic close relatives than those who settle into healthy patterns of social drinking, it is not at present possible to predict future alcoholism (Vaillant 1996).
Emotion, Motivation, and Brain Reward Circuitry
Emotions are critical to the survival of individuals and species. Constituting far more than subjective feelings, emotions are transient physiological, cognitive, and behavioral responses (p. 208) to potentially survival-relevant stimuli. Primary emotions can be crudely divided into two broad categories by their valence. Negative emotions, such as fear, anger, and disgust may be elicited by threat, present danger, pain, or foul tastes and smells. These result in avoidance or protective behaviors. Positive emotions may be elicited by food, drink, safety, comfort, or sexual opportunities and lead to approach and consummatory behaviors. Physiological responses to either threatening or life-enhancing stimuli may include increased arousal and activation of the sympathetic nervous system. Threat may also cause release of stress hormones such as corticotropin releasing hormone, adrenocorticotropin, and cortisol. Cognitive responses include alterations in attentional state, and significantly, memory formation. Learning to predict danger may save precious seconds on a future occasion that can be the margin between life and death. Learning the circumstances under which food, water, or safety can be obtained might, in a competitive world, be the difference, e.g. between eating and starvation. Thus experiences and predictive cues learned under conditions of strong negative or positive emotion, are learned rapidly (i.e. without need for much repetition), and are relatively resistant to forgetting. Neural circuits that have been highly conserved in evolution underlie these survival functions. A “fear circuit,” centered on the amygdala regulates responses to threats, and a “reward circuit” that involves dopamine-releasing neurons that project from the ventral tegmental area (VTA) of the midbrain to the nucleus accumbens (NAc), prefrontal cortex, and other forebrain structures, regulates the pursuit of positive goals.
Goals with positive survival value such as food, water, safety, and sexual opportunities act as “rewards” (Kelley and Berridge 2002). A simple operational definition of a reward is a stimulus that elicits approach and appetitive behaviors. Rewards are experienced as pleasurable, but more significantly from the point of view of survival, they are imbued with motivational or incentive properties. They are desired and activate physiological, cognitive, and behavioral responses that promote acquisition and consumption or consummation. Environmental cues that predict the availability of rewards also become imbued with motivational properties or “incentive salience” (Robinson and Berridge 2003). Such cues, like the rewards themselves, induce desire and activate responses aimed at obtaining the associated goal. Desire is intensified by motivational states of the organism, such as hunger, thirst, or perhaps, drug withdrawal symptoms. Behaviors required to obtain rewards tend to be repeated (i.e. they are reinforced), and to become automatic and highly efficient.
Natural rewards cause firing of VTA neurons and release of dopamine in the NAc and other forebrain regions. When dopamine action is blocked, whether by experimental lesions of dopamine neurons, blockade of post-synaptic dopamine receptors, or inhibition of dopamine synthesis, rewards no longer motivate approach behaviors. Dopamine release in the NAc plays the central role in binding reward-associated stimuli to reward-seeking responses including reinforcement. Dopamine release in the orbital prefrontal cortex is involved in updating of internal representations of rewards and assignment of relative values compared to other possible goals (Montague et al. 2004; Schoenbaum et al. 2006).
Dopamine neuron firing and dopamine release in response to rewards and predictive cues involves not only VTA neurons, which project to the NAc, prefrontal cortex, hippocampus, and amygdala, but also substantia nigra (SN) neurons, which project to the caudate and putamen. While dopamine is acting in the NAc to associate incentive salience with specific cues, it is acting in parallel in the caudate and putamen to consolidate programs of action aimed at efficiently obtaining rewards. Because reward seeking, a pleasurable experience if (p. 209) successful, tends to be repeated, associated motor programs become deeply ingrained (or overlearned) under the guidance of dopamine. Ultimately reward-seeking behaviors become automatic and come under the control of predictive cues. Under natural conditions, speed and efficiency in gaining food, water, and shelter improve the probability of survival. Over time responses to strong predictors of highly valued rewards can be characterized as stimulus-response habits (Everitt and Robbins 2005).
