Sleep as a training input

For most of my thirties I treated sleep like a bank overdraft. Borrow Monday, pay back Saturday. Repeat.

Then I bought a Withings sleep pad and watched my deep sleep average for a month. The number was 38 minutes a night. The textbook range is closer to 90 minutes. I had been training hard, eating well, and quietly bleeding the recovery I needed.

Sleep is not a lifestyle accessory. It is a training input. If you don't measure it, you're guessing at half your programme.

What the consensus says

The sleep research has converged on a few unromantic numbers.

• Adults need 7 to 9 hours of total sleep, with 7 as a floor not a target
• Deep sleep (slow-wave sleep) is when most physical recovery happens
• REM sleep is when most memory consolidation and emotional processing happens
• Both decline with age, deep sleep especially after 40

The night you skip an hour, you lose disproportionately from the deep stages, because they cluster in the first half of the night. Cutting sleep from 8 to 6 hours doesn't take 25% off your recovery. It takes more.

What I noticed when I started measuring

The wearable confirmed what my body already knew. The patterns were stark.

• Two beers with dinner: deep sleep dropped by 30% that night
• Late-night training (after 8pm): deep sleep dropped, REM dropped
• Stress at work: total sleep held, deep sleep collapsed
• A run in the morning sun: deep sleep up the following night

None of this was new science. It was just new evidence about my own particular body. The number on the screen made the abstract concrete.

The inputs that move the needle

The literature on sleep optimisation is large. The interventions with the strongest evidence are also the simplest.

• A consistent sleep and wake time (within 30 minutes), seven days a week
• Morning daylight exposure within 30 minutes of waking
• A cool bedroom (around 18°C is the lab consensus)
• No alcohol within 3 hours of bed (within 4 if you can manage it)
• No caffeine after lunch (caffeine has a half-life of 5 to 7 hours)
• A wind-down routine of 30 to 60 minutes with dim light

You don't need all of them. You probably need three you'll actually do.

The training-recovery loop

Here's what changed in my training when sleep stopped being an afterthought.

A week of 7+ hours with 60+ minutes of deep sleep gave me:

• Faster strength progression (the bar moved week to week instead of month to month)
• Lower resting heart rate (down 5 beats over six weeks)
• Higher HRV (a proxy for recovered nervous system)
• Better appetite regulation (less mindless snacking, especially after 9pm)

A week of 6 hours with 30 minutes of deep sleep gave me:

• Stalled lifts (same numbers, harder reps)
• Higher resting heart rate
• Cravings, especially carb cravings, especially after dinner
• A short fuse (asked my partner, she confirmed)

The training didn't change. The recovery did.

The wind-down that actually works for me

I tried elaborate routines. I gave them up. The minimal version that survives:

• Phone out of the bedroom by 9:30pm (in a kitchen drawer, not on the bedside table)
• Lights down to lamps only after 9pm (overhead lights are too bright)
• A book, paper not screen, for 20 minutes
• Bedroom at 18 to 19 degrees, window cracked if needed

That's it. No magnesium spray, no blue-light glasses, no eight-step protocol. The phone and the lights do most of the work.

When the data says something is wrong

If you measure sleep for a month and notice patterns you can't explain (chronic short total time, very low deep sleep, frequent waking, daytime exhaustion despite "enough" hours), that's worth a GP conversation. Sleep apnoea is dramatically under-diagnosed in men over 40, especially men with any neck circumference at all. The screening is simple. The treatment, if needed, is well-established.

A wearable can flag a pattern. It can't diagnose. The GP and (if relevant) a sleep clinic can.

What to do this week

Pick one input. Just one.

• Move bedtime 30 minutes earlier (and hold it for two weeks)
• Get phone out of bedroom for seven nights
• Stop caffeine after lunch for a week
• Get 10 minutes of morning sun before 9am

Measure if you can. Notice if you can't. The body keeps better records than you do.

(For chronic insomnia, suspected apnoea, or sleep that doesn't respond to the basics, see your GP. Cognitive behavioural therapy for insomnia, CBT-I, has the strongest evidence base and is usually first-line.)

Sleep is the work.