Barbiturates, Z-drugs & Other Sedative-Hypnotics
For half a century, barbiturates were the world's sleeping pills — until they became one of its most common ways to die, by accident and on purpose. Why did an entire drug class get pushed aside, and what safer pills replaced it? The answer is a single, deadly difference in how a molecule opens one channel — and it's the clearest lesson in drug safety you'll ever learn.
Flip through mid-20th-century medicine and barbiturates are everywhere — for sleep, for anxiety, for seizures. Then flip through the obituaries, and they're there too: a narrow margin between the dose that helped you sleep and the dose that stopped your breathing meant countless accidental and intentional overdoses. Today they've almost vanished from the pharmacy, replaced by safer drugs. To understand why, we return to the GABA-A channel from the last article — because barbiturates open it in a subtly, and fatally, different way.
Duration, not frequency — and no ceiling
Barbiturates also work on the GABA-A channel — differently. Where a benzodiazepine increases how OFTEN the chloride channel opens, a barbiturate increases how LONG it stays open each time — deepening inhibition. (A handy memory: benzodiazepines = Frequency, barbiturates = Duration.) But here is the killer difference: at higher doses, barbiturates can force the channel open entirely on their own, WITHOUT any GABA. That removes the safety ceiling benzodiazepines have. There is no upper limit to how far a barbiturate can shut the brain down — so a large enough dose simply stops breathing. That single fact explains the deaths, and the class's decline.
Barbiturates haven't disappeared entirely. Phenobarbital is still used for certain seizures (and in some settings for neonatal seizures). Thiopental, ultra-short-acting, was long used to induce anaesthesia — and it's the classic example of redistribution from the pharmacokinetics series, where its effect ends as it moves from brain to fat. Barbiturates also strongly INDUCE liver enzymes (speeding the metabolism of other drugs), have a narrow therapeutic index, and cause dependence — a stack of reasons they've been retired from everyday use.
The benzodiazepine-vs-barbiturate story is the textbook lesson in the therapeutic index. Both enhance GABA, but the benzodiazepine's ceiling (it needs GABA present) gives it a WIDE safety margin, while the barbiturate's ability to open the channel alone gives it a DANGEROUSLY NARROW one. When two drugs do the same job, the safer mechanism wins — which is why benzodiazepines, and then even safer hypnotics, replaced barbiturates almost entirely.
- Barbiturates increase channel-opening DURATION; benzodiazepines increase FREQUENCY.
- At high doses barbiturates open the channel WITHOUT GABA → no ceiling → fatal overdose.
- Still used: phenobarbital (seizures), thiopental (anaesthesia induction — redistribution).
- Barbiturates strongly INDUCE liver enzymes and cause dependence.
The safer replacements
Modern insomnia care leans on gentler drugs. The Z-drugs — zolpidem, zaleplon, and eszopiclone — act on the same GABA-A receptor but at a subtype that produces sleep with less of the anti-anxiety and anticonvulsant baggage, giving a cleaner hypnotic with less dependence (though they can cause odd sleep behaviours like sleep-walking or sleep-driving). For anxiety without sedation or dependence, buspirone works on serotonin (a 5-HT1A partial agonist) rather than GABA — but it takes weeks to work, so it's for ongoing anxiety, not a panic attack. Other options include melatonin-receptor agonists (like ramelteon) for insomnia, sedating antihistamines sold over the counter, and newer orexin-blockers that dial down wakefulness.
A patient with generalized anxiety could take a benzodiazepine (fast relief, but sedation and dependence) or buspirone (no sedation, no dependence, but takes weeks to build up). The choice shows a recurring trade-off: the fast GABA drugs relieve now but carry risk, while the slower serotonin route is safer for long-term daily use. Buspirone is useless for an acute panic attack precisely because it isn't fast.
- Thinking benzodiazepines and barbiturates are equally safe. Barbiturates lack the safety ceiling.
- Expecting buspirone to stop a panic attack. It takes weeks — it's not a rescue drug.
- Forgetting barbiturates induce liver enzymes, weakening other drugs (e.g., contraceptives).
- Assuming Z-drugs are harmless. They can cause complex sleep behaviours and some dependence.
What makes a barbiturate overdose more likely to be fatal than a benzodiazepine overdose?
- Barbiturates prolong GABA-A channel opening and, at high doses, open it alone — no safety ceiling.
- That narrow margin made overdoses fatal, so safer drugs replaced them.
- Z-drugs (zolpidem) are cleaner hypnotics; buspirone treats anxiety via serotonin, no dependence (slow).
- Phenobarbital (seizures) and thiopental (anaesthesia) are the surviving barbiturate uses.
- Katzung BG. Basic & Clinical Pharmacology — Sedative-Hypnotic Drugs (barbiturates, Z-drugs, buspirone).
- Brunton LL, et al. Goodman & Gilman's The Pharmacological Basis of Therapeutics — Hypnotics & sedatives.
- Rang HP, Dale MM, et al. Rang & Dale's Pharmacology — Anxiolytic & hypnotic drugs.
- Whalen K. Lippincott Illustrated Reviews: Pharmacology — Barbiturates & non-benzodiazepine hypnotics.
- Miller RD. Miller's Anesthesia — Thiopental & barbiturate pharmacology.