New insights into the role of dopamine have emerged from studies of patients with Parkinson’s disease (PD) (Dagher and Robbins 2009). PD results from the death of midbrain dopamine neurons. Neurons within the SN, which project to the caudate and putamen, are more severely affected than neurons within the VTA. L-DOPA, a dopamine precursor, is an effective neurotransmitter replacement therapy early in the illness. As the disease progresses, however, there are no longer enough SN neurons to take up the L-DOPA, convert it to dopamine, and release it in the caudate and putamen. Thus drugs that can directly bind postsynaptic dopamine receptors, such as selective D2 dopamine receptor agonists, may become necessary. A minority of patients who are treated with D2 dopamine receptor agonists, develop striking, new risky, goal directed behaviors such as compulsive gambling or compulsive shopping. These behaviors generally cease when the D2 agonist is withdrawn. It has been hypothesized that selective D2 dopamine receptor agonists act within the dopamine depleted caudate and putamen to produce therapeutic effects on motor behavior. However, when combined with dopamine from preserved VTA neurons, these drugs may overstimulate the NAc and other components of reward circuitry. These observations not only underscore the role of dopamine in motivation and reward seeking, but also are consistent with the idea that compulsive gambling and related behaviors are dependent on brain reward circuitry.
The Function of Dopamine
Current theories of dopamine action in the forebrain were initially based on electrophysiological recordings from midbrain dopamine neurons in monkeys (Schultz et al. 1997; Schultz 2006). Similar results have subsequently been obtained in human subjects using diverse rewards, including monetary rewards combined with functional magnetic resonance imaging (fMRI). Contrary to earlier ideas, dopamine does not function as the neural representation of pleasure; rather it serves as a learning signal in diverse forebrain circuits to shape behavior so as to maximize future success in obtaining rewards. Other neurotransmitters, perhaps endogenous opioid peptides, may act as hedonic signals (i.e. signaling pleasure). Additional evidence that dissociates dopamine from hedonic signaling is the action of nicotine, a substance that causes dopamine release, and is highly addictive, but does not produce significant euphoria of the sort produced by cocaine or heroin.
The shaping of behavior to maximize future reward is dependent on the precise pattern of dopamine release. Transient changes in dopamine neuron firing send a signal to the forebrain that there is a discrepancy between expectations and actual rewards. When the test animal (or test subject) is in a resting state, dopamine neurons exhibit a slow basal rate of firing, referred to as a “tonic” firing pattern. When a reward is encountered that is not expected or greater than expected (based on already learned cues), a transient or “phasic” (p. 210) burst of firing occurs, causing a transient increase in synaptic dopamine. Once a particular cue predictive of reward is fully learned, dopamine neurons produce a phasic burst when that cue appears unexpectedly, but produce no additional phasic bursts if the predicted reward appears at the time expected. If, however, the predicted reward is omitted at the time when it would have been expected, there is a pause in dopamine neuron firing. Finally, once a cue is fully learned, dopamine neurons stop responding to it if it is, in turn, predicted by a prior cue. Overall, dopamine neurons fire at the earliest reliable predictor of reward and thus influence behavior to maximize future consumption of rewards. Phasic increases in firing, and thus synaptic dopamine, signify that the world is better than expected, facilitate learning of new predictive information, and to bind the newly learned predictive cues to action. Pauses signify that the world is worse than expected.
Addictive drugs are chemically diverse and interact with different molecular targets in the nervous system (Nestler et al. 2009). Unlike natural rewards, addictive drugs have no homeostatic, reproductive, or other survival value. Given their chemical differences, it is not surprising that addictive drugs exert diverse physiological and behavioral effects. For example, cocaine and amphetamines are stimulants: they increase arousal, at lower doses they enhance cognitive performance, and at higher doses they may cause anxiety and insomnia. Alcohol, in contrast, is a depressant; it is anxiolytic at low doses, and degrades cognitive and motor performance. Despite their differences, addictive drugs share the pharmacologic property of releasing dopamine in the forebrain. They share the behavioral property of being able to cause compulsive use.
