Sleep Medication After 50: What Nobody Tells You About Long-Term Use

I have not taken sleeping medication. My sleep difficulties were metabolic, and they resolved when the metabolic inputs changed. But sleeping tablets – and the questions around them – come up constantly in the WMTWL community from people over 50 who are managing something difficult and deserve a clear-eyed look at the evidence. This is my attempt at that, with the usual caveat: specific decisions about medication belong with your GP, not with me.

---

How Sleeping Tablets Work and What They Do to Sleep

There are several classes of sleep medication in common UK use. Z-drugs (zopiclone, zolpidem – the latter sold as Stilnoct in the UK) and benzodiazepines (temazepam, nitrazepam) are the most widely prescribed for sleep. Antihistamines (diphenhydramine, found in over-the-counter products) are widely self-medicated.

The important thing to understand about all of them: they produce sedation. Sedation is not the same as the natural sleep architecture that includes slow-wave deep sleep and REM. Patrick McKeown cites research showing that sleep medications can quietly reduce slow-wave sleep – the deep restorative stage crucial for physical recovery, immune function, and memory consolidation – even as they produce apparent sleep of adequate hours. You may be in bed for eight hours, have no memory of being awake, and still be missing the slow-wave deep sleep that does the restorative work.

Matthew Walker’s group is direct about this in the sleep medicine literature: the quality of sleep induced by sleeping pills is not equivalent to natural sleep architecture. The hours count differently.

---

Rebound Insomnia: The Exit Problem

The mechanism most people encounter when they attempt to stop sleeping medication is rebound insomnia. When the brain has adapted to the sedative effect of a drug – particularly benzodiazepines and Z-drugs – it compensates by upregulating the arousal systems. Stop the drug, and the arousal system that was being suppressed rebounds, often producing worse sleep than before the medication began.

This is not a sign that medication is needed. It is a sign that the nervous system adjusted to the medication in the way nervous systems do. Rebound insomnia is predictable, temporary, and does not indicate that underlying insomnia has worsened. But it is uncomfortable enough that many people restart the medication rather than work through it, which reinforces the dependency cycle.

The practical implication is that stopping sleeping medication should be done gradually and with GP support – not because the medication is dangerous to stop quickly in a medical emergency sense, but because a managed taper produces less severe rebound and a higher success rate for eventual discontinuation.

---

The Alzheimer’s Risk Data

University of Montreal and University of Bordeaux researchers found that benzodiazepines increase Alzheimer’s risk by 32 percent after three to six months of use. Beyond six months, the risk increases by 84 percent. The association was specific to long-term use and remained after controlling for multiple confounders. This is not a minor finding and not a fringe result – it has been replicated in subsequent pharmacoepidemiological studies.

This does not mean that benzodiazepines cause Alzheimer’s in a direct causal sense – the data is observational. But combined with the known mechanism (benzodiazepines reduce slow-wave sleep, and slow-wave sleep is when the brain’s glymphatic system performs its amyloid clearance), there is a plausible biological pathway that makes the association biologically credible rather than coincidental. For adults over 50 who are already experiencing the natural decline in deep sleep that reduces glymphatic clearance, adding a medication that further reduces deep sleep warrants attention.

---

What NICE Now Says

NICE guidance on the management of insomnia in adults was updated in 2022. The guidance now recommends cognitive behavioural therapy for insomnia (CBT-I) as the first-line treatment for adults with insomnia disorder, ahead of pharmacological options. Medication is recommended only where CBT-I is not available or has not been effective, and only for short-term use.

This is a meaningful change from previous guidance that effectively positioned sleeping pills as the primary tool and CBT-I as an optional extra. The evidence base for CBT-I is now robust enough – matching medication in the short term and outperforming it at six weeks and beyond – that the guidance reflects it. If your GP is still recommending sleeping tablets without discussing CBT-I first, the 2022 NICE guidance gives you a specific, legitimate opening for that conversation.

---

A Note on the Freight Train

Robert Sapolsky’s observation about the autonomic nervous system – that it moves like a freight train and takes time to stop – is relevant here in a specific way. Sleeping medication applies the brakes to the freight train from the outside. It does not address why the train was moving too fast. When you remove the brakes (stop the medication), the train is still at speed, plus the rebound from the upregulated arousal system.

The metabolic inputs I address in the rest of this article series – food timing, light environment, cortisol conditions, morning routine – address the speed of the train itself. They are slower interventions. But they work with the biology rather than overriding it, and they do not produce dependency or rebound. For most people over 50 whose sleep difficulties are metabolic and habitual rather than clinical, they are the correct starting point.

If CBT-I outperforms sleeping pills at three years and NICE now recommends it first – is there a conversation you have not yet had with your GP?