Insomnia

What insomnia actually is

A bad night of sleep is not insomnia. Insomnia is a clinical pattern: difficulty falling asleep or staying asleep, occurring at least three nights per week, persisting for three months or longer, and producing meaningful daytime impact — fatigue, mood changes, difficulty focusing, or impaired performance at work or school.

The clinical criteria matter because they distinguish insomnia from situational sleep disruption. After a stressful week, anyone can have several bad nights in a row. That's stress, not insomnia. When the same pattern persists across months and disrupts how you function during the day, that's when it crosses into clinical territory.

Acute insomnia (less than three months) is often situational and tends to resolve when the trigger does. Chronic insomnia (three months or longer) is the form that needs structured intervention — and that's the form most readers of this page are dealing with.

The DSM-5 framework also requires that the insomnia isn't better explained by another condition: untreated sleep apnea, depression with sleep disruption as a symptom, restless leg syndrome, or substance use can all produce insomnia-like presentations. A proper evaluation rules these out first.

For the rest of this page, we'll focus on chronic primary insomnia — sleep difficulty that persists after other conditions have been ruled out. Most of what's written about insomnia online treats it as a single problem with a single set of fixes. It's not. The three patterns below behave differently and respond to different protocols.

The three patterns of insomnia

Three patterns account for most chronic insomnia. Many people show one pattern dominantly. Some show a primary pattern with a secondary contributor — our diagnostic identifies both when relevant. The patterns are:

Sleep-onset insomnia

This is the pattern where you can't fall asleep. You go to bed at a reasonable time, lights off, comfortable enough — and then nothing. Thirty minutes pass. An hour. Sometimes two. Your body feels tired but your mind won't let you go.

What it feels like

Lying in bed wide awake while exhausted. Feeling more alert as bedtime approaches, not less. Looking at the clock every fifteen minutes despite knowing better. Once you do fall asleep, you typically stay asleep — the problem is the door to sleep, not what happens after you walk through it.

What's actually happening

The sleep researchers Spielman and Glovinsky formalized the dominant model decades ago: chronic insomnia involves a hyperaroused state at the moment when most people are downshifting. Your sympathetic nervous system stays elevated; cortisol doesn't drop the way it should; heart rate variability decreases. Some people are biologically prone to this state. For others, repeated frustrating nights teach the brain that bed equals trying-and-failing-to-sleep, which builds anxiety that itself prevents sleep.

What helps

• Stimulus control: bed is only for sleep and intimacy. If you've been lying awake for more than 20 minutes, get out of bed. Return when sleepy. This breaks the bed-as-battleground association.
• Sleep restriction therapy: temporarily restrict time in bed to consolidate sleep pressure. Counterintuitive but well-supported in randomized trials of CBT-I.
• Cognitive techniques: address the catastrophic thoughts ('if I don't sleep I'll fail tomorrow') that fuel the arousal loop.
• No clock-watching: turn the clock away or remove it from the bedroom entirely.
• A consistent wake time, even after bad nights. Lie-ins amplify the pattern.

What does not help much: melatonin (it's a chronobiotic, not a sedative — only useful if your timing is off, which is a different pattern). Sleeping pills can break a bad streak but don't address the underlying mechanism.

The behavioral component of CBT-I is called stimulus control therapy. We have a full protocol guide with the six rules, the timeline, and the failure modes nobody warns you about.

The other behavioral CBT-I protocol — usually paired with stimulus control — is sleep restriction therapy. Our 4-week guide walks through the calculations and the week-by-week titration.

When this needs a professional

If it's been more than three months, if you're experiencing significant daytime impact, or if you've tried behavioral approaches for two months without improvement — see a sleep specialist. CBT-I delivered by a trained clinician is more effective than self-directed attempts.

See our medical disclaimer for guidance on when professional evaluation is the right next step.

Sleep-maintenance insomnia

This pattern is the inverse: you fall asleep fine, then 3am happens. You wake. Sometimes you know why (a sound, a thought, a temperature shift). Sometimes there's no obvious trigger — you're just suddenly awake, and then alert, and then unable to return.

What it feels like

Falling asleep within fifteen or twenty minutes is rarely the problem. Waking somewhere between 2 and 4am is. Once awake, you're not groggy — you're alert, often anxious, sometimes mind-racing. The next two or three hours feel like punishment. Morning arrives feeling like you slept half a night.

What's actually happening

Several mechanisms can produce this pattern, often together:

• A premature cortisol awakening response — the hormone that normally peaks around 6am to wake you up starts ramping earlier than your scheduled morning.
• Elevated nocturnal heart rate or core body temperature — your autonomic system isn't fully descending into sleep mode.
• Alcohol within three or four hours of bed — it speeds sleep onset but fragments sleep in the second half of the night.
• Glycemic dips — for some people, blood sugar crashes around 3-4am, triggering a cortisol response.

If you're also experiencing loud snoring, witnessed breathing pauses, or daytime sleepiness despite enough time in bed, get evaluated for sleep apnea first — it's the most common medical cause of fragmented sleep and is often missed.

If you're waking at 3am specifically and can't get back to sleep, we have a dedicated guide on what to do at 3am.