Addictive drugs can be likened to Trojan horses in the brain. All addictive drugs mimic one or another of the endogenous neurotransmitters and thus interact with neurotransmitter receptors, transporters, and other signaling proteins in the brain. Cocaine, for example, resembles dopamine in such a manner that it binds to—and blocks—the dopamine transporter (DAT), which normally clears dopamine from synapses, but does not interact with dopamine receptors. Because cocaine blocks the DAT, dopamine builds up to very high levels in synapses. Opiates, nicotine, alcohol, and cannabinoids act on different receptors in the brain, but by diverse mechanisms, all ultimately cause dopamine release (Nestler et al. 2009; Tang and Dani 2009). (Opiates and other drugs also influence reward by other mechanisms, but this is a level of detail beyond the scope of this chapter.) Psychotropic drugs, such as tricyclic or selective serotonin reuptake inhibitor (SSRI) antidepressants that do not release dopamine, are neither rewarding nor addictive.
Because of their direct pharmacologic action, addictive drugs always cause dopamine release and can cause a false reward prediction signal that cannot be corrected by experience: upon consumption, addictive drugs invariably signal that the world is better than expected, thus reinforcing further drug taking (Redish 2004). Moreover, drug-induced dopamine release masks any potential pauses in dopamine neuron firing, even when drug use proves less pleasurable than expected or even aversive. For example, when the inhalation of tobacco smoke causes painful coughing or shortness of breath in an ill smoker, it might seem that the brain would signal an experience that is worse than expected, with a resulting (p. 211) decrement in VTA neuron firing rate. Because, however, nicotine causes dopamine release pharmacologically, independent of the smoker’s actual experience, forebrain circuits, still receive a signal that reinforces tobacco use. Among other effects, this grossly abnormal dopamine signal acts in the orbital prefrontal cortex to value drug use above all other rewards (Montague et al. 2004), thus the life of the addicted person often becomes narrowed to obtaining, using, and recovering from drugs. These actions of addictive drugs within the reward circuit begin to explain why drug use continues despite negative consequences. Such mechanisms also are consistent with the notion that decision-making in addicted people is highly deranged. Instead of making choices freely, addicted individuals are powerfully influenced by a reward circuit that has been usurped by false (i.e. direct pharmacologic) signals. Because dopamine projections across the forebrain have the critical role, under normal circumstances, of directing and integrating a critical survival function, the maximization of future reward, addicted individuals finds drugs to be the chief objects of their desire and their most valued goal among all other goals. In addition to impairments in decision-making, addicted people are subject to abnormal prepotent behaviors. Cues that had previously been associated with drug use active automatic drug seeking (Berke and Hyman 2000; Everitt and Robbins 2005). If drug seeking cannot proceed to completion because of some obstacle or because the addicted person is attempting to cut down, intense drug craving is likely to result (Tiffany 1990).
Research that has compared individuals with established drug addiction to healthy control subjects has found impairments in cognitive control. Given tasks requiring cognitive control of thought or behavior, addicted people fare worse than healthy subjects, and, as ascertained by fMRI, fail to recruit their prefrontal cortex. These impairments are thought to reflect abnormalities in glutamatergic excitatory neurotransmission that develop late in the course of addiction (Kalivas and Volkow, 2005). The implication is that the ability to exert executive control over impulses is weakened just at the time when an addicted person is also experiencing powerful drives to seek and consume drugs. This research suggests that in addiction, not only is decision-making impaired as discussed earlier, but also the ability to control behavior is undermined resulting from the combination of subcortically-based drug-seeking with failures of top-down cortical control.
