As Müller & Schumann (M&S) refreshingly remind us, most people who regularly use psychoactive drugs, even highly addictive ones such as cocaine, do not go on to develop addiction. Only a minority of regular drug users eventually become addicted (Anthony Reference Anthony, Davis, Charney, Coyle and Nemeroff2002). A long-standing question in addiction research, with critical consequences for prevention and policy, is how we should interpret this individual variation, particularly the high rate of non-addictive drug use. Do environmental circumstances (e.g., economic constraints, societal regulations, cultural norms) prevent people from exposing themselves sufficiently to drugs to cross the “threshold of addiction” (Benowitz & Henningfield Reference Benowitz and Henningfield1994)? Or, alternatively, are most drug users somehow biologically resilient to addiction (e.g., genetically resistant to addiction regardless of drug exposure)? Epidemiology of drug addiction alone has been so far unable to resolve this apparent dilemma. This probably explains the prevailing default view in drug prevention and policy that considers each one of us a potential addict. Given sufficient drug exposure, in this view, each one of us could be turned into a drug addict. A biological resilience to drug addiction would not exist, or such resilience would exist only in rare individuals.
Overall, the framework of M&S – which echoes, without citing it, the “mental tool” hypothesis of psychoactive drug use that Michael Pollan initially formulated in his Botany of Desire (Pollan Reference Pollan2001) – seems to support the notion of a biological resilience to drug addiction and even tentatively suggests a possible evolutionary basis for its origin. Accordingly, most people “will never become addicted” because our ancestors would have evolved psychological mechanisms to learn to counterexploit plant neurochemical defences for their own fitness benefits while avoiding their immediate costs (Hagen et al. Reference Hagen, Sullivan, Schmidt, Morris, Kempter and Hammerstein2009). In other words, modern humans would inherit an ancient biological constitution that would predispose them to learn to use psychoactive drugs for their indirect fitness-enhancing effects, not necessarily for their direct-rewarding effects. One striking, albeit overlooked, implication of this framework is that most of the learned motivation for psychoactive drug use would not be driven by direct pharmacological activation of the midbrain dopamine system, as it is commonly believed, but indirectly through a facilitation or enhancement of other fitness-relevant behavioral pursuits (e.g., facilitation of sexual behavior). Pushing this idea to the limit, one could even imagine designing psychoactive substances with zero pharmacological efficacy on brain dopamine signalling but whose regular consumption would still be driven by their learned fitness benefits. This prediction could be tested empirically.
Although interesting and relevant, Müller & Schumann's (M&S's) framework rests on a rather fragile biological footing. M&S convincingly show how people can advantageously instrumentalize some psychoactive drug effects to better adapt to the multiple demands of modern urban life; yet, they provide no direct evidence that these effects really increase fitness (e.g., relative to people who abstain from drug use). In addition, M&S do not explain what the selective advantages of psychoactive drug effects to our ancestors, say from the Pleistocene, could have been. The pressures of modern urban life that make psychoactive drug use advantageous are vastly different from those of ancestral environments. The hypothetical scenario for the evolution of psychoactive drug instrumentalization from food selection and self-medication of parasitic diseases lacks some crucial episodes (e.g., there is a conceptual gap between using plants as medications for physical diseases and using them as “mental tools”) and, certainly, does not rule out other scenarios. For example, modern psychoactive drug use could be a relatively recent exaptation (Gould & Vrba Reference Gould and Vrba1982), not an ancient behavioral adaptation shaped by natural selection, or a relatively recent gene-shaping cultural innovation (Laland & Brown Reference Laland and Brown2006). All these possibilities are compatible with current data on the genetics of drug addiction. Hence, the ultimate cause of psychoactive drug use remains largely uncertain, and this uncertainty weakens the foundation of the proposed framework.
