Insomnia

Insomnia is dissatisfaction with sleep quality or quantity associated with 1 or more of: difficulty falling asleep, difficulty staying asleep, or early morning awakening without being able to return to sleep. The sleep disturbance must cause distress or impairment in daytime functioning to qualify as a disorder — poor sleep without daytime consequences is not, by itself, insomnia disorder. When these criteria are met and present at least 3 nights per week for at least 3 months, and are not substance-related, the diagnosis is chronic insomnia disorder; when the same criteria are met but the disturbance has lasted under 3 months, it is short-term insomnia disorder (sometimes still called acute insomnia).

Insomnia can be a standalone disorder, a symptom, or comorbid with another sleep disorder or a psychiatric or medical condition (sleep apnea, depression, anxiety, chronic pain). Identifying the presence or absence of comorbid conditions is central to management, since comorbid insomnia is common and usually needs the comorbidity addressed alongside the insomnia itself, not instead of it.

Behavioural intervention — culminating in cognitive behavioural therapy for insomnia (CBT-I) — is the foundation of treatment and the genuine first-line therapy for essentially every patient, not a fallback used only when medication fails or is declined.^[1] Hypnotic medications are comparably effective to CBT-I during acute treatment, and combining pharmacotherapy with CBT-I is an effective, reasonable approach when medication is needed — but CBT-I, not a drug, is where treatment starts.^[2]

Core symptom domains

• Sleep-onset difficulty — prolonged sleep latency (time to fall asleep).
• Sleep-maintenance difficulty — frequent or prolonged awakenings during the night (wakefulness after sleep onset).
• Early morning awakening — waking earlier than desired or intended and being unable to return to sleep.

Required daytime impairment (at least 1 domain)

• Fatigue or malaise.
• Attention, concentration, or memory impairment.
• Impaired social, family, occupational, or academic performance.
• Mood disturbance or irritability.
• Daytime sleepiness.
• Reduced motivation, energy, or initiative.

Insomnia develops and persists through 3 overlapping factors: predisposing factors that increase baseline risk (an anxious or ruminative temperament), precipitating factors that trigger an episode (an acute stressor, illness, or life change), and perpetuating factors that maintain the problem after the trigger resolves (learned negative sleep behaviours — excess time in bed, irregular schedule, sleep-related anxiety). CBT-I works primarily by dismantling the perpetuating factors.

• Anxious, ruminative, or worry-prone predisposition; family history of insomnia.
• Female sex; older age.
• Shift work, irregular sleep-wake schedule, jet lag.
• Psychiatric comorbidity — mood disorders, anxiety disorders, ADHD, post-traumatic stress disorder, alcohol or substance use disorder.
• Medical comorbidity causing nocturnal discomfort or pain — arthritis, fibromyalgia and other chronic pain, menopause, gastroesophageal reflux disease, migraine, stroke, COPD or asthma, congestive heart failure, hypothyroidism or hyperthyroidism.
• Medications that disrupt sleep — stimulants, some antidepressants, antihypertensives, decongestants and antihistamines, analgesics, corticosteroids, and sedatives themselves (via rebound/withdrawal effects).
• Caffeine, alcohol, nicotine, or recreational substance use, particularly close to bedtime.
• Poor sleep hygiene and stimulus-control breaches (irregular schedule, screens/work in bed, napping, an unfavourable sleep environment).
• Pregnancy and the postpartum period.
• An untreated primary sleep disorder — obstructive sleep apnea, restless legs syndrome, a circadian rhythm disorder, or a parasomnia — presenting with an insomnia complaint.

Sleep history

A structured sleep history is essential to quantify daytime impairment, determine what previous pharmacologic and nonpharmacologic treatment has already been tried, and screen for other sleep pathologies and substances affecting sleep.

Sleep diary

A 1-2 week sleep-wake diary tracking bedtime, sleep latency, awakenings and their likely cause, wake time, out-of-bed time, and a same-day rating of daytime functioning is useful for both initial assessment and ongoing CBT-I (particularly sleep restriction therapy, see Non-Pharmacologic Treatment). Take care not to let ongoing diary-keeping itself reinforce maladaptive preoccupation with sleep — refocus the patient on daytime function rather than nightly sleep totals.

