By Ken Niemann
Addiction was ruled a disease by the American Medical Association (AMA) in 1976 and in 1997 the National Institute on Drug Abuse adopted the Brain Disease Model of Addiction (BDMA). More recently, however, there has been a rising tide of investigators who are skeptical of describing addiction as a “disease”. It will be argued here that such skepticism is warranted in favor of addiction being understood as a difficult choice. The argument will be framed by employing Bayesian Confirmation Theory.
The process will begin with a distinction between two different arguments from probability. The first argument is that in which the premises make the conclusion probable. We shall call this argument a P-inductive argument. Second, an argument in which the premises add to the probability of a conclusion, we shall call a C- inductive argument. We will first turn our attention to the C-inductive arguments given that they are easier to demonstrate and, afterwards, they will be collectively assessed for a possible P-inductive argument. It will be also be necessary to familiarize ourselves with Bayesian nomenclature. For example, P(p/q) is to be read “the probability that p is true given q” where p and q represent propositions (propositions are denoted by lower case letters). P(p/q)= 1, if and only if q makes p certain, a deductive argument. P(p/q) = 0, if and only if q makes ~p (not p) certain. It follows that if P(p/q) > 0.5 then, P(p/q) > P(~p/q). The hypothesis under consideration is represented as (h). For example, P(h/e.k) represents the probability of a hypothesis given evidence e and background information k. A correct C-inductive argument obtains if and only if P(h/e.k) > P(h/k) where P(h/k) is the prior probability of h. Further, P(e/h.k) represents the predictive power of h, and P(e/k) is the intrinsic probability of e. A correct P-inductive argument obtains if and only if P(h/e.k) > 0.5. Here, we will be considering whether two lines of evidence (e1 and e2) support the hypothesis (h) such that P(h/e1-e2. k) > P(h/k) and which obtains if and only if P(e/h.k) >P(e/~h.k). Finally, a cumulative case will be made such that P(h/ e1-e2.k) > 0.5.
Hypothesis (h) and background information (k)
The hypothesis (h) under consideration is “addiction is choice”. Conversely, (~h) would be “addiction is not a choice”. However, as we shall see, (~h) is also presented by many workers in the field as “addiction is a disease”. In either case, the law of the excluded middle employed here holds such that a statement and its negation cannot be both true at the same time and in the same sense. And, given that “addiction”, “choice”, and “disease” are decidedly pregnant words, the careful thinker will be sure not to equivocate on their meanings but, rather, precisely and fairly define each.
What it means to have a “choice” is a controversial but fundamental topic in the field of the Philosophy of Mind and Psychology’s endless search for veridical mental models. In general, those defending the libertarian agency viewpoint envision mental substance and physical substance as two distinct kinds of things. They argue that the physical world of nerve cells and their host of ions and biochemical transmitters cannot accommodate our understanding of intension, free will, what it means to have knowledge, etc. Only mental substance may do this.(Menuge, A., 2004, Locke, E. A., 1995) For example, Steven Slate opines that the epiphenomenalism viewpoint (a form of materialism) held by NIDA Director Nora Volkow is self-referentially refuting (i.e. cannot satisfy its own criteria for truth). (Philosophical Problem with the Brain Disease Model of Addiction: Epiphenomenalism, 2014, The Clean Slate) Physicalism or materialism in its various forms, on the other hand, holds that only the physical world exists and all events, including thoughts and behaviors, take place in a physically determined fashion. There is no meaningful idea of a choice freely made by a conscious agent; it’s a reductionist understanding of the world such that every phenomenon may ultimately be explained in the language of physics, chemistry and biology. U.C. Berkeley philosopher John Searle explains:
“Our conception of physical reality simply does not allow for radical freedom…In order for us to have radical freedom, it looks as if we have to postulate that inside each of us was a self that was capable of interfering with the causal order of nature. That is, it looks as if we would have to contain some entity that was capable of making molecules swerve from their paths. I do not know if such a view is even intelligible, but it’s certainly not consistent with what we know about how the world works from physics”. (Searle, J., 1984)
There exists a number of views on the continuum between the hard determinism represented by Searle and the soulish positions of the libertarian agency advocates. Though this author sympathizes with the latter, in order to avoid the seemingly intractable and highly contentious problem of defining exactly what a choice is, the criteria for a freely willed action will be adopted from neuroscientist Patrick Haggard:
“Two characteristic features seem to define freely willed actions. The first is the well-known ‘Could have done otherwise’ feature…The important point from a neuroscientific perspective is that the system supports a range of possible actions, rather than just one, and contains a mechanism that selects definite actions, and eliminates alternative possibilities…The second characteristic of freely chosen actions…implies a conscious subject, who initiates the action, and is therefore responsible for it. That is, there must be an ‘I’ who could have done otherwise.” (Swinburne, R. ed., Free will and modern science, 2011).