The biological foundation of M&S's framework could have been made firmer by the authors better considering and integrating previous research on animal drug use in the wild (Siegel Reference Siegel1989) and in the laboratory (Campbell & Carroll Reference Campbell and Carroll2000). Overall, this research suggests that psychoactive drug instrumentalization is not unique to humans and probably corresponds to an ancient behavioral trait, at least in the mammal class. For example, rats – the most frequently used animal model in the field – can learn to regulate drug use as a function of stress, pain, and hunger. Under some circumstances, hungry rats increase their consumption of cocaine (Carroll Reference Carroll1985) probably for its powerful anorexigenic effects. Similarly, rats can regulate opiate intake when in pain (Colpaert et al. Reference Colpaert, Tarayre, Alliaga, Bruins Slot, Attal and Koek2001) or stressed (Shaham & Stewart Reference Shaham and Stewart1994). Inversely, rats can reduce their intake of drugs when the resulting effects interfere with other biological goals. For example, under some circumstances, rats learn to reduce their intake of cocaine during the light resting period, probably to sleep and recover (Bass et al. Reference Bass, Jansen and Roberts2010). Recent evidence even indicates that rats, like humans, can choose to use different drugs as a function of the demands of different environmental contexts (Caprioli et al. Reference Caprioli, Celentano, Dubla, Lucantonio, Nencini and Badiani2009). Hence, there is currently sufficient data for considering psychoactive drug instrumentalization as a behavioral trait common to humans and at least some other mammals.
This being said, I nevertheless concur with M&S that more research should be devoted to animal drug instrumentalization, particularly on instrumentalization of drug-induced cognitive enhancement that is perhaps unique to humans. For example, it would be interesting to determine whether increasing cognitive demands during access to certain cognitive-enhancing drugs that are normally poorly self-administered by rats (e.g., nicotine) could increase drug self-administration. Finally, and most important, most drug self-administering rats, apparently like most people, do not become addicted. Growing evidence indeed indicates that making animals addicted to drugs is harder than previously thought (Ahmed Reference Ahmed2010). For example, we recently found that no matter how severe the cocaine exposure, most cocaine self-administering rats do not lose control over drug use, as they can readily abstain from it when another non-drug pursuit is made available. Only a minority of rats continues to take cocaine despite choice and increasing stakes (Cantin et al. Reference Cantin, Lenoir, Augier, Vanhille, Dubreucq, Serre, Vouillac and Ahmed2010; Lenoir et al. Reference Lenoir, Serre, Cantin and Ahmed2007). These findings increase the plausibility of the hypothesis of a biological resilience to addiction to explain why most people who use psychoactive drugs do not become addicted.
As Müller & Schumann (M&S) refreshingly remind us, most people who regularly use psychoactive drugs, even highly addictive ones such as cocaine, do not go on to develop addiction. Only a minority of regular drug users eventually become addicted (Anthony Reference Anthony, Davis, Charney, Coyle and Nemeroff2002). A long-standing question in addiction research, with critical consequences for prevention and policy, is how we should interpret this individual variation, particularly the high rate of non-addictive drug use. Do environmental circumstances (e.g., economic constraints, societal regulations, cultural norms) prevent people from exposing themselves sufficiently to drugs to cross the “threshold of addiction” (Benowitz & Henningfield Reference Benowitz and Henningfield1994)? Or, alternatively, are most drug users somehow biologically resilient to addiction (e.g., genetically resistant to addiction regardless of drug exposure)? Epidemiology of drug addiction alone has been so far unable to resolve this apparent dilemma. This probably explains the prevailing default view in drug prevention and policy that considers each one of us a potential addict. Given sufficient drug exposure, in this view, each one of us could be turned into a drug addict. A biological resilience to drug addiction would not exist, or such resilience would exist only in rare individuals.
Overall, the framework of M&S – which echoes, without citing it, the “mental tool” hypothesis of psychoactive drug use that Michael Pollan initially formulated in his Botany of Desire (Pollan Reference Pollan2001) – seems to support the notion of a biological resilience to drug addiction and even tentatively suggests a possible evolutionary basis for its origin. Accordingly, most people “will never become addicted” because our ancestors would have evolved psychological mechanisms to learn to counterexploit plant neurochemical defences for their own fitness benefits while avoiding their immediate costs (Hagen et al. Reference Hagen, Sullivan, Schmidt, Morris, Kempter and Hammerstein2009). In other words, modern humans would inherit an ancient biological constitution that would predispose them to learn to use psychoactive drugs for their indirect fitness-enhancing effects, not necessarily for their direct-rewarding effects. One striking, albeit overlooked, implication of this framework is that most of the learned motivation for psychoactive drug use would not be driven by direct pharmacological activation of the midbrain dopamine system, as it is commonly believed, but indirectly through a facilitation or enhancement of other fitness-relevant behavioral pursuits (e.g., facilitation of sexual behavior). Pushing this idea to the limit, one could even imagine designing psychoactive substances with zero pharmacological efficacy on brain dopamine signalling but whose regular consumption would still be driven by their learned fitness benefits. This prediction could be tested empirically.
Although interesting and relevant, Müller & Schumann's (M&S's) framework rests on a rather fragile biological footing. M&S convincingly show how people can advantageously instrumentalize some psychoactive drug effects to better adapt to the multiple demands of modern urban life; yet, they provide no direct evidence that these effects really increase fitness (e.g., relative to people who abstain from drug use). In addition, M&S do not explain what the selective advantages of psychoactive drug effects to our ancestors, say from the Pleistocene, could have been. The pressures of modern urban life that make psychoactive drug use advantageous are vastly different from those of ancestral environments. The hypothetical scenario for the evolution of psychoactive drug instrumentalization from food selection and self-medication of parasitic diseases lacks some crucial episodes (e.g., there is a conceptual gap between using plants as medications for physical diseases and using them as “mental tools”) and, certainly, does not rule out other scenarios. For example, modern psychoactive drug use could be a relatively recent exaptation (Gould & Vrba Reference Gould and Vrba1982), not an ancient behavioral adaptation shaped by natural selection, or a relatively recent gene-shaping cultural innovation (Laland & Brown Reference Laland and Brown2006). All these possibilities are compatible with current data on the genetics of drug addiction. Hence, the ultimate cause of psychoactive drug use remains largely uncertain, and this uncertainty weakens the foundation of the proposed framework.
The biological foundation of M&S's framework could have been made firmer by the authors better considering and integrating previous research on animal drug use in the wild (Siegel Reference Siegel1989) and in the laboratory (Campbell & Carroll Reference Campbell and Carroll2000). Overall, this research suggests that psychoactive drug instrumentalization is not unique to humans and probably corresponds to an ancient behavioral trait, at least in the mammal class. For example, rats – the most frequently used animal model in the field – can learn to regulate drug use as a function of stress, pain, and hunger. Under some circumstances, hungry rats increase their consumption of cocaine (Carroll Reference Carroll1985) probably for its powerful anorexigenic effects. Similarly, rats can regulate opiate intake when in pain (Colpaert et al. Reference Colpaert, Tarayre, Alliaga, Bruins Slot, Attal and Koek2001) or stressed (Shaham & Stewart Reference Shaham and Stewart1994). Inversely, rats can reduce their intake of drugs when the resulting effects interfere with other biological goals. For example, under some circumstances, rats learn to reduce their intake of cocaine during the light resting period, probably to sleep and recover (Bass et al. Reference Bass, Jansen and Roberts2010). Recent evidence even indicates that rats, like humans, can choose to use different drugs as a function of the demands of different environmental contexts (Caprioli et al. Reference Caprioli, Celentano, Dubla, Lucantonio, Nencini and Badiani2009). Hence, there is currently sufficient data for considering psychoactive drug instrumentalization as a behavioral trait common to humans and at least some other mammals.
This being said, I nevertheless concur with M&S that more research should be devoted to animal drug instrumentalization, particularly on instrumentalization of drug-induced cognitive enhancement that is perhaps unique to humans. For example, it would be interesting to determine whether increasing cognitive demands during access to certain cognitive-enhancing drugs that are normally poorly self-administered by rats (e.g., nicotine) could increase drug self-administration. Finally, and most important, most drug self-administering rats, apparently like most people, do not become addicted. Growing evidence indeed indicates that making animals addicted to drugs is harder than previously thought (Ahmed Reference Ahmed2010). For example, we recently found that no matter how severe the cocaine exposure, most cocaine self-administering rats do not lose control over drug use, as they can readily abstain from it when another non-drug pursuit is made available. Only a minority of rats continues to take cocaine despite choice and increasing stakes (Cantin et al. Reference Cantin, Lenoir, Augier, Vanhille, Dubreucq, Serre, Vouillac and Ahmed2010; Lenoir et al. Reference Lenoir, Serre, Cantin and Ahmed2007). These findings increase the plausibility of the hypothesis of a biological resilience to addiction to explain why most people who use psychoactive drugs do not become addicted.
ACKNOWLEDGMENTS
The author is supported by grants from the French Research Council (CNRS), Université Bordeaux Ségalen, Conseil Régional Aquitaine and Mission Interministérielle de Lutte contre la Drogue et la Toxicomaine (MILDT). I thank Eric Augier for reading a previous version of this commentary.