Insomnia Severity Index (ISI)

The ISI is a brief, validated 7-item questionnaire assessing the severity of sleep-onset and sleep-maintenance difficulty, sleep satisfaction, noticeability and distress of the sleep problem to others, and interference with daily functioning — useful both to characterize severity at baseline and to track response to treatment.^[3] Each of the 7 items is rated from 0 (no problem) to 4 (very severe problem), for a total possible score of 0-28.

Total score interpretation: 0-7 no clinically significant insomnia; 8-14 subthreshold insomnia; 15-21 clinical insomnia of moderate severity; 22-28 clinical insomnia of severe severity. The Pittsburgh Sleep Quality Index is a similar validated option; either is used alongside — never in isolation from — screening for comorbidities.

Objective measurement

Actigraphy (average motor activity measured over days to weeks via a wrist-worn accelerometer, often housed in a consumer smartwatch) can supplement a sleep diary but should not replace clinical assessment — its accuracy can be affected by medications or other sleep disorders, and consumer devices can exaggerate or misrepresent findings, which can itself reinforce maladaptive beliefs about sleep if not reviewed with a clinician. Any actigraphy or commercial wearable-sleep data is a rough estimate that needs clinical context, not a standalone diagnosis.

Polysomnography (PSG) is not indicated in the initial assessment of routine, uncomplicated insomnia. Reserve referral for PSG for patients with risk factors or symptoms of another sleep disorder (sleep apnea, periodic limb movements), resistant or chronic cases, and patients being considered for long-term hypnotic therapy. Home sleep apnea testing is an accepted option for uncomplicated suspected obstructive sleep apnea and is standard practice outside large urban centres, but it can only confirm moderate-to-severe obstructive sleep apnea — it does not diagnose the insomnia itself and a negative home test does not rule out a sleep disorder requiring in-lab PSG.

Screen for comorbidities

Screen every patient with insomnia for other sleep disorders, contributing medications, and comorbid conditions, particularly those causing nocturnal discomfort or pain.

Validated screening tools for the most common comorbidities:

• Improve daytime function — the primary goal of any sleep intervention. How the patient's daytime functioning relates to sleep matters more than the raw quantity or quality of sleep reported.
• Reduce daytime impairment: fatigue, dysphoria, decreased alertness, and cognitive complaints.
• Promote subjectively sound, restorative sleep despite external (stress, noise, travel) or internal (pain, anxiety) disruptors — making sleep more resilient, not just better on a given night.
• When pharmacotherapy is used, potentiate rather than replace CBT-I.
• Use the lowest effective medication dose for the shortest duration required to meet the above goals.

Sleep hygiene

Good sleep hygiene is the cornerstone of any sleep intervention, though data indicates it is often insufficient as a standalone treatment for insomnia disorder — it is incorporated into, but does not replace, full CBT-I.

1. Keep a regular sleep-wake schedule with a consistent rising time within 60-90 minutes, 7 days per week — this reinforces the circadian sleep rhythm and prevents oversleeping.
2. Restrict time in bed to the average sleep time actually obtained over the preceding week.
3. Avoid sleeping in, extensive daytime horizontal rest, or daytime naps longer than about 15-20 minutes in the afternoon — these tend to worsen the following night's sleep.
4. Get regular exercise, about 30 minutes of activity intense enough to cause sweating; if evening exercise disrupts sleep, move it earlier in the day.
5. Avoid caffeine after mid-afternoon and alcohol after supper, and avoid recreational drugs; if smoking, stop by early evening.
6. Plan a quiet period before lights out (a warm bath can help); keep the bedroom around 16-20°C, which appears to signal the circadian rhythm toward sleep.
7. Avoid large meals late in the evening; a light carbohydrate snack before bed can be helpful.
8. Turn the clock face away — checking the time on awakening can cause performance anxiety or frustration that delays returning to sleep. Use an alarm instead.
9. Keep the bedroom dark, quiet, and free of phones, televisions, and computers; use earplugs or an eye mask if the environment can't otherwise be controlled.
10. Use the bedroom only for sleep and intimacy — using the bed as a reading, work, or media space conditions the patient to be alert there instead of drowsy. If awake and alert during the night, get up, leave the bedroom, and do something quiet until drowsy again before returning to bed.

