How to Overcome Addiction to Substances or Behaviors | Dr. Keith Humphreys

When someone says they want to quit, a good clinician asks why and helps them build their own reasons to change; it’s their journey. Change sticks best alongside others on the same path because support plus accountability makes follow-through real.

Welcome to the Huberman Lab Podcast. Today with Dr. Keith Humphreys, we dive into alcohol, cannabis, opioids, gambling, and how to think clearly about risk and recovery across an economy of addiction-for-profit; to kick us off, how do you define addiction for science and policy?

People toss around “addicted,” but clinically it’s not just frequent use; it’s continuing despite clear harm. Think of persisting with a reward even as it wrecks health, work, and relationships—that persistence is the hallmark.

I’ve framed it as a progressive narrowing of what brings pleasure—does that fit?

Yes; natural rewards fall away until the drug or behavior is the only reliable source of feeling good, with both physical and psychological dependence tightening the loop.

Is there a genetic tilt toward addiction, and can we use alcohol as the first example?

Babies aren’t born addicted, though they can be physically dependent; risk, however, can be present from birth. Family and twin studies show substantial heritability, with traits like impulsivity and sensation seeking increasing cross-substance risk, and specific variants shifting alcohol response.

On labels, I’ll use alcoholism here since many in recovery do; is there also a subgroup that experiences alcohol as energizing and is at higher risk?

“Alcohol use disorder” spans mild to severe, while addiction is the severe end; that broader category invites early intervention. Regarding risk, low sensitivity to alcohol’s negative effects—less dizziness, fewer hangovers—removes natural brakes, which can propel heavy use.

Drug liking is highly individual and often genetic; some find opioids miserable and others feel instantly complete, and those differences shape who is vulnerable to which substance.

How can someone gauge their predisposition, given individual variability and early-first-drink risk?

Zero use guarantees zero drug harm; beyond that, family history—especially parental alcoholism—is the strongest single predictor we have, stronger than genetic panels.

Does that risk cross sex as reliably?

There’s cross-sex risk, but father-to-son is the strongest pattern; exposure matters because genes can’t act without behavior.

Are women drinking more, and did industry play a role?

Yes; marketing deliberately targeted women—think mommy-wine culture—and it worked, even though harm per drink is often higher for women; heavy users drive profits, with a small slice of people consuming a massive share of alcohol.

Overall drinking looks down lately—true?

Younger cohorts are showing broadly lower risk behaviors, so they may be a drier generation.

Is cannabis up among them, or is that just a swap story?

Youth use hasn’t shifted much; growth is mainly among adults, including people returning to use.

Back to alcohol: is zero better than any, and how should people weigh heart versus cancer risks amid media flip-flops?

Red wine health claims were overhyped; the J-shaped curve was confounded by former problem drinkers in the non-drinking group. Even if there’s a small cardiac benefit, it’s outweighed by cancer risk; two standard drinks per week adds a tiny risk, but alcohol isn’t a health food.

Many drink for relaxation and social ease; how should they balance that against risk and social pressure to drink?

Adults often accept small health risks for valued experiences like shared meals and conversation; that’s a personal trade. Social pressure to explain not drinking is real, and clear health information gives people a socially acceptable reason to opt out.

Why do some feel uneasy when others abstain, beyond social anxiety?

Group drinking can feel like a mutual vulnerability pact, so a sober person may seem untrustworthy or advantaged; in everyday settings, though, abstaining should be a non-issue.

Netting it out: consider genetics, accept that zero is healthiest, weigh any heart talk against cancer risk, and factor in personal and others’ vulnerabilities; now, cannabis—recreational, medical, and misuse in the era of legalization and edibles?

Today’s cannabis is not the old drug; THC jumped from low single digits to about twenty percent on average, and many now use daily or almost daily, driving brain exposure roughly sixty-five times higher than prior patterns. Some medical uses exist, CBD already helps certain seizures, and research access has improved, but overall it’s a riskier product than decades ago.

A researcher argued smokers self-regulate their high while edibles drive overshoot; does that blunt the potency issue?

People are surprisingly poor at judging potency, and edibles add delayed onset and uneven dosing, so overshooting is common; potency and daily use have strengthened evidence for psychosis risk, especially in those with family history of psychotic disorders. Risk is far higher during adolescent brain development than with later-life initiation.

What about life outcomes—the classic failure-to-launch among heavy users?

Cannabis tends to degrade performance—memory, attention, and motivation suffer—and some heavy users lose drive for school, work, and teams; in a competitive world, that makes it easy to stall at home.

Exceptional high achievers who use can mislead others; citing them can be dangerous messaging.

Exactly; rules should not be built around outliers with unique resources and incentives.

Many industries now optimize for addiction, from booze to apps.

