Sleep Journal

How I use this journal myself

I built the journal after three months of guessing. I knew some nights left me wrecked and others didn’t, but by the time I noticed the pattern the cause was four days behind me. The fix turned out to be embarrassingly low-tech: one rating, one number, one sentence, written before I looked at my phone. Within two weeks the log was already arguing with my assumptions. I had blamed screens. The entries blamed a 4 p.m. coffee and a late, heavy dinner.

What to write in the note field

Keep it to the handful of inputs that plausibly move sleep, and record the timing, not just the fact. “Coffee” is useless; “coffee at 3:40” is data. The five worth logging almost every night:

• Last caffeine — clock time, not cup count. Caffeine’s half-life runs roughly five to six hours, so a mid-afternoon cup is still working at midnight.
• Alcohol — units and how late. Alcohol shortens sleep latency and then fragments the second half of the night as it clears.
• Last meal — a large late meal raises core temperature and reflux risk right as your body is trying to cool down.
• Exercise — type, intensity, and how many hours before bed.
• Stress or a specific worry — one clause is enough. This is the variable people skip and the one that most often explains an outlier.

Reading your own data after two weeks

Don’t analyse night by night — single nights are noise. Wait until you have fourteen entries, then sort them mentally into your three best and three worst nights and look only at what those six have in common. Real signals are unglamorous and repeat: a consistent wake time, a caffeine cutoff around eight hours before bed, a bedroom under about 19 °C, and a wind-down that starts before you feel tired rather than after.

One caution about hours. Logged time in bed is not time asleep, and most people over-report by thirty to sixty minutes. Treat the number as a relative trend, and treat your own rested rating as the more honest measure — it tracks how you actually function far better than a duration you estimated at 7 a.m.

When a journal is the wrong tool

If tracking makes you anxious — checking the log, dreading the entry, lying awake calculating — stop. That loop has a name in the clinical literature, and rumination about sleep reliably makes sleep worse. Journal for two to four weeks, extract the pattern, then close it. And if you have logged persistent trouble falling or staying asleep at least three nights a week for three months, a journal is no longer the intervention you need: that is the threshold clinicians use for chronic insomnia, and it responds to CBT-I, not to more measurement. Bring the log to the appointment — a real fourteen-night record is genuinely useful to a clinician.