Pharmacologic (or any) intervention is less effective when these guidelines are not followed. In mild insomnia, consistently applied sleep hygiene strategies alone may be enough to restore a normal sleep pattern.

Cognitive behavioural therapy for insomnia (CBT-I)

CBT-I is a multi-component intervention, typically delivered in up to 8 individual or group sessions, and is the genuine first-line treatment described above. Improvement is typically seen after 3-4 sessions, and consistent follow-up through the full course sets up successful treatment.^[1]

The Dysfunctional Beliefs and Attitudes about Sleep (DBAS) tool is a brief, useful way to assess maladaptive sleep-related cognitions before starting behaviour-change work.

Delivery and access

The strongest evidence for CBT-I comes from manualized, systematic therapy delivered by trained clinicians, but primary care providers have also successfully used modified, simplified versions. Barriers include a shortage of trained specialists, cost, and patient acceptability/adherence. More accessible delivery formats include digital CBT-I, an abbreviated 2-session version, and CBT-I print materials or mobile apps — internet-delivered CBT-I has proven as effective as other CBT-I formats and is often the most accessible option. Examples of digital/online CBT-I programs include Haleo, Sleepio, Somryst, CBT for Insomnia, Go! to Sleep, CBT-i Coach, and My Sleep Well. Even with good access, a meaningful minority of patients do not respond to CBT-I, and some drop out before completing it — pharmacotherapy remains a reasonable adjunct or alternative in these patients rather than a treatment failure on the clinician's part.

Other adjuncts

• Aerobic exercise — a useful modifier of stress and dysphoric mood that also promotes deeper, more restful sleep; encourage patients with insomnia to increase exercise.
• Mindfulness meditation and bright light therapy — both show modest improvement in sleep efficacy.
• Weighted blankets — a randomized trial demonstrated a meaningful reduction in insomnia severity; a reasonable low-risk adjunct.
• Acupuncture — efficacy has been shown across multiple studies, though the evidence quality is low.
• Other adjunct strategies (music, foot baths, sound masking) — evidence quality is very low, but risk is minimal, so these can reasonably be offered as low-cost adjuncts if the patient is interested.

Natural health products and OTC agents

Of the natural supplements, melatonin has the most evidence, with a modest overall effect — see Pharmacologic Treatment Options and Drug Tables for dosing. Valerian and L-tryptophan have inconsistent trial results but are occasionally clinically useful; L-tryptophan is in the Drug Database (Tryptophan), while valerian root extract (typically 300-600 mg of standardized extract at bedtime) does not currently have a Pharmasist Drug Database monograph.

Screen every patient for use of OTC sleep agents containing diphenhydramine, dimenhydrinate, or doxylamine (e.g., store-brand "PM" analgesics, Nytol, Sleep-Eze, Unisom, Gravol, cold/flu night formulations). Routine or long-term use of these agents is discouraged given the risk of cognitive impairment, hangover effects, dizziness, anticholinergic delirium, and falls — especially in older adults — and they should generally be reserved for occasional, short-term use, or for insomnia genuinely driven by histamine-mediated symptoms (e.g., allergies, atopic dermatitis) rather than as a routine sleep aid. See Special Populations and Drug Tables for the specific caveats.

Cannabis — given legalization and heavy marketing of cannabis products for sleep, ask every patient about cannabis use. There is insufficient evidence to support routine clinical use of cannabis for insomnia, and the effects of its many components on sleep remain unclear. Reserve cannabis consideration for resistant insomnia with a comorbidity that itself has supporting evidence for cannabis (e.g., certain pain syndromes, resistant anxiety), and only after treatments with better evidence have failed. Cannabis products are mainly CBD (cannabidiol) and THC (tetrahydrocannabinol) — THC is more sedating but is psychoactive, so CBD-dominant products are usually preferred initially if cannabis is considered.