Product teams openly chase addictiveness because an addicted customer is the best customer; that’s why temptation goods need smart regulation.

“Not as bad as alcohol” gets used to justify cannabis—why is that flawed?

Alcohol kills roughly one hundred fifty thousand Americans yearly, so using it as the bar is absurd; tighten ads, raise prices, curb youth exposure, and regulate hard. Gambling shows what happens without guardrails: sports-betting ads everywhere and young people nuking finances.

Gambling addicts say every addiction becomes a bet, and some get hooked on losing’s shame.

Casinos engineered timing and novelty to the millisecond; machine play replaces dealers to deliver perfect reinforcement, and people can dissociate for hours.

Slots now monetize novelty more than money; “wins” keep people playing even while net losing.

They call it losses disguised as wins—celebration of partial returns masks steady loss, and casinos further bundle smoking and free drinks into a dense addictive environment.

Policy-wise, what’s the difference between decriminalization and legalization, and is cannabis a gateway?

Decriminalization stops punishing users; legalization creates markets that boost consumption through marketing and access, now widespread, including hemp-derived loopholes. Any early substance—alcohol, nicotine, cannabis—can be a gateway via liking, social networks, and brain sensitization; the bigger lie is pretending alcohol isn’t a drug when it’s often the main youth risk.

Psychedelics next: let’s focus on psilocybin and LSD for depression and addiction in guided settings.

There’s hype and genuine promise; small trials in structured prep–dose–integration formats look encouraging, and classic psychedelics have very low addiction potential, though some people have rough experiences and flashbacks.

Microdosing evidence is weak; high-dose clinical work shows more signal, but real-world practice is uneven.

Oregon licenses facilitators, which helps; we need rigorous, impartial trials and sober reporting, not evangelism, and those studies are now being funded.

Practicalities shape research too—LSD sessions are long, which is why psilocybin often gets the nod.

Session length also limits healthcare scale, so shorter, effective options matter.

Teams are exploring non-hallucinogenic analogs to keep benefits without the trip.

Ketamine work is testing whether blocking the dissociation preserves antidepressant effects; some patients prefer no visions if relief remains.

Zooming out, all therapies work by driving plasticity, but undirected plasticity can cut both ways.

Plasticity is richest—and riskiest—when young, which is why addictions almost always start early and why marketers target kids to cement lifelong habits.

Ketamine is popular for depression but carries real abuse and safety risks.

It’s approved but results are mixed, addiction is real, and bladder damage is a serious concern; I’d look first to accelerated, theta-burst TMS protocols, which show strong effects with minimal downsides.

TMS is noninvasive brain stimulation and is already approved for depression; how available is it?

Clinics with the equipment can deliver it and many insurers cover it; Nolan Williams’s compressed protocol is compelling, and unlike psychedelics, sham-controlled blinding works well in these trials.

And to be clear, SSRIs have helped many, including people with severe OCD; no need to demonize effective medicines.

Agreed.

While we’re on SSRIs, do they make people violent or suicidal?

I don’t see SSRIs as the driver of mass shootings; the far easier access to high-powered weapons in the United States fits the pattern better. With teens, suicide risk is a real but small concern, and many still benefit.

There’s a loud cohort reporting lasting sexual and mood side effects after SSRIs or finasteride, and medicine can take years to catch up.

Most drugs are approved on short trials, so rare or long-term effects are hard to detect; post-marketing signals help but are imperfect.

On ibogaine, our late colleague Nolan Williams ran tightly supervised, marathon sessions for veterans with PTSD or addiction and saw striking remissions; could it be the first legal psychedelic?

Nolan did groundbreaking imaging that documented brain change, but his study was open label; we need controlled trials to separate chemistry from ceremony and context.

Stimulants: I love caffeine and run a huge tolerance—am I addicted, and how worried should we be?

Coffee is rewarding and can be habit-forming, but true, damaging caffeine addiction is rare. The bigger letdown is stimulant addiction treatment overall—no solid medications yet, and contingency management with immediate rewards is what reliably helps.

What about widespread amphetamine prescriptions for kids with ADHD, and parents’ worries about growth, sleep, and appetite?

Some kids’ lives are transformed, and others are over-medicated; we also medicalize normal variation and make parents anxious. These are tough, case-by-case calls.

A psychiatrist argued untreated ADHD raises addiction risk more than treated ADHD.

That’s plausible; ADHD is common among adults with alcohol or drug problems.

Nicotine seems to sharpen and relax, yet gums and skin suffer, and some claim neuroprotection against Parkinson’s or Alzheimer’s; what’s your take?

Nicotine is toxic and highly habit-forming; many users confuse relief of withdrawal with a drug benefit. If you can, ride out withdrawal with support, and apply the same logic to cannabis or opioids.