The Persistence of Addiction
One of the most significant features of drug addiction is its persistence. While addicted individuals may recover even after years of smoking, long periods of alcoholism, or regular heroin or cocaine use, a large fraction of individuals do not. Many profoundly addicted individuals derive only brief periods of respite during repeated episodes of treatment followed by relapse. The persistence of addiction is thought to reflect several long-lived biological processes. Some of the biological mechanisms that contribute to addiction may represent homeostatic responses to drug stimulation. These include well-documented alterations in the levels of expression of certain genes within the nervous system (Nestler et al. 2009). Some persistent drug-induced changes in gene expression are now thought to reflect epigenetic modifications of chromatin (Kumar et al. 2005), the histones, and other proteins that bind DNA in the cell nucleus and render genes either silent or available for transcription. (p. 212) Perhaps the longest-lived neural mechanism underlying addiction is the synaptic plasticity that is thought to be the substrate for associative memories. Associative memories are the key mechanisms by which specific drug-associated cues activate drug seeking and drug urges. By altering the strength of connections between neurons, drug-induced synaptic plasticity that has been documented in reward circuits, including long-term potentiation (LTP) and long-term depression (LTD), produce persistent changes in information processing (Hyman and Malenka 2001; Hyman et al. 2006). Physiologic processes such LTP and LTD are ultimately associated with alterations in the number of dendritic spines that are the critical substrates for synapse formation. The persistence of addiction is thought to reflect, in part, the drug-induced remodeling of the nervous system that results from synaptic plasticity.
Implications for Voluntary Control of Behavior
In the addicted state, neural mechanisms that evolved to motivate survival behaviors, such as the pursuit and consumption of food and water, the pursuit of safety, and of opportunities for mating, are usurped by the potent dopamine signals produced by addictive drugs. The result is a person who pathologically overvalues drugs, for whom drug cues activate drug seeking, whose impulse control is weakened, and for whom hard-won attempts to suppress drug-seeking may result in little more than intense drug craving. With respect to compulsive gambling, shopping, or other putative behavioral addictions, it must be acknowledged that far too little is known to draw firm conclusions. Early experiments, however, conducted mostly with gambling tasks that are conducive to study with fMRI, suggest that gambling situations cause dopamine release and activate brain reward circuitry (Clark 2009). Although the dopamine signals activated by gambling or other risky behaviors are not likely to be as reliable or as strong as those that occur with addictive drugs, the result may still plausibly be some degree of loss of control.
The view of addiction described here can explain how individuals continue to use drugs (and perhaps engage in other maladaptive behaviors) despite powerful health-related, social, legal, and economic disincentives, and why they remain at high risk of relapse even long after detoxification. This model helps explicate one of the questions posed at the beginning of this chapter: how can a protracted series of voluntary behaviors, such as finding money, seeking and buying drugs, preparing them, and then using them, be described as out of control? The answer lies in the remodeling of neural circuits under the influence of dopamine (and undoubtedly other neurotransmitters) so that drug associated cues come reliably to activate deeply ingrained programs of behavior (Everitt and Robbins 2005) in a person who also has impaired prefrontal cortical mechanisms of impulse control (Kalivas and Volkow 2005). This bleak picture notwithstanding, this model does not reduce addicted individuals to zombies permanently at the mercy of drug cues or stress (Hyman 2007). The function of reward circuits is to facilitate adaptive responses to external cues and bodily states (ranging from hunger and thirst to symptoms of drug withdrawal). In the addicted state, these responses are no longer adaptive, but perverted by the pathological dopamine signal (p. 213) produced by addicted drugs. Nonetheless, even the most severely addicted person has settings and times free of drug seeking, drug craving, or drug-related bodily sensations. True, even at such times and places, the addicted person’s system of valuation remains highly skewed toward drugs. Nevertheless these may represent windows of greater self-control during which treatment recommendations or other adaptive goals can be weighed more rationally than at other times when drug-related goals invariably win out. In such windows of at least modest lucidity, perhaps with a good measure of initial coercion, perhaps with family, friends, physicians, and employers acting as cognitive and emotional “prostheses” to aid in decision-making and to shore up damaged mechanisms of cognitive control, addicted individuals can commit to plans for detoxification and treatment. Because there may be many false starts, caregivers must be both patient and implacable. The caregiver role for a severely addicted person may be frustrating and exhausting, but in this role there is no place for blame or moral opprobrium that might drive the addicted person away. Help rejection and relapse cannot be accepted as a final answer, but an understanding of addiction makes help rejection and relapse understandable.