Pregnancy

Disrupted sleep is a very common complaint in pregnancy and tends to worsen as pregnancy progresses; endocrine changes, physical discomfort, and bladder distension all contribute. Obstructive sleep apnea and restless legs syndrome both commonly worsen during pregnancy and can present as an insomnia complaint — screen for both rather than assuming the complaint is primary insomnia. Cognitive and behavioural techniques (CBT-I) are the treatment of choice when distress and impairment are mild to moderate.

Breastfeeding

If pharmacotherapy is required while breastfeeding, intermittent low doses of a short-acting benzodiazepine (e.g., temazepam) or a Z-drug appear relatively safe with appropriate monitoring, though the effects of chronic exposure are unknown. Trazodone is transferred into breast milk in low amounts and is not expected to affect the breastfed infant, particularly beyond 2 months of age and at doses ≤100 mg/night. Doxepin and diphenhydramine should not be used while breastfeeding, based on adverse effects reported in case reports.

Children

Nonpharmacologic therapy is first-line for pediatric insomnia, especially in young children under 6 years. Graded extinction (gradual withdrawal of parental attention to a child crying rather than settling alone) is a useful additional behavioural technique in younger children and infants. Medication use for pediatric insomnia is not well evidence-based and is guided largely by clinical experience and treatment of comorbidities.

If pharmacotherapy is genuinely needed, melatonin can be considered, starting at 0.5-1 mg at bedtime — a reasonably sized meta-analysis found it effective and tolerable for short-term treatment of sleep-onset insomnia in children and adolescents, particularly with comorbid ADHD or autism spectrum disorder, with no significant adverse effects observed in follow-up out to 7 years. L-tryptophan has shown potential efficacy specifically for pediatric parasomnias. OTC antihistamines are not recommended in children given the risk of rapid tolerance, next-day sedation, cognitive impairment, and paradoxical reaction. Prescription medication should be used extremely judiciously and targeted at a specific comorbidity — clonidine (Clonidine) has shown efficacy in a case series of children with comorbid ADHD and insomnia. Z-drugs have not demonstrated efficacy in children, and benzodiazepines and TCAs are not recommended given an unfavourable benefit-to-risk ratio in this age group.

Older adults

Older patients face higher polypharmacy and drug-interaction risk, and the deleterious effects of hypnotics (cognitive impairment, falls) are amplified — making CBT-I as first-line therapy even more essential in this population.

Safer pharmacologic options include low-dose doxepin (under 6 mg) for sleep-maintenance insomnia, or melatonin for sleep-onset insomnia. Lemborexant has also shown a favourable 6-month safety and efficacy profile specifically in older adults. If a benzodiazepine receptor agonist genuinely must be used, zolpidem is generally preferred for its shorter half-life; low-dose eszopiclone (2 mg) has shown short-term efficacy (up to 24 weeks) with a comparatively low fall risk. Low-dose trazodone and mirtazapine are commonly used off-label but have limited evidence specifically in older adults — reserve for comorbid depression/anxiety or resistant cases, with attention to fall and fracture risk.

Deprescribing of long-term hypnotic therapy should be discussed and offered, but rigid, mandatory discontinuation in a stable, functional, long-term low-dose user is not always the right answer — an overly rigid approach can push a patient toward more harmful, unmonitored substitutes (high-anticholinergic-burden OTC drugs, cannabis, or alcohol). Remember that off-label psychotropics used for insomnia, and indicated drugs used for other conditions (antihypertensives, antidepressants, antipsychotics), carry their own fall risk in this population — the comparison is never "hypnotic vs. nothing," but "this drug vs. the realistic alternative."

Renal and hepatic impairment

Zopiclone, eszopiclone, and zolpidem all require dose reduction in hepatic impairment (see Drug Tables); zopiclone also requires dose reduction in renal impairment. Lemborexant and daridorexant require dose reduction with moderate hepatic impairment or with a co-administered CYP3A inhibitor, and are not recommended with severe hepatic impairment.