Many want to avoid addiction or wonder if they’re in an early spiral with phones, food, or substances; how do you approach early versus late stage?

Lead with empathy and hope, then clarify their own reasons to change. Map cues and high-risk situations, adjust the environment, build practical skills, and plug into a recovery community for support and accountability.

Addiction was long framed as a character flaw, and loved ones get hurt, yet sobriety often reveals the person beneath.

We must validate the harm while recognizing addiction as illness; people do things they would not otherwise do, and families deserve that reality acknowledged.

Given visible street addiction and mental illness, how do we help people living outside who won’t walk into meetings?

Pair decent housing with recovery culture, like Oxford House, and use civil commitment or drug courts when judgment is profoundly impaired, with safeguards and quality care.

Policy-wise, where are we now, and what actually helps families seeking care?

Parity laws since two thousand eight improved private coverage, and Medicaid expansion became the backbone of public addiction care; looming Medicaid cuts worry me, but insurers can no longer carve out skimpy mental health benefits.

For those without insurance or who avoid facilities, does 12-step really work, and what about the other fellowships?

AA is instantly accessible and free, and rigorous reviews show about 50 percent higher abstinence than top therapies when abstinence is the goal; drug-focused groups look positive but the evidence base is smaller.

Some call AA a cult or fear the higher power language.

AA refuses big money, never traps members, and defines higher power flexibly; the non-negotiable part is admitting you are not in control of your drinking.

Could brain stimulation retune reward circuits away from cues and toward healthy rewards?

Noninvasive RTMS is being tested for cannabis, alcohol, and cocaine; implants are rare but possible, and we should expect more protocol work because it targets the circuits directly.

GLP-1 drugs are reshaping weight loss.

Semaglutide and related GLP-1s may reduce wanting for alcohol in some, offering a two-for-one on weight and drinking; they have long safety histories, and new studies are underway.

What about drug ads everywhere?

Only the United States and New Zealand allow broad pharma ads, which can stoke demand and, via backdoor messaging, nudge risky prescribing; we called that out in a Lancet Commission.

Do you see true social media or gaming addiction, and what helps when total abstinence is unrealistic?

People feel trapped because everyone is online; societal steps like age bans or shared norms could help, and on the personal side, lockboxes, dumb phones, and no-phone zones are practical brakes.

I’ve seen young men quit games, rebuild attention, leave meds, and regain agency, then thrive in school and life.

There are many paths out—parenthood, time away, or a new routine can flip the script—and most recoveries happen outside clinic walls.

You worked in hospice; what did being close to death teach you?

Hospice staff are strikingly upbeat because the worst is accepted and the focus becomes a good death; being the last friend is a privilege, and steady, informed presence helps families meet the end with calm and care.

In places where death is out in the open, there’s oddly less fear; in wealthier, high tech settings we hide it and dread it, so you have to push past those norms to truly be with people who are dying.

Something I’ve wrestled with: is addiction a way to outrun our awareness of mortality, and could leaning into the reality of death help people recover?

It’s a compelling angle, but the drive to use is usually broader than fear of death; people are escaping pain like trauma, abuse, or a collapsing relationship, and drugs briefly lift them out of that. The hard part in recovery is that the hurt remains, which persuades some to keep using rather than face it directly.

Listener questions we didn’t cover: are men truly more prone to addiction, or do they just seek help more?

Across cultures, men use more and are overrepresented in major addictions—opioids roughly four men for every one woman, alcohol about sixty to forty—while prescription drug problems are closer to half and half.

Why is lying so entwined with addiction, beyond hiding the use itself?

It’s not a special lying circuit; people lie because the truth would expose failures at home or work, so they cover missed pickups or missing money. We also sometimes prompt dishonesty by asking judgmental questions in clinic, and people learn to answer in ways that avoid a harsh reaction.

Can relapse hit when life is going great, not just when it’s rough?

Yes; I’ve seen someone land a dream job after sobriety and immediately return to drinking with tragic consequences, because success can trick you into thinking the danger is gone. That said, stress is the most common setup—small frictions like a fight or poor sleep, or heavier burdens like a child’s addiction.

Last one from me: as a dad of two in college, what counsel do you give about avoiding addiction?

I hammer home the fentanyl risk and the reality of counterfeit pills, and I tell them never to take anything they didn’t themselves obtain because you can’t know what it is. Your strongest leverage is to never start; after that first use, no one can predict where it will lead.

Keith, thank you for bringing deep knowledge and real heart to this conversation. It’s rare to see that level of expertise paired with such compassion in day to day care, and I’m grateful we could share it with people who can use it.

Thank you, Andrew. It was a pleasure to be here.

For links to Keith’s work, check the show notes. Please share questions or guest ideas in the YouTube comments. I read them. And thanks, as always, for caring about science.