“Almost every patient arrives with a bag of things that didn’t work. Nine times out of ten the products were fine. They were just built for a different half of the night.”
There are two ways sleep fails, and they have almost nothing in common.
Either the night won’t start. You are in bed at eleven and still awake at one. The body is tired; the mind is still working.
Or the night won’t hold. You drop off in minutes, then surface between two and four, alert, heart going slightly too fast, and stay there.
The second is the more common of the two in adults, and it gets more common with age. Almost every product on the shelf is built for the first.
So before you spend another twenty pounds: which of those two is actually happening to you? Get that wrong and nothing else matters — you can buy the best product in Britain and it will do nothing, because it was made for somebody else’s problem.
These are the five things people reach for first. Each one is aimed at the wrong moment in the night.
Governs when you fall asleep, not whether you stay asleep. It has also largely cleared your system by the early hours. In the UK it is prescription-only.
They produce unconsciousness, not sleep. Architecture flattens, mornings get heavier, and tolerance builds quickly.
Sensible for falling asleep. Irrelevant to a wake-up at 3AM, because the trigger fires while you are already unconscious.
Helps you get to sleep. Does almost nothing at three, because that wake-up is physiological, not psychological.
Frequently the actual culprit. Alcohol sedates early, then rebounds as it is metabolised — spiking you awake at close to exactly 3AM.5
What is actually happening at 3AM
Cortisol — the body’s main alertness hormone — should sit at its lowest around midnight and climb gently toward morning. In people with a dysregulated stress system it starts climbing too early and too steeply, pulling them out of deep sleep in the second half of the night.1 That is why the wake-up is so clock-like. It is not random. It is a hormone with bad timing.
There is a second half to it. GABA is the brain’s brake pedal — the signal that holds you down in deep sleep. When GABA activity is low, the brain cannot keep itself settled, so even a modest cortisol bump is enough to tip you fully awake.2
Once you see the mechanism, the goal changes. You are not trying to knock the patient out. You are trying to keep the nervous system calm enough that a normal cortisol rise doesn’t wake them.
And this is precisely why sedation disappoints. A sedative carries you across the threshold at eleven. It has nothing to say about what your adrenal glands are doing at three.
What about the tablets from your GP?
They work. If sleep has collapsed entirely they have a legitimate place, and no one should feel bad about needing them.
But UK prescribing guidance is deliberately narrow — lowest effective dose, shortest possible course, usually no more than two to four weeks.3 That restriction exists for reasons worth understanding before you get there rather than afterwards.
They get weaker. The dose that worked in week one does less by week four. That creates pressure to increase it, or to move to something stronger.
Stopping is its own event. After regular use, sleep can get temporarily worse than it was before you started. Patients read that as proof they needed the medication. It is usually withdrawal.
Mornings cost you. Next-day impairment is common, matters for driving, and in older adults raises the risk of falls.
None of that is an argument against prescription medicine. It is an argument about order. The gentler options carry no tolerance and no withdrawal, so trying them first costs you three weeks and nothing else — and the stronger ones will still be there afterwards.
One more thing that rarely gets mentioned in articles like this: UK guidance puts cognitive behavioural therapy for insomnia (CBT-I) ahead of medication for persistent insomnia.3 It is unglamorous, it takes several weeks, and it has the best long-term evidence of anything on this page. It is available on the NHS and through app-based programmes.
What to look for instead
The strongest single herb in the category, and underrated. Nytol Valerian and Kalms Night both hold Traditional Herbal Registration, an MHRA standard ordinary supplements do not have to meet. What valerian will not do is reach into the small hours.
Ashwagandha is the most studied botanical for lowering the cortisol response: standardised extract reduced cortisol by around 28% over 60 days in placebo-controlled trials.4 Research uses 300–600 mg, and it builds over two to three weeks — nothing happens on night one. Alongside it, lemon balm, hops and motherwort support GABA activity and settle the hyper-alert state,2 while B6 and B12 help you wake clear-headed rather than heavy. What you want is one formula covering both pathways rather than two products stacked together; Hypnozan is one of the products that best matches this mechanism.
No melatonin and no sedatives — if the label is selling unconsciousness, it is solving the wrong half of the night. Be wary of gummies, too: many contain more sugar per serving than active ingredient.
The gap is the whole point. The best ingredient for falling asleep does nothing for staying asleep, and the one aimed at staying asleep will not help you drop off.
Before you buy anything
Spend two weeks noticing. Not tracking apps, not wearables — just noting when you wake and how long you stay awake.
If you are lying awake at the start, you have an onset problem, and a pharmacy valerian is where I would begin. If you fall asleep within twenty minutes and surface at three, stop buying things designed to help you fall asleep. And if it is both, look for one formula covering both pathways rather than stacking two products and guessing which one is working.
Then give whatever you choose three weeks before judging it. Sedatives declare themselves on night one; anything working on the stress axis does not. Most people quit on day five and never find out.
If none of that shifts it, see your GP — and ask about CBT-I before you ask about tablets. Early-morning waking can also come from low mood, thyroid problems, sleep apnoea or the perimenopause, and no supplement addresses any of those.
- Nocturnal stress-hormone levels and sleep-wake states in chronic insomnia — ScienceDirect (2024)
- Lemon balm, hops and valerian: GABAergic modulation and sleep quality — NCBI PMC11510126; double-blind placebo-controlled trials, ScienceDirect
- NICE clinical guidance on insomnia: CBT-I as first-line; hypnotics at the lowest effective dose for the shortest period
- Adaptogenic effects of ashwagandha root extract (RCT) — NCBI PMC6979308; systematic review and meta-analysis — NCBI PMC8462692
- Alcohol, cortisol and sleep-architecture fragmentation — Early Morning Awakening Syndrome review (2025)
This article is for information only and is not medical advice, diagnosis or treatment. Food supplements are not a substitute for a varied diet and healthy lifestyle. Do not start, stop or change prescribed medication without speaking to your doctor. If sleep problems persist, consult your GP.
This article is for general information and is not medical advice. If a health problem is affecting your daily life, speak to your GP.
Comments
We read every comment. Be kind, stay on topic, and please don't ask for medical advice — we can't give it.