Follow up frequently (every 3-6 weeks) and dispense limited hypnotic quantities (30-60 days), especially early in treatment. Reassess daytime functioning, not just reported sleep quantity, at every visit — this is the primary treatment target. If sleep performance and daytime functioning have not returned to normal after 2-6 weeks, re-evaluate the diagnosis and comorbidities and consider referral to a sleep specialist.

For patients continuing hypnotic therapy longer-term, periodically reassess ongoing need with a clear documented rationale and an explicit review of risks and benefits — long-term sedative use should be the exception, reserved for cases with significant comorbidity or failure/inability to complete nonpharmacologic therapy, not a default outcome of an initial short-term prescription. Polysomnography should be considered in resistant or chronic cases and in patients on long-term hypnotic therapy.

Monitor specifically for complex sleep-related behaviours with any Z-drug — discontinue immediately if these occur and do not rechallenge with the same drug or class.

Counsel on driving and machinery-operation timing after every hypnotic dose — follow the recommended pre-driving interval for the specific agent (generally ≥7-8 hours for zopiclone/eszopiclone/zolpidem/lemborexant/daridorexant, ≥12 hours for longer-acting benzodiazepines such as temazepam), while recognizing that sleep deprivation itself also impairs driving. Some habituation to a hypnotic's driving-impairment effect develops with regular use, but off-label sedatives used for insomnia have not been rigorously tested for this and may be more impairing than expected. Use clinical judgment, factor in the nature of the patient's driving (commercial drivers need more vigilance), and be aware of provincial requirements for reporting an unfit driver.

Screen for emerging mood or anxiety disorders throughout treatment, not just at diagnosis — chronic sleep disturbance lasting at least 1 year increases the risk of a subsequent mood disorder, and insomnia often occurs as a symptom, comorbidity, or prodrome of depression or anxiety. A reduced response to insomnia treatment should specifically prompt re-screening for these.

1. Confirm CBT-I components were genuinely implemented — stimulus control, sleep restriction, and cognitive therapy, not just generic sleep hygiene advice given in isolation, which is not sufficient as monotherapy.
2. For any pharmacotherapy trialed, confirm adherence, correct timing (30-60 minutes before the intended bedtime, with awareness that taking a Z-drug with food delays and blunts absorption), and correct dose.
3. Explicitly screen or re-screen for sleep apnea, restless legs syndrome, depression, anxiety, and medication-induced insomnia using the tools above — these are the most common reasons for apparent treatment failure.
4. If a comorbidity is newly identified or undertreated, treat or optimize it — insomnia often improves once the comorbidity is addressed, and CBT-I remains effective alongside comorbidity treatment rather than needing to wait for it.
5. If an adequate CBT-I trial and an adequate pharmacotherapy trial have both failed, consider switching drug class (e.g., a dual orexin antagonist if a Z-drug was ineffective or not tolerated) rather than repeatedly escalating within the same class.
6. Refer to a sleep medicine specialist for resistant or chronic cases, suspected other primary sleep disorder, or before committing to open-ended long-term hypnotic therapy.

• Treating insomnia with medication alone without ever offering CBT-I as genuine first-line therapy.
• Delivering sleep hygiene advice alone and calling it "CBT-I" — sleep hygiene is only 1 of 5 components and is not sufficient as monotherapy for insomnia disorder.
• Reflexively recommending diphenhydramine or doxylamine as a routine, indefinite "OTC-safe" solution — anticholinergic burden, tolerance, and next-day impairment make these poor choices beyond occasional, short-term use, especially in older adults.
• Prescribing a benzodiazepine or Z-drug in an older adult without discussing fall, confusion, and cognitive risk and safer alternatives first, per the Beers Criteria.
• Reaching for flurazepam, nitrazepam, or triazolam as a first-choice hypnotic — the longer-acting agents (flurazepam, nitrazepam) accumulate and cause more next-day impairment, and triazolam carries a distinct risk of amnesia, agitation, and rebound insomnia.
• Applying rigid, mandatory deprescribing to a stable, functional, long-term low-dose hypnotic user — an overly rigid approach can push the patient toward more harmful, unmonitored substitutes.
• Missing an underlying sleep apnea, restless legs syndrome, depression, or medication-induced cause and treating the insomnia symptom indefinitely in isolation.
• Failing to counsel on driving/machinery-operation timing after a hypnotic dose, particularly since patients can feel "fully awake" despite measurable next-day impairment.
• Combining any hypnotic with alcohol or another CNS depressant.
• Continuing an ineffective or poorly tolerated hypnotic past the planned reassessment point instead of stopping and reassessing the diagnosis.

