Few topics in sleep medicine generate as much confusion as sleep medications. Patients arrive in clinic either terrified of them (“I don’t want to get addicted”) or frustrated that previous prescriptions didn’t work (“The Ambien stopped helping after a month”). Both reactions are understandable, because the conversation around sleep pharmacotherapy has been messy for years.

Let’s sort through what’s actually available, what the evidence says, and when medication makes sense versus when it doesn’t.

The Older Generation: Benzodiazepines

Drugs like temazepam (Restoril), triazolam (Halcion), and lorazepam (Ativan) were the go-to sleep medications for decades. They work by enhancing GABA activity in the brain, producing sedation, muscle relaxation, and anxiolysis.

They’re effective. That’s not debatable. The problems are:

• Tolerance develops. Most patients find the same dose becomes less effective over weeks to months.
• Dependence risk is real. Stopping benzodiazepines after regular use can produce rebound insomnia that’s worse than the original problem, along with anxiety, irritability, and in rare cases, seizures.
• Cognitive effects. Particularly in older adults, benzodiazepines are associated with increased fall risk, daytime sedation, and potential links to cognitive decline.

The American Geriatrics Society’s Beers Criteria explicitly recommends against benzodiazepine use in adults over 65 for these reasons. That doesn’t mean they’re never appropriate — short courses for acute situational insomnia can be reasonable — but they’re no longer first-line therapy for chronic insomnia.

The Z-Drugs: Better, But Not Perfect

Zolpidem (Ambien), zaleplon (Sonata), and eszopiclone (Lunesta) were developed to offer the sleep benefits of benzodiazepines with fewer downsides. They’re more selective in which GABA receptors they target, which was supposed to mean fewer side effects.

In practice, the advantages are real but modest:

• They do produce less next-day hangover than most benzodiazepines
• Zaleplon, with its ultra-short duration, is particularly useful for sleep-onset insomnia without morning grogginess
• Eszopiclone has the most data supporting longer-term use

But Z-drugs share some of benzodiazepines’ problems. Tolerance can still develop. Complex sleep behaviours — sleepwalking, sleep-eating, even sleep-driving — occur more frequently than initially reported, and the FDA has added boxed warnings to reflect this.

My take: Z-drugs have a role, especially for short-term or intermittent use. But they’re not the answer for chronic, every-night insomnia.

The Newer Options: Orexin Receptor Antagonists

This is where things get interesting. Suvorexant (Belsomra) and lemborexant (Dayvigo) represent a genuinely different approach. Instead of sedating the brain, they block orexin — the neurotransmitter system that promotes wakefulness. You’re essentially quieting the wake signal rather than amplifying the sleep signal.

The clinical implications are meaningful: lower abuse potential, less cognitive impairment (important for older patients), no significant tolerance development in studies lasting up to 12 months, and effectiveness for both sleep onset and maintenance.

I’ve been increasingly reaching for these medications when pharmacotherapy is warranted for chronic insomnia. They’re not as immediately powerful as a benzodiazepine — patients expecting to be knocked out will be disappointed — but for sustainable, longer-term improvement, they’re a genuine step forward.

Melatonin and Melatonin Receptor Agonists

Over-the-counter melatonin is the most widely used sleep supplement in the world, and it’s… complicated.

For circadian rhythm disorders — delayed sleep phase, jet lag, shift work — melatonin has decent evidence. The timing of the dose matters more than the dose itself. Most people take far too much; 0.5-1mg taken 2-3 hours before desired bedtime is usually sufficient and mirrors physiological levels.

For generic insomnia, the evidence for melatonin is weak. Effect sizes in clinical trials are small. But the safety profile is excellent, which is why it remains popular.

Ramelteon (Rozerem) is a prescription melatonin receptor agonist with more targeted activity. It’s genuinely helpful for sleep-onset insomnia and has zero abuse potential. It won’t help if your main problem is waking at 3 AM and not getting back to sleep.

The Elephant in the Room: CBT-I

The American College of Physicians recommends cognitive behavioural therapy for insomnia (CBT-I) as first-line treatment for chronic insomnia. Not medication. CBT-I.

The evidence is unambiguous: CBT-I produces equivalent short-term results to medication and superior long-term outcomes. Benefits persist after treatment ends, with no tolerance or dependence.

The catch? Access is limited. There aren’t enough trained providers, and insurance coverage is inconsistent. Digital CBT-I programs are helping close the gap, but we’re not there yet.

A Practical Framework

Here’s how I think about it:

1. Acute insomnia (less than 3 months, triggered by identifiable stress): Short-course Z-drug or low-dose benzodiazepine is reasonable. Combine with sleep hygiene counselling.

2. Chronic insomnia: CBT-I first. Always. If medication is needed as an adjunct, an orexin receptor antagonist is my preference for ongoing use.

3. Insomnia with comorbid anxiety or depression: Treat the underlying condition. Sometimes a sedating antidepressant like trazodone or mirtazapine can address both problems with one medication.

4. Older adults: Avoid benzodiazepines. Avoid high-dose Z-drugs. Consider ramelteon or low-dose suvorexant. Prioritise CBT-I and non-pharmacologic approaches.

There’s no shame in needing a sleep medication. But the goal should always be the minimum effective intervention for the shortest necessary duration, with ongoing attention to the behavioural and environmental factors that medication alone can’t fix.