Sleep restriction therapy: the behavioural fix for chronic insomnia

The problem with spending too long in bed

Most people with chronic insomnia do the reasonable thing: they go to bed earlier, stay in bed later, and try harder to sleep. This makes insomnia worse.

Lying awake in bed for hours teaches the brain that bed is a place for wakefulness, worry, and frustration — not sleep. By the time someone with chronic insomnia finally falls asleep, the bed has become the most alerting place in their home. The problem is behavioural and psychological, not a chemical deficiency, which is why the Therapeutic Guidelines and RACGP both recommend behavioural treatment — CBT-I — before any medication.

Sleep restriction therapy (SRT) is the most powerful single component of CBT-I. It counterintuitively solves insomnia by compressing time in bed, not extending it. The Australasian Sleep Association, AASM 2021, and ESRS 2023 all endorse SRT as a stand-alone treatment for adults with chronic insomnia. Its effects are durable — unlike medication, which loses effectiveness within weeks.

This article explains what SRT is, how to do it, what to expect, and when it is not appropriate.

A. Core clinical — the AU general practice framework

What chronic insomnia is

Chronic insomnia disorder requires all three of (RACGP; Sleep Health Foundation):

• Difficulty initiating or maintaining sleep, or early-morning awakening
• Occurring ≥3 nights per week for ≥3 months
• Causing meaningful daytime impairment (fatigue, mood, concentration, function)

The daytime impairment criterion matters. People who sleep 6 hours and feel well are short sleepers, not insomniacs. SRT is for people whose sleep difficulty genuinely impairs their day.

About 10–15% of Australian adults meet criteria for chronic insomnia at any time. It is more prevalent in women, older adults, shift workers, and people with comorbid anxiety, depression, or chronic pain.

Why SRT works: the three mechanisms

Maurer, Espie, Kyle 2018 describe the “triple-R” model:

1. Homeostatic sleep-pressure build-up. Restricting time in bed below habitual waking duration creates mild, controlled partial sleep deprivation. Adenosine accumulates across the day, accelerating sleep onset and deepening sleep continuity once in bed.
2. Stimulus-control reconditioning. Compressing waking-in-bed time removes the pairing between bed and wakefulness. Bed once again becomes a reliable sleep-onset cue.
3. Circadian re-regulation. Anchoring a fixed wake time — regardless of the previous night — stabilises the morning light-entrainment signal and re-entrains the sleep-wake circadian phase.

Who should not attempt self-administered SRT

Screen for these before starting. Refer to your GP if any apply (ESRS 2023; Sleep Health Foundation):

• Bipolar disorder or mania/hypomania in the last 12 months — sleep deprivation is a well-established manic precipitant. A modified CBT-IBD protocol (6.5-hour floor, psychologist-supervised) exists for this group.
• Active suicidal ideation or severe untreated depression — defer until mood is stabilised under professional care. SRT in well-treated depression is safe and beneficial; it is the unstabilised, suicidal group who need to wait.
• Seizure disorder — sleep deprivation lowers the seizure threshold in most epilepsies. Defer to specialist co-management.
• Untreated obstructive sleep apnoea (OSA) — SRT compresses already-fragmented OSA sleep, worsening hypoxia. CPAP must be established and tolerated for ≥4 weeks first.
• Safety-critical occupational driving or hazardous shift work — weeks 1–3 carry measurable reaction-time impairment. Defer until a period of lighter duties.
• Pregnancy — limited data; seek GP or obstetric advice first.

The protocol, step by step

Step 1 — Two-week baseline sleep diary. Log each night: time into bed, estimated time to sleep onset, number and duration of awakenings, final wake time, time out of bed. Use the Sleep Health Foundation sleep diary or the CBT-i Coach app (free, auto-calculates). At the end, calculate:

• Time in bed (TIB) — mean minutes between lying down and final rising
• Total sleep time (TST) — mean estimated actual sleep duration
• Sleep efficiency (SE) — TST ÷ TIB × 100%

Step 2 — Set the initial sleep window. Sleep window = mean TST + 30 minutes. Floor is 5.5 hours — never go lower. A person sleeping an average of 5 hours gets a 5.5-hour window; a person sleeping 6 hours gets 6.5 hours.

Step 3 — Anchor a fixed wake time. Choose the latest realistic wake time you can sustain seven days a week — usually your earliest work-day alarm. This time is fixed, regardless of how badly you slept. Calculate bedtime by subtracting the sleep window from the wake time. Example: wake 06:30, window 6 hours → earliest bedtime 00:30.

Step 4 — Layer stimulus-control rules (non-negotiable companions to SRT).

• Bed is for sleep and sex only — no phones, TV, reading, or work
• If awake in bed for an estimated 20 minutes, get up, go to a dim quiet room, return only when sleepy
• No daytime naps (or ≤20 minutes before 3 pm only if genuinely unavoidable)
• No caffeine after 2 pm; no alcohol within 3 hours of bedtime

Step 5 — Weekly titration based on sleep efficiency (Spielman 1987; AASM 2021):

Mean SE (past 7 nights)	Action
≥90%	Extend sleep window by 15 minutes (move bedtime 15 minutes earlier)
85–89%	Hold the current window
<85%	Reduce sleep window by 15 minutes — never below the 5.5-hour floor

Step 6 — Continue until target reached. Target: TST ≥7 hours with SE ≥90% sustained across two consecutive weeks, or the Insomnia Severity Index dropping below 8. Typical course is 4–8 weeks.

What to expect week by week

Weeks 1–2: Daytime sleepiness worsens — this is the treatment working, not a sign to stop. Kyle 2014 confirmed measurable reaction-time and vigilance impairment during this phase. Avoid long, monotonous, or night-time drives. Use public transport if needed. Use caffeine before 2 pm. Stick rigidly to the wake time even after a terrible night — this is the single most important rule.

Weeks 3–4: Sleep consolidates. Sleep efficiency begins climbing. Daytime energy starts recovering.

Weeks 6–8: Most adults with uncomplicated chronic insomnia have reached a consolidated, extended sleep window with restored daytime function. Some older adults settle on a 6–6.5-hour window — this is normal and adequate if daytime function is good.

B. Evidence — what the trials actually show

Effect size

Spielman, Saskin, Thorpy 1987 — the original SRT trial — enrolled 35 adults with a mean insomnia duration of 15.4 years. Eight weeks of SRT produced statistically significant improvements in total sleep time, sleep latency, total wake time, sleep efficiency, and subjective insomnia severity. Modern meta-analyses (Maurer & Schneider 2021 Sleep Med Rev) confirm moderate-to-large effects across all sleep-diary and ISI outcomes, comparable to or exceeding multi-component CBT-I.

The Trauer et al. 2015 Ann Intern Med CBT-I meta-analysis of 20 RCTs found:

• Sleep-onset latency reduced by ~19 minutes
• Wake-after-sleep-onset reduced by ~26 minutes
• Sleep efficiency improved by ~10%
• Effects durable at 12-month follow-up — unlike medication

Self-administered versus therapist-delivered

Espie et al. 2012 randomised 164 adults to fully-automated digital CBT-I (Sleepio), imagery-based placebo, or treatment-as-usual. The CBT-I arm produced significant improvements in SE, sleep-onset latency, and wake-after-sleep-onset, with ~76% reaching healthy sleep efficiency. The Soh 2020 digital CBT-I meta-analysis confirmed digital delivery is non-inferior to face-to-face for most adults with uncomplicated insomnia.

Self-administered SRT is an appropriate first-pass for adults without the contraindications listed above. Escalate to a psychologist-delivered program if no response by weeks 6–8, or if comorbidity (PTSD, bipolar, chronic pain, OSA) is present.

SRT with comorbid depression

Manber et al. 2008 Sleep and Manber et al. 2016 J Clin Psychiatry both showed CBT-I added to antidepressant therapy improved both sleep and depressive outcomes, outperforming antidepressant alone. A 2024 meta-analysis of single-component SRT found it also improves depressive symptoms independently. In well-treated depression, SRT is safe and beneficial.

C. CBT-I delivery formats in Australia

SRT is one of five CBT-I components — alongside sleep hygiene, stimulus control, cognitive restructuring, and relaxation. The components can be delivered via:

Format	Access	Cost
This Way Up Insomnia — AU-developed 4-lesson online CBT-I (UNSW / St Vincent’s)	Free with GP referral	Free
CBT-i Coach app — sleep diary, SRT calculator, audio relaxation	No referral needed; iOS/Android	Free
Sleepio — fully-automated CBT-I; NICE-endorsed in UK	Self-pay	~AUD $400/yr
Psychologist under Mental Health Care Plan — 6–8 sessions typical	GP referral; MBS items 2715/2717	Partially subsidised
GP-delivered Brief Behavioural Therapy for Insomnia (BBTi) — stimulus control + SRT; 4 sessions	Via GP with AMHTC training	MBS items 2721–2727

This Way Up is the recommended AU first-line digital option — free with a GP referral, evidence-based, and developed locally.

D. Australian operations

Medicare items

• Standard GP consults — 23 / 36: initial protocol-card visit, 2-week and 4-week reviews
• Telehealth — 91891 / 91892 phone consults for follow-up reviews
• Mental Health Care Plan — 2715 / 2717 for GP referral to CBT-I-trained psychologist when insomnia is linked to anxiety, depression, or adjustment
• GPCCMP — 965 / 967: when chronic insomnia coexists with another chronic condition; enables direct allied-health referral (post 1 July 2025 reform)
• 75+ Health Assessment 705; ATSI Health Assessment 715: include sleep screening

Finding a CBT-I-trained psychologist

• APS Find a Psychologist — filter by “sleep disorders” or “insomnia”
• Australasian Sleep Association directory
• Sleep Health Foundation provider listings

Veterans

DVA Gold and White Card holders: CBT-I including SRT is fully covered for veteran insomnia and PTSD-related insomnia via Open Arms Veterans & Families Counselling and DVA-funded psychologists.

Driving and medico-legal documentation

Austroads Assessing Fitness to Drive 2022 requires drivers to self-report unfitness. Commercial and heavy-vehicle drivers have stricter standards — defer SRT until leave can be arranged if they cannot avoid safety-critical driving during weeks 1–3. Document the driving counsel conversation in the medical record.

E. Special populations

Older adults. Sleep need declines with age — 6–7 hours is normal in healthy adults aged ≥65. Applying SRT to a 72-year-old who sleeps 6 hours but feels well is inappropriate. Screen for genuine daytime impairment before starting. When SRT is indicated in older adults, falls risk from week-1 sleepiness is heightened — consider a 6-hour floor and closer GP monitoring. SRT in this group reduces hypnotic dependence, which is a priority given the falls-and-fracture risk of benzodiazepines and Z-drugs in the elderly.

Shift workers. Standard SRT anchor times do not fit rotating shift schedules. Pair with advice from the Australasian Sleep Association shift-work resources; consider specialist referral.

Comorbid chronic pain. SRT is effective but tolerability of week-1 sleepiness is lower in people with pain, as sleep loss worsens pain sensitivity. Pair with pain-CBT and consider a slower progression.

Pregnancy. No formal contraindication but limited data. Seek GP or obstetric advice; a modified 6-hour floor behavioural approach without strict restriction may be preferred.

When to escalate

Refer to a psychologist or sleep physician if:

• Self-administered SRT produces no improvement after 6–8 weeks
• OSA or restless legs syndrome is suspected
• Comorbid bipolar disorder, PTSD, severe MDD, or substance use disorder is present
• Parasomnia (REM sleep behaviour disorder, sleepwalking) is present — sleep deprivation worsens these
• Suicidal ideation emerges or worsens during SRT — halt immediately and seek urgent GP review

What this article is and is not

This is general health information drawn from current Australian guidelines — the Therapeutic Guidelines, AMH, Sleep Health Foundation, and Australasian Sleep Association — and major clinical trials. It is not personal medical advice and does not create a doctor–patient relationship. Decisions about your specific treatment are made with your GP and treating clinicians.

For Australian consumer resources: Sleep Health Foundation, HealthDirect — Insomnia, Better Health Channel, This Way Up.

For acute mental-health crisis: Lifeline 13 11 14, Beyond Blue 1300 22 4636.

---

Sources cited

1. Sleep Health Foundation — CBT-I
2. Australasian Sleep Association — Insomnia GP Guideline 2024
3. Sleep GP Resources AU — Bedtime / Sleep Restriction Therapy
4. RACGP — Clinical resources
5. Therapeutic Guidelines (eTG) — Psychotropic: Insomnia
6. Australian Medicines Handbook (AMH)
7. This Way Up — Insomnia online CBT-I
8. CBT-i Coach app
9. Edinger et al. AASM 2021 — Behavioural treatments for chronic insomnia (JCSM 17:255)
10. Riemann et al. ESRS 2023 — European insomnia guideline (J Sleep Res e14035)
11. Spielman, Saskin, Thorpy 1987 — Treatment of chronic insomnia by restriction of time in bed (Sleep 10:45)
12. Maurer, Espie, Kyle 2018 — How SRT works: triple-R model (Sleep Med Rev 42:127)
13. Trauer et al. 2015 — CBT for chronic insomnia meta-analysis (Ann Intern Med 163:191)
14. Espie et al. 2012 — Digital CBT-I RCT (Sleep 35:769)
15. Soh et al. 2020 — Digital CBT-I meta-analysis (Sleep Med 75:315)
16. Manber et al. 2008 — CBT-I augments depression outcomes (Sleep 31:489)
17. Manber et al. 2016 — CBT-I augments antidepressant in comorbid depression (J Clin Psychiatry)
18. Kyle et al. 2014/2018 — Daytime sleepiness and driving during SRT (Sleep Med)
19. Austroads — Assessing Fitness to Drive 2022
20. HealthDirect — Insomnia
21. Better Health Channel — Sleep hygiene