How to use this table: Doses are adult doses and require patient-specific confirmation, including renal/hepatic adjustment and reassessment against the Beers Criteria in older adults. Select any drug name to open its complete Pharmasist Drug Database page. All benzodiazepine receptor agonists (Z-drugs), benzodiazepines, and orexin antagonists require a planned sleep opportunity of at least 7-8 hours before the patient needs to be fully alert.

Benzodiazepines officially indicated for insomnia in Canada

4 benzodiazepines carry a Health Canada insomnia indication. Only temazepam is generally recommended as a first-choice agent within this class.

Benzodiazepine receptor agonists (Z-drugs)

Act at the benzodiazepine receptor but are a distinct drug class with 10-40 times less myorelaxant effect than benzodiazepines and no demonstrated worsening of sleep apnea. Complex sleep behaviours (e.g., sleep-related eating or driving with no recollection) have been reported with zopiclone and zolpidem — review this risk with every patient and counsel that the medication should be taken in bed with limited access to dangerous items (car keys, kitchen tools). Combination with alcohol increases the risk of all adverse events, including complex sleep behaviours.

Dual orexin receptor antagonists (DORA)

Reduce wakefulness and unwanted wake-sleep transitions rather than directly sedating via GABA-A. Indicated for both sleep-onset and sleep-maintenance insomnia. No evidence of withdrawal symptoms or rebound insomnia on discontinuation, efficacy data to 12 months, and minimal abuse potential — they are not controlled substances. Not recommended in combination with alcohol or other CNS depressants; contraindicated in narcolepsy.

Low-dose tricyclic antidepressant

At the very low doses used for insomnia, doxepin acts as a selective H1-receptor antagonist rather than through its antidepressant mechanism, avoiding typical TCA adverse effects. Indicated specifically for sleep-maintenance difficulty, with trial data to 3 months. Not associated with rebound or dependence, and has minimal next-day sedation.

Antihistamines (OTC)

First-generation, sedating antihistamines with a flat dose-response above the doses listed and significant anticholinergic burden. The 2023 AGS Beers Criteria specifically recommend against routine use in adults 65 and older given markedly increased anticholinergic toxicity, delirium, and CNS adverse effects. Recommend intermittent use only (no more than about 4 nights per week) and advise patients to seek professional advice if a sleep aid is needed for more than about a week.

Off-label sedating antidepressants

Rational for comorbid insomnia despite limited direct evidence — e.g., for insomnia with a depressive disorder. All are associated with next-day sedation and, for TCAs and mirtazapine, weight gain.

Off-label anticonvulsants

Rational specifically for insomnia associated with centralized pain syndromes (fibromyalgia, neuropathic pain, restless legs syndrome). Weight gain and next-day sedation are common; CNS depression and cognitive impairment can also occur.

Off-label antipsychotics

Not recommended for insomnia in the absence of a comorbidity that itself indicates antipsychotic therapy, given unacceptable anticholinergic/neurological toxicity risk (first-generation) or metabolic syndrome risk (second-generation) relative to any sleep benefit. Reserve for insomnia with very resistant bipolar disorder, schizophrenia, or (for low-dose quetiapine/olanzapine) major depressive disorder augmentation — dosed per the treatment of the underlying disorder, monitoring for metabolic effects as usual, not independently dosed for insomnia alone.

Other off-label agents

Natural health products

Melatonin has the most consistent evidence of the natural supplements, with a modest overall effect; valerian and L-tryptophan have inconsistent trial evidence but are occasionally clinically useful.