The study of addiction suggests that some apparently voluntary behaviors may not be as freely planned and executed as they first appear. While addicted individuals may not be zombies, their decision-making and behavioral control are undoubtedly severely impaired. Beyond the stance, described earlier, assumed by many experienced clinicians, of caring but implacable confrontation of maladaptive behaviors, the impairments that are central to addiction raise the question of whether an even more paternalistic approach might be appropriate. There is little question that for severely demented individuals, a significantly paternalistic approach to care is warranted. Such an approach might mandate treatment if it is safe and effective and would limit freedom in the interests of preventing self-harm. Unlike dementia, however, in which deficits are fixed and often progressive, addiction permits windows of relative lucidity that would seem, in Western societies, to make an excessive abrogation of freedom repugnant. Thus, I would argue, that treatment ethically demands the consent of an addicted person, even if it is known to be safe and effective. The need for consent is heightened when it comes to invasive treatments, such as implantation of a long-acting opiate antagonist drug such as naltrexone into the bodies of opiate addicts (Hulse et al. 2009). An important situation arises when an addicted person is convicted of a non-violent drug offense, such as possession. Much has been written about the ethics and efficacy of mandated treatment (whether behavioral or pharmacologic) in such contexts. Many ideas about mandated treatment have been implemented, often without adequate long-term information, in the growing number of drug courts in the US (Belenko 2001). This is a topic that warrants a long discussion. Here it must suffice to say that I believe it a highly defensible position, knowing what we do about addiction, to argue that a judge can ethically give a convicted offender a choice between a punishment, such as incarceration, and mandated treatment, but ethically, I believe that a choice must be offered.
Ideas emerging from cognitive neuroscience suggest that even healthy individuals exert far less control over their behavior than folk psychology recognizes or is obvious upon introspection. As a result, some scientists have argued that retribution makes little sense as a justification for punishment (Greene and Cohen 2004). Punishments such as incarceration could still be justified by the desire to incapacitate criminals, for deterrence, and for rehabilitation. For the addicted individual, with substantial impairment in decision-making and control of behavior, the justification for moral outrage (independent of criminal acts) and (p. 214) retribution would seem to be pressing questions. I agree with Morse (2004, 2007) that at this stage of knowledge, it is premature to use neuroscience as an excuse for crimes committed by addicted individuals. The open question, however, is whether the mechanistic explanations of neuroscience should cause society to retire moral outrage and retribution as justifications for punishment, at least of addicts who have committed nonviolent crimes. Strong arguments can be made that long mandatory sentences for non-violent drug crimes, likely driven in recent decades by anger and moral outrage among US legislators were neither just in proportion to punishments for other crimes, nor good policy with respect to cost and the goal of rehabilitation. The response to this historical error does not, however, mean that moral outrage and retribution have no place in justifications for punishment. Indeed, insofar as a goal of punishment is deterrence of future crimes, including non-violent drug-related crimes, a good argument can be made that moral outrage, if kept within proportion, is an important component of the instructive environment.
I have argued here, that for caregivers, a moralizing stance is neither useful nor appropriate with respect to addicted individuals. I would not argue, however, that a moral stance should be eschewed by the entire society as long as it is kept in proportion. Addicts are highly impaired, but, as I have stated, they are not zombies. If an important societal goal is to rehabilitate currently addicted people and to prevent harmful drug use and addiction in the future, it may be wise and ethically defensible for some components of society to retain a moral rather than a mechanistic, scientific stance concerning unwanted behaviors. Without demanding too much of addicted individuals, it may be wise to err, if only slightly, on the side of holding them responsible for their behavior, and to act as if they can exert at least somewhat more control than perhaps they can.