Haggard accepts the first characteristic but, based at least partly on the Libet type experiments, views the conscious ‘I’ as an epiphenomenal event that does not participate in the causal chain from brain event to action. For the sake of argument, then, it will be conceded that a choice means the individual “could have done otherwise” due to a complex set of causes. This criterion seems compatible with the Bayesian analysis in vogue among cognitive scientists. Here, the brain generates multiple predictions from memory given a certain context then scans the environment for information. The Gibbs Free Energy from biomolecules turn uncertainty into certainty “by creating patterns of functional connectivity in the brain” known as Bayesian updating. (Ridder, D., Verplaetse, J., & Vanneste, S.,2013). Severe addiction, then, would interfere with Bayesian updating processes.
The DSM-5 has eliminated the term “addiction” in favor of “substance use disorder” defined as “a cluster of cognitive, behavioral, and physiological symptoms indicating that the individual continues using the substance despite significant substance-related problems” and the diagnosis of which entails recognition of a “pathological pattern of behaviors related to use of the substance…impaired control, social impairment, risky use, and pharmacological criteria.” (DSM-5, 2013) While the DSM-5 acknowledges the intense craving of the addict and employs the term “compulsive” to describe drug seeking behavior in its most severe forms, the work skirts the issue on whether, in the most severe forms, substance use disorder precludes any meaningful choice on the part of the addict on whether to use or not use. The DSM-5 states
“Note that the word addiction is not applied as a diagnostic term in this classification, although it is in common usage in many countries to describe severe problems related to compulsive and habitual use of substances. The more neutral term substance use disorder is used to describe the wide range of the disorder, from a mild form to a severe state of chronically relapsing, compulsive drug taking. Some clinicians will choose to use the word addiction to describe more extreme presentations, but the word is omitted from the official DSM-5 substance use disorder diagnostic terminology because of its uncertain definition and its potentially negative connotation.” (DSM-5, 2013)
The term “compulsion’ is also defined on page 819 of the DSM-5 but the language remains soft on whether the addict is physically determined to use. Koob and Le Moal, both pioneers in the field, explain compulsivity as “persistent functional abnormalities in the prefrontal network involved in decision making and anticipation of reward.” (Koob, G., Le Moal, M.,2006, p367). That is, there is a loss of top down control over the dopaminergic reward system:
There exists an enormous body of literature on the neurobiological basis of compulsive substance abuse. However, and with some exception, authors in the field seem reticent to commit to the idea that the addict no longer has any choice. This is not to say the addict has merely diminished capacities to choose, but the complete loss of choice such that he or she is biologically determined toward substance abuse. Benjamin Hill points out a couple examples of such exception in Volkow and Burns:
“In this respect, such abstractions imply that “choices” are not really choices at all, because they are caused by imbalances and thus allow no possibilities. For example, Nina [sic.] Volkow (2005), the current director of the National Institute on Drug Abuse, asserts ‘… addicted individuals continue to be stigmatized by the pernicious yet enduring popular belief that their affliction stems from voluntary behavior.’ (p. 1430). Burns & Bechara (2007) support such an assertion by noting: ‘…We might conceptualize this as a ‘hijacking’ of the execution of willpower by an overactive impulsive system, where will becomes guided by the amygdala rather than by the prefrontal cortex. (pp. 263, 267, & 271)”
Indeed, stigmatization is often cited as a reason to consider addiction a disease, but the reverse is also true. A person diagnosed as a substance abuser is now understood to have a chronic, relapsing, medically defined disease which carries with it enormous disadvantages such as an indelible imprint on medical records, unfavorable bias in the insurance industry, an unfavorable impact on employment, and the learned helplessness that often follows from having such a disease. Let us now look at the two lines of evidence to possibly support the hypothesis that addiction is a choice
E1 Evidence from Brain Scans
Brain scans such as the Positron Emission Tomography (PET) scan pictured below are thought to support the disease model by its advocates, indicating structural differences in the addict brain as opposed to a normal healthy brain. (Volkow et al., 2001)