What helps

• Stimulus control applies here too: if you've been awake for 20+ minutes, leave the bed. Read in dim light. Return when sleepy. Same principle as sleep-onset — don't lie there building frustration.
• Bedroom temperature in the 65-68°F (18-20°C) range tends to support second-half sleep architecture.
• No alcohol within four hours of bed — this is the single biggest lifestyle lever for sleep-maintenance insomnia. The improvement is often visible within a week.
• A small protein-and-fat snack before bed for those with glycemic-driven waking (a small handful of nuts, for example).
• Light exposure timing: bright light in the morning anchors the cortisol curve to the right hour.

When this needs a professional

Same threshold as sleep-onset, plus this specific flag: if you wake gasping, choking, or someone has observed you stop breathing — see a doctor before changing anything else. Sleep apnea masquerades as insomnia and the home remedies don't fix it.

Hyperarousal insomnia

The third pattern overlaps with the first two but has a distinct signature: the nervous system itself can't downshift. You may have onset trouble, maintenance trouble, or both — but the underlying issue is that your body and mind stay in a low-grade fight-or-flight state when they should be moving toward parasympathetic rest.

What it feels like

Racing thoughts at bedtime, even when physically exhausted. Lying down and feeling more wired, not more relaxed. Anticipating bedtime with dread because the previous nights have been bad. Anxiety about sleep that produces — and this is the cruelest part — the very arousal that prevents sleep. Researchers call this 'sleep effort' and identify it as one of the strongest predictors of chronic insomnia.

What's actually happening

Chronic insomnia rewires the relationship between your nervous system and the act of going to bed. Most healthy sleepers cue down on the approach to bed: lights dim, body cools, breathing slows, mind quiets. In hyperarousal insomnia, the opposite happens — bedtime becomes a trigger for vigilance. The bed isn't a place of rest; it's a place where the brain expects to fail at sleep.

This is the pattern where CBT-I has its strongest evidence base. Randomized trials consistently show that an 8-session structured CBT-I protocol matches or exceeds the short-term effectiveness of sleeping pills, and unlike pills, the gains persist after treatment ends.

If the main thing keeping you awake is your mind refusing to quiet down, we have a dedicated piece on why your mind races at bedtime and what actually quiets it.

What helps

• CBT-I, ideally with a trained clinician: stimulus control plus sleep restriction plus cognitive restructuring plus relaxation training. The whole package, not just one piece.
• Acceptance and Commitment approaches (ACT): instead of trying to force sleep, learn to tolerate wakefulness without escalating arousal. The paradox: when you stop trying to sleep, sleep often comes.
• Extended-exhale breathing in the 30-60 minutes before bed: a longer exhale than inhale (for example, 4-second inhale, 6-8 second exhale) activates the parasympathetic branch and is one of the few self-administered techniques with reasonable evidence.
• Reducing 'sleep effort': paradoxically, caring less about whether you sleep tonight is one of the most effective shifts.
• A non-bedroom worry window earlier in the evening: 15 minutes of explicit problem-solving on paper, hours before bed, so it isn't waiting for you when you lie down.

A note on referral

This pattern responds well to CBT-I delivered by a clinician. If you've tried self-directed approaches for two months without improvement, that's a strong indicator to escalate to professional care.

What about supplements?

You'll see supplements everywhere in the sleep conversation. The honest summary:

• Melatonin. A chronobiotic, not a sedative. It helps shift your sleep timing forward or back — useful for jet lag and circadian-rhythm misalignment — but the evidence for its use in primary chronic insomnia is weak. Most people take too much (3-10mg when 0.3-1mg is the studied range) and use it for the wrong reason.
• Magnesium glycinate. Supportive but not definitive evidence for sleep, particularly when anxiety is part of the picture. The studied dose range is 200-400mg before bed. Magnesium oxide and citrate are less well-absorbed.
• Valerian, lavender, chamomile. Small, mixed trial data. They probably do something for some people. Don't expect dramatic effects.
• L-theanine. Decent evidence for relaxation but mixed evidence for sleep specifically.
• Cannabis (CBD or THC). Evidence is poor, dependence risk is real, and tolerance builds quickly. Not recommended as a standard approach.

What's notably absent from this list: anything we can confidently recommend as a primary insomnia treatment. Supplements are at best adjuncts to behavioral approaches. If your insomnia is significant enough that you're considering supplements, the higher-leverage move is to address the pattern itself.

If you're weighing prescription sleeping pills against behavioral treatment, we have an honest comparison of CBT-I and pills that covers the trade-offs most clinicians give privately but rarely in writing.

What about sleep tracking?

Trackers measure sleep, they don't fix it. An Oura ring or Apple Watch can tell you that your sleep is fragmented or your HRV is low. It cannot tell you why. The data is only useful if you do something with it — typically by feeding it into a structured protocol.

Orthosomnia is a real thing. The term, coined by sleep researchers in 2017, describes people whose obsessive monitoring of their own sleep tracker readings makes their sleep worse. If checking your sleep score in the morning produces anxiety, or you find yourself trying to 'fix' a low score by changing behaviors arbitrarily, you've crossed from useful tracking into counterproductive tracking.

For most people with insomnia, the order matters: address the pattern first, use tracking as a feedback mechanism second. Tracking is not a treatment.

Frequently asked questions

Tools for this topic

Continue reading