The cyclist tracks every macro, every gel, every supplement, every recovery session. Whoop strap on. Garmin synced. RecoveryPlus app paid through 2027. A meticulous training log going back two years.

He averages 6 hours 45 minutes of sleep on weeknights.

The intervention with the largest performance return in his entire stack is the one he isn't doing. And it's free.

This is the most underused intervention in serious-amateur endurance training: deliberately sleeping more than you currently do, for a sustained block, when there's a deficit to repay. Here's the data, the dose-response, and the practice that gets you the gain without inventing a new supplement budget line.

Mah2011 — the seminal study

Cheri Mah at the Stanford Sleep Disorders Clinic ran what is still the single best demonstration of the sleep-extension effect on athletic performance (Mah et al. 2011). The protocol was simple. Eleven members of the Stanford men's varsity basketball team. Two-to-four-week baseline of habitual sleep (the players slept what they normally slept; mean 6 hours 30 minutes per night). Then 5–7 weeks of sleep extension — instruction to obtain as much nocturnal sleep as possible, with a target of at least 10 hours time-in-bed each night.

Actigraphy confirmed the intervention worked: mean total sleep time during extension was 110.9 minutes longer per night than baseline. Players were sleeping nearly 2 hours more.

The performance changes:

• 282-foot sprint times improved from 16.2 seconds at baseline to 15.5 seconds after extension — a 4% improvement.
• Shooting accuracy improved roughly 9% on free throws and 9.2% on three-point field goals.
• Reaction time on the Psychomotor Vigilance Task decreased significantly (faster responses).
• Mood scores on POMS improved — vigor up, fatigue down.

None of this required a new training stimulus. The players didn't add interval work, change diet, or take supplements. They went to bed earlier and slept until they woke naturally. The basketball performance metrics moved by margins that no legal supplement on the market reliably produces.

The study has been criticised on small sample size and observational design, fairly. But the effect size is striking and the direction has replicated.

Roberts2019 — it replicates in endurance cyclists

The Mah finding was specific to power and skill sport. Roberts and colleagues at Deakin University asked whether the effect translates to endurance (Roberts et al. 2019). Counterbalanced crossover design. Nine trained endurance cyclists / triathletes. Three conditions: sleep restriction, normal sleep, sleep extension. Each condition was sustained for seven nights, with a TT performance test on multiple days within each block.

The finding: finishing time on a standardised TT was approximately 2 minutes faster after three nights of sleep extension compared to three nights of normal sleep (56.8 min vs 58.7 min) — a 3% performance improvement. Sleep restriction in the opposite direction also moved time — two nights of restricted sleep produced a 3% impairment vs normal.

The symmetry matters. Sleep is acting like training load — deficit hurts, extension helps, and the magnitude in either direction is meaningful for competitive amateurs where a 3% margin in a 60-minute TT is roughly two podium places.

Roberts also confirmed the practical recommendation: endurance athletes should sleep more than 8 hours per night to optimise performance. Below 7 hours, performance declined measurably. The 6-hour-45-minute amateur from the opening of this article is leaving a 3% performance margin on the bedroom floor every night.

Fullagar2015 — the systematic review

Fullagar and colleagues at Saarland University reviewed the available literature on sleep and athletic performance (Fullagar et al. 2015). Their conclusion: sleep loss reliably impairs exercise performance, particularly sport-specific and cognitive performance, though the effects on raw maximal physical efforts and gross motor performance are more conflicting.

What this means in practice: the inverse — extension when there's a deficit to repay — produces the cleanest gains in sport-specific performance and cognitive sharpness. Pure maximal strength tests are less responsive. But cycling is overwhelmingly sport-specific endurance, not raw maximal force production. The Fullagar systematic review's conclusion sits squarely in support of Roberts.

Section 04Why most amateurs leave this on the table

Three reasons.

One — baseline drift. Most amateurs don't actually know their current sleep duration. The subjective estimate is consistently off by 30–60 minutes. Without a wearable or actigraphy, I sleep about 7 hours is usually 6 hours 15 minutes of actual sleep. The gap to the 8-hour minimum is bigger than people think.

Two — sleep extension feels wasted. Going to bed at 9:30 instead of 10:45 "costs" you a TV episode, a podcast, a glass of wine. Those activities have a vivid immediate payoff. The performance gain from the extra 75 minutes is delayed, distributed, invisible on Tuesday morning's training file. Most riders trade the obvious cost for the invisible benefit by default.

Three — there's no app to pay for. A supplement subscription is concrete. A sleep extension protocol is just go to bed earlier. The intervention with no purchase cost feels less serious than the intervention with a credit card subscription. Both research literatures suggest the free intervention works better.

Section 05How to actually do this

The protocol from the literature, applied to a competitive amateur cyclist:

Target time-in-bed, not sleep time. Aim for 9 hours 30 minutes time-in-bed nightly for at least 7 consecutive nights. Actual sleep time will be 8–9 hours after accounting for sleep onset and wake-after-sleep-onset.

Anchor on consistent bedtime, not consistent wake time. Most amateur schedules force a wake-time anchor (kids, work). Move the bedtime earlier. Pick the new bedtime, hold it for 7+ consecutive nights, including weekends.

Eliminate the obvious sleep robbers. Caffeine cutoff at noon (caffeine half-life is ~5–6 hours). Screens off 60 minutes before bedtime, or use a hard-blue-light filter. Bedroom temperature 17–19°C. Total dark.

Look for the deficit signal. If you fall asleep within 5 minutes of getting into bed, you have a sleep debt. (Healthy sleep onset is 15–20 minutes. Falling asleep within 5 means you're sleep-deprived.) If you can wake without an alarm at your target time, your sleep duration matches your need.

Sustain for at least 7 nights before evaluating. Roberts' 3% TT improvement showed up after 3 nights of extension and grew slightly across the 7-night block. Don't bail after one night because you don't feel transformed.

Section 06Applying it with HELIOS

HELIOS measures total sleep time, sleep onset latency, wake-after-sleep-onset, time in each stage, and — critically — your personal baseline sleep need based on objective recovery markers (HRV, resting HR, body temperature) over a rolling window.

The daily call isn't did you sleep 8 hours. It's your 7-day average sleep is 7 h 12 min, your personal baseline for clean Z2 next-day performance is 8 h 5 min, you have a recurring 53-minute deficit per night. The recommendation is concrete: target a 9 h 30 min time-in-bed window starting tonight, repeat for 7 nights, retest baseline.

The ring does the math. The discipline is yours.

Section 07Bottom line

Sleep extension is the largest-return, lowest-cost, best-evidenced intervention in the entire serious-amateur cycling toolkit. Mah's data shows 4–9% performance gains in athletes with existing deficits. Roberts replicates the direction in endurance cyclists at 3% TT-time improvement after just three nights. Fullagar's systematic review confirms that sleep loss reliably degrades performance — which is the same finding in mirror image.

Most amateurs leave this on the table because the intervention is free, the cost is delayed gratification, and the gain is invisible the next morning.

An extra hour in bed for a week beats every supplement on the market.