In this case, we see the PET scan revealing either dopamine terminal damage or downregulated dopamine receptors in the methamphetamine abuser which lead to pleasure deafness and subsequent drug seeking behavior to normalize . In a follow-up study, the following scan was produced:
The accompanying caption reads: “These images showing the density of dopamine transporters in a brain area called the striatum illustrate the brain’s remarkable potential to recover, at least partially, after a long abstinence from drugs—in this case, methamphetamine.” (Volkow et. al, 2001) What is even more remarkable, however, is that the meth abuser after 14 months of abstinence, was not medically treated but merely drug tested in a California drug-court monitoring rehabilitation program. This forces the question “How did the abuser manage to stay abstinent in the absence of medical treatment and while the dopamine receptors were downregulated?” That is, this diagnosed addict must have had some sort of top down control on whether to use or not despite having the remarkably different brain scan in the early months. Down regulated dopamine receptors do not seem to be a sufficient reason to override choice. Hall, Carter, and Forlini writing in The Lancet Psychiatry are also skeptical. The authors report that:
“Neuroimaging studies of addiction report more statistically significant differences between addicted and non-addicted persons than they should, given the small samples studied and the size of average differences between groups. The excess number of significant findings reflects capitalization on chance when performing large numbers of comparisons of activation between brain regions or structures, the selective publication of positive findings, and delays in publishing failures to replicate the positive findings. In studies that do find differences between cases and controls, there are large overlaps in the size of brain structures and “hypo-” or “hyperfunctionality” of specific brain regions between addicted and control groups. Neuroimaging researchers in addiction clearly acknowledge these limitations but more popular accounts often do not.” (Hall, W., Carter, A., & Forlini, C., 2015)
Heyman generally agrees:
“In support of this interpretation brain imaging studies often reveal differences between the brains of addicts and comparison groups (e.g., Volkow et al., 1997; Martin-Soelch et al., 2001) However, these studies are cross-sectional and the results are correlations. There are no published studies that establish a causal link between drug-induced neural adaptations and compulsive drug use or even a correlation between drug-induced neural changes and an increase in preference for an addictive drug…In principle then it is possible that the drug-induced neural changes play little or no role in the persistence of drug use… Thus, drug-induced neural plasticity does not prevent quitting.” (Heyman, G. (2013).
Moreover, Cognitive Behavioral Therapies are regarded as frontline treatments for substance abuse. (McHugh, R. K., Hearon, B. A., & Otto, M. W., 2010). Each of these therapies entail top down control training over drug seeking behavior. That is, these are reason or meaning based therapies yet reason is the very thing that is supposed to be compromised (or “hijacked” according to NIDA) due to structural brain changes as demonstrated by the scans. In other words, it is generally considered in addiction treatment that, through reflection on meaning and reason, the frontal cortex may be fortified to once again govern reward related behaviors – even in the presence of drug induced neuroplastic changes. This puts considerable pressure on the epiphenominalist viewpoint of Haggard, Volkow, and others who insist the conscious ‘I’ does not participate in the causal chain from brain event to action. Moreover, meaning and reason are not physical properties of nerve cells. Meaningfulness may not be viewed on a brain scan, measured, have its temperature taken, or weighed because it is not a physical property. To conflate such a value with a brain state is to confuse smoke with fire. Correlation or even constant conjunction does not entail identity. While acknowledging the expertise of the therapeutic community, why is it not possible for the addict, at least theoretically, to perform something akin to these therapies unassisted and reason their way out of substance abuse- even despite its admitted difficulty? Given this, and the fact that untreated addicts may remain clean despite their abnormal scans and drug induced neuroplastic changes, it is highly suggestive that the addict has a choice and could have acted otherwise. Therefore, evidence E1 adds to the probability that (h) is true and a correct C-inductive argument obtains such that P(h/e.k) > P(h/k).
