Benzodiazepines Were Sold as the Safe Pill. Quitting Them Cold Can Still Kill You.

Nobody in Nabarun Dasgupta’s North Carolina cohort thought they were taking a risk that had a name. They were pain patients on record with the state’s prescription monitoring program, taking opioids exactly as prescribed. A subset of them also had a second bottle in the cabinet — a benzodiazepine, alprazolam or clonazepam, written by a different doctor for the anxiety or sleeplessness that comes with chronic pain. When Dasgupta and his co-authors tracked a year of outcomes across that population, people taking both drugs together died of overdose at roughly ten times the rate of people taking opioids alone. Nobody was doctor-shopping. Nobody was chasing a high. They were following two sets of instructions that were never designed to be followed at the same time.

Stopping a benzodiazepine cold can be more dangerous than the anxiety it was prescribed to treat — and mixed with an opioid, this “mild” pill is quietly implicated in most of a whole category of overdose deaths.

That’s the part almost nobody hears from the doctor who writes the script. Benzodiazepines belong to the depressants class of drugs — central nervous system depressants, meaning they slow down brain and body activity — and the family includes Xanax (alprazolam), Valium (diazepam), Klonopin (clonazepam), and Ativan (lorazepam), prescribed for anxiety, panic disorder, insomnia, and seizures. They get handed out like a minor convenience. Biologically, they’re not.

What’s actually happening in the brain

Your brain runs on a balance between excitatory signals telling neurons to fire and inhibitory ones telling them to settle. Gamma-aminobutyric acid, or GABA — the brain’s main “stop” chemical — binds to GABA-A receptors and calms activity down. Benzodiazepines don’t replace GABA; they’re positive allosteric modulators, meaning they sit next to the GABA-A receptor and turn up the volume on GABA’s own calming signal, the way a dimmer switch doesn’t add its own light, it just makes the existing light work harder. That’s why a benzodiazepine can quiet a racing mind or stop a seizure within minutes.

Take one daily for weeks, though, and the brain adapts to counteract that constant “stop” signal. That adaptation is tolerance — the reason the dose that worked in week one barely touches the anxiety by month three. A 2026 molecular review in the International Journal of Molecular Sciences traces the mechanism: chronic use down-regulates GABA-A receptor subunits, meaning the brain effectively pulls back some of the receptors the drug was working through, while the brain’s excitatory glutamate system ramps up to compensate. The New England Journal of Medicine’s clinical review of benzodiazepine dependence describes the same pattern from the treatment side: chronic exposure blunts the receptor system’s sensitivity to the drug, tolerance builds, and the body comes to depend on the benzodiazepine just to hold a normal, non-anxious baseline.

That’s the part that surprises people who think of this as “just an anti-anxiety pill”: quit suddenly after your nervous system has spent months quietly dialing down its own brakes to compensate for a drug doing that job externally, and you’re left with a brain over-tuned toward “go,” with the brakes it used to have partly gone missing. That’s the biological reason abrupt benzodiazepine withdrawal isn’t just uncomfortable — it’s one of the few withdrawal syndromes, alongside alcohol, that can trigger seizures severe enough to kill.

Why the opioid combination is worse than the sum of its parts

That same fact — a nervous system straining to hold its own baseline up — is why mixing a benzodiazepine with an opioid is such a bad bet. Opioids depress the drive to breathe. Benzodiazepines depress the reflex that would otherwise jolt a person awake to fight for air when breathing slows. Stack them and the two risks don’t add, they multiply.

The numbers track the biology. Federal overdose-tracking data reported by the National Institute on Drug Abuse put benzodiazepine involvement at roughly 10,870 overdose deaths in 2023, with an illicitly manufactured fentanyl also present in close to 70% of those deaths. That figure comes from mortality coding rather than a single peer-reviewed paper, so treat the exact digits as approximate rather than precise — but it lines up with the North Carolina cohort’s tenfold finding, and with newer research. A 2026 Finnish national cohort study of more than 48,000 benzodiazepine users, reported on by Mad in America, found overdose death risk climbing with dose: roughly three times higher at medium-to-high doses and better than six times higher at the highest doses studied, alongside a sharp rise in suicide risk. Dose matters. So does everything else already in the system.

A pattern with a sixty-year head start

None of this is a new discovery — it’s a pattern repeating on a long delay. When Librium arrived in 1960 and Valium followed in 1963, both were marketed explicitly as the safe replacement for barbiturates, the older sedative class then killing people in overdoses and withdrawal seizures at alarming rates. Valium became the best-selling prescription drug in America through much of the 1970s, so ubiquitous the Rolling Stones sang about “Mommy’s Little Helper.” It took until the 1980s for dependence and withdrawal risk to be widely recognized in the very drugs sold as the non-addictive fix. Yale Medicine notes that benzodiazepine prescriptions still climbed 67% between 1996 and 2013, and that roughly half of long-term users develop dependence within about a month of steady use. The marketing changed. The molecule’s behavior in a human brain never did.

None of this is a new discovery — it’s a pattern repeating on a long delay.

What a safe taper actually looks like

If you’re reading this because you or someone you love has been on a benzodiazepine for months or years, the instinct to just stop — flush the bottle, white-knuckle it — is the most dangerous option on the table. Don’t taper alone, and don’t taper fast. That’s not caution for caution’s sake; it’s the actual finding. A 2026 systematic review in Age and Ageing pooling thirty studies and roughly 11,000 older adults found that structured, gradual tapering was generally safe, with withdrawal symptoms typically mild and transient when the dose came down slowly on a fixed schedule — while it’s abrupt or unsupervised discontinuation where the seizure and delirium risk concentrates. The molecular review describes the standard clinical approach in more detail: switch to a longer-acting benzodiazepine if needed, taper the dose down over weeks to months rather than days, and pair the taper with therapy — cognitive behavioral therapy in particular has real evidence behind it for the insomnia and anxiety that tend to surge back once the drug is gone.

What that means in practice: if you’ve been on a benzodiazepine for more than a few weeks, the moment to ask your prescriber for the taper plan is now, not after a crisis forces the question. If you’re supporting someone through a taper, the mild-symptoms outcome in that Age and Ageing review didn’t happen by luck — it happened because the taper was slow, structured, and medically supervised, with someone checking in along the way. That combination is what separates a hard few weeks from a medical emergency; skipping it is a different, far riskier process entirely. This is also exactly the kind of dependence biology Rize’s science and medicine coverage keeps returning to, because the mechanism — not the drug’s reputation — is what determines how dangerous quitting actually is.

Librium and Valium were built to be the drug that finally let people put barbiturates down without risking a body. Six decades on, the risk didn’t disappear. It just moved into the taper — the one part of the prescription nobody hands out instructions for.