341 - Overcoming insomnia: improving sleep hygiene and treating disordered sleep with CBT-I

KEYPOINTS

• 01

  Insomnia is common, clinical and episodic, and what keeps it going isn't what started it.

  Mason, a UCSF psychologist who trained where CBT-I co-inventor Richard Bootzin worked, estimates that about 90% of adults struggle with insomnia at some point. At any moment, 5–10% meet the definition: roughly three months of bad sleep that is distressing and impairs life. Predisposing traits plus a trigger such as job loss, divorce or a crash start it. Coping behaviours—Benadryl or Ambien, naps, scrolling or reading in bed—keep it going after the trigger is gone.

• 02

  CBT-I treats current behaviour, not the original cause, so one recipe fits most types.

  Trouble falling asleep, waking at night and waking too early all get the same intervention; for night waking, the racing mind just shows up later. Mason doesn't ask what caused the insomnia until the end of the first session. Done well, about 50% reach remission and about 70% improve meaningfully. Failure mostly comes from not doing it—or from distressed natural short sleepers who may not need fixing.

• 03

  Stimulus control: the bed is only for sleep and sex.

  This descends from a 1970s study in which struggling students could study in only one library carrel. People with insomnia often 'camp out' in bed—snacking, TV, podcasts—hoping to catch sleep, and break the bed–sleep association. Dismantling studies show CBT-I fails without stimulus control or time-in-bed restriction.

• 04

  Time-in-bed restriction matches time in bed to the sleep your body actually produces.

  From a 7-day paper sleep diary, average sleep plus 30 minutes sets the time in bed, with a floor of about 5.5 hours. If someone sleeps about 6 hours with a 7 a.m. wake time, bedtime becomes 12:30 a.m. When sleep efficiency (time asleep ÷ time in bed) exceeds 85%, bedtime moves 15 minutes earlier each week. Patients get worse for the first couple of weeks.

• 05

  Fix the wake time first—a consistent wake time alone helps 1–2 of every 8 patients.

  In Mason's 'democracy within a dictatorship,' she sets the wake time from what the diary shows the person can actually do, not their aspirational 5 a.m. She runs one week with that wake time and no naps before calculating a bedtime. You can force a wake time with an alarm, but not a bedtime. During treatment, every day that ends in 'day' has the same wake time.

• 06

  Scheduled worry time keeps worries from 'demanding attention' at night.

  Saying 'stop worrying' doesn't work. Instead, book a non-negotiable hour a day and jot down intrusive worries until then. Mason also has patients track how much they believe a 1 a.m. or 7 a.m. thought at different times of day. A patient with MS believed 'I'll never make it through today' at 7 a.m., but 0% by 5:30 p.m. She turned 7–9 a.m. into Italian-learning time.

• 07

  Keep the body cool and the hands and feet warm.

  Mason swaps down comforters and duvets for cotton blankets ('if I had five cents...') and recommends a room in the mid-60s °F. Warm hands and feet help the core dump heat for sleep onset; socks or a small auto-off heating pad are fine. She's wary of heated settings on mattress-cooling devices, citing the electric-blanket era. Avoid a cold plunge before bed, and test whether a sauna helps or overstimulates you.

• 08

  Caffeine, including decaf, ends by 11 a.m.—but don't cut the amount.

  Caffeine's half-life is about 4–6 hours. Decaf can carry about 15% or more of regular caffeine, and one outside test found a tall Starbucks with about 310 mg against a stated 155 mg. Mason keeps the total the same to avoid withdrawal and just moves it before 11. She doubts truly caffeine-immune people exist.

• 09

  Melatonin is the starter's pistol, not a car in the race.

  Supplements often don't contain what the label says, and taking melatonin at the wrong time can push sleep before the body is ready, leading to night waking. Mason stops supplements during treatment. Exception: beta-blockers suppress melatonin, and one patient on metoprolol for 30 years needed only a small nightly melatonin dose.

• 10

  Taper hypnotics slowly, and only after stabilising.

  First come one drug, one dose, one time, with no middle-of-the-night top-ups and no 'rotating' Ambien, trazodone and Benadryl. After CBT-I, patients reduce in steps they rate as 1 out of 10 on a stress scale—say, 12.5 to 12 mg—weighed on a jewellery scale and held about three weeks, often over a year. The taper pauses during life crises, always alongside the prescriber. Mason worries least about low-dose trazodone.