Sleeping pills help many people sleep in the short term, but the best-performing ones also cause the most side-effects, and the evidence for taking them long term is thin. A 2022 analysis of 154 trials found benzodiazepines, doxylamine, eszopiclone, lemborexant, seltorexant, zolpidem and zopiclone were more efficacious than placebo for acute insomnia, [decrescenzo-2022] while a meta-analysis of older adults found the benefit small and the risks clear. [glass-2005] This guide covers what reviews found on benefits, side-effects, older adults and talking therapy, what they could not show and what to ask a prescriber. Our overview of sleep covers the wider picture.
Do sleeping pills work?
For short-term insomnia, many do, though not all drugs sold for sleep do. A 2022 network meta-analysis included 170 trials with 36 interventions and 47,950 participants, and analysed 154 double-blind randomised trials of 30 interventions and 44,089 participants. For acute treatment, benzodiazepines, doxylamine, eszopiclone, lemborexant, seltorexant, zolpidem and zopiclone were more efficacious than placebo, with standardised mean differences from 0.36 to 0.83 and certainty rated high to moderate. [decrescenzo-2022]
Benzodiazepines, eszopiclone, zolpidem and zopiclone were also more efficacious than melatonin, ramelteon and zaleplon (differences of 0.27 to 0.71). The authors’ overall verdict was that melatonin, ramelteon and non-licensed drugs did not show overall material benefits. [decrescenzo-2022] That is a finding about insomnia disorder, and it sits alongside the narrower use of melatonin for body-clock problems that our sleep overview describes.
What are the side-effects and risks?
The drugs that worked best also had the worst tolerability. In the same review, zopiclone and zolpidem caused more dropouts due to adverse events than placebo (odds ratios 2.00 and 1.79), and at the end of treatment benzodiazepines, eszopiclone, zolpidem and zopiclone left more people with side-effects than placebo, doxepin, seltorexant or zaleplon (odds ratios 1.27 to 2.78). [decrescenzo-2022]
For long-term treatment, the picture was thinner. Only eszopiclone (standardised mean difference 0.63) and lemborexant (0.41) were more effective than placebo, both with very low certainty, and zolpidem was associated with more dropouts due to side-effects than placebo (odds ratio 2.00). [decrescenzo-2022] The authors summarised that many licensed drugs can be effective in acute treatment but are associated with poor tolerability, or information about long-term effects is not available. [decrescenzo-2022]
Are sleeping pills safe for older adults?
The evidence says the risk is higher and the benefit smaller. A meta-analysis of 24 trials with 2,417 people aged 60 or over, with insomnia and no psychiatric disorders, found that sleep quality improved slightly (effect size 0.14), total sleep time rose by 25.2 minutes and night-time awakenings fell by 0.63 compared with placebo. [glass-2005]
Adverse events were more common with sedatives: cognitive events were 4.78 times more common (95% confidence interval 1.47 to 15.47), psychomotor events 2.61 times (1.12 to 6.09) and daytime fatigue 3.82 times (1.88 to 7.80). The authors concluded that in people over 60 the benefits may not justify the increased risk, particularly with additional risk factors for cognitive or psychomotor problems. [glass-2005]
Ratios for adverse events with any sedative versus placebo for at least five nights in people aged 60 or over with insomnia, Glass et al. (2005), 24 trials, 2,417 participants. 95% confidence intervals: cognitive 1.47 to 15.47, daytime fatigue 1.88 to 7.80, psychomotor 1.12 to 6.09.
How do sleeping pills compare with CBT for insomnia?
In the short term, about as well, with one difference. A meta-analysis of 21 studies with 470 people compared benzodiazepines and benzodiazepine receptor agonists (zolpidem, zopiclone, zaleplon) with behavioural treatments such as stimulus control and sleep restriction in primary insomnia. Effect sizes on subjective measures of sleep latency, awakenings, time awake after sleep onset, total sleep time and sleep quality were moderate to large for both. [smith-2002]
There were no differences in magnitude between the two on any measure except time to fall asleep, where behavioural therapy gave the greater reduction. The authors concluded that behavioural therapy and drug treatment produce similar short-term outcomes in primary insomnia. [smith-2002] Our guide to CBT for insomnia explains what the behavioural approach involves.
| Question | What the reviews found | Caveat |
|---|---|---|
| Do they work short term? | Several drugs beat placebo (SMD 0.36 to 0.83) | Same drugs cause more side-effects |
| Is melatonin better? | No overall material benefit for insomnia disorder | Not about jet lag or body-clock problems |
| Long term? | Only eszopiclone and lemborexant beat placebo | Very low certainty |
| Older adults? | 25.2 minutes more sleep, 4.78 times more cognitive events | Trials in healthy older adults, short term |
| Versus CBT-I? | Similar short-term results | Behavioural therapy faster at sleep onset |
What can this research not show?
It cannot show which drug is safe for you, or what happens when someone takes a pill for years. Five limits matter most:
- Long-term evidence is weak. The only drugs better than placebo over the long term were rated very low certainty. [decrescenzo-2022]
- Tolerability and efficacy pull in opposite directions. The review’s most effective drugs were the ones with the most side-effects, so no drug is a clear winner. [decrescenzo-2022]
- The older-adult trials were short and screened. They lasted at least five nights and included people with insomnia and no psychiatric disorders, and the search ended in 2003. [glass-2005]
- The therapy comparison is short term. It covered primary insomnia, studies up to 2000 and acute outcomes, and says nothing about which approach lasts longer. [smith-2002]
- Dependence and stopping are not measured here. None of these abstracts reports tolerance, withdrawal or rebound insomnia, so those are questions for a prescriber.
What should you ask a prescriber?
Ask what the pill is for, how long to take it and what the plan is for stopping. This section is our suggestion and is not something the reviews tested, though the cautions for older adults come from the evidence above. [glass-2005]
Useful questions are whether a course of behavioural treatment can start first or alongside, which side-effects to watch for in the morning, what to do about driving or alcohol, and whether a medicine you already take could be affecting your sleep. Our guides to why you cannot sleep and sleep hygiene cover the habits that help either way. Do not stop a prescribed sleeping pill suddenly without speaking to the prescriber.
If you are not sure whether your sleep problem is insomnia, the Insomnia Severity Index screener on this site is a short self-check. It is not a diagnosis, and it does not tell you whether a medicine is right for you.
When to seek help
Speak to a doctor if poor sleep has lasted three weeks or more, if it is affecting your days or mood, or if you are already taking a sleeping pill and want to stop or are having side-effects such as morning drowsiness or confusion. If you feel unable to stay safe, contact your local emergency services or a crisis helpline.
How MyFreud can help
A sleep record helps a prescriber more than a description from memory. Daily mood tracking in MyFreud shows how your days run over weeks, and the notepad is somewhere to note when you slept, what you took and how you felt the next day. It does not replace a doctor.
Download MyFreud and start today: App Store or Google Play.