Your Brain on Drugs

Understanding the neuroscience of addiction and drug tolerance.

Years ago, while I was studying for the GRE, the major admissions exam required by many PhD programs, a friend was also studying for her MCAT. As exams go, the GRE is unpleasant but not unendurable. Typically, you plan to spend 4 uninterrupted hours in a room with a few other students staggered at computers walled away in cubicles. By comparison, the MCAT clocks in at a monolithic 7.5 hours, spanning a dizzying range of topics. What’s more, most aspiring medical students plan on taking it at least twice.

For weeks ahead of our respective test dates, we quizzed each other and retreated to conference rooms or empty offices to labour over practice exams, trying as faithfully as possible to recreate the dull and probably uncomfortable conditions of the testing centres we were destined for on exam day. As tempting as it was to review the fundamentals of algebra and organic chemistry curled up on the couch with a glass of wine, we thought our chances would be better if we found the most austere corners of the lab to hide in, huddled over laptops in fluorescent light.

Whether or not this strategy helped boost our scores is debatable, but decades of research into the concept of state-dependent memory would strongly suggest that we were on the right track.


Much of our understanding of how this type of memory recall works comes from observations of patients who experienced trauma. Take repressed memories. The experiences so depressing or traumatic, our brains bury them deep inside where they can’t trouble us. In normal life, we don’t remember them at all. But then, one fine day, our senses experience something in the environment that remind us of the hidden memory, and up it comes to the surface, all of a sudden.

To explain this occurrence, people came up with the idea of “situationally accessible knowledge”: knowledge that you don’t ordinarily have access, unless you’re in a very specific situation. Knowledge of this phenomenon is useful to clinicians as they work with people suffering from traumatic memory loss or dissociation; depending on the context, recall might be desirable for some but detrimental to others.

Sensitivity to the pain of unearthing deeply buried trauma in people is paramount in a therapeutic setting, but one of the earliest pieces of experimental evidence for state-dependent memory came from classical conditioning studies on dogs in the 1930s.

In a 1937 paper, researchers observed muscle twitches in response to stimuli that had been administered while the animals were under the influence of a sedative called curare. The reflexes disappeared under identical conditions when the dogs were not sedated, implying an association between the reflex and the altered state of conscious under which it was learned.


Learning requires the formation of new proteins at the synapse, the space between two neurons where chemical messages are transformed into electrical impulses. Interrupting the process of protein synthesis that strengthens new memories also prevents the development of tolerance to certain drugs.

These molecular and electrical processes can be observed under a microscope, but the study of learning in real time often follows a simple formula: an animal is trained to perform a task, and is either rewarded for performing well or punished for performing poorly. If a neutral outcome is paired with the reward, an animal will start to associate that neutral element with a treat (recall Pavlov’s dog salivating for its dinner when it hears a bell ring).

The basic biology behind learning a new task or cramming for an exam also applies to drug tolerance. When you take a drug or have a drink, your brain immediately rewards you with a tiny rush of dopamine. If you do this habitually, your brain will learn to compensate to bring you back to your steady state, and over time you’ll need more and more of whatever you’ve been ingesting to feel the same kind of rush as you did initially. This is where the idea of state-dependent memory gets more complicated, and potentially deadly.


In 2005 a group of Hungarian physicians published the case report of a 26 year old man who died of an apparent overdose. He and his wife had sought treatment several times previously for heroin addiction. At the time of his death, he had been injecting the drug habitually, having steadily increased the amount and frequency of his dosage over a period of several years.

Together, he and his wife had decided to withdraw from the drug. However, the following day he bought and administered his regular dose in the bathroom of a metro station, to avoid doing so at home, where his wife was waiting. He was found dead shortly thereafter, but on autopsy his blood was found to contain no other substances, and the level of the drug in his blood was identical to the levels found on a previous hospital admission, when he’d originally sought treatment. The key element, in this case, seemed to be the context of his drug use: he’d acquired tolerance to the effects of the drug at home, but the same dose in a different setting now proved lethal.

This is just one example of dozens of similar cases. This phenomenon isn’t limited to illicit substance use either; similar deaths of apparent overdose have taken place when patients receiving regular doses of morphine for pain control receive the same dosage in their homes, or in one case, merely moving between the rooms of one’s own house.


A seminal experiment demonstrating this effect in a controlled setting was conducted in rats in the late 1970s. Rats were separated into three groups, two of which received steady daily intravenous infusions of a low dose of heroin. The third group received nothing. At the end of the training phase, all three groups were given a much higher dose of heroin that the researchers determined would likely be lethal.

Indeed, almost all of the animals who were naive to the drug (in the third group) died, but the remaining two groups differed critically. One group received this final, high dose in the same setting in which they’d received their previous daily infusions, while the other group received it in a different setting. Among the animals who received this dose in an unfamiliar location, well over half died. However, only about a third of the animals whose location remained stable died —  a much higher survival rate.


This puzzling phenomenon might be explained in part by the idea of state-dependent memory. Researchers use the term “behavioural tolerance” to distinguish environmental and social factors surrounding substance use from the direct biochemical actions of the substances themselves. The idea is that if someone indulges in a particular substance regularly, they can also adapt behaviourally and physiologically to compensate for the effects of that substance.

Part of what might drive that adaptation is, associative learning - the same principal behind classical conditioning. One line of thinking holds, for example, that if a person is a regular social drinker and is consistently “rewarded” for their apparent tolerance to the effects of alcohol, they’ll maintain those levels of consumption and behaviour patterns because of this positive reinforcement. The reward could come in the form of the approval of their peers, or merely a lack of repercussions for their consumption. Since these positive associations are specific to the people and places they’re formed among, people may not display similar levels of tolerance if they’re in an unfamiliar place or by themselves.

Of course, this is just one angle from which to approach the puzzle of drug tolerance, and this angle is distinct from the mechanisms of physical dependency and biochemical tolerance. But it underscores an important and frequently overlooked aspect of substance use and dependency, and suggests that behavioural interventions could be powerful therapeutic tools for managing substance use disorders and minimising unintentional overdose deaths.


While many clinicians or community-based support groups promote total abstinence, harm reduction has proven to be a more effective strategy for many people. This model favours supportive outlets in society, such as safe injection sites or access to opioid receptor agonists like methadone, as replacement therapy for people who are dependent on heroin.

Emphasising behavioural changes that an individual substance user can make to minimise the negative impacts of their substance use on their lives, can feel more manageable than abstinence to some, and may represent an intermediate step on the way to total abstinence.

In the United States, housing programs or court-mandated treatment programs are often abstinence focused, but in a study comparing programs for non-housed people that required evidence of total abstinence and those where housing was offered without this condition, the abstinence-contingent housing programs actually seemed to perpetuate homelessness. Considering similar studies which show that stable housing reduces the incidence of medical expenses related to alcohol and substance use and lower anecdotal reports of frequent alcohol use, harm reduction in this case is an approach that would have far-reaching benefits within the community as well as for individuals seeking treatment.


Barriers to the broader adoption of such interventions are rooted in social stigma attached to substance use disorders, as well as the criminalization of substance use and possession. These ideas encourage all-or-none approaches to substance use that assign moral judgements to dependency. These attitudes hamper our ability to talk freely about these issues as systemic societal problems that could be addressed in supportive and constructive evidence-based ways.

Understanding of phenomena like behavioural tolerance could empower clinicians to educate patients seeking treatment for substance use disorders to be mindful of their environments, and to observe their own behaviour patterns and relationship to drugs or alcohol in the context of their communities and social support structures.