E2 Evidence from Genetics
The argument for a genetically based, deterministic view of addiction succumbs to roughly the same the same criticisms as does that from neuroplastic changes in the brain. For example, Hall et al. argue that “Addiction is not a disorder confined to people who carry the small number of so-called addiction genes. A large number of alleles are involved in the genetic susceptibility to addiction and individually these alleles might very weakly predict a risk of addiction.” (Hall, W., Carter, A., & Forlini, C., 2015) That is to say, one does not have to have a given “addiction gene” to become addicted and one may possess the addiction gene without ever becoming addicted. This falls far short of any deterministic model of addiction and the best case that can be made for such is “susceptibility”. The study of monozygotic (identical) and dizygotic (fraternal) twins further rule out a genetic determinism to addiction; but just like other voluntary behaviors, genes can influence addiction behavior. (Heyman, G., 2010, p. 93) For example, Kendler’s study of 1198 male-male twin pairs (708 monozygotic and 490 dizygotic) at the Medical College of Virginia found that:
“if one member of a fraternal twin pair had been dependent, then there was about a 25% chance his co-twin was also dependent; whereas if one member of an identical twin pair were dependent, then there was a 40 percent chance that his identical brother was dependent…the correlation for addiction among identical twins was far less than 100 percent and that fewer than 20 percent of the biological sons of serious alcoholics became alcoholics themselves, even when their adoptive fathers were alcoholics.” (Heyman, G., 2010, p. 93)
If genes have a deterministic influence on addiction behavior, then we should expect nearly 100% concordance in monozygotic twins who have nearly identical genomes. Moreover, genes produce proteins and to a certain extent system architectures. To say that they determine addictive behaviors in adult human beings stretches credibility. The Baldwin Research Group explains:
“predisposition can only prove a difference in bodily processes, not a difference in thinking. ”Knowing the sequence of individual genes doesn’t tell you anything about the complexities of what life is,” said Dr. Brian Goodwin, a theoretical biologist at Schumacher College in Devon, England, and a member of the Santa Fe Institute in New Mexico. Goodwin goes on to explain single gene mutations are not accountable for, and cannot explain, complex behaviors. Genes produce proteins they do not guide behaviors. The truth is a predisposition for substance abuse, if it does exist, has no bearing on subsequent behaviors. Chemical processes do not make a person an alcoholic. The person makes the conscience choice.”(Alcoholism Is Not a Disease, n.d.)
Considerable hype at the popular level often clouds the issue for policy makers. For example, Blum and Noble publishing in JAMA in 1990 found a defective dopamine D2 receptor in 69% of an alcoholic population whereas the gene existed in only 20% of the nonalcoholic population. The finding was understood to be the discovery of the “alcoholic gene” by the popular press. However, as Stanton Peel points out:
“Several subsequent studies, however, failed to find such a remarkable occurrence of the gene defect in alcoholics. Two studies published in JAMA after Blum and Noble’s paper found little or no relationship between alcoholism and the D2 receptor gene. Other studies have found a relationship weaker than the one measured by Blum and Noble. A study reported in the October 2 issue of JAMA found the gene variation occurred in people with several disorders — including Tourette’s syndrome, hyperactivity, and autism — at least as often as alcoholics.” (Peele, S., 1992)
In finding that there is nothing deterministic about a genetic influence on addiction behavior, and that the genome is only one of two possible physiological etiologies, the evidence (E2) makes the hypothesis (h) more probable. That is, E2 adds to the probability that (h) is true and a correct C-inductive argument obtains such that P(h/e.k) > P(h/k).