American Psychiatric Association (2000). Diagnostic and Statistical Manual of Mental Disorders,4th Edition, Text Revision. Washington, DC: American Psychiatric Association.Find this resource:
Barch, D.M. (2005). The cognitive neuroscience of schizophrenia. Annual Review of Clinical Psychology, 1, 321–53.Find this resource:
Belenko, S. (2001). Research on Drug Courts. A Critical Review. 2001 Update. New York: The National Center on Addiction and Substance Abuse at Columbia University.Find this resource:
Berke J.D. and Hyman, S.E. (2000). Addiction, dopamine, and the molecular mechanisms of memory. Neuron, 25, 515–32.Find this resource:
Caplan, A. (2008). Denying autonomy in order to create it: the paradox of forcing treatment upon addicts. Addiction, 103, 1919–21.Find this resource:
Clark L., Lawrence A.J., Astley-Jones F., and Gray, N. (2009). Gambling near-misses enhance motivation to gamble and recruit win-related brain circuitry. Neuron, 61, 481–90.Find this resource:
Dagher A. and Robbins T.W., (2009). Personality, addiction, dopamine: insights from Parkinson’s disease. Neuron, 61, 502–10.Find this resource:
Everitt, B.J. and Robbins, T.W. (2005). Neural systems of reinforcement for drug addiction: from actions to habits to compulsion. Nature Neuroscience, 8, 1481–9.Find this resource:
Gilbert, C.D. and Sigman, M. (2007). Brain states: Top-down influences in sensory processing. Neuron, 54, 677–96.Find this resource:
Gold, J.I. and Shadlen, M. N. (2007). The neural basis of decision making. Annual Review of Neuroscience, 30, 535–74.Find this resource:
(p. 215) Goldman, D., Oroszi, G., and Ducci, F. (2005). The genetics of addictions: uncovering the genes. Nature Reviews Genetics, 6, 521–32.Find this resource:
Greene, J. and Cohen, J. (2004). For the law, neuroscience changes nothing and everything. Philosophical Transactions of the Royal Society B. Biological Sciences, 359, 1775–85.Find this resource:
Greene, J.D., Nystrom, L.E., Engell, A.D., Darley, J.M., and Cohen, J.D. (2004). The neural bases of cognitive conflict and control of moral judgment. Neuron, 44, 389–400.Find this resource:
Haggard, P. (2008). Human volition: towards a neuroscience of will. Nature Reviews Neuroscience, 9, 934–46.Find this resource:
Hall, W., Capps, B., and Carter, A. (2008). The use of depot naltrexone under legal coercion: the case for caution. Addiction, 103, 1922–4.Find this resource:
Haynes, J-D. and Rees, G. (2006). Decoding mental states from brain activity in humans. Nature Reviews Neuroscience, 7, 523–34.Find this resource:
Hser, Y.I., Hoffman, V., Grella, C.E., and Anglin, M.D. (2001). A 33-year follow-up of narcotics addicts. Archives of General Psychiatry, 58, 503–8.Find this resource:
Hulse, G.K., Morris, N., Arnold-Reed, D., and Tait, R.J. (2009). Improving clinical outcomes in treating heroin dependence. Randomized, controlled trial of oral or implant naltrexone. Archives of General Psychiatry, 66, 1108–15.Find this resource:
Hyman, S.E. (2007). The neurobiology of addiction: Implications for voluntary control of behavior. American Journal of Bioethics, 7, 8–11.Find this resource:
Hyman, S.E. and Malenka, R.C. (2001). Addiction and the brain: the neurobiology of compulsion and its persistence. Nature Reviews Neuroscience, 2, 695–703.Find this resource:
Hyman, S.E., Malenka, R.C. and Nestler, E.J. (2006). Neural mechanisms of addiction: The role of reward-related learning and memory. Annual Review of Neuroscience, 21, 565–98.Find this resource:
Kable, J.W. and Glimcher, P.W. (2009). The neurobiology of decision: consensus and controversy. Neuron, 24, 733–45.Find this resource:
Kalivas, P.W. and Volkow, N.D. (2005). The neural basis of addiction: a pathology of motivation and choice. American Journal of Psychiatry, 162, 1403–13.Find this resource:
Kelley, A.E. and Berridge, K.C. (2002). The neuroscience of natural rewards: relevance to addictive drugs. The Journal of Neuroscience, 22, 3306–11.Find this resource:
Koechlin, E. and Hyafil, A. (2007). Anterior prefrontal function and the limits of human decision-making. Science, 318, 594–8.Find this resource:
Koob, G. (2008). A role for brain stress systems in addiction. Neuron, 59, 11–14.Find this resource:
Koob, G.F and Le Moal, M. (2005). Neurobiology of Addiction. New York: Academic Press.Find this resource:
Knoch, D., Schneider, F., Schunk, D., Hohmann, M. and Fehr, E. (2009). Disrupting the prefrontal cortex diminishes the human ability to build a good reputation. Proceedings of the National Academy of Sciences, 106, 20895–99.Find this resource:
Kosfeld, M., Heinrichs, M., Zak, P.J., Fischbacher, U. and Fehr, E. (2005). Oxytocin increases trust in humans. Nature, 435, 673–6.Find this resource:
Kumar, A., Choi, K.-H., Renthal, W., et al. (2005). Chromatin remodeling is a key mechanism underlying cocaine-induced plasticity in striatum. Neuron, 48, 303–14.Find this resource:
Leshner, A.I. (1997). Addiction is a brain disease, and it matters. Science, 278, 45–7.Find this resource:
Mayberg, H.S., Lozano, A.M., Voon, V., et al. (2005). Deep brain stimulation for treatment-resistant depression. Neuron, 45, 651–60.Find this resource:
Merikangas, K.R., Stolar, M., Stevens, D.E., et al. (1998). Familial transmission of substance use disorders. Archives of General Psychiatry, 55, 973–9.Find this resource:
Miller, E.K. and Cohen, J.D. (2001). An integrative theory of prefrontal cortex function. Annual Review of Neuroscience, 24, 167–202.Find this resource:
(p. 216) Montague, P.R., Hyman, S.E. and Cohen, J.D. (2004). Computational roles for dopamine in behavioural control. Nature, 431, 760–7.Find this resource:
Morse, S.J. (2004). Medicine and morals, craving and compulsion. Substance Use & Misuse, 39, 437–60.Find this resource:
Morse, S.J. (2007). Voluntary control of behavior and responsibility. The American Journal of Bioethics, 7, 12–14.Find this resource:
Nestler, E.J., Hyman, S.E. and Malenka, R.J. (2009). Molecular Neuropharmacology: Foundation for Clinical Neuroscience, Chapter 15. New York: McGraw-Hill.Find this resource:
Norman, K.S., Polyn, S.M., Detre, G.J. and Haxby, J.V. (2006). Beyond mind-reading: multi-voxel pattern analysis of fMRI data. Trends in Cognitive Sciences, 10, 424–30.Find this resource:
O’Brien, C.P., Childress, A.R., Ehrman, R. and Robbins, S.J. (1998). Conditioning factors in drug abuse: Can they explain compulsion? Journal of Psychopharmacology, 12, 15–22.Find this resource:
Potenza, M.N. (2006). Should addictive disorders include non-substance-related conditions? Addiction, 101(Suppl 1), 142–51.Find this resource:
Redish, A.D. (2004). Addiction as a computational process gone awry. Science, 306, 1944–7.Find this resource:
Robbins, T.W. and Arnsten, A.F.T. (2009). The neuropsychopharmacology of fronto-executive function: monoaminergic modulation. Annual Review of Neuroscience, 32, 267–87.Find this resource:
Robinson, T.E. and Berridge, K.C. (2003). Addiction. Annual Review of Psychology, 54, 25–53.Find this resource:
Roskies, A. (2006). Neuroscientific challenges to free will and responsibility. Trends in Cognitive Sciences, 10, 419–23.Find this resource:
Roskies, A. (2010). How does neuroscience affect our concept of volition? Annual Review of Neuroscience, 33, 109–30.Find this resource:
Satel, S.L., (1999). What should we expect from drug abusers? Psychiatric Services, 50, 861.Find this resource:
Schoenbaum, G., Roesch, M.R. and Stalnaker, T.A. (2006). Orbitofrontal cortex, decision-making and drug addiction. Trends in Neurosciences, 29, 116–24.Find this resource:
Schultz, W. (2006). Behavioral theories and the neurophysiology of reward. Annual Review of Psychology, 57, 87–115.Find this resource:
Schultz, W., Dayan, P. and Montague, P.R. (1997). A neural substrate of prediction and reward. Science, 275, 1593–9.Find this resource:
Sigvardsson, S., Bohman, M. and Cloninger, C.R. (1996). Replication of the Stockholm adoption study of alcoholism confirmatory cross-fostering analysis. Archives of General Psychiatry, 53, 681–7.Find this resource:
Snead, C. (2008). Neuroimaging and capital punishment. The New Atlantis, 19, 35–63.Find this resource:
Soon, C.S., Brass, M., Heinze, H-J. and Haynes, J-D. (2008). Nature Neuroscience, 11, 543–5.Find this resource:
Sugrue, L.P., Corrado, G.S. and Newsome, W.T. (2005). Choosing the greater of two goods: neural currencies for valuation and decision making. Nature Reviews Neuroscience, 6, 363–75.Find this resource:
Tang, J. and Dani, J.A. (2009). Dopamine enables in vivo synaptic plasticity associated with the addictive drug nicotine. Neuron, 63, 673–82.Find this resource:
Thaler, R.H. and Sunstein, C.R. (2008). Nudge: Improving Decisions about Health, Wealth, and Happiness. New Haven, CT: Yale University Press.Find this resource:
Tiffany, S.T. (1990). A cognitive model of drug urges and drug-use behavior: Role of automatic and nonautomatic processes. Psychological Review, 97, 147–68.Find this resource:
Tsuang, M.T., Lyons, M.J., Eisen, S.A., et al. (1996). Genetic influences on DSM-III-R drug abuse and dependence: a study of 3,372 twin pairs. American Journal of Medical Genetics, 67, 473–7.Find this resource:
(p. 217) Vaidya, C.J., Bunge, S.A., Dudukovic, N.M., Zalecki, C.A., Elliott, G.R. and Gabrieli, J.D. (2005). Altered neural substrates of cognitive control in childhood ADHD: Evidence from functional magnetic resonance imaging. American Journal of Psychiatry, 162, 1605–13.Find this resource:
Vaillant, G.E. (1996). A long-term follow-up of male alcohol abuse. Archives of General Psychiatry, 53, 243–9.Find this resource:
Volkow, D.D., Want, G.-J, Fowler, J.D. and Telang, F. (2008). Overlapping neuronal circuits in addiction and obesity: evidence of systems pathology. Philosophical Transactions of the Royal Society B. Biological Sciences, 363, 3191–200.Find this resource:
Wallis, J.D. (2007). Orbitofrontal cortex and its contribution to decision-making. Annual Review of Neuroscience, 30, 31–56.Find this resource:
World Health Organization (1992). The ICD-10 Classification of Mental and Behavioural Disorders. Geneva: World Health Organization. (p. 218) Find this resource: