This book is a summary of what the author learned over the past twenty or so years as a researcher studying the neuroscience of addiction, and as an addict herself. She writes that her hope is that sharing this information might help loved ones, caregivers, and crafters of public policy make more informed choices.
In the United States, about 16 percent of the population twelve and older meet criteria for a substance use disorder, and about a quarter of all deaths are attributed to excessive drug use. In purely financial terms, it costs more than five times as much as AIDS and twice as much as cancer. In the United States, this means that close to 10 percent of all health-care expenditures go toward prevention, diagnosis, and treatment of people suffering from addictive diseases, and the statistics are similarly frightening in most other Western cultures. Despite all this money and effort, successful recovery is no more likely than it was fifty years ago. Although reliable estimates are hard to come by, most experts agree that no more than 10 percent of substance abusers can manage to stay clean for any appreciable time. It’s also not likely to be completely prevented by anything less than preventing learning and memory.
James Olds and Peter Milner discovered that electrical stimulation applied to the nucleus accumbens of the brains of rats led to release of the neurotransmitter dopamine. This is an area of the brain located about three inches behind the bottom of the eye sockets and is part of the limbic system, a group of structures that are primarily involved in emotion. The dopamine was released here by neurons that originate in the midbrain, following the mesolimbic pathway (so called because it goes from the midbrain, that is, meso, to the limbic system). They found that when the rats were able to stimulate this area themselves, their single-minded quest to stimulate this area of the brain resulted in starvation and sleep deprivation to the point of death.
The mesolimbic system evolved to promote behaviors such as eating and sex, and the sense of pleasure it confers is not so much a mood state as an emotional experience of “thrill” or pleasure. In recent years, new evidence has shown that dopamine in the mesolimbic pathway works not exactly by signaling pleasure but by signaling the anticipation of pleasure.
Things that cause a release of dopamine in the mesolimbic pathway may be something pleasurable (sexual stimulation, a bump of cocaine) but also something surprising (drama, whatever is in the package), novel (such as travel), potentially newsworthy (a lottery ticket), or really valuable (oxygen to a deprived organism). In other words, this system alerts us to the anticipation of a meaningful event, not to pleasure per se.
All drugs affect multiple brain circuits, and variation in their sites of neural action accounts for their different effects. However, all addictive drugs are addictive precisely because they share the ability to stimulate the mesolimbic dopamine system.
As a rule of thumb, it takes about five half-lives to get rid of about 95 percent of any drug.
There is a second dopamine circuit as well that figures in addiction. Dopamine in the nigrostriatal pathway (which connects the substantia nigra at the base of the brain to the striatum, a large piece of real estate roughly in the center of each hemisphere) enables us to take action either toward or away from a stimulus. As dopamine release in the nucleus accumbens signals something newsworthy happening in the environment, it also activates this second circuit to motivate us to move. Dopamine in both pathways naturally declines with age, partly accounting for general decreases in the eagerness to explore new things.
Solomon and Corbit’s opponent-process theory posits that any stimulus that alters brain functioning to affect the way we feel will elicit a response by the brain that is exactly opposite to the effect of the stimulus. Having a set point enables meaningful interpretation of a stream of ceaselessly changing input. The b process is the brain’s response to the a process, or the brain’s response to what the drug does to the brain, counteracting the changes in neural activity produced by the stimulus in an effort to return brain activity to its neutral, homeostatic state. The b process learns with time and exposure. Repeated encounters with the stimulus result in faster, bigger, and longer-lasting b processes that are better able to maintain homeostasis in the face of disruption. Moreover, the b process can be elicited solely by environmental stimuli that promise the a process is coming - which is what happened with Pavlov’s dogs, who learned to salivate even when food was not present. Withdrawal happens because the b process outlasts the drug’s effects.
Because of the brain’s tremendous capacity to adapt, it’s impossible for a regular user to get high, and the best a voracious appetite for more drug can hope to accomplish is to stave off withdrawal. Dramatic changes in the balance of positive and negative reinforcement and positive and negative punishment occur with regular use in a way that makes addiction more likely. In particular, we rapidly form a tolerance to positive reinforcing effects, while the negatively reinforcing effects typically become stronger as users use more to stave off symptoms associated with withdrawal.
If you want to achieve a sustained positive state, you could submit yourself to negatively charged experiences. This way the opponent process would be positive. Solomon and Corbit argued that such a pattern may be at work in an activity like skydiving. Or for Ben, lead climbing.
Tachyphylaxis: During the rising phase of the curve, the drinker experiences the pleasure resulting from mesolimbic activation. But at the same time, motor impairments such as stumbling gait and slowed speech grow increasingly pronounced. Later, when alcohol levels are falling, both pleasure and impairment are greatly reduced.
Craving happens because the brain is so well organized to counteract perturbations that it uses its exceptional learning skills to anticipate disruptions, rather than wait for the changes themselves, and begins to dampen drug effects before the drug has even been delivered.
Suppose you enjoy a few drinks with friends at a neighborhood bar every Friday after work, as you have been doing for years. It turns out that the predictability of this routine leads to changes in your experience. First, the alcohol will have less effect on you in that particular place and time and with those particular friends than it would have elsewhere. This is the same effect that partially explains why people are more likely to overdose in novel environments.
Endocannabinoid system: (endogenous cannabis). Activated exogenously by THC, found in marijuana. THC modulates activity in a huge number of synapses, in almost every brain structure. Its primary receptor in the brain is called CB1 for cannabinoid receptor 1. It also binds to anandamide and 2-AG receptors, which communicate in the opposite direction, diffusing “upstream” across the synapse to convey information from the receiving to the sending cell. Once they reach their target, CB1 receptors on the first cell’s surface, they produce their actions, which essentially are to alter the signal-to-noise relationship so that the message conveyed by classical neurotransmitters in this synapse carries more import. Thus THC makes us feel like whatever the message just transmitted across the synapse, it was important. The cannabinoid system helps to sort our experiences, indicating which are the most meaningful or salient. Anandamide and 2-AG and their receptor are all over the brain because such input might be carried in any number of pathways, depending on the exact nature of the stimulus.
There is a well-established positive correlation between exposure to weed or other natural cannabinoids and a diagnosis of schizophrenia. All sorts of things can influence risk including a biological predisposition but also living under stressful conditions and, apparently, excessive activation of CB1 signaling.
There is a large body of solid evidence that early exposure to marijuana causes changes in brain structure to embryos, children, and adolescents and that these structural changes can produce cognitive and behavioral deficits. There is also good evidence that exposure during development makes someone, among other things, permanently less sensitive to rewards, so that later, when given the ability to self-administer drugs of abuse, they take more. We call this the “gateway effect,” and a growing body of research documents enhanced drug-taking and drug-seeking behavior in humans and animals following adolescent exposure to substances including cannabis.
Opioids: There are dozens of different opioids manufactured by the brain (including actual morphine). Experiments have shown that these chemicals serve a range of critical functions including modulating activities like sex, attachment, and learning.
Anti-opiates: Rats undergoing a pain protocol were shown a flash of light immediately following the last shock each day. The light came to signal safety. Soon, the light began to reverse the rats’ analgesia so that their pain sensitivity immediately returned to normal. The safety signal led to the release of anti-opiates. Dozens of peptides (short chains of amino acids) with anti-opiate properties have been identified. Some of the better known of these are dynorphin, orphanin FQ, cholecystokinin, and NPFF, all of which may facilitate adaptation and contribute to tolerance, dependence, and craving. Cues that predict drug use lead to activation of the anti-opiate system, too. Seeing a spoon in a bathroom, for instance, is enough to cause chills and craving in any mainline opiate user, even if they haven’t picked up for years. In contrast, only about 5 percent of soldiers addicted in Vietnam ended up relapsing, likely because they left the context in which they became addicted (compared with about 90 percent of people addicted and treated in the US).
Heroin and other street opiates show about a 40 percent increase in use each year since 2010. The vast majority of new heroin users (about four out of five) start by misusing prescription painkillers and turn to street narcotics because they are cheaper and easier to get.
Tolerance: Although the number of opiate receptors isn’t dramatically decreased in a chronic narcotic user (the way CB1 receptors decrease in marijuana smokers), the ability of opiates to affect intracellular signaling is compromised with frequent use, so the receptors are downregulated in effectiveness if not in number. The brain adapts to all exogenous chemicals that alter its activity, but the degree of tolerance, dependence, and craving in opiate users is legend - stronger than for almost every other drug. The strength of these opponent processes may be so robust because the sensation of pain is so critical for survival. Opiate tolerance is mind-bogglingly robust. Addicts can administer upwards of 150 times the dose that would be lethal to naive users and, even so, just feel “right” but not really high.
Preventing overdose: Narcan occupies all of the same sites opiates do but doesn’t activate them, however, the effects are so terrible general anesthesia is often administered. Suboxone is a combination of a Narcan-like drug and an opiate drug called buprenorphine. Buprenorphine doesn’t have much street appeal for the same reason it’s a good choice here: although it occupies the same places in the brain as opiate drugs, it doesn’t do as good a job and therefore it is much less rewarding than its abused counterparts.
Alcohol: The author points out the irony of the Research Society on Alcoholism annual meeting starting with free drinks. Over 85 percent of the world’s adults drink, but only about one-tenth of these develop a problem. Normally, alcohol is degraded at a constant rate, in contrast to other drugs whose metabolism depends on their concentration, and the drug reaches equilibrium in the blood and brain in less than an hour. The targets of alcohol are quite diverse and hard to pin down, as it’s a small molecule that travels in the brain and body easily. The drug’s foremost neural consequence is facilitating GABA neurotransmission, the predominant inhibitory neurotransmitter in the brain. At moderate doses this reduces anxiety, but at higher concentrations it produces sedation and eventually sleep. Alcohol also slows neural activity by impeding neurotransmitter release through its actions on calcium channels. Calcium is a necessary catalyst for exocytosis - the process by which synaptic vesicles release neurotransmitters into the cleft.
Alas, if someone finds alcohol especially rewarding because of an inherited tendency toward anxiety and she imbibes frequently, she’ll become increasingly anxious and require more drug.
The mechanism of reward in the case of alcohol may reflect activation of opioid receptors by endogenous opioids, which then lead to dopamine release. The drug can also interact with a particular type of serotonin receptor as well as acetylcholine receptors, perhaps impacting mood and cognition. We have long known that alcohol use rapidly leads to the synthesis and release of beta-endorphin, a string of thirty-one amino acids that is thought to contribute to the drug’s euphoric and relaxing effects by increasing mesolimbic dopamine levels and inhibiting the “fight or flight” response.
At high concentrations, alcohol may also have general effects on the physical integrity of brain cells. Neuron membranes are basically made of fat. Bathing them in alcohol causes the membranes to become more and more fluid, and as the cell structure is compromised, so is the ability of neurons to conduct information - leading to stupor or unconsciousness.
Treatment: Antabuse interferes with the metabolism of alcohol and leads to a buildup of acetaldehyde, which is toxic. This metabolite produces uncomfortable physiological effects including flushing, sweating, and irregular heartbeat. However, the author indicates that in general, punishment is not an especially effective way to change behavior. One of the pharmaceutical strategies to combat alcohol abuse is naltrexone, a longer-acting and orally available cousin of naloxone, which acts to block endogenous opioid action and make alcohol less rewarding.
Data from studies on twins had demonstrated that about 50-60 percent of the risk for alcoholism came from inherited factors. High-risk individuals have about half as much beta-endorphin in their blood as those at low genetic risk. Alcohol was able to remedy this natural deficit especially in those who inherit a high risk for excessive drinking and, at higher doses, produce a surfeit of the peptide. Thus, a genetic propensity for drinking may reflect an effort to achieve normal levels of the endogenous opioid beta-endorphin.
One of the most effective ways to curtail the risk of addiction is to avoid intoxication during periods of rapid brain development. People who begin drinking in their early teens, as I did, are at least four times more likely to eventually meet the criteria for an alcohol use disorder. In fact, the lifetime risk for substance abuse and dependence decreases about 5 percent with each additional year between ages thirteen and twenty-one.8 Yet young people are especially prone to binge drinking in part because they are neurobiologically primed to seek and appreciate novel and high-risk experiences.
Alcohol killed about twice as many people in 2016 as prescription opioids and heroin overdoses combined, and even this number would be almost three times higher if it included drunk-driving-related deaths. It is the third-largest cause of preventable death.
Barbiturates: The popularity of barbiturates grew rapidly, and by the 1920s they were virtually the sole treatment for conditions benefiting from sedation. However, in 1960 Valium was introduced, initiating a second wave in this pharmacological revolution. All of these compounds are sedating in that they induce both muscular and psychic relaxation; “hypnotic” refers to their sleep-inducing properties. Phenobarbital is currently the most widely prescribed antiepileptic drug in the world, aptly dubbed “king of the barbiturates.” They are often abused because they create a sense of distance from one’s feelings. Many also began combining barbiturates with alcohol to increase the effect or took them alternately with stimulants to mitigate drowsiness. More recently, their co-prescription with opiates has been thought to contribute to a surge in lethal overdoses.
Both inventors of barbiturates, the chemists Fischer and von Mering, died of overdose after years of dependence.
Benzodiazepines: in the 1960s another class of sedative-hypnotics was introduced - benzodiazepines - and these were purported to be much safer and less addictive than their predecessors. Not surprisingly, those claims were overstated. Millions of people are now hooked on benzos, but on the bright side it’s not possible to overdose from them alone, so the market is likely to stay strong.
Both barbiturates and benzodiazepines are GABA-A agonists. There are nineteen distinct subunits of this receptor that may assort in the cell membrane to form the receptor complex and well over a thousand possible structures for this single receptor. The major difference between benzos and barbiturates is that overdose is virtually impossible with benzodiazepines alone and fairly likely with barbiturates. Typically, both drugs act in concert with GABA on its receptor, and so their effects are limited by the presence of GABA in the synapse. But at high doses, barbiturates can mimic GABA and open the chloride channel directly. This allows them to decrease excitation and generally inhibit neurotransmitter release, perhaps to such an extent that the brain activity necessary to life is stopped.
Downregulation of GABA-A receptor sensitivity is produced by decreasing the number of receptors on the cell surface as well as changes in the subunit constituency, and both of these contribute to tolerance.
Whether or not you are able to drink others under the table, or are known as a “lightweight,” has been attributed to the particular makeup of subunits of the GABA-A receptor. Structural differences may also confer individual variation in pain sensitivity, anxiety, premenstrual or postpartum depression, diagnosis on the autism spectrum, and need for sleep, among others.
Stimulants: Stimulant classification is based solely on effect. The stimulant class is large and diverse. The stimulants methylphenidate (Ritalin) and amphetamine (Adderall) have been effectively used to treat attention deficit hyperactivity disorder (ADHD) for almost sixty years. Multiple studies have shown that when these substances are administered in a controlled laboratory setting, virtually everybody enjoys their effects. Alcohol, THC, and even opiates are notorious for producing variable effects.
Caffeine is the most popular psychoactive drug in the world, although there is some debate about whether or not it is addictive. Though regular use may result in a modicum of tolerance and will likely cause dependence (that is, withdrawal upon abstinence) and craving, the drug is not considered harmful, thus skipping one of the core criterion for addiction. The mechanism by which caffeine produces its effects is not completely understood, but we do know that it doesn’t act like cocaine, amphetamine, or MDMA to directly enhance the transmission of dopamine, norepinephrine, and/or serotonin. Instead, caffeine is an antagonist at adenosine receptors. Adenosine serves as a neurotransmitter and is thought to build up over the day, accumulating in synapses where it binds to its receptors, precipitating a state of drowsiness.
On average, tobacco users lose fifteen years of life. Today, 5.7 percent of total health expenditures are spent treating smoking-related illnesses, and 12 percent of all adult deaths worldwide result from this habit. In fact, the total annual cost of smoking is almost 2 percent of global gross domestic product, which is also about 40 percent of what all the world’s governments spent on education. Fewer people in the United States are smoking, though it still remains the number one preventable cause of sickness and mortality in this country.
It is believed nicotine is highly addictive in part because of its rapid onset of action. A cigarette contains between six and eleven milligrams of nicotine, though much less than this actually reaches the bloodstream of a smoker because most of the drug is metabolized to cotinine in the liver by a specific enzyme called cytochrome P450 2A6 (CYP2A6) and excreted in the urine. The dynamic adaptations that lead to tolerance within such a short time are mirrored on the other end as tolerance partially decreases during even a few hours of withdrawal, so the first few puffs are the best of the day. The bigger lesson here is the temporal symmetry: tolerance that develops rapidly tends to reverse quickly too, while changes that take longer to accrue tend to persist.
Nicotine is classified as a stimulant, even though regular smokers report that it helps them relax and deal with stress. This is related to rapid tolerance. Nicotine works first to stimulate acetylcholine receptors, but they quickly respond by becoming insensitive, leading to the bidirectional effects of the drug. There are two major types of acetylcholine receptors, and nicotine interacts with only one: aptly named the nicotinic acetylcholine receptor, or nAChR. A cousin of the GABAA receptor, nAChRs are also made of five subunits that surround a central pore. This receptor permits the flow of sodium, rather than chloride, and because sodium ions carry a positive charge, nAChRs are excitatory. Nicotine activates these receptors by substituting for acetylcholine and thereby increasing neuronal activity.
Because acetylcholine signaling is also pervasive, and nAChR expression is extremely broad, practically every area of the brain is affected by nicotine. However, unlike the GABAA receptor that can be either closed or open, the nAChR has three states: closed, open, and desensitized. The open state is responsible for the stimulant properties of the drug, while the desensitized state produces a cigarette’s calming effects.
Genetic variation in the subunit(s) structure of the nAChR accounts for a substantial amount of the risk for nicotine dependence.
Cocaine, the amphetamines (including methamphetamine), and ecstasy all have a very similar mechanism of action. They interfere with the recycling mechanism for monoamine neurotransmitters. Transporters, like receptors, are proteins embedded in the neural cell membrane, but unlike receptors the function of transporters is to transport (or recycle) released neurotransmitter back into the presynaptic neuron, where it can be repackaged and reused. When one of these drugs occupies a spot on a transporter, it prevents the monoamines from utilizing their reuptake mechanism and prolongs their effects.
The half-life of methamphetamine is about ten hours (ten times that of cocaine), but amphetamine’s half-life varies widely - anywhere from seven to thirty hours, depending on the pH of the user’s urine. The effects from acute low to moderate doses include euphoria, “rush,” wakefulness, antifatigue, increased confidence, hyperactivity, and loss of appetite. Higher doses also cause talkativeness, aggressiveness, restlessness, and stereotypy. Amphetamines may cause excitotoxicity even in recreational and infrequent use. This pay be partially due to the ability of amphetamine to get recycled into a cell and stay in the nerve terminal.
MDMA has a relatively greater effect on the serotonin system and has recently been approved for its first clinical trial in the United States.
Lynn Taurah et al., 2014: two groups used MDMA; the first had used the drug at some point during the previous six months, but not for at least three weeks, and the other group were former ecstasy users, who’d been abstinent for at least four years. There were two major findings. First, former and current ecstasy users were virtually identical, and, second, these groups showed significantly more clinically relevant levels of depression, impulsiveness, poor sleep, and memory impairment.
Psychedelics: Psilocybin, mescaline, and DMT are natural compounds that have been used for millennia by indigenous people in sacred rituals; LSD is a synthetic compound, created by Albert Hofmann, a Swiss chemist, in 1938. LSD is one of the most potent psychoactive compounds we know and is effective at about two hundred times lower concentrations than the next strongest. The author indicates that most psychedelics should not be classified as addictive. These drugs don’t lead to the release of dopamine in the nucleus accumbens, so nonhuman animals won’t self-administer them. They can’t be taken compulsively, because tolerance is so fast and profound that it precludes regular use; there’s no evidence for dependence and no compelling evidence for harm in most people. The drugs induce activity in genes that are associated with neuroplasticity and disrupt established/default connections between groups of neurons.
Phencyclidine (PCP) and ketamine are two examples of the class of drugs known as dissociative anesthetics. Ketamine became a successful anesthetic and is still often used in humans, especially during pediatric and geriatric surgery, where the risk of overdose from barbiturates is higher, or when there isn’t time for delivering and monitoring barbiturates, such as on a battlefield. The peculiar EEG pattern generated under these drugs looks like a mix of complex waveforms, suggesting to researchers a dissociation of sensory and limbic systems, or sensing and feeling. In fact, the EEG tracings mirror the class’s name - dissociative anesthetics. These work by producing a state of separation between sensation and “self.” These drugs block the flow of ions through one of the glutamate receptors, known as the NMDA receptor. NMDA receptors are widely distributed throughout the brain and play a critical role in many functions including cognition and memory formation. They also potently stimulates mesolimbic dopamine neurons: dopamine levels rise in the nucleus accumbens following systemic administration, drug-naive volunteers roundly seem to like the drug and desire more, and given the opportunity, experimental animals will readily self-administer these drugs.
Salvinorin A is the most potent natural hallucinogenic so far discovered (LSD surpasses it, but this, of course, is synthetic). Salvia has been used for a long time in religious rituals by the Mazatec people. Until its characterization, no other molecule with a similar structural-functional profile was known, so it has opened up a new avenue of research. Salvinorin A is a naturally occurring, potent, and selective kappa opioid agonist. Kappa receptors are one of the canonical opioid receptors, often appreciated best for anti-opioid effects. Kappa receptors don’t interact well with morphine or other narcotics, and we’d thought that activation of kappa receptors produces a dysphoric state. Salvinorin A differs from psychedelics is that it appears to have abuse liability. Animals will work to obtain the drug, they form positive associations with contexts in which the drug is delivered, and most convincingly, salvinorin A increases dopamine levels in the nucleus accumbens.
What makes some people so much more susceptible to problematic drug use? The author believes the main factors are an inherited biological disposition, copious drug exposure, particularly during adolescence, and a catalyzing environment.
The first is that the more DNA one shares with an addict, the higher one’s risk. If the biological child of an addicted individual is adopted immediately after birth by a family with no history of addiction, he or she retains an elevated risk. Instead of “genes for addiction,” we have discovered dozens of locations on the genome where polymorphisms combine and interact to influence risk, and each variation in sequence may explain only a very small fraction (typically less than 1 percent) of the carrier’s inherited liability. If a particular sequence is more common in one group than the other, there are plenty in the other group with the sequence too. Additionally, a strand of DNA may be present in only a subset of those possessing what appears to be the same disorder - for instance, a stimulant addiction. These realities make it very difficult to locate influential genes.
Epigenetics may play a role. It’s increasingly looking as if exposure to drugs of abuse in our parents and grandparents predisposes us to take drugs ourselves - effectively a b process across generations.
Research indicates that most people with a substance use disorder started using during adolescence and met the criteria before age twenty-five.
But why might some people start using earlier? High dopamine tone correlates with enhanced sensitivity to potentially rewarding experiences, as if the message concerning something of potential value were being delivered at a higher volume and essentially drowning out drawbacks.
Future treatments? Vaccines are being developed that would render a drug of choice moot, and while such antibodies would have narrow effects, they would be permanent. The author believes that removing the drive to get high, as ancient, ubiquitous, and neurologically relevant as it is, is about as likely to happen as removing our desire to create and explore.
So, what might be an ideal cure? First, an easily administered formulation that would obviate withdrawal and craving, removing the biological necessity for a swift relapse. It’s been done with Suboxone/buprenorphine for opiate addicts, with Chantix/varenicline for smokers, and with benzos for alcoholics, to a lesser degree, because the drug is such a generalist. In each of these cases, the treatment is only effective when coupled with a slow reduction in dose and ample social support (such as Alcoholics or Narcotics Anonymous, which can provide such support for a lifetime).
Perpetual medications, deep brain interventions, antibodies, religious dogmas, or congressional edicts are all likely to be counterproductive in the long run. The principle missing in all these formulations is the opportunity for each of us to freely seek a meaningful life.
Those with a liability toward addiction in the twenty-first century might also be those most likely to survive and thrive in our distant past, or future. Even today, such factors might be assets for endeavors that benefit from an ability to tolerate - or even seek - uncertainty, or to push on the margins of conventional practice, such as entrepreneurial avocations or scientific research.
We support the tools of addiction, including pathological individualism that leads to alienation, widespread and enthusiastic endorsement of avoidance, and a smorgasbord of consumptive excess and self-medication. The author concludes that “Federal agents, friends’ deaths, expulsions and evictions, physical withdrawal, and myriad other tragedies weren’t enough to propel me to change; instead, it was human love and connection. My father’s willingness to be seen with me and to treat me with kindness split open my defensive shell of rationalizations and justifications.”
Ben’s MDMA deep dive:
The most interesting section to me was the somewhat alarming results presented about relatively long-lasting damaging effects on serotonergic neurons due to MDMA. MDMA, like psilocybin, has been increasing in its public reputation recently, particularly for its promise in PTSD, but also for its potential to generally facilitate connection in a disconnected time. So naturally, I did a deep dive of my own on MDMA effects on cognition, and it definitely left me more concerned about it than I was before reading this book, although probably less so than the author.
The problem with MDMA research (and much research on illegal substances) is that it is difficult to do randomized-controlled trials. There are a few RCTs using MDMA now in clinical populations with PTSD, but I have yet to see an RCT that has looked at memory performance as an outcome measure. Almost everything is focused on PTSD symptoms. Also, PTSD leads to reductions in certain kinds of cognitive performance and emotional processing, so I would expect that in that specific populations, the benefit to cognition from the reduction of PTSD symptomology could outweigh any impairment of cognition due to neuronal damage. It’s really strange to me that given the fact that long-term memory impairment and depression of mood are some of the biggest concerns that cognitive batteries have not been included in Phase 1 trials or other RCTs. We need to do this. And until we do, it’s hard to inform the risk-benefit ratios for different uses. In absence of RCTs, it would be great if someone could do a Mendelian Randomization study - but I’m not sure if there are clear alleles associated with MDMA use.
From my more in-depth study, I mostly came to same conclusions as the author. MDMA clearly causes damage to serotonergic neurons in animals at dosages that are probably relevant to humans (the metabolism of drugs is always a little different). There might be multiple mechanisms of toxicity, but the major one seems to just be that MDMA is ridiculously good at releasing and keeping most of the available serotonin in the synapse. That causes downstream neurons to be more active and causes oxidative stress through excitotoxicity. If the oxidative stress is too high, the neuron is damaged. There may also be toxicity through the uptake of MDMA into the pre-synaptic cell. MDMA is much more efficient at releasing serotonin than Adderall or Vyvanse are at releasing dopamine, and thus the long-term safety profile of these other pharmaceutical amphetamines seem to be better unless you use really high doses ~2 times as much as the max prescribed use.
Besides a clear mechanism, there are studies in humans looking at cognitive performance. These studies are mostly observational and bad, but this is the curse of needing to do observational studies. For one, most people who use MDMA also use other drugs, which strongly confounds our interpretation of results. Secondly, many people who do drugs lie about or underestimate their use. Despite this, observational studies have consistently shown small effects, mostly on memory and mood, after relatively low use (1-6 times per year), and even after fairly long periods of abstinence (over months to years).
One of the best studies out there that I saw was an observational study that did a really good job controlling for confounds via careful sample selection and drug testing (Halpern et al., 2011). When they did this, there were no differences in cognitive performance between the groups, but the sample was relatively small. There have also been some prospective studies (e.g. Schilt et al., 2007) that did show differences between the groups, which is stronger evidence than the purely cross-sectional design. Because you usually have learning effects, most “deficits” in memory refer to less improvements in memory at the second testing session - so the MDMA naïve group performs significantly better in the second session, and the MDMA group performs better, too, but significantly less so than the controls. Both groups performed in what is considered the normal range.
A now older, but quite good systematic review, (Rogers et al., 2009; The harmful health effects of recreational ecstasy: a systematic review of observation evidence) found that there is likely publication bias in this literature, because as the sample sizes of the studies increase, they find smaller effect sizes of the impairment due to MDMA use. This is thought to be because when small studies find large effects, they are able to publish the results, but when small studies find small effects, the result is insignificant and so they don’t publish their results. Thus, the effect of MDMA on cognitive impairments is likely over-inflated in the literature. But also, there does seem to be a dose effect, where increases in monthly use, amount used in the last year, and lifetime dose of ecstasy all seem to be associated with increases in impairments, mainly to declarative and prospective memory, which is at least some indication that the small effect is real.
My personal conclusion is that MDMA probably causes some damage, which effects cognitive performance, especially in declarative memory, but that the effect is quite small per use. That being said, it’s still alarming, especially because the narrative I’ve been hearing around it recently is that it is quite safe. After reading into it, I would advise against frequent MDMA use (even more than a couple times a year), and I would advise users to not take a second dose when they use it. But I also think that the risks and impairments are not so large that they would necessarily offset clinical or social benefits for many people. I also don’t think there is enough solid long-term data to determine whether or not the serotonergic neurons recover after periods of abstinence. I think the literature is suggestive that longer-term damage is possible, particularly in heavier-users, but that seems far from conclusive to me in low-users.