The two possible sources of a deterministic (i.e. no possible choice) understanding of addiction (~h) were examined and found to be more likely to be not true than true. Therefore, P(e/h.k) >P(e/~h.k). P(h/ e1-e2.k) > 0.5 obtains given demonstrable natural remission rates mentioned above and the competitive reward studies found in Gene M. Heyman’s Addiction: A Disorder of Choice. (Heyman, 2010) Steven Slate explains the latter:
“He recounts studies in which cocaine abusers were given traditional addiction counseling, and also offered vouchers which they could trade in for modest rewards such as movie tickets or sports equipment – if they proved through urine tests that they were abstaining from drug use. In the early stages of the study, 70% of those in the voucher program remained abstinent, while only 20% stayed abstinent in the control group which didn’t receive the incentive of the vouchers. This demonstrates that substance use is not in fact compulsive or involuntary, but that it is a matter of choice, because these “addicts” when presented with a clear and immediately rewarding alternative to substance use and incentive not to use, chose it.” (Slate, S., n.d.)
Thus, trivial and delayed rewards can entice the addict to stay clean. This is highly suggestive of addiction being a choice. Further, priming doses of alcohol, cocaine, or methamphetamine do not necessarily lead to uncontrollable drinking or using. (Slate, S., n.d.) Natural remission rates also indicate addiction is a disorder of choice rather than a disease. “In the case of heroin addiction it was well over 80% in the two Robins studies, and for all opioids, it was 75% in the NESARC Study, and 60% in the ECA Study. In each of those studies, only a small minority of substance dependent people was treated – most recovered without treatment. More recent results from the NESARC data show that well over 90% of those categorized as heroin abusers or heroin dependent are currently “in remission.” (Slate, S., n.d., Wu et al., 2011(2))
Coupled with the positive evidence that addiction is indeed a choice given studies into competitive rewards and natural remission rates, neither genetic studies nor investigations into drug induced neuroplastic changes in the brain have produced any reason to think addiction is a disease, understood to be a physically determined behavior. Therefore, P(h/ e1-e2.k) > 0.5 and the hypothesis (h) that addiction is a choice is successful. Public policies aimed at addiction prevention, education, and treatment should consider this conclusion in their respective strategies.
Alcoholism Is Not a Disease. (n.d.) Retrieved December 12, 2015, from http://www.baldwinresearch.com/alcoholism.cfm#sthash.XHM9f3jK.dpuf
Alcoholism Is Not a Disease. (n.d.). Retrieved December 10, 2015, from http://www.baldwinresearch.com/alcoholism.cfm
Diagnostic and statistical manual of mental disorders: DSM-5. (5th ed.). (2013). Washington, D.C.: American Psychiatric Association
Drug War Statistics. (n.d.). Retrieved December 10, 2015, from http://www.drugpolicy.org/drug-war-statistics
Hall, W., Carter, A., & Forlini, C. (2015). The brain disease model of addiction: Is it supported by the evidence and has it delivered on its promises? The Lancet Psychiatry, 2(1), 105-110.
Heyman, G. (2010). Addiction: A disorder of choice. Cambridge, Mass.: Harvard University Press.
Heyman, G. (2013). Addiction and Choice: Theory and New Data. Frontiers in Psychiatry Front. Psychiatry, 4(Article 31). doi:10.3389/fpsyt.2013.00031
Hill, Wiley Benjamin III, “An Ontological Analysis of Mainstream Addiction Theories: Exploring Relational Alternatives” (2010). All Theses and Dissertations. Paper 2473.
Kendler KS, Karkowski LM, Neale MC, Prescott CA. Illicit Psychoactive Substance Use, Heavy Use, Abuse, and Dependence in a US Population-Based Sample of Male Twins. Arch Gen Psychiatry. 2000;57(3):261-269. doi:10.1001/archpsyc.57.3.261.
Koob, G., & Le Moal, M. (2006). Neurobiology of Addiction (p. 367). London: Elsevier Academic.
Leshner, A.I.(1997) Addiction is a brain disease and it matters. Science, 278, 45.
Locke, E. A. (1995). Beyond determinism and materialism, or isn’t it time we took consciousness seriously? Journal of Behavior Therapy and Experimental Psychiatry, 26(3), 265–273.
Lopez-Quintero C., Hasin D.S., Pérez de los Cobos J. et al. Probability and predictors of remission from life-time nicotine, alcohol, cannabis or cocaine dependence: results from the National Epidemiologic Survey on Alcohol and Related Conditions. Addiction: 2011, 106(3), p. 657–669, http://findings.org.uk/count/downloads/download.php?file=Lopez_Quintero_C_1.txt
McKay, J., & Hiller-Sturmhöfel, S. (n.d.). Treating Alcoholism as a Chronic Disease: Approaches to Long-Term Continuing Care. Retrieved December 12, 2015, from http://pubs.niaaa.nih.gov/publications/arh334/356-370.htm
Menuge, A. (2004). Agents under fire: Materialism and the rationality of science. Lanham, Md.: Rowman & Littlefield.
McHugh, R. K., Hearon, B. A., & Otto, M. W. (2010). Cognitive-Behavioral Therapy for Substance Use Disorders. The Psychiatric Clinics of North America, 33(3), 511–525. http://doi.org/10.1016/j.psc.2010.04.012
NIDA Drug Facts: Understanding Drug Abuse and Addiction. (2012). Retrieved December 5, 2015, from https://www.drugabuse.gov/publications/drugfacts/understanding-drug-abuse-addiction
NIDA Prescription and Over-the-Counter Medications. Retrieved from http://www.drugabuse.gov/publications/drugfacts/prescription-over-counter-medications on November 29, 2015
Peele, S. (1986), The implications and limitations of genetic models of alcoholism and other addictions. Journal of Studies on Alcohol, 47, 63- 73.
Peele, S. (1992, March 1). The Stanton Peele Addiction Website. Retrieved December 13, 2015, from http://www.peele.net/lib/blumrev.html
Ridder, D., Verplaetse, J., & Vanneste, S. (2013). The predictive brain and the “free will” illusion. Frontiers in Psychology Front. Psychol., 4. doi:10.3389/fpsyg2013.00131
Rosner S et al. Opioid antagonists for alcohol dependence. Cochrane Reviews, 2010, Issue 12. Article No. CD001867. DOI:10.1002/14651858. http://www.cochraneprimarycare.org/pearls/naltrexone-effective-alcohol-dependence
Searle, J. (1984). Minds, brains, and science (p. 92). Cambridge, Mass.: Harvard University Press.
Slate, S., Philosophical Problem with the Brain Disease Model of Addiction: Epiphenomenalism – The Clean Slate Addiction Site. (2014, April 4). Retrieved December 2, 2015, from http://www.thecleanslate.org/philosophical-brain-disease-addiction-epiphenomenalism/
Slate, S. (n.d.). Addiction is NOT a Brain Disease, It is a Choice – The Clean Slate Addiction Site. Retrieved December 10, 2015, from http://www.thecleanslate.org/myths/addiction-is-not-a-brain-disease-it-is-a-choice/
Swinburne, R. ed. (2011). Free will and modern science. Oxford: Oxford University Press.
Treatment and Recovery. (n.d.). Retrieved December 6, 2015, from http://www.drugabuse.gov/publications/drugs-brains-behavior-science-addiction/treatment-recovery
Treatment Doesn’t Work. (2004). Retrieved December 11, 2015, from http://www.soberforever.net/currenttreatdoesnt.cfm
West, R. (2015). Using Bayesian analysis for hypothesis testing in addiction science. Addiction. doi:10.1111
Wu, L.-T., Woody, G. E., Yang, C., Mannelli, P., & Blazer, D. G. (2011). Differences in onset and abuse/dependence episodes between prescription opioids and heroin: results from the National Epidemiologic Survey on Alcohol and Related Conditions. Substance abuse and rehabilitation, 2011(2), 77–88. doi:10.2147/SAR.S18969
Volkow ND, Chang L, Wang GJ, Fowler JS, Leonido-Yee M, Franceschi D, Sedler MJ, Gatley SJ, Hitzemann R, Ding YS, Logan J, Wong C, Miller EN. Association of dopamine transporter reduction with psychomotor impairment in methamphetamine abusers. Am J Psychiatry 158(3):377-382, 2001.
Volkow ND, Chang L, Wang GJ, Fowler JS, Franceschi D, Sedler M, Gatley SJ, Miller E, Hitzemann R, Ding YS, Logan J. Loss of dopamine transporters in methamphetamine abusers recovers with protracted abstinence. J Neurosci 21(23):9414-9418, 2001